**5.1. Statistical analysis**

88 Enhancing Success of Assisted Reproduction

Sampling of venous blood was in the cubital vein in VACUETTE test tubes with the buffer solution of sodium citrate with the proportion of 9:1 (9NC Coagulation sodium citrate 3.2%). Blood was centrifuged at 1400 g and room temperature for 15 minutes. Plasma samples were generally studied within the two hours after being obtained. Prior to conducting immune-enzyme assays and estimating the endogenous thrombin potential, plasma was stored at -40 degrees Centigrade for the period of time ranging from 24 hours to 1 month.

*Measuring of the endogenous thrombin potential* was made as a part of the thrombin generation assay (TGA). We believe that this method may be regarded as a historical modification of the two-stage self-coagulogram suggested by Berkarda et al. (1965) and developed to pursue similar goals. It was demonstrated that TGA allows experts to measure the dynamics of thrombin generation and inactivation with high precision (Hemker et al., 2003; Hemker et al., 2006). To perform calibrated automated thrombography, the Fluoroskan Ascent microplate fluorometer was applied (Thermo Fisher Scientific, Finland) equiped with a dispenser with Thrombinoscope 3.0.0.26 software. Coagulation of plasma under study was conducted in the presence of tissue factor (5 μM) and phospholipids (4 μM); thrombin generation was continually registered by measuring the signal of fluorogenic substrate (Z-Gly-Gly-Arg-АМС). The following parameters were considered: endogenous thrombin potential (ETP, nM×min), calculating the area under the thrombin generation curve and taking into account specifics of the enzyme inactivation, and peak thrombin concentration -

*Activity of t-PA and PAI-1* was defined by means of the immune-enzyme analysis with sets of reagents t-PA Combi Actibind ELISA Kit and Actibind PAI-1 ELISA (Technoclone, Austria), while the collected data was estimated by mutual comparison. Due to the fact that these important participants of fibrinolytic responses are antagonistic to each other and have a common origin (vascular endothelium), we calculated the index of endothelial ability to

Activity of t PA, un / ml EAAF index, % 100%.

*Defining clot lysis time (CLT).* Many authors refer this method to the group of global assays for fibrinolysis assessment. It can be conducted in a variety of ways (Lisman et al., 2005;

Method to define spontaneous lysis time of a fibrin clot obtained from plasma euglobulins was described by Kowarzyk and Buluk (1954). It is based on the precipitation of euglobulin fraction stabilized with sodium citrate in the acid medium with the simultaneous removal of fibrinolysis inhibitors. Then, following this methodology, specialists promote clot formation by recalcifying the reconstituted euglobulin solution and register period for its complete dissolution at fixed temperature (+37 degrees Centigrade). Martinez-Zamora et al. (2011) have recently published original CLT results obtained in women participating in IVF cycles using another method described by Lisman et al. (2002). In particular, this method implies the study of CLT for clots obtained from blood plasma by means of activitaing fibrinolysis with

Activity of PAI 1, un / ml 

Peak thrombin (nM/l), maximal thrombin concentration per time unit.

Martinez-Zamora et al., 2011; Wichers et al., 2009).

activate fibrinolysis (EAAF index). To do that we applied the following formula:

Statistical processing of the obtained data was made with the following software: Microsoft Office Excel 2003, Statistica 6.1, and Medcalc 12.2.1. Validity of the differences in mean values was defined with the Student's t-test (t). Group distribution normalcy was estimated with the Shapiro-Wilk test. In cases when the distribution deviated from the norm, we used the non-parametric Mann-Whitney U test for two independent groups and the Spearman's rank correlation coefficient (R). For experimental data presented in percentages or rates the Fisher's exact test was used. Odds ratio (OR) and log-linear rate analysis were calculated as a measure of predictor impact. To estimate the accuracy of obtained values, we defined a 95% confidence interval. Differences P<0.05 were considered statistically significant. We assessed the efficacy of the chosen treatment methods with conventional criteria applied in evidencebased medicine, including Absolute Risk Reduction (ARR), Relative Risk (RR), Relative Risk Reduction (RRR), Number Needed to Treat (NNT), and Confidence Interval (CI).

#### **5.2. Study results**

*5.2.1. Estimation of hemostatic and fibrinolytic indications in the IVF cycle and definition of threshold values occurring in the assay results for the selection of women who need therapeutic intervention during the controlled ovarian hyperstimulation* 

At the beginning of the study, it was important to investigate shifts in the hemostatic and fibrinolytic systems that may affect IVF outcomes (in terms of the pregnancy rate) and to define their quantitative level that may help in selecting those women who need the correction of disturbed hemostatic and fibrinolytic responses. As a result, we defined that

thrombin generation in the observational group had considerably increased after the start of controlled ovarian hyperstimulation in response to a sharp increase in the level of estrogens in blood, which is consistent with recently published data (Westerlund et al., 2012). We have also found that the degree of the increase in thrombin generation is different at successful and unsuccessful IVF outcomes (Table 2). In particular, the values of such thrombin generation factors as ETP and peak thrombin concentration at the 2nd observation point were higher in case of IVF failures (P<0.001) as compared with the similar data in patients with pregnancies.

The Means of Progress in Improving the Results of *in vitro* Fertilization Based on the Identification and Correction of the Pathology of Hemostasis 91

**Figure 2.** Definition of the threshold values for TGA indications (ETP and peak thrombin

concentration) at different IVF outcomes in the 1st (observational) group (at the 2nd observation point).


t – test; \* - in this table, as well as in tables 5, 8-11, the validity of differences P<0.05 between the groups characterized by different outcomes occurring within the same observation periods of the IVF cycle; \*\* - the same, P<0.01.

**Table 2.** Quantitative level and change dynamics of hemostatic and fibrinolytic indications (mean ± SD) in women of the 1st (observational) group (n=163)

The most significant adverse indication for this reproductive technology was ETP value, which range limits at different IVF outcomes in the middle of the cycle did not intersect even with the M ± 3SD limit (Fig. 2). Peack thrombin concentration also changed, but its values in the compared groups did not intersect within the M ± SD limit.

With regard to the conducted studies, the threshold value for the positive decision on administering heparin prophylaxis was set according to the following criteria: ETP exceeds 1900 nM/min and/or increased Peack thrombin concentration of over 360 nM/l.

1st observation point

in women of the 1st (observational) group (n=163)

with pregnancies.

Peak thrombin concentration,

nM/l

PAI-1, activity, un/ml

t-PA, activity,

EAAF index,

Clot lysis time, min

D-dimers,

thrombin generation in the observational group had considerably increased after the start of controlled ovarian hyperstimulation in response to a sharp increase in the level of estrogens in blood, which is consistent with recently published data (Westerlund et al., 2012). We have also found that the degree of the increase in thrombin generation is different at successful and unsuccessful IVF outcomes (Table 2). In particular, the values of such thrombin generation factors as ETP and peak thrombin concentration at the 2nd observation point were higher in case of IVF failures (P<0.001) as compared with the similar data in patients

Indication At IVF failure (n=107) At pregnancy (n=56)

3rd observation point

328.6±24.8\*\* 394.1±25.6\*\* 378.4±25.5\*\* 313.2±24.6 338.3±25.5 310.1±26.0

4.22±2.98 4.08±2.76 4.10±2.16 3.67±2.68 3.41±2.87 3.25±2.31

13.5±3.7\* 14.3±3.9\* 14.9±3.8\* 11.2±3.5 11.9±2.9 12.5±3.3

ETP, nМ/min 1655±39.2\*\* 2060.5±52.7\*\* 1853.9±55.6\*\* 1574.8±36.9 1723±54.2 1632.7±56.4

un/ml 0.36±0.16 0.37±0.18 0.38±0.18 0.37±0.14 0.36±0.16 0.35±0.17

% 8.5±3.4 9.1±4.1 9.3±4.2 10.0±4.1 10.5±4.9 10.7±4.6

ng/ml 223.7±31.7\*\* 266.4±27.8\* 321.0±31.4\*\* 198.5±25.4 251.6±26.1 298.4±28.4

t – test; \* - in this table, as well as in tables 5, 8-11, the validity of differences P<0.05 between the groups characterized by different outcomes occurring within the same observation periods of the IVF cycle; \*\* - the same, P<0.01.

**Table 2.** Quantitative level and change dynamics of hemostatic and fibrinolytic indications (mean ± SD)

The most significant adverse indication for this reproductive technology was ETP value, which range limits at different IVF outcomes in the middle of the cycle did not intersect even with the M ± 3SD limit (Fig. 2). Peack thrombin concentration also changed, but its

With regard to the conducted studies, the threshold value for the positive decision on administering heparin prophylaxis was set according to the following criteria: ETP exceeds

values in the compared groups did not intersect within the M ± SD limit.

1900 nM/min and/or increased Peack thrombin concentration of over 360 nM/l.

1st observation point

2nd observation point

3rd observation point

2nd observation point

**Figure 2.** Definition of the threshold values for TGA indications (ETP and peak thrombin concentration) at different IVF outcomes in the 1st (observational) group (at the 2nd observation point).

The level of D-dimers, a known marker for fibrin generation and fibrinolysis, showed less distinctive differences in subgroups 1.1 and 1.2, even though it was slightly decreased, according to the mean data, in case of IVF failures (Table 2). Mean values of this indication in virtually healthy women of fertile age registered in our Center were equal to 205.3 ng/ml, with М ± 2SD 148.5 - 262.1 ng/ml. Respectively, registered results of the D-dimer level at the 2nd observation point were within the allowed value limits or slightly exceeded them.

The Means of Progress in Improving the Results of *in vitro* Fertilization Based on the Identification and Correction of the Pathology of Hemostasis 93

Impregnation

% of pregnant women

Аbs. (n=56)

49 30.1 47 83.9

46 28.2 3 5.4

35 21.5 4 7.1

33 20.2 2 3.6

% of the whole study populatio n

nadroparin calcium (Sanofi-Aventis): 0.3 ml twice a day for 12-14 days. Decision to begin the therapy was based on marking a suprathreshold increase in the major indications of thrombin generation - ETP (over 1900 nM/min) and/or Peack thrombin concentration (over

(n=163)

fibrinolytic pathologies 114 69.9 9 16.1

**Table 3.** IVF results in women of the 1st (observational) group depending on the presence or absence of

Impact on the vessel wall to increase fibrinolytic activity was made by means of IPC. In the publication byKakkos et al. (2005) a comparative description of the two widely used compression machines - SCD Express™ Compression System (Tyco Healthcare Group LP, Mansfield, MA, USA) was introduced with a rapid inflation device that delivers uniform compression and VenaFlow® (Aircast Inc, Summit, NJ, USA). However neither of them is normally equipped with proper braces to provide mechanical invasion for arm vessels. Still, we found it important to compress this vessel area to exclude even the hypothetic possibilty

360 nM/l) at the 2nd observation point.

1.1. Without target indicators of hemostatic and fibrinolytic

1.2.1. With ETP exceeding 1900 nM/min and/or increased Peack thrombin concentration of over 360 nM/l (at the 2nd

1.2. With hemostatic and

observational point)

than 11% (at the 1st observational point)

1.2.2. With decreased EAAF index of less than 11% and prolonged clot lysis time of over 12 min. (at the 1st observational point)

1.2.3. With ETP exceeding 1900 nM/min and/or increased Peack thrombin concentration of over 360 nM/l (at the 2nd observational point) and decreased EAAF index of less

hemostatic and fibrinolytic pathologies

pathologies

Patient subgroups Аbs.

It was more difficult to define the threshold values that reflect the decrease in the activity of fibrinolytic responses and allow specialists to select patients in need of hypofibrinolysis correction. Recent important publications devoted to this field demonstrate that hypofibrinolysis may be typical of some women who participate in the IVF procedure, but this pathology is original and not triggered in the course of controlled hormonal stimulation as a part of the IVF program (Martinez-Zamora et al., 2011; Westerlund et al. 2012). We received similar data proving that the suppression of fibrinolytic responses was actually typical of a number of women before the beginning of the IVF cycle. Suppression was steady throughout the cycle. Dynamics analysis of the changes in t-PA and PAI-1 activities, their EAAF index, and clot lysis time revealed more dramatic shifts at IVF failures, though the difference between the mean values of the parameters under study turned out to be invalid (P<0.05) (Table 2). Nevertheless, we have recorded two facts. First, mean values of the EAAF index defined in the group of 10 virtually healthy female volunteers (20-23 years) were equal (М±SD) to 11.0±3.3%. Second, in case of IVF failures we registered decreased EAAF indexes (less than 11%) in 93.5% (43 out of 46) of women in the 1st (observational) group before the start of the IVF program (1st observational point) as compared to 4,1% (2 out of 49) of women after successful impregnation (P<0.000001) (Table 3).

We used the values of the applied CLT assay to be the method of general fibrinolysis monitoring and refused to consider it as a potential criterion for the selection of women in need of therapeutic invasion in relation to its laboratory standardization. Thus, to select females to undergo IPC procedure we chose EAAF index calculation rate, which records the correlation between t-PA and PAI-1 activities, with the value of less than 11%.

Back to the data presented above and obtained during the study of the hemostasis and fibrinolysis systems in the patients of the 1st or observational group (n=163), one can see significant correlation between the detected pathologies and certain IVF outcomes (Table 3). Adverse shifts in these systems had records in 114 out of 163 women (70%) in the 1st group. In general, the IVF cycle efficacy at this stage of the study was equal to 34.4%, however, certain hemostatic and fibrinolytic pathologies facilitated the decrease in the number of successful impregnations from 95.9% (47 out of 49 women in subgroup 1.1) to 7.9% (9 out of 114 women in subgroup 1.2), i.e. in 12.1 times (P<0,000001).
