Human Papillomavirus Types 16 and 18 in Adenocarcinoma of the Uterine Cervix

Transcription

1 ANATOMIC PATHOLOGY Human Papillomavirus Types 16 and 18 in Adenocarcinoma of the Uterine Cervix ARTO LEMINEN, M.D., JORMA PAAVONEN, M.D., ERVO VESTERINEN, M.D., TORSTEN WAHLSTROM, M.D., IMMO RANTALA, PH.D., AND MATTI LEHTINEN, M.D. Many reports have shown a link between human papillomavirus (HPV) and cervical squamous neoplasia. However, the association of HPV with cervical adenocarcinoma has been studied less extensively. The authors evaluated the presence of HPV- DNA in 106 patients with adenocarcinoma of the uterine cervix by in situ hybridization, using 35 S-labeled probes for HPV 16 DNA and HPV 18 DNA. The overall prevalence of HPV-DNA was 18% (19 of 106). HPV 16 was present in 2 (2%) cases, HPV 18 was observed in 15 (14%) cases, and both HPV 16 and HPV 18 were found in 2 (2%) cases. There was a correlation between HPV-DNA positivity and tumor stage (P < 0.01) and tumor size (P < 0.05), but there was no relationship between HPV-DNA positivity and tumor differentiation, proliferation (S-phase fraction), ploidy, lymph node metastases, or five-year survival rate. These results suggest that HPV 18 DNA is associated with cervical adenocarcinoma but the presence of HPV 18 has no influence on overall survival. (Key words: HPV-DNA; Cervical adenocarcinoma) Am J Clin Pathol 1991;95: The major risk factors for cervical carcinoma are early sexual activity and a large number of sexual partners, suggesting that sexually transmitted infection may be an important etiologic factor. 1 ' 2 The association of human papillomavirus (HPV) infection with cervical carcinoma is well documented,'" 3 but its role in pathogenesis has not yet been established. 4 " 6 Several recent studies have demonstrated HPV-DNA in most cervical carcinomas. 2 HPV 16 has been the predominant type in cervical squamous cell carcinomas, whereas HPV 18 has dominated in cervical adenocarcinomas. 4 ' 7 "" In previous evaluations, the prevalence of HPV-DNA in cervical adenocarcinoma has ranged from 25 to 80%. 4 ' 7 "" However, the study populations in these reports have been relatively small, and histologic and clinical criteria have been too heterogeneous The tumors were classified according to criteria of the From the Departments of Obstetrics and Gynecology and Pathology, Helsinki University Central Hospital; the Department of Pathology, International Federation of Gynecology and Obstetrics Tampere University Central Hospital; and the Department of Biomedical (FIGO). The principles of treatment and follow-up of these Sciences, University of Tampere, Finland. patients have been described in detail elsewhere. 15 Received May 1, 1990; received revised manuscript and accepted for publication August 16, Supported by grants from Paulo Foundation, the Finnish Cancer Foundation, The University of Helsinki, The University of Tampere, and the Medical Research Council, Academy of Finland. Address reprint requests to Dr. Leminen: Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Haartmaninkatu 2. SF Helsinki, Finland. for one to determine whether HPV-DNA indeed plays a role in the development of cervical adenocarcinoma. In the current study, we examined tissue samples from a large number of patients with cervical adenocarcinoma using in situ hybridization, to determine the prevalence of HPV 16 and HPV 18 DNA and to assess the prognostic significance of HPV-DNA. Patient MATERIALS AND METHODS Material Records were reviewed of patients with cervical adenocarcinoma seen at the University Central Hospital of Helsinki between 1956 and Formalin-fixed, paraffin-embedded tissue blocks from 116 patients were available for analysis of HPV-DNA content. The original histopathologic diagnoses were verified by one of us (T.W.). In situ Hybridization The in situ hybridization of paraffin-embedded tissues was performed as described by Syrjanen and Syrjanen,' 2 with some modifications. Briefly, three parallel 5-^m sections from formalin-fixed tissue blocks were mounted on 647

2 648 ANATOMIC PATHOLOGY glass slides coated with aminopropoxymethyl-silane. The sections were heated to 60 C overnight, deparaffinized, and rehydrated by sequential immersion in xylene, graded ethanols, and phosphate-buffered saline (PBS). Thereafter, they were digested with proteinase-k (1.0 mg/ml, dissolved in PBS) at 37 C for 10 minutes. After this step, the slides were washed with PBS and air dried. DNA probes (whole genome of HPV types 16 and 18 cloned in pbr322 vector, obtained from Dr. Lutz Gissman) were nick-translated with 35 S-labeled nucleotides. The probes and sample DNA were heat-denatured concomitantly at 95 C for 10 minutes and immediately cooled thereafter on ice. Hybridization was accomplished at 42 C in a humidified chamber for two days. After hybridization, the slides were dipped in liquid photographic emulsion (Kodak Nucler Tract NTB-3, Eastman Kodak Company, BDR) and allowed to dry overnight. They were then exposed for ten days, developed with D-3 solution (Eastman Kodak), and counterstained with hematoxylin. In ten cases (9%), analyses could not be performed because tissue sections became detached from the glass slides. Flow Cytometric Analysis Flow cytometric analysis was performed as described in detail previously. 13 Thirty-micron sections from paraffin-embedded tumors were dewaxed, rehydrated, and digested overnight with trypsin. Nuclear suspensions were stained with ethidium bromide, treated with RNAase, and analyzed by flow cytometry using 488 nm excitation. Only histograms with diploid DNA peak coefficients of variation (CVs) less than 7.5% were regarded as satisfactory, and were subjected to additional analyses. TABLE 1. CORRELATION OF HPV-16 AND HPV-18 POSITIVITY WITH AGE, DNA PLOIDY, AND HISTOLOGIC FINDINGS IN 106 PATIENTS WITH CERVICAL ADENOCARCINOMA Age <40 years >40 years DNA ploidy Diploidy Aneuploidy Histologic differentiation High to intermediate Low Not known Histologic type Adenocarcinoma Adenosquamous carcinoma Total Negative 8(89) 79(81) 62 (82) 25 (83) 51 (70) 17(81) 9(75) 78 (80) 9(82) HPV 16/18 Positive 1(11) 18(19) 14(18) 5(17) 12 (30) 4(19) 3(25) 17(20) 2(18) TABLE 2. CORRELATION OF HPV-16 AND HPV-18 POSITIVITY WITH TUMOR PROGRESSION CHARACTERISTICS IN 106 PATIENTS WITH CERVICAL ADENOCARCINOMA Stage I II III-IV Tumor size <4 cm >4 cm Tumor proliferation SPF< 14% SPF> 14% Lymph node metastases No Yes Not known * />< t P < Statistical Methods Total Negative 74 (88) 13(59) 70 (86) 17(68) 71 (82) 16(79) 62 (84) 13(81) 12(75) HPV 16/18 Positive 10(12)* 9(41)* H(14)t 8 (34)t 15(18) 4(21) 12(16) 3(19) 4(25) The association between HPV-DNA and clinical and histopathologic findings was analyzed by chi-square analysis and Fischer's exact test. Survival curves and the log rank test for statistical significance were performed as described by Peto and associates. 14 Characteristics of the Study RESULTS Population The mean age of the patients was 56 years (range, years). Nine patients (8%) were younger than 40 years of age (Table 1). Histopathologic findings are summarized in Table 1; 95 patients (90%) had pure adenocarcinomas, whereas the remaining 11 (10%) had adenosquamous carcinomas. Tumor differentiation could be determined in 94 cases: 73 (69%) were well or moderately differentiated, and 21 (20%) were poorly differentiated. Eighty-four patients (79%) had stage I II (early) and 22 patients (21%) had stage III-IV (advanced) disease (Table 2). Primary tumors measured less than 4 cm in size in 81 patients (76%). Lymph node metastases were found in 16 patients (15%). S-phase fractions (SPFs), reflecting tumor proliferation, and tumor ploidy were calculated in all patients by flow cytometry. The median SPF was 14%. Belowmedian SPF values were found in 86 (81%) and abovemedian SPF values in 20 (19%) of the patients. Diploid tumors were found in 76 of the cases (72%) and aneuploid tumors in 30 cases (28%). A.J.C.P.-May 1991

3 649 LEMINEN ET AL. HPV-DNA in Cervical Adenocarcinoma Correlations with HPV-DNA The overall prevalence of HPV-DNA was 18%. HPV 16 was detected in 2 cases (2%), HPV 18 in 15 cases (14%), and both HPV 16 and HPV 18 in 2 cases (2%) (Fig. 1). The rate of HPV-DNA positivity was 21.7% (5 of 23 cases) during and 16.9% (14 of 87 cases) during One HPV 16-positive case was a stage I lesion, and the other was stage III. The same was also true of the two cases positive for both HPV 16 and 18. Eight of the HPV 18-positive cases (53%) represented earlystage tumors, and seven (47%) presented in advanced stages. HPV-DNA positivity correlated with tumor stage (P < 0.01) and tumor size (P < 0.05) (Table 2) but not W%W 3t"_ A Jmt FIG. 1. Demonstration of HPV-18 DNA by in silu hybridization in formalin-fixed, paraffin-embedded, hematoxylin-counterstained tissue sections from cervical adenocarcinoma. 35S-labeled HPV-18 or HPV-16 DNA (whole genome) probes were hybridized at high stringency conditions for 48 hours. Autoradiography exposure time was 10 days. HPV-18 specific grains are indicated by large arrows. A. An HPV-18 negative cell is indicated by a small arrow HPV-18-positive cervical adenocarcinoma in a 56-year-old woman (X100). B. HPV-18 positive cervical adenocarcinoma in a 47year-old woman (X100). Vol. 95 No. 5

4 650 ANATOMIC PATHOLOGY TABLE 3. CORRELATION OF HPV DNA FINDINGS WITH FIVE-YEAR SURVIVAL RATES IN 106 PATIENTS WITH CERVICAL ADENOCARCINOMA HPV Finding HPV negative HPV positive HPV 16 HPV 18 HPV 16 and 18 Total No Survival Rate 55 (63) 12(63) 1(50) 9(60) 2(100) 67 (63) with age, ploidy, histologic type, grade (Table 1), tumor proliferation (SPF), or the presence of lymph node metastases (Table 2). The follow-up period for the study population was five years. Thirty-nine patients died during this time, indicating that the overall five-year survival rate was 63% (67 of 106 patients survived). Fifty-five HPV-negative patients (63%) survived for five years. The survival rate for HPVpositive patients also was 63% (12 of 19 patients) (Table 3). Thus, HPV-DNA status had no effect on the survival of patients with either early (Fig. 2) or advanced (Fig. 3) disease. The survival rate was 60% (5 of 15) in HPV 18- positive patients, suggesting specifically that HPV-DNA typing had no effect on survival. DISCUSSION The relative incidence of cervical adenocarcinoma has increased since the early 1970s, whereas the incidence of squamous cell carcinoma in this location has decreased.' 5 Also, the absolute incidence of adenocarcinoma of the uterine cervix seems to be increasing concomitantly with increasing frequency of HPV infections, especially in young women. 16 Most reports have concerned HPV typing in squamous cell carcinomas, adenosquamous carcinomas, and adenocarcinomas of the uterine cervix without discriminating between these different histologic types. The study populations in such analyses have ranged from 5 to 84 patients. 47 "" In our study, the overall prevalence of HPV 16 or HPV 18 DNA in cervical adenocarcinomas was 18%. Prolonged storage of tissue blocks had no effect on DNA positivity. The prevalence in previous reports using in situ hybridization has ranged from 25 to 50% (Table 4). The remarkable variability between different reports probably results from differences in laboratory technique and, perhaps, geographic differences in the distribution of specific HPV types. The exact sensitivity of in situ assays is difficult to determine. DNA is not extracted from the tissue samples, x " so "- 10 -" " 60 - so - 40 " " 10 " HPV*(N 10J80.0X HPV - (N 74) 64.9X and the HPV-DNA copy number varies considerably between specimens, especially in carcinomatous tissues. The estimated limit of the sensitivity of the in situ hybridization method using 35 S-labeled DNA probes is copies per cell. 3 Southern blot and "dot blot" hybridizations are more sensitive, and are capable of detecting copy of HPV-DNA per cell.' Caussy and associates 17 found that the specificity of different hybridization techniques was comparable (87% with Southern blot analysis; 93% with in situ hybridization), but the sensitivity in carcinomatous tissues was 30% lower for the in situ technique than for the Southern blot method. In cervical adenocarcinomas and adenosquamous carcinomas, the mean frequency of HPV 16 and HPV 18 DNA, as detected by in situ hybridization, was 33% (range, 18-50%) (Table 4), whereas the mean frequency with the use of Southern blot detection was 56% (range, 33- H I I Time (months> HPV-(N 13)53.a«HPV (N 9) 44.4X r Time (months) FlG. 2 (.upper). Correlation of five-year survival rates with HPV results in patients with stage I 11 cervical adenocarcinoma. FIG. 3 (lower). Correlation of five-year survival rates with HPV results in patients with stage III IV cervical adenocarcinoma. A.J.C.P.-May 1991

5 LEMINEN ET AL. 651 HPV-DNA in C ' Adenocarcinoma TABLE 4. FREQUENCY OF HPV-16 AND HPV-18 DNA IN ADENOCARCINOMA AND ADENOSQUAMOUS CARCINOMA OF THE UTERINE CERVIX BY IN SITU HYBRIDIZATION HPV-16 HPV-18 Total No. of Reference (Country) Cases King and colleagues 8 (USA) 50 7(14) 18(36) 25 (50) Ostrow and colleagues 9 (USA) 8 1 (12.5) 1 (12.5) 2 (25) Okagaki and colleagues 10 (USA) 40 1 (2.5) 16(40) 17 (42.5) Tase and colleagues'' (USA) 84 9(11) 24(28) 33 (39) Leminen and colleagues (Finland) 106 4(4) 17(16) 19* (18) Total (8) 76(26) 96 (33) * Includes two cases positive for both HPV-16 and HPV %). 718 " 21 The most sensitive technique for detection of HPV-DNA is the polymerase chain reaction (PCR). We evaluated ten patients from our study population with PCR, and nine of them were HPV 18 positive. From those nine cases, three were HPV 18 DNA positive with in situ hybridization. 22 Among all HPV-DNA-positive cases in our study, 88% were HPV 18 positive. This finding is in agreement with results of previous studies in which HPV 18 has predominated (range, 40 to 94%) " HPV 18 infection clearly is associated with cervical adenocarcinoma and may be considered a potential etiologic factor for this neoplasm. However, the pathogenesis of cervical adenocarcinoma is poorly understood in general and might be different from that of squamous cell carcinoma. For example, infectious agents are not thought to play a significant role in the genesis of the former tumor, in contrast to the latter. However, up to 70% of adenocarcinomas in situ (AIS) of the uterine cervix have been reported to contain HPV- DNA. 923 Thus, it would follow that if HPV 18 is an etiologic factor for cervical adenocarcinoma, AIS may be a true precursor of this neoplasm. We found no evidence that the presence of HPV-DNA has any influence on the clinical behavior of cervical adenocarcinoma. If HPV-DNA is amplified within the tumor cells before it becomes detectable by in situ hybridization, our results suggest that HPV 16 and 18 are amplified in the advanced stages of cervical adenocarcinoma. Recently, amplification and overexpression of the oncogene c-myc have been associated with poor prognosis among patients with cervical squamous carcinoma. 24 Our study did show a correlation between HPV-DNA and tumor size and stage, but not with SPF. One might therefore suggest that HPV-DNA has an effect on tumoral invasion rather than on proliferation. The survival-stage unadjusted rates for HPV-negative and HPV-positive patients were equal in our study. This observation contrasts with the results of Barnes and associates. 7 According to their study, HPV 18 was associated with a more aggressive course of cervical carcinoma. However, behavior of the tumors they evaluated was estimated using only the grade of differentiation and the presence of lymph node metastases. Furthermore, no data were provided on whether patients with lymph node metastases had adenocarcinomas or squamous cell carcinomas. Moreover, there was no evidence that the metastatic tissues contained HPV-DNA. In addition to our study, there are others in which no association between HPV-DNA and age, histologic tumor type, grade, lymph node metastases, or survival has been found. 8 In summary, HPV 18 DNA was seen frequently in tissue sections from patients with cervical adenocarcinoma. Considering the fact that the in situ hybridization technique with 35 S-labeled probes is rather insensitive, the presence of HPV-DNA in cervical adenocarcinoma may, in fact, be even greater than our data indicate. On the other hand, the number of copies of the HPV genome may be relatively high in cervical adenocarcinomas. Eighty-eight percent of the HPV-DNA-positive cases in this assessment were HPV 18 positive. The presence of HPV 18 DNA was associated with an advanced stage of disease but had no influence on survival. Our findings do support the hypothesis that HPV 18 may play a role in the pathogenesis of adenocarcinoma of the uterine cervix. Acknowledgment. The authors thank Inkeri Lehtimaki for technical assistance. REFERENCES 1. Koutsky LA. Galloway DA. Holmes KK. Epidemiology of genital human papillomavirus infection. Epidemiol Rev 1988:10: Paavonen J. Koutsky LA. Kiviat N. Cervical neoplasia and other STD-relaled genital and anal neoplasias. In: Holmes KK. Mardh P-A. Sparling PF, Wiesner PJ. eds. Sexually transmitted diseases. New York: McGraw-Hill. 1989: Roman A. Fife KH. Human papillomaviruses: are we ready to type? Clin Microbiol Rev 1989:2: Meanwell CA. Cox MF. Blackledge G, Maitland NJ. HPV 16 DNA in normal and malignant cervical epithelium: implications for aetiology and behaviour of cervical neoplasia. Lancet 1987:1: Munoz N, Bosh X, Kaldo JM. Does human papillomavirus cause cervical carcinoma? The state of epidemiological evidence. Br J Cancer 1988;57: Reeves WC, Brinton LA, Garcia M, et al. Human papillomavirus infection and cervical cancer in Latin America. N Engl J Med 1989;320: Barnes W, Delgado G, Kurman RJ, et al. Possible prognostic significance of human papillomavirus type in cervical cancer. Gynecol Oncol 1988;29: Vol. 95 No. 5

6 652 ANATOMIC PATHOLOGY 8. King LA, Tase T, Twiggs LB. Prognostic significance of the presence of human papillomavirus DNA in patients with invasive carcinoma of the cervix. Cancer 1989;63: Okakagi T, Tase T, Twiggs LB, Carson LF. Histogenesis of cervical adenocarcinoma with reference to human papillomavirus-18 as a carcinogen. J Reprod Med 1989:34: Ostrow RS, Manias DA, Clark BA, Okagaki T, Twiggs LB, Faras AJ. Detection of human papillomavirus DNA in invasive carcinomas of the cervix by in situ hybridization. Cancer Res 1987;47: Tase T, Okagaki T, Clark BA, et al. Human papillomavirus types and localization in adenocarcinoma and adenosquamous carcinoma of the uterine cervix: a study by in situ DNA hybridization. Cancer Res 1988;48: Syrjanen S, Syrjanen K. An improved in situ DNA hybridization protocol for detection of human papillomavirus (HPV) DNA sequences in paraffin-embedded biopsies. J Virol Methods 1986; 14: Leminen A, Paavonen J, Vesterinen E, et al. DNA How cytometric analysis of cervical adenocarcinoma: prognostic significance of DNA ploidy and S-phase fraction. Am J Obstet Gynecol 1990,162: Peto R. Pike MC, Armitage P. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer 1977;35: Leminen A, Paavonen J, Forss M, Wahlstrom T, Vesterinen E. Adenocarcinoma of the uterine cervix. Cancer 1990:65: Schwartz SM, Weiss NS. Increased incidence of adenocarcinoma of the uterine cervix in young women in the United States. Am J Epidemiol 1986:124: Caussy D, Orr W. Daya AD, Roth P, Reeves W, Rawls WE. Evaluation of methods for detecting human papillomavirus deoxyribonucleotide sequences in clinical specimens. J Clin Microbiol 1988;26: Smotkin D, Berek JS, Fu YS, Hacker NF, Major FJ. Wcttstcin FO. Human papillomavirus deoxyribonucleic acid in adenocarcinoma and adenosquamous carcinoma of the uterine cervix. Obstet Gynecol 1989;68: Wilczynski SP. Walker J, Liao S-Y, Bergen S, Berman M. Adenocarcinoma of cervix associated with human papillomavirus. Cancer 1986;62: de Villiers E-M, Schneider A, Gross G, zur Hausen H. Analysis of benign and malignant urogenital tumors for human papillomavirus infection by labelling cellular DNA. Med Microbiol Immunol (Berl) 1986:174: Yoshikawa H, Matsukura T, Yamamoto E. Kawana T, Mizuno M. Yoshiike K. Occurrence of human papillomavirus types 16 and 18 DNA in cervical carcinomas from Japan: age of patients and histological type of carcinomas. Jpn J Cancer Res I985;76: Cone R, Beckmann A, Galloway D, Paavonen J. HPV typing of cervical adenocarcinomas by PCR. J Cell Biochem [Suppl] 1989;13C: Tase T, Okagaki T, Clark BA, Twiggs LB, Ostrow RS. Faras AJ. Human papillomavirus DNA in adenocarcinoma in situ, microinvasive adenocarcinoma of the uterine cervix, and coexisting cervical squamous intraepithelial neoplasia. Int J Gynecol Pathol 1989:8: Riou G, Barrois M, Le MG, George M, Le Doussal V, Haic C. C- myc proto-oncogene expression and prognosis in early carcinoma of the uterine cervix. Lancet 1987;2: AJ.C.P. -May 1991