Secondary Logo

The predictive value of C-erbB-2 and p53 overexpression in the management of molar pregnancy

Fayed, Salah T.a; Naguib, Ahmad H.a; Hemeda, Hosama; Ahmad, Naglaa S.b; Elsheikh, Abdelfattaha; Dakrony, Monaa

Egyptian Journal of Pathology: July 2012 - Volume 32 - Issue 1 - p 129–135
doi: 10.1097/01.XEJ.0000415933.54088.b3
ORIGINAL ARTICLES
Free

Objective To test the usefulness of immunostaining for the tissue markers C-erbB-2 and p53 in predicting the course of the disease in molar pregnancy in spontaneous remission or development of gestational trophoblastic neoplasia.

Methods A total of 71 cases, comprising a partial mole (34 cases), a complete mole (33 cases), an invasive mole (two cases), and choriocarcinoma (two cases), were analyzed immunohistochemically on paraffin blocks using anti-p53 and anti C-erbB-2 antibodies. The clinical data were retrieved from the follow-up sheets of these patients and correlated with the results of immunostaining.

Results The results showed that the expression of p53 and C-erbB-2 was markedly enhanced in the gestational trophoblastic diseases. There was a trend toward a higher expression of protein p53 in cases of complete moles, invasive moles, and choriocarcinoma, and a trend toward a negative expression in cases of partial moles. The two markers together had a poor positive predictive value but a good negative predictive value.

Conclusion Positive immunostaining for p53 and C-erbB-2 could not predict the requirement for chemotherapy in patients with molar pregnancy, whereas the concomitant negative staining for both markers could predict spontaneous remission.

Departments of aObstetrics & Gynecology

bPathology, Faculty of Medicine, Ain Shams University, Cairo, Egypt

Correspondence to Naglaa S. Ahmad, MD, Department of Pathology, Faculty of Medicine, Ain Shams University, Cairo, Egypt Tel: +002 022 472 0733; e-mail: naglaasamier@yahoo.com

Received March 27, 2012

Accepted April 12, 2012

Back to Top | Article Outline

Introduction

Hydatidiform moles (HMs) are gestational trophoblastic tumors that result from abnormal fertilization. They are categorized into partial and complete forms on the basis of morphologic, genetic, and clinical characteristics (Hussein, 2005; Yamauchi et al., 2007). HMs, although a rare kind of pregnancy, has received considerable attention because ∼20% of complete hydatidiform moles (CHMs) and 5% of partial hydatidiform moles (PHMs) progress to a gestational trophoblastic tumor (Greene et al., 2002). It has been observed that 8–30% of cases develop persistent trophoblastic disease requiring therapeutic interventions; hence, the molecular mechanisms behind this disorder need to be explored (Fong et al., 2005). The effectiveness of chemotherapy treatment using methotrexate, which was established by MCupon et al. (1956), allowed gestational trophoblastic tumors to be curable in almost 100% of cases (Hancock and Tidy, 2002). In clinical practice, indications of chemotherapy include increasing or plateauing (β-hCG), evidence of metastases, or a histological diagnosis of choriocarcinoma. Currently, the challenge consists in finding markers that can predict malignant transformation of the HMs before an increase in hCG serum levels occurs (Yazaki-Sun et al., 2006). It would be beneficial if the histopathological assessment of HMs could provide further prognostic information on the likelihood of developing subsequent persistent trophoblastic disease in addition to a simple distinction between CHM and PHM. Several studies have attempted to use a range of immunohistochemical markers to aid diagnosis or prognostication in HM, but none are part of the current routine practice (Sebire and Seckl, 2010)

Apoptosis is an intrinsic cellular self-destruction mechanism that is essential for a variety of biological events including the removal of unwanted cells (Jeong and Seol, 2008). The TP53 gene encodes a 53-kDa oncosuppressive protein critical for the maintenance of genomic stability. TP53 can induce cell-cycle arrest in G1, apoptosis, or affect DNA replication in response to DNA damage. Alterations in the TP53 gene are among the most common genetic changes in human cancers (Shaarawy and Sheiba, 2004). Some reports suggest that p53 expression in HMs that progressed to gestational trophoblastic disease (GTD) was significantly higher than that in HMs that achieved spontaneous remission (Xuan et al., 2008). In contrast, an earlier study did not detect the p53 gene mutation in GTD (Halperin et al., 2000).

The C-erbB-2 oncogene produces a protein with a structural characteristic similar to epidermal growth factor receptor (EGFR), and its overexpression has been related to several types of cancers (Kay et al., 1994). In a relatively recent study, C-erbB-2 expression was significantly higher in CMs that progressed to GTD when compared with those with spontaneous remission (Yang et al., 2002). HMs with C-erbB-2 overexpression in association with DNA hyperploidy also showed a more aggressive behavior (Jelincic et al., 2003). When compared with normal placentas and partial moles (PMs), complete moles (CMs) also showed a higher C-erbB-2 expression (Fulop et al., 1998). Therefore, it seems that C-erbB-2 and p53 genes may be involved in the pathogenesis of trophoblastic disease (Yazaki-Sun et al., 2006)

The aim of the present study is to compare the protein expression of p53 and C-erbB-2 among HMs in patients who progressed to GTD with those who attained spontaneous remission in an attempt to predict the course of the disease.

Back to Top | Article Outline

Materials and methods

The paraffin blocks of 71 patients with a diagnosis of either molar or postmolar disease were retrieved from the archives of Early Cancer Detection Unit, Ain Shams University Maternity Hospital. The clinical data of patients were retrieved from their files in the Gynecologic Oncology Unit of the same hospital. Relevant data included obstetric history, serial hCG titer, course of the disease, the need for chemotherapy and its regimens, and the number of courses required to induce remission in a patient. Four-micrometer histological sections were prepared and subjected to routine hematoxylin and eosin staining and immunostaining using p53 and C-erbB-2.

Back to Top | Article Outline

Immunohistochemistry

The sections were deparaffinized and rehydrated in graded alcohol and distilled water. The slides were rinsed with PBS and then incubated for 20 min at room temperature in 0.3% H2O2 in methanol. Nonspecific antibody binding was blocked with PBS and 2% goat serum for 20 min. Subsequently, they were incubated with the primary antibody [p53, a monoclonal rabbit antibody (Labvision; catalog no. #MS-1290-R7, Fremont, California, USA)], [C-erbB-2, a monoclonal mouse antibody (Labvision; catalog no. #MS-730-R7, Fremont, California, USA)] for 2 h at room temperature, followed by rinsing in PBS. The slides were incubated with biotinylated secondary antibody (Biogenex; catalog no. AD 000-SL, Fremont, California, USA) for 30 min at room temperature. They were rinsed with PBS and incubated with peroxidase-labeled streptavidin for 20 min at room temperature. They were again rinsed with PBS, and developed with a chromogenic substance, diaminobenzidine. They were rinsed with distilled water. Counterstaining with hematoxylin was performed.

Appropriate positive and negative controls were included in each run. For p53 and C-erbB-2-stained slides, all sections were examined by light microscopy to assess the positivity of immunostaining semiquantitatively.

For p53, only nuclear staining was considered positive. Here, the percentages of immunoreactive nuclei were graded as follows: less than 10%; (+1) 10–25%; (+2) 25–50%; and (+3) more than 50% positive nuclei (Rath et al., 2011).

C-erbB-2 protein was located in the membrane or the cytoplasm with brown staining. C-erbB-2 expression was graded as follows:

  • (0) Negative, absence of positive or rare cells presenting immunoexpression.
  • (+) Weak, immunoexpression in less than 50% of the sample, usually with an incomplete membranous pattern and a weak to moderate intensity.
  • (++) Strong, immunoexpression in more than 50% of the sample, usually with a complete membranous pattern and a strong intensity (Yazaki-Sun et al., 2006).

Data obtained from the study groups were compared for C-erbB-2 and p53 expressions using Fisher’s exact test (two-sided) and χ2 analysis. The data were analyzed using the statistical package for the social sciences version 10.0 (SPSS Inc., Chicago, Illinois, USA). The minimum significance level was set at P value less than 0.05.

Back to Top | Article Outline

Results

A total of 71 women were included in the study. Their mean age was 27.18±7.34 years (range: 16–49 years). The median parity was 1 (range: 0–9; interquartile range: 0–3). The median number of previous abortions was 1 (range: 0–4; interquartile range: 0–1). Only three women (4.23%) had a history of previous gestational trophoblastic disease. Of the 71 women included, 33 (46.5%) had CMs, 34 (47.9%) had PMs, two (2.8%) had an invasive mole, and two (2.8%) had choriocarcinoma.

Of the 71 women included, 43 (60.6%) achieved spontaneous remission of the disease, whereas 28 (39.4%) received chemotherapy; of these, 20 (28.2%) received one regimen (single agent) chemotherapy (first line), six (8.4%) received two regimens of chemotherapy (second line), and two (2.8%) required three regimens of chemotherapy (third line). Figure 1 shows the regimens of chemotherapy in the treated patients. There were no significant differences in terms of age, parity, number of previous abortions, or a positive history of previous gestational trophoblastic disease between women who achieved spontaneous remission and those who received chemotherapy. Sixteen of 33 (57.1%) women with a histological diagnosis of a complete mole and eight out of 34 (28.6%) patients who had a partial mole required chemotherapy. All patients with a histological diagnosis of an invasive mole or choriocarcinoma received chemotherapy.

Fig. 1

Fig. 1

Only four patients (5.6%) underwent hysterectomy as a part of the treatment, either primary (two) or secondary (two) after initial resistance to chemotherapy. The median duration until remission in women not requiring chemotherapy was 6.4 weeks (range: 2–12.29 weeks; interquartile range: 4.6–7.9 weeks). The mean duration of chemotherapy in women who received chemotherapy was 13.9 weeks (range: 8–51.71 weeks; interquartile range: 11.8–20.2 weeks). The median number of total doses of chemotherapy was 9 doses (range: 5–24 doses; interquartile range: 6–13 doses).

Immunohistochemical staining of the tissue specimens showed a negative expression for the tumor suppressor protein p53 in 14 (19.7%) women, a weak expression in 22 (31%) women, a moderate expression in 19 (26.8%) women, and a strong expression in 16 (22.5%) women. However, immunohistochemical staining of the same specimens showed a negative expression for the C-erbB-2 oncogene in 19 (26.8%) women, a weak expression in 10 (14.1%) women, and a strong expression in 42 (59.1%) women.

p53 expression in CMs was significantly stronger than in PMs (Figs 1–4) [31/33 (93.3%) vs. 23/34 (67.6%), respectively] and significantly weaker than in invasive mole and choriocarcinoma (Fig. 5) [31/33 (93.3%) vs. 2/2 and 2/2, respectively]. There was a trend toward a higher expression for p53 protein in cases of CMs, invasive moles, and choriocarcinoma, and a trend toward a negative expression in cases of PMs.

Fig. 2

Fig. 2

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Fig. 5

Fig. 5

C-erbB-2 expression in CMs was stronger than in PMs (Figs 6–9) [25/33 (75.75%) vs. 23/34 (67.6%) respectively] and weaker than in invasive moles and choriocarcinoma (Fig. 10) [25/33(75.75%) vs. 2/2 and 2/2, respectively]. C-erbB-2 expression, however, showed no significant difference in distribution among the different histological types of the gestational trophoblastic disease (Table 1). Of note, the two cases of choriocarcinoma among the women included showed the highest expression for both tumor suppressor p53 and the C-erbB-2 oncogene (3+ and 2+, respectively).

Fig. 6

Fig. 6

Fig. 7

Fig. 7

Fig. 8

Fig. 8

Fig. 9

Fig. 9

Fig. 10

Fig. 10

Table 1

Table 1

Both tumor suppressor protein p53 and the C-erbB-2 oncogene were expressed in 42 (59.2%) specimens; both were not expressed in four (5.6%) specimens. Tumor suppressor protein p53 was expressed exclusively in 15 (21.1%) specimens, whereas C-erbB-2 was expressed exclusively in 10 (14.1%) specimens. There was, therefore, poor and insignificant agreement between both markers (κ=0.02, P>0.05) (Table 2). Of note, all of the four cases who showed a negative expression for both markers achieved spontaneous remission with no need for chemotherapy. However, 19 (45.2%) of the 42 patients who showed positive expression for both markers received chemotherapy.

Table 2

Table 2

There was no significant association between the expression of the p53, the C-erbB-2 oncogene, and their combinations, and either the need for chemotherapy, need for more than one regimen of chemotherapy, need for combination chemotherapy, and the total dose of chemotherapy or duration until spontaneous remission (Table 3). The strength of expression of either marker showed no significant correlation with the duration of spontaneous remission or the duration and the total dose of chemotherapy.

Table 3

Table 3

As a predictor of the need for chemotherapy versus spontaneous remission, both tumor suppressor protein p53 and the C-erbB-2 oncogene had comparable validity, showing a relatively high sensitivity and negative predictive value (NPV) and poor specificity and positive predictive value (PPV), making both markers good negative, but poor positive tests. The combination of both markers increased the sensitivity and NPV to 100% (thus, completely excluding the need for chemotherapy in the absence of expression of both markers) (Table 4).

Table 4

Table 4

Back to Top | Article Outline

Discussion

Tumor suppressor protein p53 is involved in the development and progression of various tumor types through its role in apoptosis (Cardoso et al., 2009). C-erbB-2 is a member of the epidermal growth factor receptor family, the products of which are transmembrane signaling molecules that share close structural homology, and have all been implicated in cell transformation and tumor pathogenesis (Maurer et al., 1998). The present study attempted to test the expression pattern and the possible predictive value of expression of both markers in the management of molar pregnancy.

In the present study, we observed a relatively higher incidence of a persistent mole and the need for chemotherapy (about 40%), a finding that could be explained by the fact that our gynecologic oncology unit is a referral center to which patients who failed to attain spontaneous remission are referred.

In terms of the results of p53 immunostaining, p53 expression in CMs was significantly stronger than in PMs and significantly weaker than that in invasive moles and choriocarcinoma. In an early study by Cheville et al. (1996) and Qiao et al. (1998), the average percentage of p53-positive cells was around 10% for nonmolar tissue, 30% for PMs, and 40% for CMs. A more recent study by Petignat et al. (2006) and Hussein (2009) reported increased expression in HMs than in normal placentas, with greater expression in CMs versus PMs.

Rath et al. (2011) noted that p53 overexpression in CMs was presumed to be because of an upregulation of p53 (wild type). Mutation of p53 is the most frequent genetic alteration detected in human cancer, which inactivates its growth-regulatory function and causes loss of tumor-suppressive activity. Hence, the loss of the underlying p53 mutation in CMs suggests the restoration of p53 tumor suppressor function.

In terms of C-erbB-2 immunostaining, C-erbB-2 expression in CMs was stronger than that in PMs and weaker than that in invasive moles and choriocarcinoma. C-erbB-2 expression, however, showed no significant difference in distribution among the different histological types of the gestational trophoblastic disease. Fulop et al. (1998) found significantly stronger expression of C-erbB-2 protein in CMs and choriocarcinoma when compared with normal placenta or PMs. Moreover, Yang et al. (2002) found that the expression of C-erbB-2 and p53 products was significantly increased in CMs with subsequent malignant transformation.

Yuxia et al. (2002) reported that the positive rate of C-erbB-2 expression in HMs was 47.6%, significantly lower than that in invasive moles (81%) and in choriocarcinoma (95%). Yazaki-Sun et al. (2006) observed that C-erbB-2 immunoexpression occurred exclusively in cells of the extravillous trophoblast. They suggested that the function of C-erbB-2 is to promote the growth of these cells, which have a physiologic invasive property, and that, when enhanced in the HMs, will result in its progression to a gestational trophoblastic tumor.

In terms of the prediction of clinical behavior, our results showed that tumor suppressor protein p53 had comparable validity, showing a relatively high sensitivity (85.7%) and NPV (71.4%) and poor specificity (23.3%) and PPV (42.1%). In contrast, Cheung et al. (1994) reported that immunohistochemical expression of p53 in patients who progressed to persistent gestational trophoblastic disease did not show a significant difference from that with spontaneous remission. Yang et al. (2002) examined 50 CMs that progressed to GTD and 32 CMs with spontaneous remission, and reported that p53 expression was greater in those CMs that progressed to GTD, but logistic regression analysis taking into account various factors showed that p53 had no significant independent predictive value for malignant transformation. Similarly, another study by Yazaki-Sun et al. (2006) examined 35 CMs with progression to GTD and 32 with spontaneous remission and reported that p53 expression was significantly greater in those with GTD, with a suggested cut-off value of 40% for the prediction of malignant transformation. Sebire and Seckl (2010) concluded that p53 expression is increased in HMs compared with nonmolar gestations, especially CHM, and a highly increased expression may be associated with GTD. However, on the basis of current evidence, this marker does not have sufficient predictive value to be of use in routine clinical practice. Kale et al. (2001) suggested that the increased overexpression of p53 in gestational trophoblastic disease seems to correlate with the higher trophoblastic proliferation rate found mainly in complete and invasive HMs. Also, the staining pattern was intense and extensive, involving mainly the cytotrophoblastic cells in the invasive HMs group.

Our study showed that C-erbB-2 oncogene had high sensitivity (82.1%) and NPV (73.7%) and poor specificity (32.6%) and PPV (44.2%). In a study of Yang et al. (2002), the expression of C-erbB-2 had a sensitivity of 80%, a specificity of 70%, PPV of 84%, and NPV of 75%. They concluded that C-erbB-2 expression was significantly increased in CMs that progressed to GTD compared with those that required no treatment. Similarly, Yazaki-Sun et al. (2006) reported that C-erbB-2 expression was significantly increased in patients with GTD. They added that at a cut-off value of 10.8%, the test has a PPV of 73% and an NPV of 71%.

In the present study, concomitant negative expression of both markers was associated with spontaneous remission in 100% of our studied group (NPV 100%). This finding, if confirmed by further larger studies, may allow a shorter follow-up period.

Targeted therapy using monoclonal antibodies against the C-erbB-2 protein trastuzumab represents a well-established modality in the management of women with breast cancer with C-erbB-2 overexpression (Kauraniemi et al., 2004; Toi et al., 2004). Yang et al. (2002) observed that in almost half of the cases, C-erbB-2 expression increased after malignant transformation. This finding may provide a new avenue of treatment for the rare category of patients with gestational trophoblastic neoplasia who develop resistance to the standard chemotherapy to try this targeted therapy as a salvage treatment if their tumor shows overexpression of C-erbB-2.

Our data indicate that altered expression of p53 and C-erbB-2 products is associated with the malignant transformation of molar pregnancy. The increased expression of p53 and C-erbB-2 products has no prognostic value for predicting the outcome of molar pregnancy. Yet, negative expression may predict spontaneous remission. We recommend large future studies using several markers to assess their usefulness in routine practice and to aid the selection of high-risk patients for closer follow-up or possible administration of prophylactic chemotherapy.

Back to Top | Article Outline

Acknowledgements

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline

References

Cardoso SV, Silveira-Junior JB, De Machado VC, De-Paula AMB, Loyola AM, De Aguiar MCF. Expression of metallothionein and p53 antigens are correlated in oral squamous cell carcinoma. Anticancer Res. 2009;29:1189–1194
Cheung ANY, Srivastava G, Chung LP, Ngan HYS, Man TK, Liu YT, et al. Expression of the p53 gene in trophoblastic cells in hydatidiform moles and normal human placentas. J Reprod Med. 1994;39:223–227
Cheville JC, Robinson RA, Benda JA. p53 expression in placentas with hydropic change and hydatidiform moles. Mod Pathol. 1996;9:392–396
Fong PY, Xue WC, Ngan HYS, Chan KYK, Khoo US, Tsao SW, et al. Mcl-1 expression in gestational trophoblastic disease correlates with clinical outcome: a differential expression study. Cancer. 2005;103:268–276
Fulop V, Mok SC, Genest DR, Szigetvari I, Cseh I, Berkowitz RS. c-myc, c-erbb-2, c-fms and bcl-2 oncoproteins: expression in normal placenta, partial and complete mole and choriocarcinoma. J Reprod Med. 1998;43:101–110
Greene FL, Page DL, Fleming IDGreene FL, Page DL, Fleming ID, Fritz A, Balch CM, Haller DG, et al. Gestational trophoblastic tumors. AJCC cancer staging manual. 20026th ed. New York, USA Springer:323–328
Halperin R, Peller S, Sandbank J, Bukovsky I, Schneider D. Expression of the p53 gene and apoptosis in gestational trophoblastic disease. Placenta. 2000;21:58–62
Hancock BW, Tidy JA. Current management of molar pregnancy. J Reprod Med. 2002;47:347–354
Hussein MR. Apoptosis in the ovary: molecular mechanisms. Hum Reprod Update. 2005;11:161–177
Hussein MR. Analysis of p53, BCL-2 and epidermal growth factor receptor protein expression in the partial and complete hydatidiform moles. Exp Mol Pathol. 2009;87:63–69
Jelincic D, Hudelist G, Singer CF, Bauer M, Horn LC, Bilek K, et al. Clinicopathologic profile of gestational trophoblastic disease. Wien Klin Wochenschr. 2003;115:29–35
Jeong SY, Seol DW. The role of mitochondria in apoptosis. J Biochem Mol Biol. 2008;41:11–22
Kale A, Söylemez F, Ensari A. Expressions of proliferation markers (Ki-67, proliferating cell nuclear antigen, and silver-staining nucleolar organizer regions) and of p53 tumor protein in gestational trophoblastic disease. Am J Obstet Gynecol. 2001;184:567–574
Kauraniemi P, Hautaniemi S, Autio R, Astola J, Monni O, Elkahloun A, et al. Effects of Herceptin treatment on global gene expression patterns in HER2-amplified and nonamplified breast cancer cell lines. Oncogene. 2004;23:1010–1013
Kay EW, Walsh CJB, Cassidy M, Curran B, Leader M. C-erbB-2 immunostaining: problems with interpretation. J Clin Pathol. 1994;47:816–822
Maurer CA, Friess H, Kretschmann B, Zimmermann A, Stauffer A, Baer HU, et al. Increased expression of erbB3 in colorectal cancer is associated with concomitant increase in the level of erbB2. Hum Pathol. 1998;29:771–777
MCupon Li, Hertz R, Spencer DB. Effect of methotraxate therapy in choriocarcinoma and chorioadenoma. Proc Soc Exp Biol Med. 1956;99:361–366
Petignat P, Laurini R, Goffin F, Bruchim I, Bischof P. Expression of matrix metalloproteinase-2 and mutant p53 is increased in hydatidiform mole as compared with normal placenta. Int J Gynecol Cancer. 2006;16:1679–1684
Qiao S, Nagasaka T, Harada T, Nakashima N. p53, Bax and Bcl-2 expression and apoptosis in gestational trophoblast of complete hydatidiform mole. Placenta. 1998;19:361–369
Rath G, Soni S, Prasad CP, Salhan S, Jain AK, Saxena S. Bcl-2 and p53 expressions in Indian women with complete hydatidiform mole. Singapore Med J. 2011;52:502–507
Sebire NJ, Seckl MJ. Immunohistochemical staining for diagnosis and prognostic assessment of hydatidiform moles: current evidence and future directions. J Reprod Med. 2010;55:236–246
Shaarawy M, Sheiba M. Diagnostic and prognostic significance of circulating tumor suppressor gene p53 autoantibodies in patients with gestational trophoblastic tumors. Acta Oncol. 2004;43:43–48
Toi M, Takada M, Bando H, Toyama K, Yamashiro H, Horiguchi S, et al. Current status of antibody therapy for breast cancer. Breast Cancer. 2004;11:10–14
Xuan YH, Kim KH, Choi YL, Ahn GH, Chae SW, Lee HC, et al. The expression of G1-S cell cycle inhibitors in normal placenta and gestational trophoblastic diseases. Korean J Pathol. 2008;42:67–74
Yamauchi H, Katayama KI, Ueno M, He XJ, Mikami T, Uetsuka K, et al. Essential role of p53 in trophoblastic apoptosis induced in the developing rodent placenta by treatment with a DNA-damaging agent. Apoptosis. 2007;12:1743–1754
Yang X, Zhang Z, Jia C, Li J, Yin L, Jiang S. The relationship between expression of c-ras, c-erbB-2, nm23 and p53 gene products and development of trophoblastic tumor and their predictive significance for the malignant transformation of complete hydatidiform mole. Gynecol Oncol. 2002;85:438–444
Yazaki-Sun S, Daher S, De Souza Ishigai MM, Alves MTS, Mantovani TM, Mattar R. Correlation of c-erbB-2 oncogene and p53 tumor suppressor gene with malignant transformation of hydatidiform mole. J Obstet Gynaecol Res. 2006;32:265–272
Yuxia W, Yang C, Yongyu S. Expression of C-erbB-2 in gestational trophoblastic disease and its clinical significance. J of Huazhong Univ Sci Technolog Med Sci. 2002;22:123–125
©2012Egyptian Journal of Pathology