**Validation Approaches**

**Chapter 7**

Provisional chapter

**Validation of Analytical Methods**

Validation of Analytical Methods

Additional information is available at the end of the chapter

Additional information is available at the end of the chapter

Method validation is a key element in the establishment of reference methods and within the assessment of a laboratory's competence in generating dependable analytical records. Validation has been placed within the context of the procedure, generating chemical data. Analytical method validation, thinking about the maximum relevant processes for checking the best parameters of analytical methods, using numerous relevant overall performance indicators inclusive of selectivity, specificity, accuracy, precision, linearity, range, limit of detection (LOD), limit of quantification (LOQ), ruggedness, and robustness are severely discussed in an effort to prevent their misguided utilization and ensure scientific correctness and consistency among publications.

DOI: 10.5772/intechopen.72087

Analytical method validation is an essential requirement to perform the chemical evaluation [1–3]. Method validation is a procedure of performing numerous assessments designed to verify that an analytical test system is suitable for its intended reason and is capable of providing beneficial and legitimate analytical data [4–8]. A validation examine includes testing multiple attributes of a method to determine that it may provide useful and valid facts whilst used robotically [9–11]. To accurately investigate method parameters, the validation test ought to consist of normal test conditions, which includes product excipients [11–14]. Therefore, a

> © The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.

distribution, and reproduction in any medium, provided the original work is properly cited.

© 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use,

Keywords: method validation, accuracy, precision, linearity, LOD, LOQ

http://dx.doi.org/10.5772/intechopen.72087

method validation examine is product-specific.

2.1. Parameters to be checked for method validation

Tentu Nageswara Rao

Tentu Nageswara Rao

Abstract

1. Introduction

2. Procedure

• Precision

• Selectivity/Specificity

#### **Chapter 7** Provisional chapter

### **Validation of Analytical Methods** Validation of Analytical Methods

### Tentu Nageswara Rao Tentu Nageswara Rao

Additional information is available at the end of the chapter Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/intechopen.72087

#### Abstract

Method validation is a key element in the establishment of reference methods and within the assessment of a laboratory's competence in generating dependable analytical records. Validation has been placed within the context of the procedure, generating chemical data. Analytical method validation, thinking about the maximum relevant processes for checking the best parameters of analytical methods, using numerous relevant overall performance indicators inclusive of selectivity, specificity, accuracy, precision, linearity, range, limit of detection (LOD), limit of quantification (LOQ), ruggedness, and robustness are severely discussed in an effort to prevent their misguided utilization and ensure scientific correctness and consistency among publications.

DOI: 10.5772/intechopen.72087

Keywords: method validation, accuracy, precision, linearity, LOD, LOQ

### 1. Introduction

Analytical method validation is an essential requirement to perform the chemical evaluation [1–3]. Method validation is a procedure of performing numerous assessments designed to verify that an analytical test system is suitable for its intended reason and is capable of providing beneficial and legitimate analytical data [4–8]. A validation examine includes testing multiple attributes of a method to determine that it may provide useful and valid facts whilst used robotically [9–11]. To accurately investigate method parameters, the validation test ought to consist of normal test conditions, which includes product excipients [11–14]. Therefore, a method validation examine is product-specific.

### 2. Procedure

#### 2.1. Parameters to be checked for method validation


© 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

© The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.


#### 2.1.1. Selectivity/specificity

Selectivity of an analytical method is its ability to measure accurately an analyte in the presence of interferences that may be expected to be present in the sample matrix.

Selectivity is checked by examining chromatographic blanks (from a sample that is known to contain no analyte) in the expected time window of the analyte peak. And the raw data for selectivity will be recorded in the raw data in approved formats.

#### 2.1.2. Precision

Precision of a method is the degree of agreement among individual test results when the procedure is applied repeatedly to multiple samplings.

Precision is measured by injecting a series of standards or analyzing series of samples from multiple samplings from a homogeneous lot. From the measured standard deviation (SD) and Mean values, precision as relative standard deviation (% rsd) is calculated.

$$\% \text{rssd orCV} = \frac{\text{SD}}{\text{Mean}} \times 100 \tag{1}$$

modified Horwitz values for repeatability CV given under may be used for guidance. If measured repeatability is outside those values, suggested explanation must be submitted for

Validation of Analytical Methods

133

http://dx.doi.org/10.5772/intechopen.72087

The accuracy of an analytical method is the degree of agreement of test results generated by

Accuracy is measured by spiking the sample matrix of interest with a known concentration of analyte standard and analyzing the sample using the "method being validated." The procedure and calculation for Accuracy (as% recovery) will be varied from matrix to matrix and it

The linearity of an analytical method is its capability to elicit check consequences which might be at once, or with the aid of well described mathematical adjustments, proportional to the

Linearity is determined by injecting a series of standards of stock solution/diluted stock solution using the solvent/mobile phase, at a minimum of five different concentrations in the range of 50–150% of the expected working range. The linearity graph will be plotted manually/ using Microsoft Excel or software of the computer (Concentration vs. Peak Area Response) and

The range of an analytical method is the interval between the upper and lower levels that have been demonstrated to be determined with precision, accuracy and linearity using the set

Note: The unmodified Horwitz equation is used as a criterion of acceptability for methods collaboratively tested by

(Horwitz value 0.67)

method. This range will be the concentration range in which the Linearity test is done.

Percent of analyte Proposed acceptable % RSDr

will be given in respective study plan or amendment to the study plan.

consideration. The details were presented in Table 1.

concentration of analytes in within a given range.

which will be attached to respective study files.

100.00 1.340 50.00 1.490 20.00 1.710 10.00 1.900 5.00 2.100 2.00 2.410 1.00 2.680 0.25 3.300

2.1.3. Accuracy

2.1.4. Linearity

2.1.5. Range

CIPAC.

Table 1. Details of Horwitz values.

the method to the true value.

The raw data for precision will be recorded in the approved format and the acceptance criteria for precision will be given in the respective study plan or amendment to the study plan.

#### OR

Precision can be also calculated by using Horwitz equation:

The acceptable percent of relative standard deviation results for precision may be based on the Horwitz equation, an exponential relationship between the among-laboratory relative standard deviation (RSDR) and Concentration (C): [15]

$$2\,\% \text{RSD}\_{\text{R}} = 2^{(1-0.5 \text{logC})} \tag{2}$$

For estimation of repeatability (RSDr), is modified to:

$$\% \text{RSD}\_{\text{f}} = \% \text{RSD}\_{\text{R}} \times 0.67 \tag{3}$$

The Horwitz curve has been empirically derived and has been proven to be more or less independent of analyte, matrix and method of evaluation over the concentration range C = 1 (100%) to C = 10�<sup>9</sup> by the evaluation of vast numbers of method precision studies. The modified Horwitz values for repeatability CV given under may be used for guidance. If measured repeatability is outside those values, suggested explanation must be submitted for consideration. The details were presented in Table 1.

#### 2.1.3. Accuracy

• Accuracy • Linearity

• Range • Stability

2.1.2. Precision

OR

2.1.1. Selectivity/specificity

• Limit of Detection (LOD) and Limit of Quantitation (LOQ)

132 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

selectivity will be recorded in the raw data in approved formats.

Mean values, precision as relative standard deviation (% rsd) is calculated.

procedure is applied repeatedly to multiple samplings.

Precision can be also calculated by using Horwitz equation:

dard deviation (RSDR) and Concentration (C): [15]

For estimation of repeatability (RSDr), is modified to:

Selectivity of an analytical method is its ability to measure accurately an analyte in the

Selectivity is checked by examining chromatographic blanks (from a sample that is known to contain no analyte) in the expected time window of the analyte peak. And the raw data for

Precision of a method is the degree of agreement among individual test results when the

Precision is measured by injecting a series of standards or analyzing series of samples from multiple samplings from a homogeneous lot. From the measured standard deviation (SD) and

The raw data for precision will be recorded in the approved format and the acceptance criteria for precision will be given in the respective study plan or amendment to the study plan.

The acceptable percent of relative standard deviation results for precision may be based on the Horwitz equation, an exponential relationship between the among-laboratory relative stan-

The Horwitz curve has been empirically derived and has been proven to be more or less independent of analyte, matrix and method of evaluation over the concentration range C = 1 (100%) to C = 10�<sup>9</sup> by the evaluation of vast numbers of method precision studies. The

Mean

� 100 (1)

%RSDR <sup>¼</sup> <sup>2</sup>ð Þ <sup>1</sup>�0:5logC (2)

%RSDr ¼ %RSDR � 0:67 (3)

%rsd or CV <sup>¼</sup> SD

presence of interferences that may be expected to be present in the sample matrix.

The accuracy of an analytical method is the degree of agreement of test results generated by the method to the true value.

Accuracy is measured by spiking the sample matrix of interest with a known concentration of analyte standard and analyzing the sample using the "method being validated." The procedure and calculation for Accuracy (as% recovery) will be varied from matrix to matrix and it will be given in respective study plan or amendment to the study plan.

#### 2.1.4. Linearity

The linearity of an analytical method is its capability to elicit check consequences which might be at once, or with the aid of well described mathematical adjustments, proportional to the concentration of analytes in within a given range.

Linearity is determined by injecting a series of standards of stock solution/diluted stock solution using the solvent/mobile phase, at a minimum of five different concentrations in the range of 50–150% of the expected working range. The linearity graph will be plotted manually/ using Microsoft Excel or software of the computer (Concentration vs. Peak Area Response) and which will be attached to respective study files.

#### 2.1.5. Range

The range of an analytical method is the interval between the upper and lower levels that have been demonstrated to be determined with precision, accuracy and linearity using the set method. This range will be the concentration range in which the Linearity test is done.


Note: The unmodified Horwitz equation is used as a criterion of acceptability for methods collaboratively tested by CIPAC.

Table 1. Details of Horwitz values.

#### 2.1.6. Stability

Many analytes readily decompose prior to chromatography investigations, for example during the preparation of the sample solutions, during extraction, clean-up, phase transfer, and during storage of prepared vials. Under these circumstances, method development should investigate the stability of the analyte. Accuracy test takes care of stability. It is required to mention in the method how long a sample after extraction can be stored before final analysis, based on the duration taken for accuracy test.

• Using the concentrations and corresponding instrument response, LOD and LOQ can be

Validation of Analytical Methods

135

http://dx.doi.org/10.5772/intechopen.72087

Where, X and Y are the variables (data of two parameters). Generally, X is called the indepen-

"a" and "b" are the regression constants. Further, "a" is known as the intercept and "b," the

Let (X1, Y1), (X2, Y2), (X3, Y3)…(Xn, Yn) be the set of values required to be fit in the linear

<sup>Σ</sup>xx <sup>¼</sup> <sup>Σ</sup> <sup>X</sup> � <sup>X</sup> � �<sup>2</sup> <sup>¼</sup> <sup>Σ</sup>X<sup>2</sup> � ð Þ <sup>Σ</sup><sup>X</sup> <sup>2</sup>

<sup>Σ</sup>yy <sup>¼</sup> <sup>Σ</sup> <sup>Y</sup> � <sup>Y</sup> � �<sup>2</sup> <sup>¼</sup> <sup>Σ</sup>Y<sup>2</sup> � ð Þ <sup>Σ</sup><sup>Y</sup> <sup>2</sup>

Σxy ¼ ΣXY � ð Þ ΣX ð Þ ΣY =n

b ¼

Pxy Pxx

a ¼ Y � bX

=n

=n

calculated as follows:

slope of the line.

equation.

Let the linear regression equation be Y ¼ a þ bX.

Take concentration on X-axis and instrument response on Y-axis.

dent variable and Y, the dependent variable.

a. Method of arriving at "a" and "b" y

X1 Y1 X2 Y2 . . . . . .

Xn Yn

ii. Calculate the following parameters:

\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_

Mean, ¼ X ¼ ΣX=n Y ¼ ΣX=n \_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_

iii. Calculate the slope "b," and intercept "a" as given below:

b. Method of calculation r (correlation coefficient)

i. Tabulate as given below:

#### 2.1.7. Limit of detection and limit of quantitation

The term LOD is defined as the lowest concentration at which the instrument is able to detect but not quantify and the noise to signal ratio for LOD should be 1:3. The term LOQ is defined as the lowest concentration at which the instrument is able to detect and quantify. The noise to signal ratio for LOQ should be 1:10.

Determination of Limit of Detection (LOD) and Limit of Quantitation (LOQ) from Detector Linearity experiments (applicable to only instrument sensitivity).

LOD and LOQ values are calculated manually by taking Noise to signal ratio of a lowest/ known concentration of linearity samples and it will be expressed in μg/ml or ppm. To calculate in %, values of LOD and LOQ will be multiplied by 100/lowest or known concentration of test item (mg/L) taken for analysis of that particular a.i. or impurity analysis.

#### Calculations of LOD and LOQ values for instrument sensitivity:

$$\begin{aligned} \text{LOD} \,(\text{mg/L}) &= 3 \times \frac{\text{Noise}}{\text{Signal}} \times \text{Lowest concentration of the linearity samples} \\ \text{LODQ} \,(\text{mg/L}) &= 10 \times \frac{\text{Noise}}{\text{Signal}} \times \text{Lowest concentration of the linearity samples} \end{aligned}$$

#### Calculations of LOD and LOQ values for method:

$$\text{LOD} \left( \% \right) = \frac{\text{LOD} \left( \text{mg}/\text{L} \right)}{\text{Test item conc.used} \quad \text{for quantitative} \times 100} \times 100$$

$$\text{LOD} \left( \% \right) = \frac{\text{LOD} \left( \text{mg}/\text{L} \right)}{\text{Test item conc.used} \quad \text{for quantitative} \times 100}$$

OR

#### 2.1.8. Mathematical derivations

#### 2.1.8.1. Determination of limit of detection (LOD) and limit of quantitation (LOQ)

Prepare a series of standard solutions (minimum five concentrations covering working concentrations used for routine analysis) and analyze each solution minimum twice and record the instruments response.

• Using the concentrations and corresponding instrument response, LOD and LOQ can be calculated as follows:

Let the linear regression equation be Y ¼ a þ bX.

Where, X and Y are the variables (data of two parameters). Generally, X is called the independent variable and Y, the dependent variable.

Take concentration on X-axis and instrument response on Y-axis.

"a" and "b" are the regression constants. Further, "a" is known as the intercept and "b," the slope of the line.

Let (X1, Y1), (X2, Y2), (X3, Y3)…(Xn, Yn) be the set of values required to be fit in the linear equation.

a. Method of arriving at "a" and "b" y

i. Tabulate as given below:

2.1.6. Stability

based on the duration taken for accuracy test.

2.1.7. Limit of detection and limit of quantitation

Linearity experiments (applicable to only instrument sensitivity).

134 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Calculations of LOD and LOQ values for instrument sensitivity:

Noise

Noise

LODð Þ¼ % LOD mg ð Þ <sup>=</sup><sup>L</sup>

LOQð Þ¼ % LOD mg ð Þ <sup>=</sup><sup>L</sup>

2.1.8.1. Determination of limit of detection (LOD) and limit of quantitation (LOQ)

Prepare a series of standard solutions (minimum five concentrations covering working concentrations used for routine analysis) and analyze each solution minimum twice and record

signal ratio for LOQ should be 1:10.

LOD mg ð Þ¼ =L 3 �

LOQ mg ð Þ¼ =L 10 �

2.1.8. Mathematical derivations

the instruments response.

OR

Calculations of LOD and LOQ values for method:

Many analytes readily decompose prior to chromatography investigations, for example during the preparation of the sample solutions, during extraction, clean-up, phase transfer, and during storage of prepared vials. Under these circumstances, method development should investigate the stability of the analyte. Accuracy test takes care of stability. It is required to mention in the method how long a sample after extraction can be stored before final analysis,

The term LOD is defined as the lowest concentration at which the instrument is able to detect but not quantify and the noise to signal ratio for LOD should be 1:3. The term LOQ is defined as the lowest concentration at which the instrument is able to detect and quantify. The noise to

Determination of Limit of Detection (LOD) and Limit of Quantitation (LOQ) from Detector

LOD and LOQ values are calculated manually by taking Noise to signal ratio of a lowest/ known concentration of linearity samples and it will be expressed in μg/ml or ppm. To calculate in %, values of LOD and LOQ will be multiplied by 100/lowest or known concentra-

Signal � Lowest concentration of the linearity samples

Signal � Lowest concentration of the linearity samples

Test item conc:used for quantification � <sup>100</sup>

Test item conc:used for quantification � <sup>100</sup>

tion of test item (mg/L) taken for analysis of that particular a.i. or impurity analysis.


\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_

ii. Calculate the following parameters:

$$\begin{aligned} \Sigma \ltimes &= \Sigma \left( X - \overline{X} \right)^2 = \Sigma X^2 - \left( \Sigma X \right)^2 / \mathbf{n} \\ \Sigma \, \mathbf{y} &= \Sigma \left( Y - \overline{Y} \right)^2 = \Sigma Y^2 - \left( \Sigma Y \right)^2 / \mathbf{n} \\ \Sigma \, \mathbf{x} &= \Sigma XY - \left( \Sigma X \right) \left( \Sigma Y \right) / \mathbf{n} \end{aligned}$$

iii. Calculate the slope "b," and intercept "a" as given below:

$$b = \frac{\sum xy}{\sum x \mathbf{x}}$$

$$a = \overline{Y} - b\overline{X}$$

b. Method of calculation r (correlation coefficient)

$$r = \frac{\sum xy}{\sqrt{\sum x x. \sum y y}}$$

3. Example

0.2–3.0 ppm.

Table 2.

area counts) in a linear equation

Let the equation be Y ¼ a þ bX.

The calculations were presented in Table 3.

Table 2. Calculation details of mean, SD, and %CV.

Table 3. Calculation details of additional parameters.

In this example, the linear regression equation is employed to find out the extent of linear response of an Detector to a reference analytical standard in the concentration range of about

Each of these working standards is injected thrice (1 μl per injection), and the peak area counts

From the peak areas corresponding to each concentration level, the mean, standard deviation (SD) and coefficient of variation (%CV) are also calculated. The details were presented in

Fitting the data of concentration of standard solution and mean detector response (peak

(n � 1)

Validation of Analytical Methods

137

http://dx.doi.org/10.5772/intechopen.72087

%CV

Where, Y = Mean peak area counts and X = Concentration of standard solution, μg/ml.

123 0.1956 32,827 33,299 32,731 32,952 304 0.923 0.4890 87,783 88,480 87,446 87,903 527 0.600 0.9780 176,037 174,675 177,203 175,972 1265 0.719 1.467 246,212 250,786 246,849 247,949 2477 0.999 1.956 319,143 319,615 315,316 318,025 2358 0.741 2.934 415,059 410,773 418,407 414,746 3827 0.923

Conc. of standard solution (μg/ml) Peak area Mean SD

%CV = SD � 100/Mean: The coefficient of variation (CV) shows that the Injection variation is less than 1%.

Sl. no. Y X 1. 32952 0.1956 2. 87903 0.4890 3. 175972 0.9780 4. 247949 1.4670 5. 318025 1.9560 6. 414746 2.9340

corresponding to the active ingredient peak are given below.

c. Method of calculation standard deviation for "a" and "b"

The standard deviation of the individual deviations of measured values in Y, above and below the linear line (fitted line) is:

$$Sy.x = \sqrt{\frac{\left(\sum yy - \left\{ \left(\sum xy\right)^2 / \sum x \right\} \right)}{n-2}}$$

From this, the standard deviation for "a" and "b" are calculated.

Standard deviation

$$\text{for } \text{''a\textquotedblleft a\textquotedblright respectright} = Sy.x \sqrt{\frac{\sum {X^2}}{n \sum x}}$$

as Sa

Standard deviation.

$$\text{For } \text{''b} \text{'' represented} = Sy.x \sqrt{\frac{1}{n \sum x^n}}$$

as Sb

#### 2.1.8.2. Application of a, b, and Sa to obtain limit of detection and limit of quantitation

When Sa is obtained for a linear calibration line, then it provides a clear information on the standard deviation of the "Blank" (or Control) response from the instruments.

The LOD and LOQ can be worked out, as given below:

$$\begin{aligned} \text{LOD} &= \frac{|\mathbf{a}| + 3 \mathbf{S\_a}}{\mathbf{b}} \\ \text{LOD} &= \frac{|\mathbf{a}| + 10 \mathbf{S\_a}}{\mathbf{b}} \end{aligned}$$

Note:


### 3. Example

r ¼

c. Method of calculation standard deviation for "a" and "b"

136 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Sy:x ¼

From this, the standard deviation for "a" and "b" are calculated.

ffiffiffiffiffiffiffiffiffiffiffiffi PX<sup>2</sup> n Pxx

n Pxx q

The LOD and LOQ can be worked out, as given below:

2.1.8.2. Application of a, b, and Sa to obtain limit of detection and limit of quantitation

standard deviation of the "Blank" (or Control) response from the instruments.

When Sa is obtained for a linear calibration line, then it provides a clear information on the

LOD <sup>¼</sup> j jþ <sup>a</sup> 3 Sa b

LOQ <sup>¼</sup> j j <sup>a</sup> <sup>þ</sup> 10 Sa b

• The above calculations can be programmed in a computer but before every use, the

• The above procedure can also be used for obtaining LOD and LOQ of the method from recovery test results by taking fortified concentration on X-axis and obtained concentra-

computer program must be validated using the example given in section

r

vuut

the linear line (fitted line) is:

Standard deviation

Standard deviation.

as Sa

as Sb

Note:

tions on Y-axis.

for "a," represented = Sy:x

For "b," represented = Sy:x ffiffiffiffiffiffiffiffiffiffiffiffi <sup>1</sup>

Pxy ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pxx: Pyy p

The standard deviation of the individual deviations of measured values in Y, above and below

<sup>P</sup>yy � <sup>P</sup>ð Þ xy

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

� � n o

n � 2

2 = Pxx In this example, the linear regression equation is employed to find out the extent of linear response of an Detector to a reference analytical standard in the concentration range of about 0.2–3.0 ppm.

Each of these working standards is injected thrice (1 μl per injection), and the peak area counts corresponding to the active ingredient peak are given below.

From the peak areas corresponding to each concentration level, the mean, standard deviation (SD) and coefficient of variation (%CV) are also calculated. The details were presented in Table 2.

#### Fitting the data of concentration of standard solution and mean detector response (peak area counts) in a linear equation

Let the equation be Y ¼ a þ bX.

Where, Y = Mean peak area counts and X = Concentration of standard solution, μg/ml.

The calculations were presented in Table 3.


%CV = SD � 100/Mean: The coefficient of variation (CV) shows that the Injection variation is less than 1%.

Table 2. Calculation details of mean, SD, and %CV.


Table 3. Calculation details of additional parameters.

$$\begin{array}{llll} \sum \text{Y} = 1277547 & \sum \text{X} = 8.0196 & \sum \text{XY} = 2424193.441\\ \overline{\text{Y}} = 212924.5 & \overline{\text{X}} = 1.3366 & \text{n} = 6\\ \sum \text{Y}^2 = 3.7441177 \times 10^{11} & \sum \text{X}^2 = 15.820245 \end{array}$$

Sy:x ¼

¼

The standard deviation for a is calculated as:

The standard deviation for b is calculated as

Category I

Category II

endotoxin tests.

Category III

servatives in finished goods.

s

s

¼ 20731:806

Sa ¼ Sy:x

Sb ¼ Sy:x

¼ 20731:806

¼ 20731:806

¼ 14905

Pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi yy � <sup>P</sup>ð Þ xy

n � 2

2 = <sup>P</sup>xx n o

<sup>1</sup>:0239070X1011 � � � ð Þ <sup>716624</sup>:<sup>12</sup> <sup>2</sup>

ffiffiffiffiffiffiffiffiffiffiffiffiffiffi PX<sup>2</sup> n <sup>P</sup>xx <sup>s</sup>

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 n � <sup>P</sup>xx <sup>r</sup>

Note: Assay procedures vary from highly exacting analytical determinations to subjective evaluations of attributes. Therefore different test methods require different validation schemes.

Analytical methods for quantitation of major excipients and/or active ingredients, and pre-

Analytical methods for determination of impurities or degradation compounds in finished goods. These methods include quantitative assays and limit tests, titrimetric and bacterial

Analytical methods for determination of performance characteristics, e.g., sterility testing,

dissolution and drug release for pharmaceutical products.

Details of required validation parameters of assay presented in Table 4.

Data Elements Required for Assay Validation.

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

6 � 2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 15:820245 <sup>6</sup> � <sup>5</sup>:<sup>101248</sup> <sup>r</sup>

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 <sup>6</sup> � <sup>5</sup>:<sup>101248</sup> <sup>r</sup>

<sup>=</sup>ð Þ <sup>5</sup>:<sup>101248</sup> n o

Validation of Analytical Methods

139

http://dx.doi.org/10.5772/intechopen.72087

#### Using the above parameters, calculate the following

$$\begin{array}{rcl} \Sigma \text{ x} &=& \Sigma \text{X}^2 - (\sum \text{X})^2/\text{n} \\ &=& 15.820245 - (8.0196)^2/6 \\ &=& 5.101248 \\\\ \Sigma \text{ y} \mathbf{y} &=& \sum \mathbf{Y}^2 - (\sum \mathbf{Y})^2/\text{n} \\ &=& 3.7441176 \times 10^{11} - (1277547)^2/6 \\ &=& 1.0239070 \times 10^{11} \\\\ \Sigma \text{ xy} &=& \sum \mathbf{X}\mathbf{Y} - (\sum \mathbf{X})(\sum \mathbf{Y})/\text{n} \\ &=& 2424193.441 - (1277547)(8.0196)/6 \end{array}$$

¼ 716624:12

Calculation of a, b, and r

$$\begin{aligned} \mathbf{b} &= \frac{\sum \mathbf{xy}}{\sum \mathbf{x}} \\ &= \frac{716624.12}{5.101248} \\ &= 140480.16 \\ \mathbf{b} &= \frac{\sum \mathbf{xy}}{\sum \mathbf{x}} \\ &= \frac{716624.12}{5.101248} \\ &= 140480.16 \\ \mathbf{a} &= \frac{\mathbf{\bar{y}} - \mathbf{b}\overline{\mathbf{x}}}{212924.5 - 140480.16 \times 1.3366} \\ &= \frac{212924.5 - 140480.16 \times 1.3366}{25158.718} \\ \mathbf{r} &= \frac{\sum \mathbf{x}\mathbf{y}}{\sqrt{\sum x \mathbf{x} \cdot \sum y}} \\ \mathbf{r} &= \frac{716624.12}{\sqrt{1.0239070310^{11} \mathbf{X} 5.101248}} = 0.99157 \end{aligned}$$

Note: Sometimes r2 is also used to express the goodness of fit. Calculation of standard deviation for a and b:

#### Validation of Analytical Methods http://dx.doi.org/10.5772/intechopen.72087 139

$$\begin{array}{rcl} \text{Sy.x} &=& \sqrt{\frac{\sum yy - \left\{ (\sum xy)^2 / \sum x \right\}}{n-2}} \\ &=& \sqrt{\frac{\left(1.0239070 \text{X} 10^{11} \right) - \left\{ (716624.12)^2 / (5.101248) \right\}}{6-2}} \\ &=& 20731.806 \end{array}$$

The standard deviation for a is calculated as:

$$\begin{aligned} S\_a &= -S \text{yr} \sqrt{\frac{\sum X^2}{n \sum \text{rx}}} \\ &= -20731.806 \sqrt{\frac{15.820245}{6 \times 5.101248}} \\ &= -14905 \end{aligned}$$

The standard deviation for b is calculated as

$$\begin{aligned} S\_b &= -S \text{y.x} \sqrt{\frac{1}{n \cdot \sum \text{xx}}} \\ &= -20731.806 \sqrt{\frac{1}{6 \times 5.101248}} \end{aligned}$$

Note: Assay procedures vary from highly exacting analytical determinations to subjective evaluations of attributes. Therefore different test methods require different validation schemes.

#### Category I

<sup>P</sup><sup>Y</sup> <sup>¼</sup> <sup>1277547</sup> <sup>P</sup><sup>X</sup> <sup>¼</sup> <sup>8</sup>:<sup>0196</sup> <sup>P</sup>XY <sup>¼</sup> <sup>2424193</sup>:<sup>441</sup>

<sup>¼</sup> <sup>15</sup>:<sup>820245</sup> � ð Þ <sup>8</sup>:<sup>0196</sup> <sup>2</sup>

<sup>¼</sup> <sup>3</sup>:<sup>7441176</sup> � 1011 � ð Þ <sup>1277547</sup> <sup>2</sup>

¼ 2424193:441 � ð Þ 1277547 ð Þ 8:0196 =6

=n

=n

=6

=6

Y ¼ 212924:5 X ¼ 1:3366 n ¼ 6

<sup>P</sup>xx <sup>¼</sup> <sup>P</sup>X2 � <sup>P</sup>ð Þ <sup>X</sup> <sup>2</sup>

¼ 5:101248

<sup>¼</sup> <sup>1</sup>:<sup>0239070</sup> � 1011

<sup>P</sup>xy <sup>¼</sup> <sup>P</sup>XY � <sup>P</sup>ð Þ <sup>X</sup> <sup>P</sup>ð Þ <sup>Y</sup> <sup>=</sup><sup>n</sup>

b ¼

b ¼

a ¼ Y � bX

r ¼

Calculation of standard deviation for a and b:

Note: Sometimes r2 is also used to express the goodness of fit.

¼ 25158:718

<sup>r</sup> <sup>¼</sup> <sup>716624</sup>:<sup>12</sup> ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Pxy ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi <sup>P</sup>xx � Pyy p

Pxy Pxx <sup>¼</sup> <sup>716624</sup>:<sup>12</sup> 5:101248 ¼ 140480:16

Pxy Pxx <sup>¼</sup> <sup>716624</sup>:<sup>12</sup> 5:101248 ¼ 140480:16

¼ 212924:5 � 140480:16 � 1:3366

<sup>1</sup>:0239070X1011X5:<sup>101248</sup> <sup>p</sup> <sup>¼</sup> <sup>0</sup>:<sup>99157</sup>

<sup>P</sup>yy <sup>¼</sup> <sup>P</sup>Y2 � <sup>P</sup>ð Þ <sup>Y</sup> <sup>2</sup>

¼ 716624:12

<sup>P</sup>Y2 <sup>¼</sup> <sup>3</sup>:<sup>7441177</sup> � 1011 <sup>P</sup>X<sup>2</sup> <sup>¼</sup> <sup>15</sup>:<sup>820245</sup>

138 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Using the above parameters, calculate the following

Calculation of a, b, and r

Analytical methods for quantitation of major excipients and/or active ingredients, and preservatives in finished goods.

#### Category II

Analytical methods for determination of impurities or degradation compounds in finished goods. These methods include quantitative assays and limit tests, titrimetric and bacterial endotoxin tests.

#### Category III

Analytical methods for determination of performance characteristics, e.g., sterility testing, dissolution and drug release for pharmaceutical products.

#### Data Elements Required for Assay Validation.

Details of required validation parameters of assay presented in Table 4.


[5] Horwitz W. Evaluation of analytical methods used for regulation of foods and drugs.

Validation of Analytical Methods

141

http://dx.doi.org/10.5772/intechopen.72087

[6] Thompson M, Ellison SLR, Wood R. Harmonised guidelines for single laboratory valida-

[7] European Commission. Annex 15. EU guide to good manufacturing practice: Qualifica-

[8] Ravichandran V, Shalini S, Sundram KM, Rajak H. Validation of analytical methods— Strategies & importance. International Journal of Pharmacy and Pharmaceutical Sciences.

[9] Tangri P, Rawat PS, Jakhmola V. Validation: A critical parameter for quality control of

[10] Sharma A, Sharma R. Validation of analytical procedures: A comparison of ICH Vs Pharmacopoiea (USP) Vs FDA. International Research Journal of Pharmacy. 2012;3(6):39-42

[11] Lambert J. Validation guidelines for pharmaceutical dosage forms. Health Canada Health

[12] Prabh SS, Gagan S. Analytical method development and validation. Journal of Pharmacy

[13] Ramamurthy M, Sarvanakumar K. Pharmaceutical validation. The Eastern Pharmacist.

[14] Agalloco J. Validation: An unconventional review and reinvention. PDA. J. Pharm Sci

[15] INRA Quality Policy and Quality Guidelines for the Research and Experimental Units.

[16] Haider I. Section VAL 1100.00. In: Validation Standard Operating Procedures. A Step by Step Guide for Achieving Compliance in the Pharmaceutical Medical Device and BiotechIn-

pharmaceuticals. Journal of Drug Delivery & Therapeutics. 2012;2(3):34-40

Products and Food Branch Inspectorate. 2004;26:7-15

Analytical Chemistry. 1982;54(1):67A-76A. DOI: 10.1021/AC00238A765

tion and validation. 2010;4:1–10

Research. 2011;4(7):2330-2332

dustries. Boca Raton: CRC Press LLC; 2001

1997;476:45-47

2013

Tech. 1995;49:175-179

2010;2(3):340-345

tion of method of analysis. Pure and Applied Chemistry. 2008;74(5):835-855

\*May be required depending on the specific test.

Table 4. Validation parameters of assay [16].

### 4. Conclusions

Analytical validation data playing a fundamental role in pharmaceutical industry, pesticide industry for releasing the economic batch and long term stability information consequently, the records must be produced to suited regulatory authority requirements.

### Author details

Tentu Nageswara Rao

Address all correspondence to: tentu6581@rediffmail.com

Department of Chemistry, Krishna University, Machilipatnam, Andhra Pradesh, India

### References


4. Conclusions

Analytical parameters Assay

\*May be required depending on the specific test.

Table 4. Validation parameters of assay [16].

category 1

140 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Author details

References

Tentu Nageswara Rao

reproducibility

Analytical validation data playing a fundamental role in pharmaceutical industry, pesticide industry for releasing the economic batch and long term stability information consequently,

Assay category 2 quantitative

Assay accuracy Yes Yes \* \* Precision Yes Yes No Yes Specificity Yes Yes Yes \* Limit of detection Yes Yes Yes \* Limit of quantitation Yes Yes No \* Linearity Yes Yes No \* Range Yes Yes \* \* Robustness \* \* \* \*

Limit Test Assay

category III

Department of Chemistry, Krishna University, Machilipatnam, Andhra Pradesh, India

[1] Validation of analytical procedure: Methodology Q2B. In: ICH Harmonized Tripartite

[2] CIPAC Document No. 3807—Guidelines on method Validation to be performed in support of analytical methods for agrochemical Formulations, cipac.org on 28 July 2003 [3] International Standard ISO 5725. 1986. Precision of test methods—Repeatability and

[4] ISO standard 11095. 1996. Linear calibration using reference material

the records must be produced to suited regulatory authority requirements.

Address all correspondence to: tentu6581@rediffmail.com

Guidelines. Geneva, Switzerland. 1996. pp. 1-8


**Chapter 8**

Provisional chapter

**Method Validation Approaches for Pharmaceutical**

Method Validation Approaches for Pharmaceutical

**Layer Chromatographic (HPTLC) Techniques**

Layer Chromatographic (HPTLC) Techniques

David Jenkins, Cherif Diallo, Ed Bethea, Eliangiringa Kaale and Thomas Layloff

David Jenkins, Cherif Diallo, Ed Bethea, Eliangiringa Kaale and Thomas Layloff

http://dx.doi.org/10.5772/intechopen.71765

Abstract

ingredient, impurities

1. Introduction

Additional information is available at the end of the chapter

Additional information is available at the end of the chapter

**Assessments – Highlights with High Performance Thin**

DOI: 10.5772/intechopen.71765

Assessments – Highlights with High Performance Thin

Method validation is an important activity for pharmaceutical evaluations to ensure that analytical methods are suitable for their intended use. With particular focus on active ingredient and impurities, the implementation of different categories of method validation are explained for qualitative and quantitative methods. Detailed explanations with example approaches are provided for the key aspects of method validation, namely specificity, accuracy, linearity, limits of detection/quantitation, precision, robustness, and method range. While all of the sections outlined for method validation are generally applicable for a variety of techniques commonly used in pharmaceutical analysis (i.e., UV and HPLC instrumentation), focused attention is provided for examples that have been implemented using high performance thin layer chromatographic techniques.

Keywords: method validation, pharmaceuticals, HPTLC, assay, active pharmaceutical

Method Validation (MV) is a development process undertaken to establish, within acceptable statistical bounds, that an assessment procedure or method consistently yields a "true" result both in "within laboratory" and "among laboratories" testing. Pharmaceutical product quality assessments are focused on methods for the active pharmaceutical ingredient (API) and related impurities. Being able to perform methods of analysis to assess product quality is critical in law enforcement and regulating commerce. In addition, for new drug products, these quality determinations are surrogate performance indicators for assuring the safety and efficacy of a

> © The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.

distribution, and reproduction in any medium, provided the original work is properly cited.

© 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use,

#### **Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin Layer Chromatographic (HPTLC) Techniques** Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin Layer Chromatographic (HPTLC) Techniques

DOI: 10.5772/intechopen.71765

David Jenkins, Cherif Diallo, Ed Bethea, Eliangiringa Kaale and Thomas Layloff David Jenkins, Cherif Diallo, Ed Bethea,

Additional information is available at the end of the chapter Eliangiringa Kaale and Thomas Layloff

http://dx.doi.org/10.5772/intechopen.71765 Additional information is available at the end of the chapter

#### Abstract

Method validation is an important activity for pharmaceutical evaluations to ensure that analytical methods are suitable for their intended use. With particular focus on active ingredient and impurities, the implementation of different categories of method validation are explained for qualitative and quantitative methods. Detailed explanations with example approaches are provided for the key aspects of method validation, namely specificity, accuracy, linearity, limits of detection/quantitation, precision, robustness, and method range. While all of the sections outlined for method validation are generally applicable for a variety of techniques commonly used in pharmaceutical analysis (i.e., UV and HPLC instrumentation), focused attention is provided for examples that have been implemented using high performance thin layer chromatographic techniques.

Keywords: method validation, pharmaceuticals, HPTLC, assay, active pharmaceutical ingredient, impurities

### 1. Introduction

Method Validation (MV) is a development process undertaken to establish, within acceptable statistical bounds, that an assessment procedure or method consistently yields a "true" result both in "within laboratory" and "among laboratories" testing. Pharmaceutical product quality assessments are focused on methods for the active pharmaceutical ingredient (API) and related impurities. Being able to perform methods of analysis to assess product quality is critical in law enforcement and regulating commerce. In addition, for new drug products, these quality determinations are surrogate performance indicators for assuring the safety and efficacy of a

© 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

© The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.

pharmaceutical product. The safety and efficacy of a pharmaceutical product are established with a "pivotal lot" production of the product and the characterization of this lot with well validated methods with acceptable performance characteristics is critical to assure that future production lots have the same quality characteristics as the "pivotal lot", thereby assuring they have equivalent safety and efficacy.

validation. Although the approaches are generally applicable to common techniques used in pharmaceutical analysis (such UV-VIS and HPLC quantifications), particular emphasis will be

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

145

Specificity is the ability of a method to distinguish an analyte from all substances that are present or likely to be present in test samples [3, 4]. When possible, these substances should include future degradation products and other ingredients (i.e., excipients). An analytical procedure is specific when placebo and impurity spots do not overlap partially with and are not buried under the analyte spot. In addition, the calculated amount of analyte does not depend on the quantity

Various approaches are possible when evaluating method specificity [5–8]. Ideal demonstration of specificity for an HPTLC analytical procedure requires chromatographing simultaneously three types of samples: sample type 1 is the pure analyte or its reference standard, sample type 2 is the analyte mixed with a representative blank and all likely impurities, and sample type 3 is the representative blank mixed with all likely impurities. Likely impurities include degradation products, reagents, intermediates, excipients, side products, and analyte isomers. The mixtures can be created by spiking test samples (API substances or finished

In practice, the unavailability of one or more of these types of samples can pose a significant challenge. In some cases, it is often difficult to know all likely impurities. There can be several sample deficiency scenarios. If the pure analyte or its reference standard is not available, demonstration of specificity can be quite challenging if not impossible. If a representative blank is available, but some or all likely impurities are missing, the typical test sample is subjected to stress testing environments. It should be noted however that stress testing is unlikely to produce some analyte isomers, reagents, intermediates and side products. If a representative blank is unavailable, but some likely impurities are available, spiking the typical test sample with impurities can show that increasing impurities will not change analyte signal. In addition, efforts should be made (perhaps by contacting the manufacturer), whenever possible, to create a representative blank even if it's not exactly in the same dosage form as the test sample. If neither a representative blank nor impurity standards are available, the typical test sample is subjected to stress testing to alleviate some of the deficiencies. Once again, the limitations of stress testing should be acknowledged because it may not produce all likely impurities, it may not account for impurities that are completely buried under the analyte signal, and it may not

In general, stress testing, impurity spiking, and peak-purity analysis are the common tools used to address certain sample deficiencies. To demonstrate method specificity, validation reports typically discuss several measures of performance. One measure of specificity is resolution of the analyte spot relative to the closest non-analyte spot. For HPTLC, the resolution

indicate whether some excipients or impurities can react with the analyte.

placed on high performance thin layer chromatography (HPTLC) techniques.

2. Specificity

of other substances.

should be a least 1 [5].

products) or placebos with likely impurities.

In the United States of America (USA), there are both private and public standards; the private standards are created through a USA Food and Drug Administration (FDA) approval process of industry method submissions that can be used for law enforcement, and public standards, which are promulgated in the monographs of the United States Pharmacopeia (USP) [1], that may be used in law enforcement or to support commercial agreements. The private standards, which are not publicly available, are private agreements between the approving government body and the submitting industry on the methods and standards to be used in law enforcement. The method validation protocols for the establishment of private standards are provided in the guidance of the "International Conference On Harmonisation Of Technical Requirements For Registration Of Pharmaceuticals For Human Use" (ICH) which have been incorporated into the laws and regulations in the European Union, Japan and the USA; these procedures are required for the assessments of new drug entities [2]. The method validation protocols for the establishment of monographs to support public standards are provided in USP <1225> [3] and ICH Q2 [4]. Both protocols cite the same analytical performance characteristics and test procedures except that the public standard must be able to be applied to all legally marketed products containing the specific API whereas the private standard applies only to the approved API in the specific product.

The analytical performance characteristics which must be assessed in both the ICH and USP are Accuracy, Precision (both Repeatability and Intermediate in ICH), and Specificity. Detection Limit, Quantitation Limit, Linearity and Range depending on which attributes are to be assessed. The USP presents the characteristics as noted below [3]:

"Category I — Analytical procedures for quantitation of major components of bulk drug substances or active ingredients (including preservatives) in finished pharmaceutical products.

Category II — Analytical procedures for determination of impurities in bulk drug substances or degradation compounds in finished pharmaceutical products. These procedures include quantitative assays and limit tests.

Category III — Analytical procedures for determination of performance characteristics (e.g., dissolution, drug release, etc.).

Category IV — Identification tests."

These can be categorized further into Assay procedures for Category I, Category II, impurity determinations, and Category III (dissolution and drug release are different procedures for preparing a solution of the API), all performance characteristics except Detection Limit must be validated. For the Category II limit tests only the Specificity and Detection Limits must be validated and for Category IV, Identification Tests, only the Specificity needs to be validated. The following sections will provide approaches toward the various aspects of method validation. Although the approaches are generally applicable to common techniques used in pharmaceutical analysis (such UV-VIS and HPLC quantifications), particular emphasis will be placed on high performance thin layer chromatography (HPTLC) techniques.

### 2. Specificity

pharmaceutical product. The safety and efficacy of a pharmaceutical product are established with a "pivotal lot" production of the product and the characterization of this lot with well validated methods with acceptable performance characteristics is critical to assure that future production lots have the same quality characteristics as the "pivotal lot", thereby assuring they

144 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

In the United States of America (USA), there are both private and public standards; the private standards are created through a USA Food and Drug Administration (FDA) approval process of industry method submissions that can be used for law enforcement, and public standards, which are promulgated in the monographs of the United States Pharmacopeia (USP) [1], that may be used in law enforcement or to support commercial agreements. The private standards, which are not publicly available, are private agreements between the approving government body and the submitting industry on the methods and standards to be used in law enforcement. The method validation protocols for the establishment of private standards are provided in the guidance of the "International Conference On Harmonisation Of Technical Requirements For Registration Of Pharmaceuticals For Human Use" (ICH) which have been incorporated into the laws and regulations in the European Union, Japan and the USA; these procedures are required for the assessments of new drug entities [2]. The method validation protocols for the establishment of monographs to support public standards are provided in USP <1225> [3] and ICH Q2 [4]. Both protocols cite the same analytical performance characteristics and test procedures except that the public standard must be able to be applied to all legally marketed products containing the specific API whereas the private standard applies

The analytical performance characteristics which must be assessed in both the ICH and USP are Accuracy, Precision (both Repeatability and Intermediate in ICH), and Specificity. Detection Limit, Quantitation Limit, Linearity and Range depending on which attributes are to be

"Category I — Analytical procedures for quantitation of major components of bulk drug substances or active ingredients (including preservatives) in finished pharmaceutical products. Category II — Analytical procedures for determination of impurities in bulk drug substances or degradation compounds in finished pharmaceutical products. These procedures include

Category III — Analytical procedures for determination of performance characteristics (e.g.,

These can be categorized further into Assay procedures for Category I, Category II, impurity determinations, and Category III (dissolution and drug release are different procedures for preparing a solution of the API), all performance characteristics except Detection Limit must be validated. For the Category II limit tests only the Specificity and Detection Limits must be validated and for Category IV, Identification Tests, only the Specificity needs to be validated. The following sections will provide approaches toward the various aspects of method

have equivalent safety and efficacy.

only to the approved API in the specific product.

quantitative assays and limit tests.

Category IV — Identification tests."

dissolution, drug release, etc.).

assessed. The USP presents the characteristics as noted below [3]:

Specificity is the ability of a method to distinguish an analyte from all substances that are present or likely to be present in test samples [3, 4]. When possible, these substances should include future degradation products and other ingredients (i.e., excipients). An analytical procedure is specific when placebo and impurity spots do not overlap partially with and are not buried under the analyte spot. In addition, the calculated amount of analyte does not depend on the quantity of other substances.

Various approaches are possible when evaluating method specificity [5–8]. Ideal demonstration of specificity for an HPTLC analytical procedure requires chromatographing simultaneously three types of samples: sample type 1 is the pure analyte or its reference standard, sample type 2 is the analyte mixed with a representative blank and all likely impurities, and sample type 3 is the representative blank mixed with all likely impurities. Likely impurities include degradation products, reagents, intermediates, excipients, side products, and analyte isomers. The mixtures can be created by spiking test samples (API substances or finished products) or placebos with likely impurities.

In practice, the unavailability of one or more of these types of samples can pose a significant challenge. In some cases, it is often difficult to know all likely impurities. There can be several sample deficiency scenarios. If the pure analyte or its reference standard is not available, demonstration of specificity can be quite challenging if not impossible. If a representative blank is available, but some or all likely impurities are missing, the typical test sample is subjected to stress testing environments. It should be noted however that stress testing is unlikely to produce some analyte isomers, reagents, intermediates and side products. If a representative blank is unavailable, but some likely impurities are available, spiking the typical test sample with impurities can show that increasing impurities will not change analyte signal. In addition, efforts should be made (perhaps by contacting the manufacturer), whenever possible, to create a representative blank even if it's not exactly in the same dosage form as the test sample. If neither a representative blank nor impurity standards are available, the typical test sample is subjected to stress testing to alleviate some of the deficiencies. Once again, the limitations of stress testing should be acknowledged because it may not produce all likely impurities, it may not account for impurities that are completely buried under the analyte signal, and it may not indicate whether some excipients or impurities can react with the analyte.

In general, stress testing, impurity spiking, and peak-purity analysis are the common tools used to address certain sample deficiencies. To demonstrate method specificity, validation reports typically discuss several measures of performance. One measure of specificity is resolution of the analyte spot relative to the closest non-analyte spot. For HPTLC, the resolution should be a least 1 [5].

Analyte peak purity is another measure of specificity that is typically reported. Often, analyte peak purity in the analyte reference standard is compared to analyte peak purity in the other test samples mentioned above. The analysis is performed by comparison of peak spectra at the start, apex, and end of the analyte peak. Some authors use correlation coefficients [9] as a measure of peak purity, and others rely on software algorithms that may involve Matrix Algebra. It should be noted that while peak purity can detect the presence of some impurities in the analyte peak, it does have some limitations. For example, peak-purity analysis does not account for missing impurities that could overlap with the analyte peak, and it does not account for impurities having a spectrum that is similar to that of the analyte. In addition, peak-purity analysis is not applicable for detectors that do not register the entire analyte spectrum for each time point.

in an API substance, the reference standard could contain only the API. However, if the method is intended for finished product testing or for impurity quantitation in API substance, the reference standard should contain the appropriate amount of all the substances typically found in the finished product or API substance. The reference standard should be prepared by

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

147

If a representative certified reference standard is available, the accepted value is the certified

To obtain the method value, at least nine reference standard solutions are prepared and tested as if they were unknown samples, using the candidate HPTLC method. The average and standard deviations of the results will represent the method value. The standard solutions

• The first three standard solutions should contain analyte concentrations between 50 and

• The next three standard solutions should contain analyte concentrations between 90 and

• The last three standard solutions should contain analyte concentrations between 110 and

• Due to the unavailability of representative certified reference standards for most pharma-

The second option involves using a representative blank, which means a substance or mixture that contains all the chemical components of a typical unknown sample except the analyte. It is important to note that the chemical composition of a representative blank depends on both the analyte and the composition of a typical unknown. For assay, or content uniformity, the representative blank is a placebo. For API quantitation in the API substance, the representative blank is typically the solvent used to dissolve the standard. For impurity quantitation in an API substance, the representative blank is typically the API substance. For impurity quantitation in the finished product, the representative blank is a mixture of the placebo plus all the APIs plus all the typical impurities. During testing, the representative blank should be treated the same way as an unknown sample would be. Care must be taken so that only the absence of

If a representative blank is available, at least nine samples are prepared by spiking the blank with various amounts of analyte. The accepted value can be represented as the average and

90% of the analyte's label claim (or quantitation limit for an impurity).

ceutical products, option 1 is rarely used in method validation.

analyte distinguishes the representative blank from a typical unknown sample.

amount of analyte (e.g. API) per given sample of the reference standard.

an ISO certified reference material manufacturer.

should contain the following analyte concentrations:

110% of the analyte's label claim.

3.1.2. Option 2 (Using a representative blank)

3.1.2.1. Determining the accepted value

150% of the label.

3.1.1.1. Determining the accepted value

3.1.1.2. Determining the method value

A third measure of peak purity is an overlay of chromatograms. This measure is especially useful for showing the analyte peak stability during impurity spiking or stress testing. For example, the chromatograms of a finished pharmaceutical product, before and after accelerated aging, can be overlaid to support method specificity.

### 3. Accuracy

A succinct definition of accuracy is "nearness to truth". The ICH guidelines [4] provide the following definition:

"The accuracy of an analytical procedure expresses the closeness of agreement between the value which is accepted either as a conventional true value or an accepted reference value and the value found."

In other words, accuracy of a method represents the agreement between an expected value and the value generated by the candidate method (the method value). Therefore, accuracy determination involves determining the expected value, finding the method value and calculating the agreement between the two values [3, 4].

In pharmaceutical testing, accuracy is mainly relevant to quantitative methods, such as assay, content uniformity, dissolution, and impurity quantitation. To determine the accuracy of a quantitative HPTLC method, there are typically four major options, which differ mainly on how the expected value is determined. Unfortunately, the most preferable options are not always feasible due to the non-availability of appropriate reference standards or placebo samples. For each option, we will explain how to determine the expected value and the method value. Agreement between the two values will be addressed later.

#### 3.1. Options for determining the expected value and the method value

#### 3.1.1. Option 1 (using a certified reference standard)

The first option involves using a representative, certified reference standard. We say representative because the certified reference standard needs to have a chemical matrix that is the same as the matrix of a typical unknown sample. So, if the method is intended for API quantitation in an API substance, the reference standard could contain only the API. However, if the method is intended for finished product testing or for impurity quantitation in API substance, the reference standard should contain the appropriate amount of all the substances typically found in the finished product or API substance. The reference standard should be prepared by an ISO certified reference material manufacturer.

#### 3.1.1.1. Determining the accepted value

Analyte peak purity is another measure of specificity that is typically reported. Often, analyte peak purity in the analyte reference standard is compared to analyte peak purity in the other test samples mentioned above. The analysis is performed by comparison of peak spectra at the start, apex, and end of the analyte peak. Some authors use correlation coefficients [9] as a measure of peak purity, and others rely on software algorithms that may involve Matrix Algebra. It should be noted that while peak purity can detect the presence of some impurities in the analyte peak, it does have some limitations. For example, peak-purity analysis does not account for missing impurities that could overlap with the analyte peak, and it does not account for impurities having a spectrum that is similar to that of the analyte. In addition, peak-purity analysis is not applicable for detectors that do not register the entire analyte

A third measure of peak purity is an overlay of chromatograms. This measure is especially useful for showing the analyte peak stability during impurity spiking or stress testing. For example, the chromatograms of a finished pharmaceutical product, before and after acceler-

A succinct definition of accuracy is "nearness to truth". The ICH guidelines [4] provide the

"The accuracy of an analytical procedure expresses the closeness of agreement between the value which is accepted either as a conventional true value or an accepted reference value and

In other words, accuracy of a method represents the agreement between an expected value and the value generated by the candidate method (the method value). Therefore, accuracy determination involves determining the expected value, finding the method value and calculating

In pharmaceutical testing, accuracy is mainly relevant to quantitative methods, such as assay, content uniformity, dissolution, and impurity quantitation. To determine the accuracy of a quantitative HPTLC method, there are typically four major options, which differ mainly on how the expected value is determined. Unfortunately, the most preferable options are not always feasible due to the non-availability of appropriate reference standards or placebo samples. For each option, we will explain how to determine the expected value and the

The first option involves using a representative, certified reference standard. We say representative because the certified reference standard needs to have a chemical matrix that is the same as the matrix of a typical unknown sample. So, if the method is intended for API quantitation

method value. Agreement between the two values will be addressed later.

3.1. Options for determining the expected value and the method value

spectrum for each time point.

3. Accuracy

following definition:

the value found."

ated aging, can be overlaid to support method specificity.

146 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

the agreement between the two values [3, 4].

3.1.1. Option 1 (using a certified reference standard)

If a representative certified reference standard is available, the accepted value is the certified amount of analyte (e.g. API) per given sample of the reference standard.

### 3.1.1.2. Determining the method value

To obtain the method value, at least nine reference standard solutions are prepared and tested as if they were unknown samples, using the candidate HPTLC method. The average and standard deviations of the results will represent the method value. The standard solutions should contain the following analyte concentrations:


#### 3.1.2. Option 2 (Using a representative blank)

The second option involves using a representative blank, which means a substance or mixture that contains all the chemical components of a typical unknown sample except the analyte. It is important to note that the chemical composition of a representative blank depends on both the analyte and the composition of a typical unknown. For assay, or content uniformity, the representative blank is a placebo. For API quantitation in the API substance, the representative blank is typically the solvent used to dissolve the standard. For impurity quantitation in an API substance, the representative blank is typically the API substance. For impurity quantitation in the finished product, the representative blank is a mixture of the placebo plus all the APIs plus all the typical impurities. During testing, the representative blank should be treated the same way as an unknown sample would be. Care must be taken so that only the absence of analyte distinguishes the representative blank from a typical unknown sample.

#### 3.1.2.1. Determining the accepted value

If a representative blank is available, at least nine samples are prepared by spiking the blank with various amounts of analyte. The accepted value can be represented as the average and standard deviation of all the amounts of analyte spiked to the representative blank samples. At least nine difference samples should be tested covering a minimum of three different concentrations across the expected range of analyte concentration (50–150% of label claim or impurity limit).

3.2. Agreement between expected and method values

tions, it does not take the standard deviations into account.

the candidate method is considered accurate.

for more details on t-test and hypothesis testing.

the upper and lower levels anticipated during an analysis.

 0.06, 0.06, 0.06 0.12, 0.12, 0.11 0.17, 0.17, 0.16 0.22, 0.23, 0.22 0.28, 0.28, 0.28

Table 1. Data to demonstrate a linear calibration model.

4. Linearity

linear regression curve.

Several calculation methods are used to determine the agreement between the expected and method values. The percent recovery method is the simplest. It involves dividing the average method value by the average expected value and multiplying the result by 100. Although this method is considered acceptable in the ICH guidelines [3, 4], and it is found in many publica-

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

149

One method that takes variation into account is the expanded uncertainty interval method [11]. It involves combining the expected and method uncertainties to obtain the expanded uncertainty, which is then compared to the difference between the average expected value and the average method value. If the expanded uncertainty is greater or equal to the difference,

A more statistically rigorous method to calculate accuracy is the t-test for two equal means [12]. It can be performed using MS Excel or other statistical software, but it requires an understanding of hypothesis testing. Interested readers can consult any general Statistics book

The concept of accuracy profile, which is different from the concept of accuracy described herein, is described by Shewiyo et al. It aims to describe method performance using a single statistic [13].

Linearity evaluations demonstrate measurements from a test method are proportional to the amount of analyte within a particular concentration range [3, 4]. Responses from samples containing different amounts of analyte are obtained from the test method. Generally, a minimum of five different concentrations should be used where multiple (i.e., ≥3) responses are obtained at each analyte level. The method response (y-axis) is plotted as a function of the analyte concentration (x-axis) for subsequent analysis with linear regression techniques, where slope, intercept, and correlation coefficient are reported. The concentration range should cover

In the following example, a graphical representation of a linear calibration model is demonstrated, where the raw data is provided in Table 1 and Figure 1 shows the corresponding

[Concentration], w/v [Response], Instrument reading (triplicate results)

#### 3.1.2.2. Determining the method value

Once the spiked blank samples are prepared, they can be analyzed in parallel using the candidate HPTLC method. The average and standard deviation of the results (expressed in the same unit as the accepted value) can represent the method value.

#### 3.1.3. Option 3 (Using a reference method)

If options 1 and 2 are not feasible, a reference method can be used to determine the accuracy of a candidate method. The reference method must be independent of the candidate method, have been well validated with a stated accuracy, and have the same intended use as the candidate method.

#### 3.1.3.1. Determining the accepted value

To obtain the accepted value for option 3, the reference method can be used to test 6 or more unknown samples. The average and standard deviation of the results will represent the accepted value.

#### 3.1.3.2. Determining the method value

To obtain the method value, each of the samples used to determine the accepted value is tested using the candidate method. The average and standard deviation of the results will represent the method value.

#### 3.1.4. Option 4 (Using standard addition to unknown)

In lieu of option 3, method accuracy can be estimated using the standard addition method [10]. In this case the test sample is an unknown finished product or an API substance, whose analyte amount has been predetermined using the candidate method.

#### 3.1.4.1. Determining the accepted value

To obtain the accepted value for option 4, at least 6 or more stock solutions of unknown samples should be prepared and tested per the candidate method. The average and standard deviation of the results will represent the accepted value.

#### 3.1.4.2. Determining the method value

To obtain the method value, each of the stock solutions used to determine the accepted value is tested once again using the standard addition method [10]. So, each stock solution should have its own standard addition curve with 5 or more data points. The average and standard deviations of the absolute values of the x-intercepts will represent the method value.

#### 3.2. Agreement between expected and method values

Several calculation methods are used to determine the agreement between the expected and method values. The percent recovery method is the simplest. It involves dividing the average method value by the average expected value and multiplying the result by 100. Although this method is considered acceptable in the ICH guidelines [3, 4], and it is found in many publications, it does not take the standard deviations into account.

One method that takes variation into account is the expanded uncertainty interval method [11]. It involves combining the expected and method uncertainties to obtain the expanded uncertainty, which is then compared to the difference between the average expected value and the average method value. If the expanded uncertainty is greater or equal to the difference, the candidate method is considered accurate.

A more statistically rigorous method to calculate accuracy is the t-test for two equal means [12]. It can be performed using MS Excel or other statistical software, but it requires an understanding of hypothesis testing. Interested readers can consult any general Statistics book for more details on t-test and hypothesis testing.

The concept of accuracy profile, which is different from the concept of accuracy described herein, is described by Shewiyo et al. It aims to describe method performance using a single statistic [13].

### 4. Linearity

standard deviation of all the amounts of analyte spiked to the representative blank samples. At least nine difference samples should be tested covering a minimum of three different concentrations across the expected range of analyte concentration (50–150% of label claim or

Once the spiked blank samples are prepared, they can be analyzed in parallel using the candidate HPTLC method. The average and standard deviation of the results (expressed in

If options 1 and 2 are not feasible, a reference method can be used to determine the accuracy of a candidate method. The reference method must be independent of the candidate method, have been well validated with a stated accuracy, and have the same intended use as the

To obtain the accepted value for option 3, the reference method can be used to test 6 or more unknown samples. The average and standard deviation of the results will represent the

To obtain the method value, each of the samples used to determine the accepted value is tested using the candidate method. The average and standard deviation of the results will represent

In lieu of option 3, method accuracy can be estimated using the standard addition method [10]. In this case the test sample is an unknown finished product or an API substance, whose

To obtain the accepted value for option 4, at least 6 or more stock solutions of unknown samples should be prepared and tested per the candidate method. The average and standard

To obtain the method value, each of the stock solutions used to determine the accepted value is tested once again using the standard addition method [10]. So, each stock solution should have its own standard addition curve with 5 or more data points. The average and standard

deviations of the absolute values of the x-intercepts will represent the method value.

the same unit as the accepted value) can represent the method value.

148 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

impurity limit).

candidate method.

accepted value.

the method value.

3.1.2.2. Determining the method value

3.1.3. Option 3 (Using a reference method)

3.1.3.1. Determining the accepted value

3.1.3.2. Determining the method value

3.1.4.1. Determining the accepted value

3.1.4.2. Determining the method value

3.1.4. Option 4 (Using standard addition to unknown)

deviation of the results will represent the accepted value.

analyte amount has been predetermined using the candidate method.

Linearity evaluations demonstrate measurements from a test method are proportional to the amount of analyte within a particular concentration range [3, 4]. Responses from samples containing different amounts of analyte are obtained from the test method. Generally, a minimum of five different concentrations should be used where multiple (i.e., ≥3) responses are obtained at each analyte level. The method response (y-axis) is plotted as a function of the analyte concentration (x-axis) for subsequent analysis with linear regression techniques, where slope, intercept, and correlation coefficient are reported. The concentration range should cover the upper and lower levels anticipated during an analysis.

In the following example, a graphical representation of a linear calibration model is demonstrated, where the raw data is provided in Table 1 and Figure 1 shows the corresponding linear regression curve.


Table 1. Data to demonstrate a linear calibration model.

Figure 1. Graphical representation of linear calibration model data.

However, a linear model may not be the best calibration fit for the data as is the case of the data listed in Table 2, and plotted in Figure 2. When the linear model is applied to the data, the resulting correlation coefficient (R2 = 0.98) is less than ideal.

Further examination of the data indicates that a polynomial fit can provide a better calibration model from the data (Figure 3). It should be noted that most pharmaceutical analysis methods commonly use a one-point standard during routine use of the method (after validation has been established).


Table 2. Data to demonstrate a non-linear calibration model.

If the analysis range for the method only requires concentrations from 1 to 4 (w/v), a linear model for just that concentration range provides a r<sup>2</sup> of 0.9994 (Figure 4) and would be easier to implement in future analysis (note-an additional standard should be added within that

Figure 4. Graphical representation of the more linear range of the preceding polynomial calibration model.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Concentration (w/v)

Concentration (w/v)

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

151

y = -0.506x2 + 12.47x - 2.3393 R² = 0.9979

Figure 3. Graphical representation of polynomial approach to calibration model.

y = 10.3x - 0.5 R² = 0.9994

Various options are possible for determining limits of detection (LOD) and limits of quantitation (LOQ) [3, 4, 14, 15]. The section below will provide some key example approaches for tests

range during final validation).

0

10

20

30

40

50

60

) gni dae Rt ne m

urtsnI( esnopse R

) gni dae Rt ne murtsnI( esnopse R

70

5. Limits of detection/quantitation

that generate instrument based responses.

Figure 2. Graphical representation of a less than ideal linear calibration model.

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin… http://dx.doi.org/10.5772/intechopen.71765 151

Figure 3. Graphical representation of polynomial approach to calibration model.

However, a linear model may not be the best calibration fit for the data as is the case of the data listed in Table 2, and plotted in Figure 2. When the linear model is applied to the data, the

0 2 4 6 8 10 12

Concentration (w/v)

Further examination of the data indicates that a polynomial fit can provide a better calibration model from the data (Figure 3). It should be noted that most pharmaceutical analysis methods commonly use a one-point standard during routine use of the method (after validation has

[Concentration], w/v 1 2 3 4 5 6 7 8 [Response], Instrument reading 10 20 30 41 46 55 60 65

0123456789

Concentration (w/v)

resulting correlation coefficient (R2 = 0.98) is less than ideal.

Figure 1. Graphical representation of linear calibration model data.

y = 0.0279x + 0.0008 R² = 0.9987

150 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

0.00

Variable Data

) gni dae Rt ne murtsnI( esnopse R

Table 2. Data to demonstrate a non-linear calibration model.

y = 7.9167x + 5.25 R² = 0.9819

Figure 2. Graphical representation of a less than ideal linear calibration model.

0.05

0.10

0.15

Response (Instrument Reading)

0.20

0.25

0.30

been established).

Figure 4. Graphical representation of the more linear range of the preceding polynomial calibration model.

If the analysis range for the method only requires concentrations from 1 to 4 (w/v), a linear model for just that concentration range provides a r<sup>2</sup> of 0.9994 (Figure 4) and would be easier to implement in future analysis (note-an additional standard should be added within that range during final validation).

#### 5. Limits of detection/quantitation

Various options are possible for determining limits of detection (LOD) and limits of quantitation (LOQ) [3, 4, 14, 15]. The section below will provide some key example approaches for tests that generate instrument based responses.

The signal to noise ratio can be used to determine both the LOD and LOQ, where responses are obtained from blank and from an array of samples at lower concentrations. A ratio of signal (from analyte samples) to noise (from blank) of 3 is an accepted concentration level for the LOD. Likewise, a concentration level that provides a signal to noise of 10 can be used as the LOQ.

where xi represents the individual replicates measurements. The standard deviation (s) of a

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn i¼1

ð Þ n � 1

s x

ð Þ xi � x 2

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

vuuut (2)

� � (3)

http://dx.doi.org/10.5772/intechopen.71765

153

s ¼

%RSD ¼ 100 �

The %RSD is often used in method validation assessments because it normalizes the standard

From Eqs. (1)–(3), an evaluation of repeatability can be determined. In the following example, assume that an analyst has performed six replicate analysis (within the same laboratory) from a method capable of quantifying the amount of active ingredient in a pharmaceutical product in units of % label claim (assay) and obtained the following results (102.1%, 100.5%, 98.2%, 99.1%, 101.8%, 99.8%). Using Eqs. (1)–(3), the average (x), standard deviation (s), and %RSD would be 100.25%, 1.52%, and 1.52%, respectively, where s (or more commonly %RSD), is a

Intermediate precision involves an evaluation of variations "within runs" and "between runs" [17]. Consider the data in Table 3 containing replicate runs (n = 3, indexed with j) obtained on each of multiple days (p = 5, indexed with i), where each day used a different analyst with separate solution preparations but using the same method as above for the repeatability analysis. With each entry in the data representing a separate xij, the repeatability or within-run standard

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

p nð Þ � 1

xij � xi � �<sup>2</sup>

vuuuut (4)

deviation (days) can be determined using Eq. (4), where xi is provided in Eq. (5).

P p

i¼1 Pn j¼1

n = 3(j)/p = 5(i) Day 1 Day 2 Day 3 Day 4 Day 5 Replicate 1 102.2 98.7 99.3 101.9 102.1 Replicate 2 100.3 101.8 98.1 100.1 101.4 Replicate 3 99.9 102.3 98.7 99.1 101.2 xi 100.8 100.9 98.7 100.4 101.6

sr ¼

x 100.5

Table 3. Example data for intermediate precision determination.

and the % relative standard deviation (%RSD) is provided in Eq. (3).

data set can be determined through Eq. (2),

deviation to the average.

measure of method repeatability.

Another approach first involves the determination of the standard deviation of the response and the slope of calibration (linearity) curve. Although other options are possible [3, 4, 14, 15], the standard deviation of the response can be estimated from replicate injections from blank samples or from the standard deviation of y-intercepts from multiple regression lines. Multiplying the ratio of the standard deviation of the responses to the slope of the curve by 3.3 or 10 provides the LOD or LOQ, respectively.

### 6. Precision

For an analytical method, precision is an assessment of the consistency of results obtained with multiple measurements from the same sample [3, 4, 16]. There are three categories of precision for an analytical method, namely repeatability, intermediate precision, and reproducibility, which can be assessed through variations with different equipment, testing times (conducted on different days), analysts, and/or laboratories.

Repeatability is often evaluated with replicate measurements of a sample on the same day in the same laboratory, where the analyst and equipment are not changed. Intermediate precision can be evaluated from replicate measurements of a sample within the same laboratory, but with systematic variations with different analysts, times of analysis, and equipment (such as different instruments). Reproducibility is commonly determined from replicate measurements of the same sample but within different labs, which will inherently incorporate different analysts, equipment, and time of analysis.

ICH Q2 provides several recommendations for number of replicates and concentration levels for each of the three types of precision [4]. Recommended approaches for repeatability are at least nine measurements that span the method's range (such as three replicates for each of three analyte levels) or at least six measurements at the target analyte level. ICH Q2 does not specify a minimum number of samples for intermediate precision and reproducibility but encourages that the effects of variables (i.e., analysts, days, instruments) be systematically evaluated. The following section will provide possible approaches for evaluating repeatability and intermediate precision, followed by references for examples for reproducibility will be provided.

To perform the appropriate precision assessments, the following equations are indicated [16] and will be used for further development of subsequent examples. The average (x) of n replicates is provided in Eq. (1),

$$\overline{\mathbf{x}} = \frac{\sum\_{i=1}^{n} \mathbf{x}\_i}{n} \tag{1}$$

where xi represents the individual replicates measurements. The standard deviation (s) of a data set can be determined through Eq. (2),

$$s = \sqrt{\frac{\sum\_{i=1}^{n} \left(\chi\_i - \overline{\chi}\right)^2}{\left(n - 1\right)}}\tag{2}$$

and the % relative standard deviation (%RSD) is provided in Eq. (3).

The signal to noise ratio can be used to determine both the LOD and LOQ, where responses are obtained from blank and from an array of samples at lower concentrations. A ratio of signal (from analyte samples) to noise (from blank) of 3 is an accepted concentration level for the LOD. Likewise, a concentration level that provides a signal to noise of 10 can be used as the LOQ.

152 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Another approach first involves the determination of the standard deviation of the response and the slope of calibration (linearity) curve. Although other options are possible [3, 4, 14, 15], the standard deviation of the response can be estimated from replicate injections from blank samples or from the standard deviation of y-intercepts from multiple regression lines. Multiplying the ratio of the standard deviation of the responses to the slope of the curve by 3.3 or 10

For an analytical method, precision is an assessment of the consistency of results obtained with multiple measurements from the same sample [3, 4, 16]. There are three categories of precision for an analytical method, namely repeatability, intermediate precision, and reproducibility, which can be assessed through variations with different equipment, testing times (conducted

Repeatability is often evaluated with replicate measurements of a sample on the same day in the same laboratory, where the analyst and equipment are not changed. Intermediate precision can be evaluated from replicate measurements of a sample within the same laboratory, but with systematic variations with different analysts, times of analysis, and equipment (such as different instruments). Reproducibility is commonly determined from replicate measurements of the same sample but within different labs, which will inherently incorporate different

ICH Q2 provides several recommendations for number of replicates and concentration levels for each of the three types of precision [4]. Recommended approaches for repeatability are at least nine measurements that span the method's range (such as three replicates for each of three analyte levels) or at least six measurements at the target analyte level. ICH Q2 does not specify a minimum number of samples for intermediate precision and reproducibility but encourages that the effects of variables (i.e., analysts, days, instruments) be systematically evaluated. The following section will provide possible approaches for evaluating repeatability and intermediate precision, followed by references for examples for reproducibility will be

To perform the appropriate precision assessments, the following equations are indicated [16] and will be used for further development of subsequent examples. The average (x) of n

> Pn i¼1 xi

<sup>n</sup> (1)

x ¼

provides the LOD or LOQ, respectively.

on different days), analysts, and/or laboratories.

analysts, equipment, and time of analysis.

6. Precision

provided.

replicates is provided in Eq. (1),

$$\% \text{RSD} = 100 \times \left(\frac{s}{\overline{\text{x}}}\right) \tag{3}$$

The %RSD is often used in method validation assessments because it normalizes the standard deviation to the average.

From Eqs. (1)–(3), an evaluation of repeatability can be determined. In the following example, assume that an analyst has performed six replicate analysis (within the same laboratory) from a method capable of quantifying the amount of active ingredient in a pharmaceutical product in units of % label claim (assay) and obtained the following results (102.1%, 100.5%, 98.2%, 99.1%, 101.8%, 99.8%). Using Eqs. (1)–(3), the average (x), standard deviation (s), and %RSD would be 100.25%, 1.52%, and 1.52%, respectively, where s (or more commonly %RSD), is a measure of method repeatability.

Intermediate precision involves an evaluation of variations "within runs" and "between runs" [17]. Consider the data in Table 3 containing replicate runs (n = 3, indexed with j) obtained on each of multiple days (p = 5, indexed with i), where each day used a different analyst with separate solution preparations but using the same method as above for the repeatability analysis.

With each entry in the data representing a separate xij, the repeatability or within-run standard deviation (days) can be determined using Eq. (4), where xi is provided in Eq. (5).

$$s\_r = \sqrt{\frac{\sum\_{i=1}^{p} \left(\mathbf{x}\_{\overline{\mathbf{u}}} - \overline{\mathbf{x}\_i}\right)^2}{p(n-1)}}\tag{4}$$


Table 3. Example data for intermediate precision determination.

$$\overline{\mathbf{x}\_i} = \frac{\sum\_{j=1}^n \mathbf{x}\_{ij}}{n} \tag{5}$$

evaluated to assess method robustness; solution stability (to heat and or time), extraction conditions during sample preparation (time, temperature, mechanical shaking time, sonication time), type of filters used during final standard/sample preparation, minor adjustments in mobile phase composition, and adjustments in other chromatographic conditions (flow rate, different suppliers of columns, temperature). Commonly, robustness is evaluated during the

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

155

An approach for evaluating robustness could be to compare an analysis using the primary method compared to an analysis where a certain parameter is adjusted. Depending on the method and sample type, adjustments in parameters that generate less than ~2% difference relative to the primary method can provide a reasonable measure of how sensitive the method is to various types of adjustments. During the development of the method for example, a study could be conducted to evaluate the sensitivity of the method on the type of filter by comparing the results from a sample solution that was centrifuged (without filtration) to those filtered with different filter types (PTFE, PVDF, nylon) from different manufacturers. Solution stability could be evaluated by comparing results from freshly prepared solutions compared to the

There are a variety of approaches that can be incorporated to evaluate method robustness. Dejaegher and Vander Heyden provide an extensive review for a variety of approaches to

The range of the method corresponds to the lower and upper analyte concentration where satisfactory levels of linearity, precision, and accuracy have been achieved during the method validation process. The range is indicated in the same units as that of the results obtained from

For analysis of pharmaceutical products [3, 4], the following ranges (in percentage relative to the target level) are often required for the respective types of tests; assay (80–120%), content uniformity (approximately 70–130%), impurities (approximately 50–120% of the acceptance

High-performance thin layer chromatography (HPTLC), an extension of TLC, is a robust, simple, rapid, and efficient tool in quantitative and qualitative analysis of compounds [21], and a variety of applications can be found in the literature [22–29]. In this section an overview of applications of HPTLC in typical pharmaceutical testing protocol is highlighted with examples. HPTLC is one of the sophisticated instrumental techniques based on the full capabilities of thin layer chromatography. The advantages of automation, scanning, full optimization,

development stages of the method.

8. Range

the method.

same solutions stored at room temperature over several days.

systematically evaluate method robustness (ruggedness) [20].

limit), dissolution ( 20% of the required range).

9. Application with HPTLC techniques

The between-run standard deviation (days) can be calculated with Eq. (6), where x is provided in Eq. (7).

$$s\_{\mathcal{B}} = \sqrt{\frac{\sum\_{i=1}^{p} \left(\overline{\mathbf{x}}\_{i} - \overline{\overline{\mathbf{x}}}\right)^{2}}{p - 1}} - \frac{s\_{r}^{2}}{n} \tag{6}$$

$$\overline{\mathbf{x}} = \frac{\sum\_{i=1}^{p} \sum\_{j=1}^{n} \mathbf{x}\_{\overline{\mathbf{u}}\_{\overline{\mathbf{u}}}}}{\mathbf{p} \mathbf{n}} \tag{7}$$

Subsequently, the intermediate precision standard deviation can be calculated with Eq. (8).

$$\mathbf{s\_{IP}} = \sqrt{\mathbf{s\_r^2} - \mathbf{s\_B^2}} \tag{8}$$

From the data presented in Table 3 and using Eqs. (4)–(8) [17], the standard deviations for repeatability (within-run), between-run, and intermediate precision are calculated as 1.26, 0.80, and 1.49, respectively.

Evaluations for reproducibility utilize interlaboratory trials, and are commonly employed when a procedure requires further standardization for use among a more extended array of laboratories. ISO 5725 [18] provides the necessary approach and management structure needed to properly plan, conduct, and interpret the results of an interlaboratory trial that will involve multiple laboratories conducting replicate analysis of a sample(s) at a particular analyte level(s). Approaches are provided to graphically (Mandel's statistics) and quantitatively (Cochran/ Grubb) identify outliers so that the most accurate assessments of repeatability and reproducibility variance (standard deviations) are possible. The calculations involved in these types of trials are fairly extensive. Several examples are provided within ISO 5725, and Vander Heyden et al. provides a detailed example for an interlaboratory trial for an HPLC procedure [19].

Overall, desired levels for precision for pharmaceutical analysis are commonly on the order of ~2% RSD. However, different ranges can be necessary depending on the concentration level of the analyte (i.e., higher levels of %RSD can be allowed as the analyte concentration decreases) [17].

#### 7. Robustness

Robustness is a measure of how much a method is impacted by deliberate (small) changes in method conditions [3, 4]. The following are a listing of the types of parameters that can be evaluated to assess method robustness; solution stability (to heat and or time), extraction conditions during sample preparation (time, temperature, mechanical shaking time, sonication time), type of filters used during final standard/sample preparation, minor adjustments in mobile phase composition, and adjustments in other chromatographic conditions (flow rate, different suppliers of columns, temperature). Commonly, robustness is evaluated during the development stages of the method.

An approach for evaluating robustness could be to compare an analysis using the primary method compared to an analysis where a certain parameter is adjusted. Depending on the method and sample type, adjustments in parameters that generate less than ~2% difference relative to the primary method can provide a reasonable measure of how sensitive the method is to various types of adjustments. During the development of the method for example, a study could be conducted to evaluate the sensitivity of the method on the type of filter by comparing the results from a sample solution that was centrifuged (without filtration) to those filtered with different filter types (PTFE, PVDF, nylon) from different manufacturers. Solution stability could be evaluated by comparing results from freshly prepared solutions compared to the same solutions stored at room temperature over several days.

There are a variety of approaches that can be incorporated to evaluate method robustness. Dejaegher and Vander Heyden provide an extensive review for a variety of approaches to systematically evaluate method robustness (ruggedness) [20].

### 8. Range

xi ¼

P p

i¼1

P p

i¼1 Pn j¼1 xij

Subsequently, the intermediate precision standard deviation can be calculated with Eq. (8).

q

From the data presented in Table 3 and using Eqs. (4)–(8) [17], the standard deviations for repeatability (within-run), between-run, and intermediate precision are calculated as 1.26, 0.80,

Evaluations for reproducibility utilize interlaboratory trials, and are commonly employed when a procedure requires further standardization for use among a more extended array of laboratories. ISO 5725 [18] provides the necessary approach and management structure needed to properly plan, conduct, and interpret the results of an interlaboratory trial that will involve multiple laboratories conducting replicate analysis of a sample(s) at a particular analyte level(s). Approaches are provided to graphically (Mandel's statistics) and quantitatively (Cochran/ Grubb) identify outliers so that the most accurate assessments of repeatability and reproducibility variance (standard deviations) are possible. The calculations involved in these types of trials are fairly extensive. Several examples are provided within ISO 5725, and Vander Heyden et al.

Overall, desired levels for precision for pharmaceutical analysis are commonly on the order of ~2% RSD. However, different ranges can be necessary depending on the concentration level of the analyte (i.e., higher levels of %RSD can be allowed as the analyte concentration decreases) [17].

Robustness is a measure of how much a method is impacted by deliberate (small) changes in method conditions [3, 4]. The following are a listing of the types of parameters that can be

x ¼

sIP ¼

provides a detailed example for an interlaboratory trial for an HPLC procedure [19].

sB ¼

154 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

in Eq. (7).

and 1.49, respectively.

7. Robustness

Pn j¼1 xij

The between-run standard deviation (days) can be calculated with Eq. (6), where x is provided

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffiffiffiffiffiffiffiffiffiffiffiffiffiffi s2 <sup>r</sup> � <sup>s</sup><sup>2</sup> B

r n

xi � <sup>x</sup> � �<sup>2</sup> <sup>p</sup> � <sup>1</sup> � <sup>s</sup><sup>2</sup>

<sup>n</sup> (5)

vuuut (6)

pn (7)

(8)

The range of the method corresponds to the lower and upper analyte concentration where satisfactory levels of linearity, precision, and accuracy have been achieved during the method validation process. The range is indicated in the same units as that of the results obtained from the method.

For analysis of pharmaceutical products [3, 4], the following ranges (in percentage relative to the target level) are often required for the respective types of tests; assay (80–120%), content uniformity (approximately 70–130%), impurities (approximately 50–120% of the acceptance limit), dissolution ( 20% of the required range).

### 9. Application with HPTLC techniques

High-performance thin layer chromatography (HPTLC), an extension of TLC, is a robust, simple, rapid, and efficient tool in quantitative and qualitative analysis of compounds [21], and a variety of applications can be found in the literature [22–29]. In this section an overview of applications of HPTLC in typical pharmaceutical testing protocol is highlighted with examples. HPTLC is one of the sophisticated instrumental techniques based on the full capabilities of thin layer chromatography. The advantages of automation, scanning, full optimization, selective detection principle, minimum sample preparation, hyphenation, and so on enable it to be a powerful analytical tool for chromatographic information of complex mixtures of pharmaceuticals, natural products, clinical samples, and food stuffs [21]. HPTLC is one of the ideal TLC technique for the analytical purposes because of its increased accuracy, reproducibility, and ability to document the results, compared with standard TLC. Because of this, HPTLC technologies are also the most appropriate TLC technique for conformity with GMPs [30].

9.2. Assay content determination

9.3. Impurities and related substances

gel 60 F254 as the stationary phase.

A second most important critical quality attribute for pharmaceutical products testing is assay or determination of content. The procedure intended to measure the analyte present in a sample. In this context, the assay represents a quantitative measurement of the major component(s) in the drug substance. This is done by comparing the area under the peak of test substance to that of reference standard material. For a drug product, containing paracetamol an overlaid densitogram is presented in Figure 6. Similar validation characteristics also apply when assaying for the active or other selected component(s). The same validation characteristics may also apply

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

157

The principal requirement is that an analytical method for assessing impurities should be a stability indicating and meeting specificity criteria described in Section 2 above. Stability indicating method (SIM) is defined as a validated analytical procedure that accurately and precisely measures the active ingredients (drug substance or drug product) free from process impurities, excipients and degradation products. This can be demonstrated by forced degradation study of the drug substance and subjecting the resultant solution to the chromatographic conditions [34]. Testing for impurities can be either a quantitative test or a limit test for the impurity in a sample. Either test is intended to accurately reflect the purity characteristics of the sample. Quantitative tests for impurities are meant to quantify the exact amount of impurity. This is

Figure 6. An example of overlaid densitograms for assay of paracetamol in sample tracks 2, 3, 5, 8 and 9, and a reference in tracks 1, 4 and 7. Conditions: Mobile Phase: Acetone; Methanol; Toluene: 6:6:16 v/v/v acidified with three drops of Glacial Acetic Acid, Detection Wavelength: 274 nm, Application Volume: 5 μl and aluminum plates precoated with silica

to assays associated with other analytical procedures (e.g., dissolution) [33].

#### 9.1. Identification test

In a pharmaceutical testing protocol, identification tests are intended to ensure the identity of an analyte in an API or finished pharmaceutical product sample. This is normally achieved by comparison of a chromatographic behavior of unknown sample to that of a reference standard. The identity of the test substance is confirmed if the migration distance of the test substance matches that of the reference substance. Thin layer chromatography experiments are among the key identity tests in most pharmacopeia monographs. Pharmacopeia standards are typically used by industry as a basis for meeting QC requirements and current good manufacturing practices (cGMPs). Many identification tests in the major pharmacopeia (e.g., USP, Ph. Int., and Ph. Eur. [1, 31, 32]) use planar chromatography (TLC), however HPTLC is a superior technology. Figure 5 below represent a typical densitogram obtained in the identification of sulfamethoxazole (SMX) and trimethoprim (TPM). In this example, the migration distances are 0.35 and 0.90 for TMP and SMX respectively.

Figure 5. An example of overlaid densitogram for identification of sample 1 and a reference 2 containing sulfamethoxazole (SMX) and trimethoprim (TPM). Conditions Mobile Phase: (Methanol: Ethyl Acetate: Toluene 6: 9:15 v/v) Detection Wavelength: 275 nm and Application Volume: 5 μl and aluminum plates precoated with silica gel 60 F254 as the stationary phase.

#### 9.2. Assay content determination

selective detection principle, minimum sample preparation, hyphenation, and so on enable it to be a powerful analytical tool for chromatographic information of complex mixtures of pharmaceuticals, natural products, clinical samples, and food stuffs [21]. HPTLC is one of the ideal TLC technique for the analytical purposes because of its increased accuracy, reproducibility, and ability to document the results, compared with standard TLC. Because of this, HPTLC technologies are also the most appropriate TLC technique for conformity with GMPs

156 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

In a pharmaceutical testing protocol, identification tests are intended to ensure the identity of an analyte in an API or finished pharmaceutical product sample. This is normally achieved by comparison of a chromatographic behavior of unknown sample to that of a reference standard. The identity of the test substance is confirmed if the migration distance of the test substance matches that of the reference substance. Thin layer chromatography experiments are among the key identity tests in most pharmacopeia monographs. Pharmacopeia standards are typically used by industry as a basis for meeting QC requirements and current good manufacturing practices (cGMPs). Many identification tests in the major pharmacopeia (e.g., USP, Ph. Int., and Ph. Eur. [1, 31, 32]) use planar chromatography (TLC), however HPTLC is a superior technology. Figure 5 below represent a typical densitogram obtained in the identification of sulfamethoxazole (SMX) and trimethoprim (TPM). In this example, the migration distances are

Figure 5. An example of overlaid densitogram for identification of sample 1 and a reference 2 containing sulfamethoxazole (SMX) and trimethoprim (TPM). Conditions Mobile Phase: (Methanol: Ethyl Acetate: Toluene 6: 9:15 v/v) Detection Wavelength: 275 nm and Application Volume: 5 μl and aluminum plates precoated with silica gel 60 F254 as the

[30].

9.1. Identification test

stationary phase.

0.35 and 0.90 for TMP and SMX respectively.

A second most important critical quality attribute for pharmaceutical products testing is assay or determination of content. The procedure intended to measure the analyte present in a sample. In this context, the assay represents a quantitative measurement of the major component(s) in the drug substance. This is done by comparing the area under the peak of test substance to that of reference standard material. For a drug product, containing paracetamol an overlaid densitogram is presented in Figure 6. Similar validation characteristics also apply when assaying for the active or other selected component(s). The same validation characteristics may also apply to assays associated with other analytical procedures (e.g., dissolution) [33].

#### 9.3. Impurities and related substances

The principal requirement is that an analytical method for assessing impurities should be a stability indicating and meeting specificity criteria described in Section 2 above. Stability indicating method (SIM) is defined as a validated analytical procedure that accurately and precisely measures the active ingredients (drug substance or drug product) free from process impurities, excipients and degradation products. This can be demonstrated by forced degradation study of the drug substance and subjecting the resultant solution to the chromatographic conditions [34].

Testing for impurities can be either a quantitative test or a limit test for the impurity in a sample. Either test is intended to accurately reflect the purity characteristics of the sample. Quantitative tests for impurities are meant to quantify the exact amount of impurity. This is

Figure 6. An example of overlaid densitograms for assay of paracetamol in sample tracks 2, 3, 5, 8 and 9, and a reference in tracks 1, 4 and 7. Conditions: Mobile Phase: Acetone; Methanol; Toluene: 6:6:16 v/v/v acidified with three drops of Glacial Acetic Acid, Detection Wavelength: 274 nm, Application Volume: 5 μl and aluminum plates precoated with silica gel 60 F254 as the stationary phase.

done by comparing the response from a single or multi-level calibration curve. Whereas the limit test is an estimative test where the impurity is controlled not to exceed certain limit. In this case an impurity standard is prepared at the control level and compared to the response from the sample (which should not exceed this level).

important examples are provided using HPTLC techniques that provide high accuracy/preci-

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

This work was funded through support from USAID Contract No. AID-OAA-C-15-00001,

2 Muhimbili University of Health and Allied Sciences, School of Pharmacy, Dar es Salaam,

[1] USP 40/NF 35, United States Pharmacopeia. 40th ed., National Formulary. 35th ed. Rockville, MD: The United States Pharmacopeia Convention; United Book Press, Inc.; 2017 [2] International Conference for Harmonization (ICH) [Internet]. 2017. Available from:

[3] USP <1225> Validation of Compendial Procedures. In: USP 40/NF 35, United States Pharmacopeia. 40th ed., National Formulary. 35th ed. Rockville, MD: The United States

[4] Validation of Analytical Procedures: Text and Methodology Q2(R1) [Internet]. 2005. Available from: http://www.ich.org/fileadmin/Public\_Web\_Site/ICH\_Products/Guidelines/

[5] Ferenczi-Fodor A, Vigh Z, Nagy-Turak A. Validation and quality assurance of planar chromatographic procedures in pharmaceutical analysis. Journal - Association of Official

[6] Maggio RM, Vignaduzzo SE, Kaufman TS. Practical and regulatory considerations for stability-indicating methods for the assay of bulk drugs and drug formulations. TRAC-

[7] Bonfilio R, Laignier Cazedey EC, de Araujo MB, Nunes Salgado HR. Analytical validation of quantitative high-performance liquid chromatographic methods in pharmaceutical

Trends in Analytical Chemistry. 2013;49:57-70. DOI: 10.1016/j.trac.2013.05.008

, Eliangiringa Kaale2 and Thomas Layloff<sup>1</sup>

http://dx.doi.org/10.5772/intechopen.71765

159

sion with minimal use of reagents and other resources.

\*, Cherif Diallo<sup>1</sup>

\*Address all correspondence to: djenkins@fhi360.org

Global Health Supply Chain Quality Assurance Program (GHSC-QA).

, Ed Bethea1

1 FHI 360, Product Quality and Compliance, Durham, NC, USA

http://www.ich.org/home.html [Accessed: Sep, 25 2017]

Pharmacopeia Convention; United Book Press, Inc.; 2017

Analytical Chemists. 2001;84(4):1265-1276

Quality/Q2\_R1/Step4/Q2\_R1\_\_Guideline.pdf [Accessed: Sep, 25 2017]

Acknowledgements

Author details

David Jenkins<sup>1</sup>

Tanzania

References

With the improved resolution powers of HPTLC (enhanced by reduced particle sizes), it is possible to perform both tests by using HPTLC. In the literature, there are many stability indicating method for various drug substances for example pseudoephedrine and cetirizine in pharmaceutical formulations [35], clopidogrel bisulphate [36] timolol maleate [37] simultaneous determination of ezetimibe and simvastatin [38], piroxicam [39], and estradiol [40].

#### 9.4. Dissolution testing

Dissolution testing is a performance characterizing test and a requirement for all solid oral dosage forms and is used in all phases of development for product release and stability testing [41–43]. It is a key analytical test used for detecting physical changes in an active pharmaceutical ingredient (API) and in the formulated product. It is a multi-unit test and multi-point sampling, making it very tedious. HPTLC offers a multi-channel capability where a total of 18– 25 samples can be applied on one plate in form of bands and analyzed simultaneously. One lot of a product will require 6 units (tested in duplicate), plus calibrators (in triplicates at single or multiple levels). HPTLC methods have been successfully deployed for monitoring dissolution profile of diclofenac and acetaminophen [44], and the stability of rifampicin in dissolution medium in presence of isoniazid [45].

#### 9.5. Content uniformity

The test for Content Uniformity (CU) is the assay of the individual content of drug substance(s) in a number of individual dosage units to determine whether the individual contents are within the set limits [46]. Multiple capsules or tablets are selected at random and each are analyzed to determine the active ingredient in each capsule or tablet. The performance efficiency of this method can benefit from the HPTLC multi-channel capabilities. HPTLC has been successful applied in content uniformity of atorvastatin calcium tablets [47], diazepam tablets [48], diosgenin and levodopa [49], nicorandil tablets [50] and rosiglitazone in tablets [51]. All of these HPTLC method examples provide a faster, more cost efficient approach to quantitative testing for routine analysis.

### 10. Conclusions

Classic method validations for pharmaceuticals involve techniques such as UV-VIS, TLC, and HPLC. This chapter highlights different examples with High Performance Thin Layer Chromatography (HPTLC). General approaches are provided for method validation, as applicable to pharmaceutical assessments, outlined for each of the key aspects (i.e., specificity, accuracy, linearity, limits of detection/quantitation, precision, robustness, and range). Although classical application of pharmaceutical method validation uses techniques such as UV-VIS or HPLC, important examples are provided using HPTLC techniques that provide high accuracy/precision with minimal use of reagents and other resources.

### Acknowledgements

done by comparing the response from a single or multi-level calibration curve. Whereas the limit test is an estimative test where the impurity is controlled not to exceed certain limit. In this case an impurity standard is prepared at the control level and compared to the response

With the improved resolution powers of HPTLC (enhanced by reduced particle sizes), it is possible to perform both tests by using HPTLC. In the literature, there are many stability indicating method for various drug substances for example pseudoephedrine and cetirizine in pharmaceutical formulations [35], clopidogrel bisulphate [36] timolol maleate [37] simultaneous determination of ezetimibe and simvastatin [38], piroxicam [39], and estradiol [40].

Dissolution testing is a performance characterizing test and a requirement for all solid oral dosage forms and is used in all phases of development for product release and stability testing [41–43]. It is a key analytical test used for detecting physical changes in an active pharmaceutical ingredient (API) and in the formulated product. It is a multi-unit test and multi-point sampling, making it very tedious. HPTLC offers a multi-channel capability where a total of 18– 25 samples can be applied on one plate in form of bands and analyzed simultaneously. One lot of a product will require 6 units (tested in duplicate), plus calibrators (in triplicates at single or multiple levels). HPTLC methods have been successfully deployed for monitoring dissolution profile of diclofenac and acetaminophen [44], and the stability of rifampicin in dissolution

The test for Content Uniformity (CU) is the assay of the individual content of drug substance(s) in a number of individual dosage units to determine whether the individual contents are within the set limits [46]. Multiple capsules or tablets are selected at random and each are analyzed to determine the active ingredient in each capsule or tablet. The performance efficiency of this method can benefit from the HPTLC multi-channel capabilities. HPTLC has been successful applied in content uniformity of atorvastatin calcium tablets [47], diazepam tablets [48], diosgenin and levodopa [49], nicorandil tablets [50] and rosiglitazone in tablets [51]. All of these HPTLC method examples provide a faster, more cost efficient approach to quantitative

Classic method validations for pharmaceuticals involve techniques such as UV-VIS, TLC, and HPLC. This chapter highlights different examples with High Performance Thin Layer Chromatography (HPTLC). General approaches are provided for method validation, as applicable to pharmaceutical assessments, outlined for each of the key aspects (i.e., specificity, accuracy, linearity, limits of detection/quantitation, precision, robustness, and range). Although classical application of pharmaceutical method validation uses techniques such as UV-VIS or HPLC,

from the sample (which should not exceed this level).

158 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

9.4. Dissolution testing

medium in presence of isoniazid [45].

9.5. Content uniformity

testing for routine analysis.

10. Conclusions

This work was funded through support from USAID Contract No. AID-OAA-C-15-00001, Global Health Supply Chain Quality Assurance Program (GHSC-QA).

### Author details

David Jenkins<sup>1</sup> \*, Cherif Diallo<sup>1</sup> , Ed Bethea1 , Eliangiringa Kaale2 and Thomas Layloff<sup>1</sup>

\*Address all correspondence to: djenkins@fhi360.org

1 FHI 360, Product Quality and Compliance, Durham, NC, USA

2 Muhimbili University of Health and Allied Sciences, School of Pharmacy, Dar es Salaam, Tanzania

### References


analysis: A practical approach. Critical Reviews in Analytical Chemistry. 2012;42(1):87- 100. DOI: 10.1080/10408347.2012.630926

[19] Vander Heyden Y, Saevels J, Roets E, Hoogmartens J, Decolin D, Quaglia M, Van den Bossche W, Leemans R, Smeets O, Van de Vaart F, Mason B, Taylor G, Underberg W, Bult A, Chiap P, Crommen J, De Beer J, Hansen S, Massart D. Interlaboratory studies on two high-performance liquid chromatographic assays for tylosin (tartrate). Journal of

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

161

[20] Dejaegher B, Vander Heyden Y. Ruggedness and robustness testing. Journal of Chroma-

[21] Attimarad M, Ahmed KKM, Aldhubaib BE, Harsha S. High-performance thin layer chromatography: A powerful analytical technique in pharmaceutical drug discovery.

[22] Abourashed EA, Mossa JS. HPTLC determination of caffeine in stimulant herbal products and power drinks. Journal of Pharmaceutical and Biomedical Analysis. 2004;36(3):

[23] Ansari MJ, Ahmad S, Kohli K, Ali J, Khar RK. Stability-indicating HPTLC determination of curcumin in bulk drug and pharmaceutical formulations. Journal of Pharmaceutical

[24] Dar AA, Sangwan PL, Khan I, Gupta N, Qaudri A, Tasduq SA, et al. Simultaneous quantification of eight bioactive secondary metabolites from Codonopsis ovata by validated high performance thin layer chromatography and their antioxidant profile. Journal of Pharmaceutical and Biomedical Analysis. 2014;100:300-308. DOI: 10.1016/j.jpba.2005.03.021

[25] Kaul N, Dhaneshwar SR, Agrawal H, Kakad A, Patil B. Application of HPLC and HPTLC for the simultaneous determination of tizanidine and rofecoxib in pharmaceutical dosage form. Journal of Pharmaceutical and Biomedical Analysis. 2005;37(1):27-38. DOI: 10.1016/

[26] Koba M, Marszall M, Sroka W, Tarczykowska M, Bucinski A. Application of HPTLC and LC-MS methods for determination of topiramate in pharmaceutical formulations. Cur-

[27] Mahgoub H, Youssef RM, Korany MA, Khamis EF, Kamal MF. Development and validation of spectrophotometric and HPTLC methods for simultaneous determination of rosiglitazone maleate and metformin hydrochloride in the presence of interfering matrix excipients. Drug Development and Industrial Pharmacy. 2014;40(9):1190-1198. DOI:

[28] Minz S, Kaurav M, Sahu KK, Mandal V, Pandey RS. Development and validation of TLCdensitometric method for determination of lipid A adjuvant as a bulk and in solid fat nanoemulsions. Biomedical Chromatography. 2015;29(10):1473-1479. DOI: 10.1002/bmc.

[29] Shirode AR, Ghuge AD, Kadam VJ. HPTLC method development and validation of cinnarizine in bulk and marketed formulation. International Journal of Pharmaceutical Sciences and Research. 2016;7(6):2416-2422. DOI: 10.13040/ijpsr.0975-8232.7(6).2416-22

and Biomedical Analysis. 2005;39(1–2):132-138. DOI: 10.1016/j.jpba.2005.03.021

Chromatography, A. 1999;830:3-28. DOI: 10.1016/S0021-9673(98)00840-1

Pharmaceutical Methods. 2011;2(2):71-75. DOI: 10.4103/2229-4708.84436

tography, A. 2007;1158:138-157. DOI: 10.1016/j.chroma.2007.02.086

617-620. DOI: 10.1016/j.jpba.2004.06.029

rent Pharmaceutical Analysis. 2012;8(1):44-48

10.3109/03639045.2013.810634

j.jpha.2004.09.036

3444


[19] Vander Heyden Y, Saevels J, Roets E, Hoogmartens J, Decolin D, Quaglia M, Van den Bossche W, Leemans R, Smeets O, Van de Vaart F, Mason B, Taylor G, Underberg W, Bult A, Chiap P, Crommen J, De Beer J, Hansen S, Massart D. Interlaboratory studies on two high-performance liquid chromatographic assays for tylosin (tartrate). Journal of Chromatography, A. 1999;830:3-28. DOI: 10.1016/S0021-9673(98)00840-1

analysis: A practical approach. Critical Reviews in Analytical Chemistry. 2012;42(1):87-

[8] Rashmin P, Mrunali P, Nitin D, Nidhi D, Bharat P. HPTLC method development and validation: Strategy to minimize methodological failures. Journal of Food and Drug

[9] Shaalan R, Belal T, Yazbi F, Elonsy S. Validated HPTLC methods for determination of some selected antihypertensive mixtures in their combined dosage forms. Bulletin of

[10] Standard-Addition Methods [Internet]. 2012. Available from: https://chemistry.cos.ucf. edu/redesign/wp-content/uploads/2012/10/Overheads-28-34\_Spring\_14.pdf. [Accessed:

[11] Linsinger T. Comparison of a measurement result with the certified value. European Commission - Joint Research Centre Institute for Reference Materials and Measurements (IRMM) [Internet]. 2010. Available from: http://www.erm-crm.org/ERM\_products/

[12] Belal T, Mahrous M, Abdel-Khalek M, Daabees H, Khamis M. Validated HPTLC method for the simultaneous determination of alfuzosin, terazosin, prazosin, doxazosin and finasteride in pharmaceutical formulations. Analytical Chemistry Research. 2014;1:23-31.

[13] Shewiyo D, Kaale E, Risha P, Dejaegher B, Beer J, Smeyers-Verbeke J, Vander HY. Accuracy profiles assessing the validity for routine use of high-performance thin-layer chromatographic assays for drug formulations. Journal of Chromatography, A. 2013;1293:159-

[14] Shrivastava A, Gupta VB. Methods for the determination of limit of detection and limit of

[15] Paino T, Moore A. Determination of the LOD and LOQ of an HPLC method using four

[16] USP <1010> Analytical Data – Interpretation and Treatment. In: USP 40/NF 35, United States Pharmacopeia. 40th ed., National Formulary. 35th ed. Rockville, MD: The United

[17] Gonzalez A, Herrador M, Asuero A. Intra-laboratory assessment of method accuracy (trueness and precision) by using validation standards. Talanta. 2010;82:1995-1998. DOI:

[18] International Organization for Standardization (ISO), Accuracy (trueness and precision) of Measurement Methods and Results – Part 2: Basic Method for the Determination of Repeatability and Reproducibility of a Standard Measurement Method, ISO 5725-2: 1994/

quantitation of the analytical methods. Chron Young Sci. 2011;2:21-25

different techniques. Pharmaceutical Technology. Oct. 1999:86-90

States Pharmacopeia Convention; United Book Press, Inc.; 2017

application\_notes/application\_note\_1/Documents/erm\_application\_note\_1\_en-

100. DOI: 10.1080/10408347.2012.630926

glish\_rev3.pdf. [Accessed: Aug, 29 2017]

169. DOI: 10.1016/j.chroma.2013.03.074

DOI: 10.1016/j.jep.2015.08.001

10.1016/j.talanta.2010.07.071

Cor.1:2002(E)

Aug, 27 2017]

Analysis. 2021;20(4):794-804. DOI: 10.6227/jfda.2012200408

160 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Faculty of Pharmacy, Cairo University. 2014;52(2):225-237


[30] Sudberg S, Sudberg EM, Terrazas J, Sudberg S, Patel K, Pineda J, et al. Fingerprint analysis and the application of HPTLC to the determination of identity and quality of botanicals, from an industry perspective. Journal of AOAC International. 2010;93(5):1367- 1375

[43] U.S. FDA/CDER. Guidance for Industry: Dissolution testing of immediate release solid oral dosage forms [Internet]. 1997. Available from: https://www.fda.gov/downloads/

Method Validation Approaches for Pharmaceutical Assessments – Highlights with High Performance Thin…

http://dx.doi.org/10.5772/intechopen.71765

163

[44] Shah H, Rathod I, Shah S, Savale S, Shishoo C. Sensitive HPTLC method for monitoring dissolution profiles of diclofenac from different tablets containing combined diclofenac and acetaminophen. JPC Journal of Planar Chromatography Modern TLC. 2003 Feb 1;

[45] Shishoo C, Shah S, Rathod I, Savale S, Kotecha J, Shah P. Stability of rifampicin in dissolution medium in presence of isoniazid. International Journal of Pharmaceutics.

[46] USP <905> Uniformity of Dosage Units In: USP 40/NF 35, United States Pharmacopeia. 40th ed., National Formulary. 35th ed. Rockville, MD: The United States Pharmacopeia

[47] Yadav SS, Mhaske DV, Kakad A, Patil B, Kadam S, Dhaneshwar S. A simple snd sensitive HPTLC for the determination of content uniformity of atorvastatin calcium tablets.

[48] Machale V, Gatade A, Sane R. Validated HPTLC content uniformity test for the determination of diazepam in tablet dosage form. International Journal of Research and Devel-

[49] Kshirsagar VB, Deokate UA, Bharkad VB, Khadabadi SS. HPTLC method development and validation for the simultaneous estimation of diosgenin and levodopa in marketed

[50] Patel CN, Shah SA, Rathod IS, Savale SS. A simple and sensitive HPTLC method for the determination of content uniformity of nicorandil tablets. Indian Journal of Pharmaceuti-

[51] Walode SG, Chaudhari HK, Saraswat MS, Kasture AV, Wadodkar SG. Validated high performance thin layer chromatographic determination and content uniformity test for rosiglitazone in tablets. Indian Journal of Pharmaceutical Sciences. 2010 Mar;72(2):249

formulation. Asian Journal of Research in Chemistry. 2008 Jul;1(1):36-39

drugs/guidances/ucm070237.pdf [Accessed: Sep, 26 2017]

1999;190(1):109-123. DOI: 10.1016/S0378-5173(99)00286-0

Indian Journal of Pharmaceutical Sciences. 2005;67(2):182

16(1):36-44. DOI: 10.1556/JPC.16.2003.1.8

Convention; United Book Press, Inc.; 2017

opment. 2011;3:34-41

cal Sciences. 2002;64(4):362


[43] U.S. FDA/CDER. Guidance for Industry: Dissolution testing of immediate release solid oral dosage forms [Internet]. 1997. Available from: https://www.fda.gov/downloads/ drugs/guidances/ucm070237.pdf [Accessed: Sep, 26 2017]

[30] Sudberg S, Sudberg EM, Terrazas J, Sudberg S, Patel K, Pineda J, et al. Fingerprint analysis and the application of HPTLC to the determination of identity and quality of botanicals, from an industry perspective. Journal of AOAC International. 2010;93(5):1367-

[31] World Health Organization. The International Pharmacopoeia. 6th ed. Geneva: World Health Organization, Dept. of Essential Medicines and Pharmaceutical Policies [Internet]. 2008. Available from: http://apps.who.int/phint/en/p/about/. [Accessed: Sep, 26 2017]

[33] Shewiyo D, Kaale E, Risha P, Dejaegher B, Smeyers-Verbeke J, Vander HY. HPTLC methods to assay active ingredients in pharmaceutical formulations: A review of the method development and validation steps. Journal of Pharmaceutical and Biomedical

[34] Farid NF, Abdelwahab NS. Stability-indicating HPTLC method for studying stress degradation behavior of sulbutiamine HCl. Journal of Chromatographic Science. 2016;54(4):

[35] Makhija SN, Vavia PR. Stability indicating HPTLC method for the simultaneous determination of pseudoephedrine and cetirizine in pharmaceutical formulations. Journal of Pharmaceutical and Biomedical Analysis. 2001;25(3):663-667. DOI: 10.1016/S0731-7085

[36] Agrawal H, Kaul N, Paradkar A, Mahadik K. Stability indicating HPTLC determination of clopidogrel bisulphate as bulk drug and in pharmaceutical dosage form. Talanta.

[37] Kulkarni SP, Amin PD. Stability indicating HPTLC determination of timolol maleate as bulk drug and in pharmaceutical preparations. Journal of Pharmaceutical and Biomedical

[38] Dixit RP, Barhate CR, Nagarsenker MS. Stability-indicating HPTLC method for simultaneous determination of ezetimibe and simvastatin. Chromatographia. 2008;67(1–2):101-

[39] Puthli S, Vavia P. Stability indicating HPTLC determination of piroxicam. Journal of Pharmaceutical and Biomedical Analysis. 2000;22(4):673-677. DOI: 10.1016/S0731-7085(99)00300-3

[40] Kotiyan P, Vavia P. Stability indicating HPTLC method for the estimation of estradiol. Journal of Pharmaceutical and Biomedical Analysis. 2000;22(4):667-671. DOI: 10.1016/

[41] USP <711> Dissolution. In: USP 40/NF 35, United States Pharmacopeia. 40th ed., National Formulary. 35th ed. Rockville, MD: The United States Pharmacopeia Convention; United

[42] Zhang H, Yu L. Dissolution testing for solid oral drug products: Theoretical considerations. American Pharmaceutical Review. 2004;7(5):26-31. DOI: 10.1023/A:1011984216775

[32] The Council of Europe. European Pharmacopoeia 8.0: Strasbourg, France; 2014

Analysis. 2012;66:11-23. DOI: 10.1016/j.jpba.2012.03.034

162 Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

2003;61(5):581-589. DOI: 10.1016/S0039-9140(03)00364-3

Analysis. 2000;23(6):983-987. DOI: 10.1016/S0731-7085(00)00389-7

609-617. DOI: 10.1093/chromsci/bmv226

107. DOI: 10.1365/s10337-007-0459-4

S0731-7085(99)00301-5

Book Press, Inc.; 2017

1375

(00)00586-0


## *Edited by Mark T. Stauffer*

This book seeks to introduce the reader to current methodologies in analytical calibration and validation. This collection of contributed research articles and reviews addresses current developments in the calibration of analytical methods and techniques and their subsequent validation. Section 1, "Introduction," contains the Introductory Chapter, a broad overview of analytical calibration and validation, and a brief synopsis of the following chapters. Section 2 "Calibration Approaches" presents five chapters covering calibration schemes for some modern analytical methods and techniques. The last chapter in this section provides a segue into Section 3, "Validation Approaches," which contains two chapters on validation procedures and parameters. This book is a valuable source of scientific information for anyone interested in analytical calibration and validation.

Published in London, UK © 2018 IntechOpen © Daniele Levis Pelusi / unsplash

Calibration and Validation of Analytical Methods - A Sampling of Current Approaches

Calibration and Validation

of Analytical Methods

A Sampling of Current Approaches