**3.1.1 Ovarian cancer**

Ovarian cancer is the most maligne tumor among all malignant tumors of the reproductive system, with the pathogenic factors unclear yet but possibly relating to the reproductive and hereditary factors. Most ovarian cancer cases are detected in the late stage and seldom cured. For the moment, the universally-ratified specificity biomarker CA125 for ovarian cancer diagnosis, as the single biomarker, sees the positive predicted value less than 10%. Therefore it is urgent to develop an approach for the early-stage clinical diagnosis so as to enhance the survival rate. The serum based peptidomics has seen great progress in researches of ovarian cancer diagnosis.

Pleasantly, based on findings of the SELDI system, the US Vermilion Company and Quest Diagnostics worked together to develop the OVA1, which can determine the onset of ovarian cancer by detecting the cavum pelvis enclosed mass and decide whether operations

Serum Peptidomics 379

the molecular mass of 11700 Da, which was obviously higher than the density peak value of the control group. Then, the affinity chromatography method was used for purification of it and the protein sequence was then determined with the fluid chromatography and mass spectrometer. A polypeptide chain was synthesized. Finally, it was identified that this polypeptide fragment was haptoglobin chain and corresponding antibody was derived. Combined with CA125, this method saw the ovarian tumor diagnosis sensitivity of 95.7%, specificity 82.6%, enhanced the early diagnosis rate of ovarian cancers and lowered false positive of CA125. Currently, relevant technologies started being applied in screening

In addition, theses published on Lancet indicated the MS technology was used to analyze the composition pattern of serum peptidome and based on the pattern, found out differential points for detection of ovarian cancer clinically. Results showed that 50 ovarian cancer patients were all detected, with 18 patients suffering stage-I ovarian cancer; 63 of 66 gynecologic benign tumors were diagnosed correctly. In detection of ovarian cancer, the sensitivity, specificity and positive predicted value of this method were 100%, 95% and 94% respectively, much better than the conventional CA125 detection. In particular, it was successful in diagnosing the early-stage ovarian cancer (stage I), indicating that this technology may be used for the early-stage or early warning detection of ovarian cancer hopefully. Thereafter, a series of theses on the serum peptide profiling technology for disease diagnosis were published in the world. For instance, in 2004, Soltys et al. with the Stanford University published results of using serum peptidome profiling to diagnose squamous carcinoma on Clinical Cancer Research; in 2005, Kaz'ufumiHonda et al. with the Japanese National Cancer Center published thesis on utilizing low-resolution MS and highresolution MS to diagnose pancreatic cancer on Cancer Research. At the beginning of 2006, Villanueva et al. with the US Sloan-Kettering Cancer Center published an article on using serum peptide profiling technology to diagnose bladder cancer, breast cancer and prostatic cancer; at the end of 2006, Nature published remarks on this article, signaling its support for this technical development and application. Lately, Lancet published an article on using the serum peptide profiling technology to diagnose tuberculosis, which produced sensitivity

We employed the ClinTOF system in the research of breast cancer, with the experimental design as shown in Fig.8. The serum polypeptides were extracted with magnetic beads and then the polypeptide profiling was obtained with MS detection, to build a disease model. Significant difference was found with the software analysis and the reliability of the model was validated with a certain number of blind samples. At the same time, the reliability of the difference was confirmed. Multiple tandem mass spectrometry were used to identify differential polypeptides. FlexAnalysss2.4, BioExplorer, SIMCAP+ and ClinProTools2.0 were used to analyze results, as shown in Fig.9, finding out 4 differential peaks;

Most of the polypeptide sequences identified lacked of only one or several amino acids. After comparing them with results and discussions in literatures, it was concluded that under different disease response conditions, the activity of external peptidase would undergo changes, thereby to obtain a series of polypeptides by cutting different protein loca.

identification and functional forecast were performed on differential polypeptides.

ovarian cancers.

and specificity greater than 950k.

These polypeptides might share the same modif.

**3.1.2 Breast cancer** 

are needed, what operations are needed and who shall perform the operations if applicable. OVA1 ovarian cancer qualitative serum test forms a single digital grading system by combining the five immunoassay combinations. The test has shown that female older than 18, requiring operations of ovarian enclosed mass and not examined by the oncologist can undergo examination with OVA1. This immune test has determined five tried and true biomarkers, namely, thyroxin, apolipoprotein A-1, β2-microglobulin, transferin and cancer antigen 125. One algorithm is decided to determine the probability of tumor onset of female cavum pelvis enclosed mass. Quest Diagnostics exclusively provides OVA1 to the US clinical lab for three years. Food and Drug Administration (FDA) has approved OVA1 to be used for the high-sensitivity testing of ovarian cancer, with effect better than biopsy or operation examination, even if the radiation test results cannot show whether malignant tumors exist. Vermilion is devoted to discovering, developing and commercializing newstyle high-value diagnostic tests, to help doctors diagnose, cure and perfect prognoses of the patient. Vermilion provides approaches for diagnosis of tumor, blood and heart disease and for ensuring women health (OVA1, Fremont, CA: Vermillion, Inc; 2009).

Petricoin et al., with the fund from the clinical proteome plan of the US FDA/NIH, has successfully applied the serum polypeptide profiling technology to make diagnosis of earlystage ovarian cancer. The ovarian cancer serum proteomics findings with the hydrophobic chips were reported: they discover that the content of five types of proteins in the serum of ovarian cancer patients and the healthy people is changing at the same time, which is of far reaching importance to the diagnosis of ovarian cancer. This proteome worked as the specificity biomarker for diagnosing ovarian cancer and double-blind researches were done to the serum of the healthy people, ovarian cancer patients and ovarian benign pathological changes, with results showing that the sensitivity of this method is 100%, specificity 95% and positive predicted value 94%; by contrast, the positive predicted value of CA125 was only 35% (Petricoin et al, 2002).

Katherine et al. utilized the protein fingerprint technology to load 184 serum samples on the strong anion exchange (SAX) chips. The 184 cases included 109 ovarian cancer cases, 19 benign ovarian tumor cases and 56 healthy people. The univariate and multivariate statistics method has been applied for analysis. From the protein fragments from the 140 serum samples, 3 groups of protein markers of diagnosis significance were obtained: the first group included 5 candidate protein markers, with the ovarian cancer sensitivity of 95.7%, specificity of 82.6% and accuracy of 89.2%; the other two groups of protein markers included 5 and 4 candidate protein markers respectively, with the sensitivity of 81.7% and 72.8%, specificity 94.9% and of 94.9% and accuracy of 88.2% and 83.9% respectively. After screening out 3 groups of protein markers, the rest 44 unknown serum samples were used for blind validation, showing that when the three groups were employed together, 41 serum cases were diagnosed correctly, to be detailed, 21 ovarian cancers were correctly diagnosed, 11 ovarian cancers of the progressive stage were correctly diagnosed and 10 of the 11 ovarian cancers of the early stage were confirmed; 6 low-malignancy potential tumors were diagnosed, 5 of the 6 benign tumors were excluded the possibility of ovarian cancer, 1 of the 10 serum cases of healthy people had a wrong diagnosis. In this sense, Katherine et al. believed that they had found ovarian cancer protein marker groups of diagnosis significance and can effectively distinguish healthy people from the benign/malignant ovarian tumors.

Ye et al. utilized the protein chip technology to analyze the serum of 80 ovarian cancer patients and 91 healthy people, discovering the differential protein peak at the position with

are needed, what operations are needed and who shall perform the operations if applicable. OVA1 ovarian cancer qualitative serum test forms a single digital grading system by combining the five immunoassay combinations. The test has shown that female older than 18, requiring operations of ovarian enclosed mass and not examined by the oncologist can undergo examination with OVA1. This immune test has determined five tried and true biomarkers, namely, thyroxin, apolipoprotein A-1, β2-microglobulin, transferin and cancer antigen 125. One algorithm is decided to determine the probability of tumor onset of female cavum pelvis enclosed mass. Quest Diagnostics exclusively provides OVA1 to the US clinical lab for three years. Food and Drug Administration (FDA) has approved OVA1 to be used for the high-sensitivity testing of ovarian cancer, with effect better than biopsy or operation examination, even if the radiation test results cannot show whether malignant tumors exist. Vermilion is devoted to discovering, developing and commercializing newstyle high-value diagnostic tests, to help doctors diagnose, cure and perfect prognoses of the patient. Vermilion provides approaches for diagnosis of tumor, blood and heart disease and

Petricoin et al., with the fund from the clinical proteome plan of the US FDA/NIH, has successfully applied the serum polypeptide profiling technology to make diagnosis of earlystage ovarian cancer. The ovarian cancer serum proteomics findings with the hydrophobic chips were reported: they discover that the content of five types of proteins in the serum of ovarian cancer patients and the healthy people is changing at the same time, which is of far reaching importance to the diagnosis of ovarian cancer. This proteome worked as the specificity biomarker for diagnosing ovarian cancer and double-blind researches were done to the serum of the healthy people, ovarian cancer patients and ovarian benign pathological changes, with results showing that the sensitivity of this method is 100%, specificity 95% and positive predicted value 94%; by contrast, the positive predicted value of CA125 was

Katherine et al. utilized the protein fingerprint technology to load 184 serum samples on the strong anion exchange (SAX) chips. The 184 cases included 109 ovarian cancer cases, 19 benign ovarian tumor cases and 56 healthy people. The univariate and multivariate statistics method has been applied for analysis. From the protein fragments from the 140 serum samples, 3 groups of protein markers of diagnosis significance were obtained: the first group included 5 candidate protein markers, with the ovarian cancer sensitivity of 95.7%, specificity of 82.6% and accuracy of 89.2%; the other two groups of protein markers included 5 and 4 candidate protein markers respectively, with the sensitivity of 81.7% and 72.8%, specificity 94.9% and of 94.9% and accuracy of 88.2% and 83.9% respectively. After screening out 3 groups of protein markers, the rest 44 unknown serum samples were used for blind validation, showing that when the three groups were employed together, 41 serum cases were diagnosed correctly, to be detailed, 21 ovarian cancers were correctly diagnosed, 11 ovarian cancers of the progressive stage were correctly diagnosed and 10 of the 11 ovarian cancers of the early stage were confirmed; 6 low-malignancy potential tumors were diagnosed, 5 of the 6 benign tumors were excluded the possibility of ovarian cancer, 1 of the 10 serum cases of healthy people had a wrong diagnosis. In this sense, Katherine et al. believed that they had found ovarian cancer protein marker groups of diagnosis significance and can effectively distinguish healthy people from the benign/malignant ovarian tumors. Ye et al. utilized the protein chip technology to analyze the serum of 80 ovarian cancer patients and 91 healthy people, discovering the differential protein peak at the position with

for ensuring women health (OVA1, Fremont, CA: Vermillion, Inc; 2009).

only 35% (Petricoin et al, 2002).

the molecular mass of 11700 Da, which was obviously higher than the density peak value of the control group. Then, the affinity chromatography method was used for purification of it and the protein sequence was then determined with the fluid chromatography and mass spectrometer. A polypeptide chain was synthesized. Finally, it was identified that this polypeptide fragment was haptoglobin chain and corresponding antibody was derived. Combined with CA125, this method saw the ovarian tumor diagnosis sensitivity of 95.7%, specificity 82.6%, enhanced the early diagnosis rate of ovarian cancers and lowered false positive of CA125. Currently, relevant technologies started being applied in screening ovarian cancers.

In addition, theses published on Lancet indicated the MS technology was used to analyze the composition pattern of serum peptidome and based on the pattern, found out differential points for detection of ovarian cancer clinically. Results showed that 50 ovarian cancer patients were all detected, with 18 patients suffering stage-I ovarian cancer; 63 of 66 gynecologic benign tumors were diagnosed correctly. In detection of ovarian cancer, the sensitivity, specificity and positive predicted value of this method were 100%, 95% and 94% respectively, much better than the conventional CA125 detection. In particular, it was successful in diagnosing the early-stage ovarian cancer (stage I), indicating that this technology may be used for the early-stage or early warning detection of ovarian cancer hopefully. Thereafter, a series of theses on the serum peptide profiling technology for disease diagnosis were published in the world. For instance, in 2004, Soltys et al. with the Stanford University published results of using serum peptidome profiling to diagnose squamous carcinoma on Clinical Cancer Research; in 2005, Kaz'ufumiHonda et al. with the Japanese National Cancer Center published thesis on utilizing low-resolution MS and highresolution MS to diagnose pancreatic cancer on Cancer Research. At the beginning of 2006, Villanueva et al. with the US Sloan-Kettering Cancer Center published an article on using serum peptide profiling technology to diagnose bladder cancer, breast cancer and prostatic cancer; at the end of 2006, Nature published remarks on this article, signaling its support for this technical development and application. Lately, Lancet published an article on using the serum peptide profiling technology to diagnose tuberculosis, which produced sensitivity and specificity greater than 950k.

### **3.1.2 Breast cancer**

We employed the ClinTOF system in the research of breast cancer, with the experimental design as shown in Fig.8. The serum polypeptides were extracted with magnetic beads and then the polypeptide profiling was obtained with MS detection, to build a disease model. Significant difference was found with the software analysis and the reliability of the model was validated with a certain number of blind samples. At the same time, the reliability of the difference was confirmed. Multiple tandem mass spectrometry were used to identify differential polypeptides. FlexAnalysss2.4, BioExplorer, SIMCAP+ and ClinProTools2.0 were used to analyze results, as shown in Fig.9, finding out 4 differential peaks; identification and functional forecast were performed on differential polypeptides.

Most of the polypeptide sequences identified lacked of only one or several amino acids. After comparing them with results and discussions in literatures, it was concluded that under different disease response conditions, the activity of external peptidase would undergo changes, thereby to obtain a series of polypeptides by cutting different protein loca. These polypeptides might share the same modif.

Serum Peptidomics 381

Some researchers have also shown that protein MS peaks obtained by the SELDI system can effectively screen and distinguish the breast cancer and non-breast cancer patients. Van Winden et al. adopted the SELDI technology to determine the density difference between the breast cancer group and the control group, making new progress compared with the previous breast cancer marker protein reseraches (van Winden et al, 2009). Gast et al. utilized the SELDI technology to detect the protein profiling in tissues and serum to diagnose breast cancer, showing that 3 peaks are significantly correlated to breast cancer. 27 tissues were detected to have differential peaks. These protein fragments presented potential pathological and physiological mechanisms related to breast cancer, which is helpful for raising the diagnosis rate of the breast cancer (Gast et al., 2009). Taku et al employed the SELDIS technology to detect 65 breast cancer patients as a group, concluding that the overexpression of a non-identified protein and the underexpression of the other are

In the early stage, if nasopharyngeal carcinoma fails to show some disease features due to disturbance of some factors, the test method is so insensitive that the patient cannot get timely and correct diagnosis, it will delay the treatment and cause unfavorable prognosis. In the research on nasopharyngeal carcinoma serum peptidome, some progress has been

We employed the ClinTOF system to research the marker of early-stage nasopharyngeal carcinoma. The magnetic bead system was SPE-C. By comparing 40 nasopharyngeal carcinoma cases and 61 healthy people, 65 significantly differential peaks were observed in the mass range of 1,000Da-10,000Da. The peak with the most obvious difference of expression has the mass-to-charge ratio of 1262.67 and is identified with the LTQ -Orbitrap to be the fragment of fibrinopeptide A. Fibrinogen is a kind of glycoprotein rich in blood and a symmetrical dimer composed of 6 polypeptide chains; it plays a major role in blood coagulation and hemostasis. So far, more than 300 fibrinogen natural mutants have been discovered, among which, about 55% are subclinical, 25% have hemorrhagic tendency and 20% have thrombophilia. Among these mutants, the most common expression is that one amino acid is replaced by another. At the molecular biology level, expressions are point mutation, deletion, insertion and nonsense mutation. Moreover, the polymorphism of the fibrinogen results in the overexpression, which raises the fibrinogen level of plasma and

Researchers have also indicated that this system has a higher repeatability than the IMAC Cu2+ chips (Ciphergen Company) and is capable of finding more differential peaks in

Wei et al., using the SELDI system, screened out the mass-to-charge ratios of four proteins, to be detailed, 4 097Da, 4 180 Da, 5 912Da and 8 295Da, which make a marker combination to build a classification tree model for diagnosis of nasopharyngeal carcinoma. In diagnosing nasopharyngeal carcinoma, the sensitivity and specificity of this model were 94.5 % and 96.7% respectively and for the blind screening group, the two figures were 92% and 92.9% respectively. Using the protein MS peaks with the mass-to-charge ratio of 4 581Da and 7802Da to predict the stage I and stage II nasopharyngeal carcinoma, the accuracy proved to be 80%

proteins with the molecular weight ranging 2,000Da -5,000Da.

and for the stage III and stage IV, this figure was 86% (Wei et al, 2008).

related to the lymphatic metastasis of breast cancer (Taku et al., 2006).

**3.1.3 Nasopharyngeal carcinoma** 

closely relates to diseases.

made.

Fig. 8. Experimental Design of ClinTOF System for Breast Cancer Research

A: sample distribution diagram; B: average profiling of significantly differential peaks; C: distribution of samples obtained by SIMCAP+; D: results of MS identificdatin of differential polypeptides; E: description of mechanism according to distribution of identified polypeptides

Fig. 9. Findings of Using the ClinTOF System to Breast Cancer Research

Fig. 8. Experimental Design of ClinTOF System for Breast Cancer Research

Breast Cancer group Control group

A: sample distribution diagram; B: average profiling of significantly differential peaks; C: distribution of

samples obtained by SIMCAP+; D: results of MS identificdatin of differential polypeptides;

E: description of mechanism according to distribution of identified polypeptides Fig. 9. Findings of Using the ClinTOF System to Breast Cancer Research Some researchers have also shown that protein MS peaks obtained by the SELDI system can effectively screen and distinguish the breast cancer and non-breast cancer patients. Van Winden et al. adopted the SELDI technology to determine the density difference between the breast cancer group and the control group, making new progress compared with the previous breast cancer marker protein reseraches (van Winden et al, 2009). Gast et al. utilized the SELDI technology to detect the protein profiling in tissues and serum to diagnose breast cancer, showing that 3 peaks are significantly correlated to breast cancer. 27 tissues were detected to have differential peaks. These protein fragments presented potential pathological and physiological mechanisms related to breast cancer, which is helpful for raising the diagnosis rate of the breast cancer (Gast et al., 2009). Taku et al employed the SELDIS technology to detect 65 breast cancer patients as a group, concluding that the overexpression of a non-identified protein and the underexpression of the other are related to the lymphatic metastasis of breast cancer (Taku et al., 2006).

### **3.1.3 Nasopharyngeal carcinoma**

In the early stage, if nasopharyngeal carcinoma fails to show some disease features due to disturbance of some factors, the test method is so insensitive that the patient cannot get timely and correct diagnosis, it will delay the treatment and cause unfavorable prognosis. In the research on nasopharyngeal carcinoma serum peptidome, some progress has been made.

We employed the ClinTOF system to research the marker of early-stage nasopharyngeal carcinoma. The magnetic bead system was SPE-C. By comparing 40 nasopharyngeal carcinoma cases and 61 healthy people, 65 significantly differential peaks were observed in the mass range of 1,000Da-10,000Da. The peak with the most obvious difference of expression has the mass-to-charge ratio of 1262.67 and is identified with the LTQ -Orbitrap to be the fragment of fibrinopeptide A. Fibrinogen is a kind of glycoprotein rich in blood and a symmetrical dimer composed of 6 polypeptide chains; it plays a major role in blood coagulation and hemostasis. So far, more than 300 fibrinogen natural mutants have been discovered, among which, about 55% are subclinical, 25% have hemorrhagic tendency and 20% have thrombophilia. Among these mutants, the most common expression is that one amino acid is replaced by another. At the molecular biology level, expressions are point mutation, deletion, insertion and nonsense mutation. Moreover, the polymorphism of the fibrinogen results in the overexpression, which raises the fibrinogen level of plasma and closely relates to diseases.

Researchers have also indicated that this system has a higher repeatability than the IMAC Cu2+ chips (Ciphergen Company) and is capable of finding more differential peaks in proteins with the molecular weight ranging 2,000Da -5,000Da.

Wei et al., using the SELDI system, screened out the mass-to-charge ratios of four proteins, to be detailed, 4 097Da, 4 180 Da, 5 912Da and 8 295Da, which make a marker combination to build a classification tree model for diagnosis of nasopharyngeal carcinoma. In diagnosing nasopharyngeal carcinoma, the sensitivity and specificity of this model were 94.5 % and 96.7% respectively and for the blind screening group, the two figures were 92% and 92.9% respectively. Using the protein MS peaks with the mass-to-charge ratio of 4 581Da and 7802Da to predict the stage I and stage II nasopharyngeal carcinoma, the accuracy proved to be 80% and for the stage III and stage IV, this figure was 86% (Wei et al, 2008).

2008).

**3.1.4 Cervical carcinoma** 

**3.1.5 Leukemia** 

Serum Peptidomics 383

were performed over serum of the nasopharyngeal carcinoma metastasis group, nasopharyngeal carcinoma non-metastasis group and the control group. Then, analysis was made to the 3 groups of serum profiling, obtained 29 differential protein spots and identified 23 proteins. Through comparison of the cancer profiling and the profiling of the control group, it is discovered that the transferrin, zinc finger protein 544, thyroid hormone binding protein, NM 23H 1 protein and FAD synthetase showed underexpression in nasopharyngeal carcinoma patients. Yet lipoxygenase (LOX), serum amyloid protein A1, cytochrome P450, sICAM1, cathepsin G and histone lysine specific demethylase 1 showed overexpression in the two groups of nasopharyngeal carcinoma patients; the LOX, sICAM1, cathepsin G and histone lysine specific demethylase 1 showed higher expression in the nasopharyngeal carcinoma metastasis group than in the non-metastasis group, while the heat shock protein 70 was only expressed in the nasopharyngeal carcinoma metastasis group. In addition, by combining the immune histochemical technique and ELISA method, it is concluded that HSP70, sICAM1 and serum amyloid protein A (SAA) were potential serological markers for mediating nasopharyngeal carcinoma metastasis and also exerted an important role in clinical detection and control of nasopharyngeal carcinoma (Liao et al,

The serum peptidomics technology has been used in the research of endometrial carcinoma and is of guidance significance for the early diagnosis and clinical grading of diseases.

Our researches using the ClinTOF system have indicated that in the cervical carcinoma group and the control group, the protein peak intensify difference of only 21 proteins had statistical significance (P<0.05). M1450.35Da, M1778.7Da, M1896.65Da and M5520.42Da were taken as the classified variables to build a classification predictive model, to make classification diagnosis over the cervical carcinoma group and the control group, with the identification rate and predictive ability of 90.45% and 81.75% respectively. By contrasting the proteome of different pathological differentiation degrees, it is discovered that the protein peak intensity difference of 2 proteins had statistical significance (P<0.05). M5904.14Da and M5264.26Da were taken as the classified variables to build a classification model, to make classification diagnosis on different differentiation degrees, with the

Yoshizaki et al. adopted the SELDI system to analyze the protein profiling of 19 endometrial carcinoma cases and 20 normal tunica intima tissues, discovering two differential proteins, namely EC1 and EC2. The former had overexpression in the endometrial carcinoma tissues and the latter on the contrary. These differential proteins may be used for diagnosing

The treatment of acute leukemia (AL) depends on minimal residual disease (MRD) diagnosis and detection. However there has been no serum biomarker that can be used by the clinicians for diagnosing AL and evaluating MRD. Yang et al. analyzed serum of AL patients, discovering two peptides (m/z of 1778 and 1865) reduced with the rise in the modification degree and that 1865 was related to AL types. After further FT-ICR-MS

identification rate and predictive ability of 81.48% and 78.89% respectively.

endometrial carcinoma hopefully (Yoshizaki et al., 2005).

Guo et al. applied the SELDI system and the artificial neural network technology to analyze serum of 58 type-A (cranial nerve type: featured by basicranial destruction and cranial nerve violation, no lymphonodi cervicales metastasis) and type-D (lymphonodi cervicales metastasis type: widespread metastasis of one- or two-sided lymphonodi cervicales, no cranial nerve violation or basicranial bone destruction) nasopharyngeal carcinoma patients and obtained 11 potential biomarkers, with the mass-to-charge ratio of 4 053, 5 885, 4 072, 5 798, 4 209, 8 689, 2 382, 9 357, 2 221, 4 230Da and 5 901Da respectively. The MS peaks of these proteins saw accuracy for distinguishing type A and type D reach 90% (Guo et al, 2005).

Cho et al. employed IMAC3-Cu protein chips to detect the serum of the healthy people, modified nasopharyngeal carcinoma patients after treatment and the relapse nasopharyngeal carcinoma patients before and after chemotherapy, discovering 13 meaningful protein MS peaks. Protein MS peaks with the mass-to-charge ratio of 2950Da and 6 701Da were selected to build a classification tree model for predicting chemotherapy response, with the sensitivity and specificity of 80% and 87% respectively. In detection and analysis of serum protein profiling of the healthy people and the relapse nasopharyngeal carcinoma patients, it is newly discovered that the MS peaks with the mass-to-charge ratio of 3803Da and 3953Da were found in the serum of both patients taking the initial diagnosis and treatment, and the modified nasopharyngeal carcinoma patients after treatment. In addition, the protein MS peaks with the mass-to-charge ratio of 3953Da and 7765Da were identified to be the inter-alpha trypsin inhibitor heavy chain H4 precursor (ITIH4) fragments and platelet factor 4 (PF4) with the method of MS/MS sequence and immune affinity capture test, thus presuming that they are related to the onset, development, metastasis and relapse of nasopharyngeal carcinoma (Cho et al., 2004).

Huang et al. adopted the CM 10 chip and SELDI technology to detect squamous epithelial cells of the nasopharyngeal carcinoma patients and made contrast with serum of the healthy people. 94 protein MS peaks were detected, among which, 26 were obviously different from the serum of healthy people (P<0.05, Mean>SD), 5 of overexpression and 21 underexpression. Three protein MS peaks with the mass-to-charge ratio of 3159 83, 5 187 65 and 1 3738.6Da were selected to build a diagnosis model. Verified by the blind method, this diagnosis model has accuracy of 90.63%, sensitivity 95.00%, specificity 83.33%, positive predicted value 90.48% and negative predicted value 90.90%, indicating that this model has a higher diagnosis value for NPC (Huang et al, 2008).

Doustjalali adopted the ClinProt system and the MASCOT database retrieval, discovering that in the serum of nasopharyngeal carcinoma patients, the content of CPL showed overexpression compared with the control group; further, after the ELISA method validation, it was discovered that in the nasopharyngeal carcinoma tissues, CPL showed overexpression. After 6 months' treatment, the CPL content in tissues reduced. CPL is a kind of copper-bearing glycoprotein and considered to be a key molecule for activating angiogenesis factor, which can promote growth of tumors, as shown by studies. This research monitored the changes of ceruloplasmin in company with the ease of the disease with the proteome technology, which would play a role in diagnosis, curative effect observation and prognoses evaluation, observation and monitoring of nasopharyngeal carcinoma (Doustjalali et al., 2006).

Liao et al. adopted 2-DE /MALDI-TOF-MS technology to screen tumor specific antigens of nasopharyngeal carcinoma. High-abundance protein elimination and desalting pretreatment

Guo et al. applied the SELDI system and the artificial neural network technology to analyze serum of 58 type-A (cranial nerve type: featured by basicranial destruction and cranial nerve violation, no lymphonodi cervicales metastasis) and type-D (lymphonodi cervicales metastasis type: widespread metastasis of one- or two-sided lymphonodi cervicales, no cranial nerve violation or basicranial bone destruction) nasopharyngeal carcinoma patients and obtained 11 potential biomarkers, with the mass-to-charge ratio of 4 053, 5 885, 4 072, 5 798, 4 209, 8 689, 2 382, 9 357, 2 221, 4 230Da and 5 901Da respectively. The MS peaks of these proteins saw accuracy for distinguishing type A and type D reach 90% (Guo et al, 2005).

Cho et al. employed IMAC3-Cu protein chips to detect the serum of the healthy people, modified nasopharyngeal carcinoma patients after treatment and the relapse nasopharyngeal carcinoma patients before and after chemotherapy, discovering 13 meaningful protein MS peaks. Protein MS peaks with the mass-to-charge ratio of 2950Da and 6 701Da were selected to build a classification tree model for predicting chemotherapy response, with the sensitivity and specificity of 80% and 87% respectively. In detection and analysis of serum protein profiling of the healthy people and the relapse nasopharyngeal carcinoma patients, it is newly discovered that the MS peaks with the mass-to-charge ratio of 3803Da and 3953Da were found in the serum of both patients taking the initial diagnosis and treatment, and the modified nasopharyngeal carcinoma patients after treatment. In addition, the protein MS peaks with the mass-to-charge ratio of 3953Da and 7765Da were identified to be the inter-alpha trypsin inhibitor heavy chain H4 precursor (ITIH4) fragments and platelet factor 4 (PF4) with the method of MS/MS sequence and immune affinity capture test, thus presuming that they are related to the onset, development,

Huang et al. adopted the CM 10 chip and SELDI technology to detect squamous epithelial cells of the nasopharyngeal carcinoma patients and made contrast with serum of the healthy people. 94 protein MS peaks were detected, among which, 26 were obviously different from the serum of healthy people (P<0.05, Mean>SD), 5 of overexpression and 21 underexpression. Three protein MS peaks with the mass-to-charge ratio of 3159 83, 5 187 65 and 1 3738.6Da were selected to build a diagnosis model. Verified by the blind method, this diagnosis model has accuracy of 90.63%, sensitivity 95.00%, specificity 83.33%, positive predicted value 90.48% and negative predicted value 90.90%, indicating that this model has

Doustjalali adopted the ClinProt system and the MASCOT database retrieval, discovering that in the serum of nasopharyngeal carcinoma patients, the content of CPL showed overexpression compared with the control group; further, after the ELISA method validation, it was discovered that in the nasopharyngeal carcinoma tissues, CPL showed overexpression. After 6 months' treatment, the CPL content in tissues reduced. CPL is a kind of copper-bearing glycoprotein and considered to be a key molecule for activating angiogenesis factor, which can promote growth of tumors, as shown by studies. This research monitored the changes of ceruloplasmin in company with the ease of the disease with the proteome technology, which would play a role in diagnosis, curative effect observation and prognoses evaluation, observation and monitoring of nasopharyngeal

Liao et al. adopted 2-DE /MALDI-TOF-MS technology to screen tumor specific antigens of nasopharyngeal carcinoma. High-abundance protein elimination and desalting pretreatment

metastasis and relapse of nasopharyngeal carcinoma (Cho et al., 2004).

a higher diagnosis value for NPC (Huang et al, 2008).

carcinoma (Doustjalali et al., 2006).

were performed over serum of the nasopharyngeal carcinoma metastasis group, nasopharyngeal carcinoma non-metastasis group and the control group. Then, analysis was made to the 3 groups of serum profiling, obtained 29 differential protein spots and identified 23 proteins. Through comparison of the cancer profiling and the profiling of the control group, it is discovered that the transferrin, zinc finger protein 544, thyroid hormone binding protein, NM 23H 1 protein and FAD synthetase showed underexpression in nasopharyngeal carcinoma patients. Yet lipoxygenase (LOX), serum amyloid protein A1, cytochrome P450, sICAM1, cathepsin G and histone lysine specific demethylase 1 showed overexpression in the two groups of nasopharyngeal carcinoma patients; the LOX, sICAM1, cathepsin G and histone lysine specific demethylase 1 showed higher expression in the nasopharyngeal carcinoma metastasis group than in the non-metastasis group, while the heat shock protein 70 was only expressed in the nasopharyngeal carcinoma metastasis group. In addition, by combining the immune histochemical technique and ELISA method, it is concluded that HSP70, sICAM1 and serum amyloid protein A (SAA) were potential serological markers for mediating nasopharyngeal carcinoma metastasis and also exerted an important role in clinical detection and control of nasopharyngeal carcinoma (Liao et al, 2008).
