Microsatellite instability and alteration of the expression of hMLH1 and hMSH2 in ovarian clear cell carcinoma

Microsatellite Instability and Alteration of the Expression of hMLH1 and hMSH2 in Ovarian Clear Cell Carcinoma KATHY QI CAI, MD, PHD, CONSTANCE ALBARRACIN, MD, PHD, DANIEL ROSEN, MD, ROCKSHENG ZHONG, MD, WENXIN ZHENG, MD, RAJYALAKSHMI LUTHRA, PHD, RUSSELL BROADDUS, MD, PHD, AND JINSONG LIU, MD, PHD Microsatellite instability (MSI) is commonly seen in tumors associated with the hereditary nonpolyposis colorectal cancer syndrome and is caused by defects in the DNA mismatch repair genes. MSI has also been observed in various sporadic cancers, including colorectal, gastric, and endometrial. The role and incidence of MSI in ovarian clear cell carcinoma remain unknown. This study was conducted to evaluate the frequency of MSI in ovarian clear cell carcinomas and to evaluate the sensitivity and specificity of immunohistochemistry in predicting mismatch-repair gene deficiency. A total of 42 ovarian clear cell carcinomas were analyzed for MSI using a panel of 5 microsatellite markers (BAT25, BAT26, D5S346, D2S123, and D17S250). Alterations in the expression of hMLH1 and hMSH2 proteins in these tumors were examined. Of the 42 ovarian clear cell tumors analyzed, 6 demonstrated a high level of MSI (MSI-H), 3 demonstrated a low level of MSI (MSI-L), and the remaining 33 exhibited microsatellite stability (MSS). No correlation was found

between MSI level and patient age or tumor stage or size (P >0.05). Loss of expression of either hMLH1 or hMSH2 was observed in 4 of the 6 (67.7%) MSI-H tumors, whereas 34 of the 36 (94.4%) MSI-L or MSS tumors expressed both the hMLH1 and hMSH2 gene products. Our results indicate that MSI-H is involved in the development of a subset of ovarian clear cell carcinomas. A strong correlation exists between alterations in the expression of hMLH1 and hMSH2 and the presence of MSI-H in these tumors. However, immunohistochemical testing alone may miss a small fraction of cases with MSI-H. HUM PATHOL 35:552-559. © 2004 Elsevier Inc. All rights reserved. Key words: ovarian clear cell carcinoma, microsatellite instability, hMLH1, hMSH2, immunohistochemistry. Abbreviations: HNPCC, hereditary nonpolyposis colorectal cancer; MMR, mismatch repair; MSI, microsatellite instability; MSS, microsatellite stability; PCR, polymerase chain reaction.

Ovarian carcinoma is one of the most lethal cancers of women in the United States, accounting for more than 23,000 newly diagnosed cases and approximately 14,000 deaths each year.1 Moreover, although ovarian carcinoma constitutes approximately 25% of the cancers arising from the female genital organs, it accounts for approximately 50% of all deaths from gynecologic cancers. This relatively low survival rate is due to the intra-abdominal location of these tumors, the relative paucity of early symptoms, and the lack of a screening method for early detection. To help improve diagnostic capability, it is critical to understand molecular events involved in the development of ovarian tumors. In 1993, the genetic phenomenon known as microsatellite instability (MSI) was first observed in 90% of

hereditary nonpolyposis colorectal cancers (HNPCCs) and in 10% to 15% of sporadic colorectal cancers, leading to the identification of an important genetic pathway in the development of cancer. MSI is defined as frequent variations in the size of simple-sequence nucleotide repeats in tumor DNA compared with normal DNA of the same individual.2-5 It is caused by a failure of the DNA mismatch-repair system (MMR) to correct errors occurring during DNA replication. Such MMR defects may be caused either by a germline MMR gene mutation, affecting mainly hMLH1 or hMSH2 (2 major MMR genes), or by somatic MMR gene inactivation, most commonly through epigenetic silencing via methylation of the hMLH1 promoter. MMR gene defects have also been observed in other sporadic cancers, including those of the stomach (13% to 44%)6 and endometrium (17% to 23%).7,8 The role and incidence of MSI in ovarian cancer are matters of great controversy. Studies in human ovarian cancer cell lines have revealed alterations in MMR genes and MSI.9,10 In some studies, however, MSI frequencies have ranged from 0 to 53% in sporadic ovarian cancers.11-18 These studies varied greatly in terms of the type and number of loci studied, the criteria used for defining MSI, and the histological subtypes of ovarian cancer analyzed, however. There were no commonly accepted criteria for the number, type, and identity of microsatellites to used in MSI testing until 1997, when the National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome proposed specific markers for MSI assess-

From the Department of Pathology, University of Texas M. D. Anderson Cancer Center, Houston, TX and Department of Pathology, Yale University, New Haven, CT. Accepted for publication December 19, 2003. Supported by a grant from the American Cancer Society (RSG04-028-CCE, to D.R.), the National Cancer Institute (P01CA64602-1) and institutional start-up funds, an Institutional Research Grant, and a Career Development Award from the M.D. Anderson Cancer Center SPORE on Ovarian Cancer (to J.L.). Address correspondence and reprint requests to Jinsong Liu, MD, PHD, University of Texas M.D. Anderson Cancer Center, Department of Pathology, Box 85, 1515 Holcombe Blvd., Houston, TX 77030. 0046-8177/$—see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2003.12.009

552

MISMATCH REPAIR IN OVARIAN CARCINOMA (Cai et al)

ment in HNPCC, consisting of 2 mononucleotide repeats (BAT26 and BAT25) and 3 dinucleotide repeats (D5S346, D2S123 and D17S250), collectively known as the NCI panel.19 This standard panel was subsequently adapted for use in assessing ovarian cancer.20 Several recent studies have indicated that immunohistochemistry offers an alternative method for assessing MSI status, and that there is a close association between MSI status and altered immunohistochemical expression of hMLH1 and hMSH2.21-27 The purpose of this study was to investigate the involvement of MMR in the development of ovarian clear cell carcinoma. Toward this end, MSI status and immunohistochemical expression of hMLH1 and hMSH2 were assessed in 42 cases of ovarian clear cell carcinoma.

MATERIALS AND METHODS Patients and Tissue Samples Matched pairs of formalin-fixed, paraffin-embedded normal and tumor tissue samples were obtained from 42 women with ovarian clear cell cancer. Twenty-four of these patients were from the Department of Pathology at the University of Texas M. D. Anderson Cancer Center, and 18 were from the Department of Pathology at Yale University. The samples were used with the approval of the respective institutional review boards. Clinical information was obtained from pathology reports and/or patient charts. All histopathologic diagnoses were confirmed by 2 pathologists (K.Q.C. and J.L.). Tumor stages were assigned according to the classification system of the International Federation of Gynecology and Obstetrics (FIGO). Histological diagnosis was assigned based on the World Health Organization (WHO) classification

TABLE 1. Clinicopathologic Features, MSI Status, and Immunohishochemical Results in 42 Ovarian Clear Cell Carcinomas IHC expression Case no.

Age

Stage

Tumor size (cm)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

59 56 59 53 59 43 59 60 60 67 57 46 74 78 72 57 42 56 42 71 55 65 44 52 35 71 69 47 87 59 73 56 65 50 80 54 61 37 52 69 48 40

IIIC IC IA IIIC IIIC IIIC IV IIIB IIC IIC IIIC IC IA NA IV IA IC IB IA IC IC IB IA IC IC IA IC IB IIB IA IC IA IIIA IB IA IIIC IIIC IC IC IIIC IV IA

NA 7 NA 7.2 15 NA 15 19.5 12 7.5 3 NA 13 NA 1.5 11.2 10 3.5 7 17 15 5 15 13 17 3 25 4 NA 8 11.5 12 3 1.8 2.5 NA NA 8 11 6 12 20

BAT25

BAT26

D2S123

D5S346

⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ NA NA ⫺ ⫺ ⫺ NA NA ⫺ ⫺ ⫹ NA ⫺ ⫺ ⫺ ⫺ NA ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ NA ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

Abbreviation: NA, not available.

553

D17S250

MSI status

hMSH2

hMLH1

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ NA ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

MSS MSS MSS H(3/4) MSS H(3/5) MSS H(4/4) MSS L(1/5) MSS H(4/5) MSS MSS MSS MSS H(3/5) H(3/4) MSS MSS MSS MSS MSS MSS MSS MSS MSS MSS MSS L(1/5) MSS MSS MSS MSS MSS L(1/4) MSS MSS MSS MSS MSS MSS

(⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫺) (⫺) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹)

(⫹) (⫹) (⫹) (⫺) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫺) (⫹) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹)

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

FIGURE 1. Analysis of MSI with the 5 microsatellite markers in a representative sample. A sample of an MSI-positive tumor showed extra alleles (A) compared with the matched normal sample (B).

scheme. None of the patients had a personal or familial history suggestive of HNPCC syndrome. Of the 42 study subjects, 11 were under age 50 and 31 were over 50; 25 had stage I disease, 3 had stage II, 10 had stage III, and 3 had stage IV. Table 1 presents clinicopathologic features of the 42 ovarian clear cell carcinomas studied. Hematoxylin and eosin–stained sections of available paraffin blocks were examined, and the most representative samples were selected for further analysis. From selected paraffin blocks, 5-␮m sections were cut and placed on coated slides for immunohistochemical analysis, and 2 10-␮m sections were

obtained for DNA extraction. DNA extracts were prepared as described previously.28

Analysis of MSI Tumor tissues were manually microdissected from adjacent normal tissue to ensure that each tumor sample contained at least 70% neoplastic cells. Specimens from a nonmetastatic lymph node, the appendix, or segments of normal fallopian tube from the same patient were used as normal tissue controls. Tumor and normal DNAs were extracted and

554

MISMATCH REPAIR IN OVARIAN CARCINOMA (Cai et al)

analyzed for MSI using the NCI panel of 5 microsatellite markers: BAT26, BAT25, D5S346, D2S123, and D17S250. Oligonucleotide forward primers were fluorescently 5' labeled. Multiple polymerase chain reactions (PCRs) were carried out in a 12-␮L mixture containing approximately 40 ng DNA, 1 ⫻ PCR buffer, 0.5 ␮mol of each PCR primer (Life Technologies, Gaithersburg, MD), 2.5 mmol of MgCl2, 200 ␮mol of deoxynucleotide triphosphate, and 2 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA). PCR amplification was conducted in a PerkinElmer Gene Amp Thermo Cycler 9600 (PerkinElmer, Boston, MA). Samples were denatured at 95°C for 7 minutes and then subjected to 3 cycles consisting of denaturation (94°C for 1 minute), annealing (58°C for 30 seconds), and extension (72°C for 45 seconds), and then to 42 cycles each consisting of 45 seconds at 94°C, 30 seconds at 54°C, 40 seconds at 72°C, and a final extension step at 72°C for 30 minutes. Then 1 ␮L of fluorescently labeled PCR products of paired normal and tumor tissues were mixed with 12 ␮L of deionized formamide and 1 ␮L of GeneScan TAMRA 500 Size Standard (Applied Biosystems), respectively. The mixtures were denatured at 94°C for 5 minutes, cooled on ice, and then loaded on the ABI PRISM 310 Genetic Analyzer (Applied Biosystems). The data were collected automatically and analyzed using the GeneScan 3.1 and Genotyper Analysis software (Applied Biosystems), which automatically determined the actual size of the PCR products and the amount of fluorescent signal. MSI was indicated by the presence of novel peaks in the tumor tissue that were not seen in normal control tissue from the same patient or by a difference in microsatellite lengths in the 2 samples. Tumors exhibiting MSI at 2 or more markers were defined as MSIhigh (MSI-H). Tumors showing instability at only 1 marker were defined as MSI-low (MSI-L). Tumors in which no markers exhibited MSI were considered to exhibit microsatellite stability (MSS). All positive samples were assayed at least twice to confirm the results.

Immunohistochemical Staining of hMSH2 and hMLH1 For immunohistochemical analysis, mouse monoclonal antibodies against hMLH1 (PharMingen, San Diego, CA) at 1:30 dilution and against hMSH2 (Ab-2; Oncogene, La Jolla, CA) at 1:100 dilution were used. The immunostaining was performed with the Envison-horseradish peroxidase kit (Dako, Carpinteria, CA) for hMLH1 and the LSAB horseradish peroxidase kit (Dako) for hMSH2, using diaminobenzidine as a chromogen. Sections were lightly counterstained with hematoxylin. The normal staining pattern of both hMLH1 and hMSH2 was nuclear. Lost expression of hMSH2 or hMLH1 in cancer tissues was demonstrated by the total absence of detectable nuclear staining of neoplastic cells. Infiltrating lymphocytes as well as stromal cells served as internal positive controls. Two pathologists assessed all of the cases without any knowledge of MSI results. Assessment of sensitivity and specificity was performed against the classification of MSI-H or MSI-L/MSS as standard. Sensitivity referred to the ratio of abnormal immunohistochemical results found in the total number of MSI-H tumors. Specificity was calculated as the ratio of normal immunohistochemical results in the total number of MSI-L/MSS tumors.

Statistical Analysis The sensitivity and specificity of hMLH1 or hMSH2 immunostaining in identifying MSI tumors were calculated. The significance of the differences between groups was analyzed

TABLE 2. Association Between MSI Status and Clinicopathologic Variables in 42 Ovarian Clear Cell Carcinomas Variable Age in years ⱕ50 ⬎50 FIGO stage I/II III/IV Tumor size ⱕ5 cm ⬎5 cm

n

MSI-H (%)

MSI-L/MSS (%)

P value

11 31

3 (27.3%) 3 (9.7%)

8 (72.7%) 28 (90.3%)

0.31

28 13

3 (10.7%) 3 (23.1%)

25 (89.3%) 10 (76.9%)

0.36

9 25

1 (11.1%) 3 (12%)

8 (88.9%) 22 (88%)

1.0

using Fisher’s exact test. A P value ⬍0.05 was considered statistically significant.

RESULTS Among the 42 clear cell carcinomas analyzed, 6 (14.3%) were MSI-H, 3 (7.2%) were MSI-L, and the remaining 33 were MSS. The 5 NCI markers used to detect the MSI-H tumors had the following sensitivities: BAT25 (5 of 6, 83.3%), BAT26 (6 of 6, 100%), D5S346 (5 of 6, 83.3%), D2S123 (1 of 2, 50%), and D17S250 (3 of 6, 50%). One example of MSI-H observed in clear cell carcinoma is shown in Figure 1. MSI was present at 4 of the 5 NCI markers tested (Table 2). Of the 6 tumors that displayed MSI-H, 3 were stage I/II and 3 were stage III/IV. No association was observed between tumor stage and presence of MSI-H. There were no significant differences in patient age or tumor size between patients with MSI-H tumors and those with MSI-L/MSS tumors. Of the 6 MSI-H cases, 4 showed loss of hMLH1 or hMSH2 protein expression. Of the 3 MSI-L cases, none showed loss of hMLH1 or hMSH2 expression. Of the 33 MSS cases, 2 showed loss of hMLH1 or hMSH2 expression. The sensitivity and specificity rates for immunohistochemical testing assessed against the MSI results were 67.7% (4 of 6) and 94.4% (34 of 36), respectively. One tumor specimen with positive hMSH2 staining but negative hMLH1 staining is shown in Figure 2. DISCUSSION On the basis of morphological criteria, there are 4 major types of primary ovarian adenocarcinomas: serous, mucinous, endometrioid, and clear cell. Several studies indicate that the different histological types of ovarian adenocarcinomas probably represent distinct disease entities and involve different molecular pathways.29 For example, serous adenocarcinomas demonstrate frequent p53 gene mutations, whereas K-ras activation occurs more frequently in mucinous adenocarcinomas.30,31 Therefore, understanding the molecular basis of each morphological type and its biological behavior is very important and will eventually lead to the development

555

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

FIGURE 2. Representative immunohistochemical staining of hMLH1 and hMSH2. Immunohistochemical staining for hMLH1 and hMSH2 in the tumor from patient 6. (A) Tumor with normal nuclear staining of hMSH2. (B) Tumor with loss of expression for hMLH1, but stromal cells showed positive staining. (Immunohistochemical stain; original magnification ⫻ 200.)

556

MISMATCH REPAIR IN OVARIAN CARCINOMA (Cai et al)

TABLE 3. Comparison of the Predictive Values of MMR Gene Expression in This Study With Those in Published Reports Especially on Colorectal Tumors

Reference Our present study Liu et al 200328 Cawkwell et al (1999)21 Dieumegard et al (2000) 22 Marcus et al (1999) 24 Terdiman, et al (2001) 25 Lindor et al (2002) 27 Chiaravalli et al (2001)37

Chaves et al (2000)26 Ericson et al (2003)38 Halvarsson et al (2003)39

Tissue

Total number of patients

Frequency of MSI-H

Sensitivity

6/42

4/6 (67.7%)

PPV

Correlation of MSI status and IHC

34/36 (94.4%)

66.7%

66.7%

81.8% N/A

Specificity

Ovary

42

Ovary Colon

74 421 (sporadic)

15/74 49/421

9/15 (60%) 66/66 (100%)

57/59 (96.6%) N/A

Colon

34

15/17

15/17 (88%)

17/17 (100%)

100%

100%

Colon

72

37/38 (97.3%)

34/34 (100%)

100%

97%

Colon

109†

29/38 (76.3%)

40/40 (100%)

100%

94%

Colon

1144 (sporadic)

323/350 (95%)

794/794 (100%)

100%

96.7%

Colorectal, gastric, endometrial and ovarian carcinomas Colon

201§

51/55 (92.8%)

145/146 (99.3%)

98%

92.8%

N/A

75%

Colon¶ Colon

76 (sporadic)

47/109 350/1144 55/201

N/A

60% 100%

8/76

6/8 (75%)

47

59/154

52/59 (88%)

87/89 (97.8%)

96%

96%

128 (HNPCC)

59/128

54/59 (92%)

69/69 (100%)

100%

100%

Abbreviations: NA, not available; PPV, positive predictive value. †In patients that fulfill the Amsterdam criteria or Bethesda guidelines. §All are sporadic tumors. ¶With at least one other primary malignancy.

of more specific and effective treatments for ovarian cancer.32 Assessment of MMR defects has recently become an important tool in tumor molecular pathology and clinical practice. MSI-H and MSI-L/MSS phenotypes appear to characterize 2 different pathways of carcinogenesis. Patients who present with MSI-H carcinomas have a better prognosis than those with MSI-L/MSS tumors. The status of MMR may be important in predicting tumor response to clinical therapy.19 Several lines of indirect evidence suggest that MSI might play a role in the genesis of ovarian carcinoma. Ovarian carcinoma and endometrial carcinomas are 2 common exocolonic tumors occurring in patients with HNPCC, which is characterized by a high MSI frequency. MSI occurs in approximately 20% of endometrial carcinomas, a tumor closely related to ovarian carcinoma.33 Nevertheless, there is no general agreement about the frequency of MSI in sporadic ovarian carcinomas. Some studies have reported very low frequencies, whereas others have reported high frequencies (as high as 53%).11-18 In addition, several studies have suggested an association between MSI and certain histological types of ovarian carcinoma. For example, Fujita et al12 demonstrated that the incidence of MSI was significantly higher in the endometrioid subtype of ovarian tumor (50%) than in other histological subtypes (8%). Ohwada et al14 reported finding MSI more frequently in mucinous adenocarinoma (38%) than in serous car-

cinoma (13%). King et al11 found a high frequency of MSI in endometrioid carcinoma (33.3%) and a relatively lower frequency in invasive serous carcinoma (8%). Tangir et al13 and Haas et al18 reported MSI-H occurring with very low frequency in patients with serous adenocarcinoma (6.5% and 0, respectively) but relatively often in epithelial borderline ovarian tumors (27.8% and 30%). Shih et al16 did not find any MSI in 31 ovarian neoplasms of low malignant potential. Arzimanoglou et al,34 however, demonstrated that MSI in ovarian cancer is not definitely associated with a specific histopathologic subtype. In most of these studies, unfortunately, different kinds of microsatellite markers were used and MSI was assessed with the demonstration of instability at only 1 locus, which is no longer considered an adequate criterion for MSI positivity. Most of these studies were done before the development of a standard panel of microsatellite markers and standardized criteria for determining MSI. This may at least partly explain the controversial frequency of MSI observed in these studies. Until now, the NCI markers and criteria have been used in few studies of MSI in ovarian cancer. Sood et al20 found MSI-H in 12% of invasive ovarian tumors, and Gras et al35 reported that MSI-H was limited to endometrioid and clear cell carcinomas (12.5%). Our previous study demonstrated that 20% of endometrioid carcinomas were MSI-H.28 Reports of MSI investigation in clear cell carcino-

557

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

mas are rare in the literature. In the current study, we found MSI-H in 6 of 42 (14.3%) 42 ovarian clear cell carcinomas using the new-standard NCI markers and criteria. The incidence of MSI-H showed no correlation with patient age, tumor stage, or other clinicopathologic features. This result suggests that MSI plays an important role in the pathogenesis of a subset of clear cell carcinomas (Table 2). Molecular testing for MSI is relatively expensive and time-consuming. Immunohistochemical analysis, in contrast, is relatively inexpensive, fast, and simple, and has become routine for the vast majority of pathology laboratories. Several reports have demonstrated that immunohistochemical staining of hMLH1 and hMSH2 may be a highly sensitive and specific screening method for detecting MSI-H in colorectal cancers (Table 3).36 Two studies found 100% correlation between MSI status and immunohistochemical results,21,22 whereas other studies reported not so perfect correlation, ranging from 75% to 95%.23-26,37 In a large multicenter study, Lindor et al27 studied MSI and immunohistochemical findings in colorectal tumors from 1144 patients and found a sensitivity of 92.3% and specificity of 100% for immunohistochemical detection of hMLH1 and hMSH2 for MSI. In our studies, hMLH1 or hMSH2 expression was lost in 4 of 6 MSI-H tumors but in only 2 of 36 MSS/MSI-L tumors. The sensitivity and specificity rates for immunohistochemical testing assessed against the MSI-H results were 67.7% (4 of 6) and 94.4% (34 of 36), not as high as those seen in colorectal cancer; however, the sample size in our study is small. In addition, there may be other reasons why 2 MSI-H tumors exhibited normal expression of hMLH1 and hMSH2, including (1) other MMR genes (eg, hMLH3, hMSH3, hMSH6, hPMS1, hPMS2) can cause MSI-H in ovarian carcinoma and (2) functionally inactive MMR protein is still sometimes detectable with immunohistochemical analysis. In the 4 cases in which hMLH1 or hMSH2 expression was lost, hMLH1 expression was lost in 3 and expression of hMSH2 was lost in 1. This indicates that hMLH1 might play a more important role than hMSH2 in MSI-H tumors. Several studies have shown that patients with MSI-H colorectal tumors seem to have an improved prognosis, and in vitro data suggest that MSI-H cell lines may be relatively resistant to certain types of chemotherapy. If that is also the case for ovarian cancer, then a group of patients with MSI-H phenotype should able to be identified and offered more appropriate therapy. Because simple, rapid, and cost-effective immunohistochemistry enables us to determine the status of hMLH1 and hMSH2 highly correlated with MSI-H and can be implemented in any diagnostic histopathology laboratory, this method can be very useful in routine clinical practice. In conclusion, our study suggests that MSI-H has a role in the pathogenesis of a subset of ovarian clear cell carcinomas (14%). There is correlation between alterations in the expression of hMLH1 and hMSH2 and the presence of a high level of MSI in tumor tissues. Testing

with immunohistochemistry alone, however, may miss a small portion of MSI-positive cases. REFERENCES 1. Greenlee RT, Murray T, Bolden S, et al: Cancer statistics, 2000. CA Cancer J Clin 50:7-33, 2000 2. Ionov Y, Peinado MA, Malkhosyan S, et al: Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 363:558-561, 1993 3. Aaltonen LA, Peltomaki P, Leach FS, et al: Clues to the pathogenesis of familial colorectal cancer. Science 260:812-816, 1993 4. Thibodeau SN, Bren G, Schaid D: Microsatellite instability in cancer of the proximal colon. Science 260:816-819, 1993 5. Liu B, Nicolaides NC, Markowitz S, et al: Mismatch repair gene defects in sporadic colorectal cancers with microsatellite instability. Nat Genet 9:48-55, 1995 6. Halling KC, Harper J, Moskaluk CA, et al: Origin of microsatellite instability in gastric cancer. Am J Pathol 155:205-211, 1999 7. Burks RT, Kessis TD, Cho KR, et al: Microsatellite instability in endometrial carcinoma. Oncogene 9:1163-1166, 1994 8. Risinger JI, Berchuck A, Kohler MF, et al: Genetic instability of microsatellites in endometrial carcinoma. Cancer Res 53:51005103, 1993 9. Boyer JC, Umar A, Risinger JI, et al: Microsatellite instability, mismatch repair deficiency, and genetic defects in human cancer cell lines. Cancer Res 55:6063-6070, 1995 10. Orth K, Hung J, Gazdar A, et al: Genetic instability in human ovarian cancer cell lines. Proc Natl Acad Sci U S A 91:9495-9499, 1994 11. King BL, Carcangiu ML, Carter D, et al: Microsatellite instability in ovarian neoplasms. Br J Cancer 72:376-382, 1995 12. Fujita M, Enomoto T, Yoshino K, et al: Microsatellite instability and alterations in the hMSH2 gene in human ovarian cancer. Int J Cancer 64:361-366, 1995 13. Tangir J, Loughridge NS, Berkowitz RS, et al: Frequent microsatellite instability in epithelial borderline ovarian tumors. Cancer Res 56:2501-2505, 1996 14. Ohwada M, Suzuki M, Saga Y, et al: DNA replication errors are frequent in mucinous cystadenocarcinoma of the ovary. Cancer Genet Cytogenet 117:61-65, 2000 15. Sood AK, Buller RE: Genomic instability in ovarian cancer: A reassessment using an arbitrarily primed polymerase chain reaction. Oncogene 13:2499-2504, 1996 16. Shih YC, Kerr J, Hurst TG, et al: No evidence for microsatellite instability from allelotype analysis of benign and low malignant potential ovarian neoplasms. Gynecol Oncol 69:210-213, 1998 17. Allen HJ, DiCioccio RA, Hohmann P, et al: Microsatellite instability in ovarian and other pelvic carcinomas. Cancer Genet Cytogenet 117:163-166, 2000 18. Haas CJ, Diebold J, Hirschmann A, et al: Microsatellite analysis in serous tumors of the ovary. Int J Gynecol Pathol 18:158-162, 1999 19. Boland CR, Thibodeau SN, Hamilton SR, et al: A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: Development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 58:5248-5257, 1998 20. Sood AK, Holmes R, Hendrix MJ, et al: Application of the National Cancer Institute international criteria for determination of microsatellite instability in ovarian cancer. Cancer Res 61:4371-4374, 2001 21. Cawkwell L, Gray S, Murgatroyd H, et al: Choice of management strategy for colorectal cancer based on a diagnostic immunohistochemical test for defective mismatch repair. Gut 45:409-415, 1999 22. Dieumegard B, Grandjouan S, Sabourin JC, et al: Extensive molecular screening for hereditary non-polyposis colorectal cancer. Br J Cancer 82:871-880, 2000 23. Debniak T, Kurzawski G, Gorski B, et al: Value of pedigree/ clinical data, immunohistochemistry and microsatellite instability analyses in reducing the cost of determining hMLH1 and hMSH2 gene mutations in patients with colorectal cancer. Eur J Cancer 36:49-54, 2000

558

MISMATCH REPAIR IN OVARIAN CARCINOMA (Cai et al) 24. Marcus VA, Madlensky L, Gryfe R, et al: Immunohistochemistry for hMLH1 and hMSH2: A practical test for DNA mismatch repair-deficient tumors. Am J Surg Pathol 23:1248-1255, 1999 25. Terdiman JP, Gum JR Jr, Conrad PG, et al: Efficient detection of hereditary nonpolyposis colorectal cancer gene carriers by screening for tumor microsatellite instability before germline genetic testing. Gastroenterology 120:21-30, 2001 26. Chaves P, Cruz C, Lage P, et al: Immunohistochemical detection of mismatch repair gene proteins as a useful tool for the identification of colorectal carcinoma with the mutator phenotype. J Pathol 191:355-360, 2000 27. Lindor NM, Burgart LJ, Leontovich O, et al: Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors. J Clin Oncol 20:1043-1048, 2002 28. Liu J, Albarracin CT, Chang KH, et al: Microsatellite instability and expression of hMLH1 and hMSH2 proteins in ovarian endometrioid cancer. Mod Pathol 17:75-80, 2004 29. Feeley KM, Wells M: Precursor lesions of ovarian epithelial malignancy. Histopathology 38:87-95, 2001 30. Aunoble B, Sanches R, Didier E, et al: Major oncogenes and tumor-suppressor genes involved in epithelial ovarian cancer (review). Int J Oncol 16:567-576, 2000 31. Fujita M, Enomoto T, Inoue M, et al: Alteration of the p53 tumor-suppressor gene occurs independently of K-ras activation and more frequently in serous adenocarcinomas than in other common epithelial tumors of the human ovary. Jpn J Cancer Res 85:1247-1256, 1994 32. Schwartz DR, Kardia SL, Shedden KA, et al: Gene expression

in ovarian cancer reflects both morphology and biological behavior, distinguishing clear cell from other poor-prognosis ovarian carcinomas. Cancer Res 62:4722-4729, 2002 33. Staebler A, Lax SF, Ellenson LH: Altered expression of hMLH1 and hMSH2 protein in endometrial carcinomas with microsatellite instability. Hum Pathol 31:354-358, 2000 34. Arzimanoglou II, Lallas T, Osborne M, et al: Microsatellite instability differences between familial and sporadic ovarian cancers. Carcinogenesis 17:1799-1804, 1996 35. Gras E, Catasus L, Arguelles R, et al: Microsatellite instability, MLH-1 promoter hypermethylation, and frameshift mutations at coding mononucleotide repeat microsatellites in ovarian tumors. Cancer 92:2829-2836, 2001 36. de La Chapelle A: Microsatellite instability phenotype of tumors: Genotyping or immunohistochemistry? The jury is still out. J Clin Oncol 20:897-899, 2002 37. Chiaravalli AM, Furlan D, Facco C, et al: Immunohistochemical pattern of hMSH2/hMLH1 in familial and sporadic colorectal, gastric, endometrial and ovarian carcinomas with instability in microsatellite sequences. Virchows Arch 438:39-48, 2001 38. Ericson K, Halvarsson B, Nagel J, et al: Defective mismatchrepair in patients with multiple primary tumours including colorectal cancer. Eur J Cancer 39:240-248, 2003 39. Halvarsson B, Lindblom A, Rambech E, et al: Microsatellite instability analysis and/or immunostaining for the diagnosis of hereditary nonpolyposis colorectal cancer? Virchows Arch 444:135-141, 2004

559