GYNECOLOGIC
ONCOLOGY
38, 377-382
(1990)
Human lmmunodeficiency MITCHELL MAIMAN, Division
of Gynecologic
Virus Infection and Cervical Neoplasia’
M.D., RACHEL G. FRUCHTER, PH.D., ELI SERUR, M.D., JEAN CLAUDE REMY, M.D., GERALD FEUER, M.D., AND JOHN BOYCE, M.D.
Oncology,
Department at Brooklyn,
of Obstetrics 450 Clarkson
and Gynecology, State University of New Avenue, Brooklyn, New York 11203
York-Health
Science
Center
Received January 23, 1990
To determine the relationship between human immunodeficiency virus (HIV) infection and cervical neoplasia, the characteristics of invasive and preinvasive cervical diseasein 114 patients of known HIV status were assessed.Seven of thirty-seven patients (19%) under age 50 with invasive cervical carcinoma were HIVpositive, including a 16-year-old with stage IIIB disease. HIVpositive patients had more advanced invasive cancer than HIVnegative patients. Diseasepersisted or recurred in all HIV-positive patients compared to 37% of HIV-negative patients. In HIVpositive patients, the median times to recurrence and death were 1 and 10 months, respectively. No HIV-positive patient had HIVrelated symptoms. The mean T4: T8 cell ratio in HIV-positive patients was 0.49, compared to 1.86 in HIV-negative patients. The mean T4 cell count was 362/mm3 in HIV-positive and 775/mm3 in HIV-negative patients. Colposcopic evaluations of the lower genital tract of 77 patients with abnormal smears revealed higher-grade cytology and histology in 25 HIV-positive than in 52 HIV-negative patients. HIV-positive patients had significantly more multifocal/extensive lesions, multisite involvement, perianal involvement, evidence of human papillomavirus (HPV) infection, and associated gynecologic infections than HIVnegative patients. In areas at high risk for HIV infection, we must anticipate a high prevalence of HIV seropositivity in women with invasive cervical cancer. In the HIV-infected, cervical cancer is of advanced stage and responds poorly to therapy. Intraepithelial neoplasia in HIV-positive patients may be of higher grade than in HIV-negative patients, with more extensive involvement of the lower genital tract. 0 1990 Academic PWSS, IIIC. INTRODUCTION Human immunodeficiency health problem of epidemic
virus (HIV) infection is a proportions in New York
City, with over 24,000 cases of acquired immune deficiency syndrome (AIDS) reported. Current estimates are
that 200,000 individuals, or 3-4% of the city’s population, are HIV-positive. In Brooklyn, 20% of AIDS patients are female, compared to 13% in the city as a whole and 9% nationally. Women in Brooklyn have one of the highest HIV seropositivity rates in the nation, and one in 76 newborns in the borough is infected. North-central Brooklyn, by virtue of its low-income population, the high prevalence of intravenous drug users, and the many immigrants from the Caribbean basin, is an area where sexually active women are at particular risk. Neoplasms with unusually high incidence in AIDS patients include Kaposi’s sarcoma and non-Hodgkin’s Bcell lymphoma. Squamous cell carcinomas of the lung, epiglottis, oral cavity, skin, anorectal region, and cervix have also been noted in HIV-infected persons. Cervical neoplasia has many characteristics of a sexually transmitted disease, and therefore, an association with HIV can be anticipated on the basis of common behavioral risk factors. Additionally, since immunosuppressed women are known to be at high risk for lower genital tract neoplasia, immunodeficient HIV-infected women may also be at such high risk. For these reasons, we anticipated an association between HIV infection and cervical neoplasia in central Brooklyn, where the incidence of cervical cancer is above the national average. In 1983 we recognized AIDS as a cause of death in our patients with lower genital neoplasia, and noted that increasing numbers of women with HIV disease had been referred for evaluation of abnormal Pap smears. This study was undertaken to compare the characteristics of cervical neoplasia in HIV-seropositive and HIVseronegative women. MATERIAL AND METHODS
’ Presented at the annual meeting of the Society of Gynecologic Oncologists San Francisco, February 4-7, 1990. This work was supported in part by American Cancer Society Clinical Oncology Career Development Award 88-139.
The subjects of this investigation were 37 women under age 50 with histologically confirmed invasive carci377
0090-8258/90 %I.50 Copyright 0 1990by AcademicPress,Inc. All rights of reproduction in any form reserved.
378
MAIMAN
TABLE 1 HIV-Infected Patientswith Invasive Cervical Carcinoma Patient
Age
IVDU”
Clinical stage
Surgical stage
1 2 3 4 5 6 I
16 33 34 40 44 45 47
No Yes Yes Yes No No No
IIIB IIIB IIB IB, bulky IB, bulky IV IV
IIIB IIIB
-
IV IV IV
Status Persistent diseae Dead from disease Dead from disease Dead from disease Persistent disease Persistent disease Dead from disease
ET AL.
method. Individual pretest and post-test counseling was provided to patients, using standards prescribed by the New York City Department of Health. Enumeration of T4 (helper/inducer) and T8 (suppressor/cytotoxic) cells was performed using commercial monoclonal and polyclonal antibodies directed against the T-cell surface markers (the Orthomune system). The differences between proportions were estimated using the Fisher exact test and x2 test with continuity correction. Survival rates were computed by the KaplanMeier product-limit method.
” Intravenous drug use
RESULTS noma of the cervix and 77 women referred for evaluation of abnormal Papanicolaou (Pap) smears at the Kings County Hospital Center and the University Hospital of Brooklyn between 1986 and 1989. Thirty-seven patients under the age of 50 with invasive disease were tested for HIV after the diagnosis of cervical carcinoma. Patients over age 50 were not tested because of the expected low rate of HIV positivity. Of the 37 patients, 33 had squamous cell carcinoma, 3 adenocarcinoma, and 1 clear cell carcinoma. All patients were clinically staged before treatment, according to the International Federation of Gynecology and Obstetrics (FIGO) classification. Twenty-five of the thirty-seven patients had initial surgery, either radical hysterectomy with lymphadenectomy or exploratory staging laparotomy, and nine were upstaged based on histologic findings. Advanced stage was defined as bulky stage IB or higher. Treatment was planned on the basis of clinical and pathologic findings without reference to HIV status. All patients were interviewed about their medical histories and sexual and substance abuse behaviors. The mean follow-up was 20 months, with a maximum of 30 months. In addition, 77 women referred for evaluation of abnormal Pap smears were studied including 18 referred with known HIV disease. Of 59 tested nonselectively for HIV after referral, 7 were HIV-positive and 52 HIVnegative. HIV-positive patients were classified as having asymptomatic infections, AIDS-related complex (ARC), or AIDS. All women had colposcopic evaluation of the lower genital tract and colposcopically directed biopsies. Cervical lesions were categorized as multifocal if two or more distinct sites separated by normal tissue were visualized on the cervix, or as extensive if two or more quadrants of the cervix were involved. Multisite involvement, perianal involvement, and the presence of human papillomavirus (HPV) and other gynecologic infections were noted. Blood samples were analyzed for HIV antibodies utilizing the enzyme-linked immunosorbent assay (ELISA). Seropositivity was confirmed by the Western blot
Seven of thirty-seven patients (19%) with invasive cervical carcinoma were HIV-positive; all seven were asymptomatic with regard to HIV infection. HIV-positive and HIV-negative patients were compared on multiple sociodemographic and behavioral variables, including age, birthplace, transfusion history, smoking, alcohol and intravenous drug use, parity, contraceptive use, age at first coitus, and number of sexual partners. The only significant difference beween the groups was in intravenous drug use, with HIV-positive patients more likely to have used intravenous drugs than HIV-negative patients. Specifics of the seven HIV-positive patients are presented in Table 1. The 16-year-old patient with stage IIIB disease had a history of sexual abuse at age 8, after which diffuse vulvar condylomata developed and persisted over the next 8 years. She had one other partner at age 15, and no history of intravenous drug use or blood transfusion. All seven HIV-positive patients had squamous cell carcinomas, compared to 87% of the HIV-negative patients. No HIV-positive patient presented with early clinical disease (defined as IA and nonbulky IB), compared to one-half of the HIV-negative patients (P < 0.05) (Table 2). Seventy-one percent of HIV-positive patients had clinical stage II or more, compared to 37% of HIV-negative patients. When patients were upstaged based on surgical-pathologic findings, all but one HIV-positive patient had stage II or more, including one patient with grossly metastatic paraaortic nodes, one with metastatic scalene nodes, and one with malignant ascites. Four HIV-positive patients and ten HIV-negative patients had pretreatment T-cell studies (Fig. 1). The mean T4:T8 cell ratio was 0.49 in the HIV-positive group, compared to 1.86 in the HIV-negative group (1.40 in HIV-negative patients with advanced disease and 1.98 in HIV-negative patients with early disease). The mean T4 cell count was 362 cells/mm3 in HIV-positive patients compared to 775 cells/mm3 in HIV-negative patients. All but one HIV-positive patient completed primary
HIV AND CERVICAL
379
NEOPLASIA
TABLE 2 Stage at Diagnosis of Invasive Carcinoma of the Cervix by HIV Status HIV-positive Stage
N
IA/IB IB, bulky II-IV Total
0 2 5 7
Percent
N
HIV-negative
..Percent
A. Clinical stage 0 15 29 4 71 11 100 30
50 13 31 100
IA/IB vs IS bulky: P < .05 in xz test; I dJ IA/IB IB, bulky II-IV Total
0 I 6 7
B. Surgical-pathologic stage 0 II 14 3 86 I6 100 30
Months After Diagnosis FIG. 2. Recurrence-free survival by HIV status and stage. 31 IO 53 100
patients gave a history of intravenous drug abuse, compared to 2% of the HIV-negative group (P < 0.05). IA/lB vs IB bulky-IV: P < 0.10 in ,$ test; I & The cytologic and histologic findings are shown in Table 3. More than half the HIV-positive patients had highgrade cytologic abnormalities (CIN 11 or worse) comtreatment and were followed from 10 to 30 months. All pared to less than one-quarter of HIV-negative patients six treated HIV-positive patients had persistent or re- (P < 0.05). This relationship was confirmed on biopsy, current disease. compared to 58% of HIV-negative pa- when more than half of HIV-positive patients had CIN II-III lesions, compared to less than one-third the HIVtients with advanced cancer and none of the HIV-negative patients with early disease. The median time to negative group. The majority of HIV-positive patients had multifocal recurrence was 1 month in HIV-positive patients and 9 months in the HIV-negative patients with advanced dis- or extensive cervical lesions, in contrast to the HIVease (Fig. 2). The median time to death was 10 months negative patients of whom 63% had focal cervical lesions in HIV-positive and 23 months in HIV-negative patients (Table 4). In addition, the neoplasia involved multiple (Fig. 3). All four HIV-positive patients who died were sites of the lower genital tract significantly more often asymptomatic with regard to specific HIV disease and in the HIV-positive patients than in the HIV-negative patients. Perianal disease was found in almost half the their deaths were recorded as due to cervical cancer. HIV-positive patients with multiple-site involvement, but Among the 77 women with abnormal smears and in none of the HIV-negative patients. known HIV status, 25 were HIV-positive, of whom 16 Overall, there was either cytologic or histologic eviwere asymptomatic, 1 had ARC, and 8 had AIDS. The dence of HPV infection in 97% of HIV-positive patients, mean age of both HIV-positive and HIV-negative women compared to 56% of HIV-negative patients (P < 0.01). was 28.6 years. Thirty-two percent of the HIV-positive In addition, almost half the HIV-positive patients had other gynecologic infections, compared to one-fifth of the HIV-negative group.
T4:T8
1.0 1
STAGE
Advanced
HIV STATUS
Positive
Advanced
Early
Negative
FIG. 1. T4:TS cell ratio by HIV status.
0 0
I,,,,,.,,,,,,,~,,~,~ 4 6
12
16
20
24
28
Months After Diagnosis FIG. 3. Survival by HIV status and stage.
380
MAIMAN
TABLE 3 HIV Infection and Intraepithelial Neoplasia: Cytology and Histology by HIV Status
Cytology HPV only CIN I CIN II-III Invasive carcinoma Total
HIV-positive
HIV-negative
N
Percent
N
Percent
2 9 13 I 25
8 36 52 4 100
5 35 12 0 52
10 67 23 0 loo
HPV vs CIN I vs CIN II +: p < 0.05 in x2 test; 2 df Histology No pathology HPV only CIN I CIN II-III Total
0 7 4 14 25
0 28 16 56 100
4 15 17 16 52
7 29 33 31 loo
ET AL.
TABLE 4 HIV Infection and Intraepithelial Neoplasia: Multifocality and Multiple-Site Involvement by HIV Status HIV-positive
Cervical Lesion Focal Multifocal/extensive Total Genital Sites Single site Multiple sites Total Perianal involvement
HIV-negative
N
Percent
N
Percent Significance
5 20 25
20 80 100
33 19 52
63 37 100
14 11 25 5/l 1
54 46 100 45
47 5 52 O/5
90 10 100 0
P < 0.01
P < 0.01 N.S.
The potential association between HIV-associated immunodeficiency and cervical neoplasia is, however, confounded by the fact that cervical neoplasia and HIV are sCIN I vs CIN II-III: p < 0.05 in x2 test: I df both associated with high-risk sexual behavior and with Note. All categories of neoplasia may or may not include HPV. several genital infections including HPV. Intravenous drug use is associated with both HIV and high-risk sexual Although there was a trend toward higher-grade lebehavior. sions in AIDS/ARC patients compared to asymptomatic Given the multiple common risk factors, our finding HIV-positive patients, there were no statistically signifof a 19% prevalence of asymptomatic HIV infection in icant differences between the two groups (Table 5). young patients with invasive cervical carcinoma was not surprising, but extremely disturbing. We have previously DISCUSSION reported a HIV seropositivity rate of 11% in women Immunodeficient individuals are at increased risk for attending a colposcopy clinic for evaluation of abnormal developing certain malignancies, including Kaposi’s sar- Pap smears in the same high-risk population [8]. These coma, non-Hodgkin’s lymphoma, and squamous cell car- HIV prevalence rates of 19 and 11% in patients with cinomas of the skin and genitalia. This has been dem- neoplasia are significantly higher than those reported in onstrated in congenitally immunodeficient individuals [ 11 obstetrical populations in the same high-risk areas [91. There have been several reports of cytologic abnorand in iatrogenic immunodeficiency, especially organ transplant patients on immunosuppressive therapy [23. Data from the Denver Transplant Tumor Registry [3] TABLE 5 and from Schneider et al. [41 demonstrate the increased HIV Infection and Intraepithelial Neoplasia: Comparison of risk for anogenital neoplasia in women. Halpert et al. Cytology and Histology, in Asymptomatic and Symptomatic HIV[5] showed a 9-fold increased risk of cervical neoplasia Infected Women and a 17-fold increased risk of HPV infection in renal ARC/AIDS Asymptomatic transplant patients on immunosuppressive drugs. Percent More recently, AIDS has been demonstrated to be N Percent N associated with Kaposi’s sarcoma and non-Hodgkin’s lymphomas and there are numerous reports of squamous Cytology 0 HPV only 2 13 0 cell neoplasia developing in HIV-infected persons [6]. It 33 CIN I 38 3 6 is probable that, as in congenitally and iatrogenically 67 CIN II-III 50 6 8 loo Total 9 16 100 immunodeficient populations, the immunodeficiency resulting from HIV infection allows an oncogenic virus to Histology 11 HPV only 6 37 1 flourish [7], and inhibits the body’s natural immunologic 11 CIN I 3 I9 1 defense mechanisms that suppress the development of 78 CIN II-III I 7 44 neoplasia. In anogenital neoplasia in women, there is 100 9 Total 16 100 evidence that the oncogenic virus is one or more type of HPV. Note. No associations are statistically significant.
HIV AND CERVICAL
malities in HIV-positive patients. Provencher et al. [lo] found only 37% “normal” Pap smears in HIV-positive patients, compared to 95% in HIV-negative patients, with a higher rate of abnormalities after seroconversion than before seroconversion. In their study however, “abnormal” smears included hyperkeratosis, parakeratosis, Trichomonas, Herpes, and atypia. Schrager et al. [ 1I] found a 3 1% increase in abnormal smears (defined as atypical or worse) among 35 HIV-positive patients in New York City. Our findings demonstrate higher-grade dysplastic lesions, not only on cytology, but also in the colposcopically directed biopsies of HIV-positive women compared to HIV-negative. We also consistently found more cytologic and histologic evidence of HPV in the HIV-infected women and high rates of other genital infections. A high prevalence of HIV infection in women with cervical neoplasia and even a greater severity of neoplasia in HIV-infected women could be explained by the common risk factors of sexual behavior and intravenous drug use. However, we also found that HIVpositive patients had higher rates of multifocal cervical disease, multiple site involvement, and perianal involvement, characteristic of the lower genital tract field effect in iatrogenically immunosuppressed women [ 121. This suggests that immunodeficiency may make an independent contribution to the development of genital neoplasia in HIV-infected women. In addition, the lack of significant differences between asymptomatic HIV-positive patients and those with AIDS in the severity of neoplasia suggests that overt signs of immunodeficiency are not necessary to manifest the consequences of HIV on lower genital tract neoplasia. In HIV-infected patients with invasive cervical cancer, young age combined with advanced stage of several patients stood out as sentinel indicators of an alarming problem. The 16-year-old patient with stage IIIB disease is the youngest patient ever reported with advanced squamous cell cervical carcinoma known to the authors, but her young age is less anomalous in view of her 8year history of HPV infection. Kling and Buchsbaum [I31 reported eight patients under 21 years of age, the youngest of whom was 18 with stage IA disease. In our 40-year tumor registry, the only invasive cervical carcinomas in women under 20 years have been adenocarcinemas, several associated with maternal diethylstilbestrol use. The more advanced stage of presentation, the higher recurrence and death rates, and the shorter interval to recurrence and death in asymptomatic HIV-positive compared to HIV-negative patients have many implications. The findings suggest that occult HIV-related immunodeficiency may contribute heavily to the natural history of cervical carcinoma, and that HIV-positive patients need not demonstrate other signs of immuno-
381
NEOPLASIA
suppression, such as opportunistic infections, for their neoplasia to be adversely affected by HIV. Our finding that our HIV-infected patients, although asymptomatic with respect to HIV disease, had depressed T-cell function suggests that the first indication of HIV positivity may be the advanced stage and rapid progression of cervical cancer. We, as others [14,1.5], have found poorer survival in younger women with cervical cancer than in older women and believe the contribution of latent HIV infection to this trend in high-risk populations warrants investigation. We certainly recommend that young patients with cervical cancer in areas of high HIV prevalence be tested for the virus. Kaposi’s sarcoma in AIDS patients differs from that seen in transplant [16] and other patients, in that lymph node involvement, systemic disease, and fatalities are much more frequent in the AIDS patients. Our data suggest a similar pattern may exist in squamous cell carcinoma of the cervix in the HIV-infected, and imply that unique treatment approaches are needed to combat the more advanced and aggressive form of disease resulting from the interaction of these two disease processes. A multidisciplinary approach to treatment is essential since conventional radiation therapy and chemotherapy may be inadequate for invasive disease, as may standard therapies for preinvasive disease. Laboratory services for monitoring the immune system and consultation with medical and infectious disease specialists on the treatment of the HIV disease are essential. Psychological and counseling services that address the needs of patients with two life-threatening disease processes are needed. Decisions concerning treatment priorities must be made: Should the HIV infection be treated first? Should it be treated at all? Or, should oncologic therapy be instituted initially? The association between HIV infection and cervical neoplasia is undoubtedly complex. As new drugs that suppress, but do not destroy, HIV are developed with wider indications for use, life will be prolonged and we expect to see cervical neoplasia in HIV-infected women with greater frequency. We believe that careful study of this disease interaction will provide invaluable insight into carcinogenesis, immunology, and neoplasia, and that development of research and treatment protocols should proceed in a scientific and compassionate manner. REFERENCES 1.
Gatti, R. A., and Gard, R. A. Occurrence of malignancy in immunodeficient diseases, Cancer 28, 89 (1971).
2.
Penn,I. Malignanciesassociatedwith immunosuppressive or cy-
3.
Penn,I. Cancersof the anogenitalregionin renaltransplantre-
totoxic therapy, Surgery 83, 492-502 (1978). cipients, Cancer 58, 611 (1986).
382
MAIMAN
4. Schneider, V., Kay, S., and Lee, H. M. Immunosuppression as a high risk factor in the development of condyloma accuminatum and squamous neoplasia of the cervix, Acta Cytol. 97, 220-224 (1983).
5. Halpert, R., Fruchter, R. G., Sedlis, A., Butt, K., Boyce, J. G., and Sillman, F. H. Human papillomavirus and lower genital neoplasia in renal transplant patients, Obstet. Gynecol. 68, 251-258 (1986). 6. Fahey, J. C., and Emmanuel, 0. Acquired immune deficiency syndrome (AIDS) and neoplasia, Amer. J. Pediatr. Hematol./Oncol. 9, 193-195 (1987). 7. Purtilo, D. T., Manolov, G., Manolova, S., Harada, S., and Lip-
scomb, H. Squamous cell carcinoma, Kaposi’s sarcoma and Burkitt’s lymphoma are consequences of impaired immune surveillance of ubiquitous virus in acquired immune deficiency syndrome, allograft recipients and tropical African patients, IARC Pub. 83,749770 (1984).
8. Maiman, M., Fruchter, R. G., Serur, E., and Boyce, J. G. Prevalence of human immunodeficiency virus in a colposcopy clinic, J. Amer. Med. Assoc. 260, 2214-2215 (1988). 9. Shapiro, C. N., Schulz, S. L., Lee, N. C., and Dondero, T. J.
Review of human immunodeficiency virus in women in the United States, Obstet. Gynecol. 74, 800-808 (1989).
ET AL. 10. Provencher, D., Valme, B., Averette, H. E., Ganjei, P., Donata, D., Penalver, M., and Sevin, B. U. HIV status and positive Papanicolau screening: Identification of a high-risk population, Gynecol. Oncol. 31, 184-188 (1988). Il. Schrager, L., and Friedland, G. Increased risk of cervical and/or vaginal squamous atypia in women infected with HIV, in Proceedings, dejiciency
Third International Syndrome (AIDS),
Conference on Acquired ImmunoWashington, DC, p. 86 (1987).
12. Sillman, F. H., Stanek, A., and Sedlis, A. The relationship between human papillomavirus and lower genital neoplasia in immunosuppressed women, Amer. J. Obstet. Gynecol. 150, 300 (1984). 13. Kling, T., and Buchsbaum, H. Cervical carcinoma in women under twenty-one years of age, Obstet. Gynecol. 42, 205-207 (1973). 14. Chapman, G. W., Abreo, F., and Thompson, H. E. Carcinoma of the cervix in young females (35 years and younger), Gyecol. Oncol. 31, 430-434 (1988). 15. Fedorkow,
D. M., Robertson, I., Duggan, M. A., McGregor, S. E., and Stuart, G. C. E. Invasive squamous cell carcinoma of the cervix in women less than 35 years old. Recurrent versus nonrecurrent disease, Amer. .I. Obstet. Gynecol. 158, 307-311 (1988). 16. Penn, I. Cancer is a complication of severe immunosuppression, Surg. Gynecol.
Obstet. 162, 603-609 (1986).
ONCOLOGY
38, 377-382
(1990)
Human lmmunodeficiency MITCHELL MAIMAN, Division
of Gynecologic
Virus Infection and Cervical Neoplasia’
M.D., RACHEL G. FRUCHTER, PH.D., ELI SERUR, M.D., JEAN CLAUDE REMY, M.D., GERALD FEUER, M.D., AND JOHN BOYCE, M.D.
Oncology,
Department at Brooklyn,
of Obstetrics 450 Clarkson
and Gynecology, State University of New Avenue, Brooklyn, New York 11203
York-Health
Science
Center
Received January 23, 1990
To determine the relationship between human immunodeficiency virus (HIV) infection and cervical neoplasia, the characteristics of invasive and preinvasive cervical diseasein 114 patients of known HIV status were assessed.Seven of thirty-seven patients (19%) under age 50 with invasive cervical carcinoma were HIVpositive, including a 16-year-old with stage IIIB disease. HIVpositive patients had more advanced invasive cancer than HIVnegative patients. Diseasepersisted or recurred in all HIV-positive patients compared to 37% of HIV-negative patients. In HIVpositive patients, the median times to recurrence and death were 1 and 10 months, respectively. No HIV-positive patient had HIVrelated symptoms. The mean T4: T8 cell ratio in HIV-positive patients was 0.49, compared to 1.86 in HIV-negative patients. The mean T4 cell count was 362/mm3 in HIV-positive and 775/mm3 in HIV-negative patients. Colposcopic evaluations of the lower genital tract of 77 patients with abnormal smears revealed higher-grade cytology and histology in 25 HIV-positive than in 52 HIV-negative patients. HIV-positive patients had significantly more multifocal/extensive lesions, multisite involvement, perianal involvement, evidence of human papillomavirus (HPV) infection, and associated gynecologic infections than HIVnegative patients. In areas at high risk for HIV infection, we must anticipate a high prevalence of HIV seropositivity in women with invasive cervical cancer. In the HIV-infected, cervical cancer is of advanced stage and responds poorly to therapy. Intraepithelial neoplasia in HIV-positive patients may be of higher grade than in HIV-negative patients, with more extensive involvement of the lower genital tract. 0 1990 Academic PWSS, IIIC. INTRODUCTION Human immunodeficiency health problem of epidemic
virus (HIV) infection is a proportions in New York
City, with over 24,000 cases of acquired immune deficiency syndrome (AIDS) reported. Current estimates are
that 200,000 individuals, or 3-4% of the city’s population, are HIV-positive. In Brooklyn, 20% of AIDS patients are female, compared to 13% in the city as a whole and 9% nationally. Women in Brooklyn have one of the highest HIV seropositivity rates in the nation, and one in 76 newborns in the borough is infected. North-central Brooklyn, by virtue of its low-income population, the high prevalence of intravenous drug users, and the many immigrants from the Caribbean basin, is an area where sexually active women are at particular risk. Neoplasms with unusually high incidence in AIDS patients include Kaposi’s sarcoma and non-Hodgkin’s Bcell lymphoma. Squamous cell carcinomas of the lung, epiglottis, oral cavity, skin, anorectal region, and cervix have also been noted in HIV-infected persons. Cervical neoplasia has many characteristics of a sexually transmitted disease, and therefore, an association with HIV can be anticipated on the basis of common behavioral risk factors. Additionally, since immunosuppressed women are known to be at high risk for lower genital tract neoplasia, immunodeficient HIV-infected women may also be at such high risk. For these reasons, we anticipated an association between HIV infection and cervical neoplasia in central Brooklyn, where the incidence of cervical cancer is above the national average. In 1983 we recognized AIDS as a cause of death in our patients with lower genital neoplasia, and noted that increasing numbers of women with HIV disease had been referred for evaluation of abnormal Pap smears. This study was undertaken to compare the characteristics of cervical neoplasia in HIV-seropositive and HIVseronegative women. MATERIAL AND METHODS
’ Presented at the annual meeting of the Society of Gynecologic Oncologists San Francisco, February 4-7, 1990. This work was supported in part by American Cancer Society Clinical Oncology Career Development Award 88-139.
The subjects of this investigation were 37 women under age 50 with histologically confirmed invasive carci377
0090-8258/90 %I.50 Copyright 0 1990by AcademicPress,Inc. All rights of reproduction in any form reserved.
378
MAIMAN
TABLE 1 HIV-Infected Patientswith Invasive Cervical Carcinoma Patient
Age
IVDU”
Clinical stage
Surgical stage
1 2 3 4 5 6 I
16 33 34 40 44 45 47
No Yes Yes Yes No No No
IIIB IIIB IIB IB, bulky IB, bulky IV IV
IIIB IIIB
-
IV IV IV
Status Persistent diseae Dead from disease Dead from disease Dead from disease Persistent disease Persistent disease Dead from disease
ET AL.
method. Individual pretest and post-test counseling was provided to patients, using standards prescribed by the New York City Department of Health. Enumeration of T4 (helper/inducer) and T8 (suppressor/cytotoxic) cells was performed using commercial monoclonal and polyclonal antibodies directed against the T-cell surface markers (the Orthomune system). The differences between proportions were estimated using the Fisher exact test and x2 test with continuity correction. Survival rates were computed by the KaplanMeier product-limit method.
” Intravenous drug use
RESULTS noma of the cervix and 77 women referred for evaluation of abnormal Papanicolaou (Pap) smears at the Kings County Hospital Center and the University Hospital of Brooklyn between 1986 and 1989. Thirty-seven patients under the age of 50 with invasive disease were tested for HIV after the diagnosis of cervical carcinoma. Patients over age 50 were not tested because of the expected low rate of HIV positivity. Of the 37 patients, 33 had squamous cell carcinoma, 3 adenocarcinoma, and 1 clear cell carcinoma. All patients were clinically staged before treatment, according to the International Federation of Gynecology and Obstetrics (FIGO) classification. Twenty-five of the thirty-seven patients had initial surgery, either radical hysterectomy with lymphadenectomy or exploratory staging laparotomy, and nine were upstaged based on histologic findings. Advanced stage was defined as bulky stage IB or higher. Treatment was planned on the basis of clinical and pathologic findings without reference to HIV status. All patients were interviewed about their medical histories and sexual and substance abuse behaviors. The mean follow-up was 20 months, with a maximum of 30 months. In addition, 77 women referred for evaluation of abnormal Pap smears were studied including 18 referred with known HIV disease. Of 59 tested nonselectively for HIV after referral, 7 were HIV-positive and 52 HIVnegative. HIV-positive patients were classified as having asymptomatic infections, AIDS-related complex (ARC), or AIDS. All women had colposcopic evaluation of the lower genital tract and colposcopically directed biopsies. Cervical lesions were categorized as multifocal if two or more distinct sites separated by normal tissue were visualized on the cervix, or as extensive if two or more quadrants of the cervix were involved. Multisite involvement, perianal involvement, and the presence of human papillomavirus (HPV) and other gynecologic infections were noted. Blood samples were analyzed for HIV antibodies utilizing the enzyme-linked immunosorbent assay (ELISA). Seropositivity was confirmed by the Western blot
Seven of thirty-seven patients (19%) with invasive cervical carcinoma were HIV-positive; all seven were asymptomatic with regard to HIV infection. HIV-positive and HIV-negative patients were compared on multiple sociodemographic and behavioral variables, including age, birthplace, transfusion history, smoking, alcohol and intravenous drug use, parity, contraceptive use, age at first coitus, and number of sexual partners. The only significant difference beween the groups was in intravenous drug use, with HIV-positive patients more likely to have used intravenous drugs than HIV-negative patients. Specifics of the seven HIV-positive patients are presented in Table 1. The 16-year-old patient with stage IIIB disease had a history of sexual abuse at age 8, after which diffuse vulvar condylomata developed and persisted over the next 8 years. She had one other partner at age 15, and no history of intravenous drug use or blood transfusion. All seven HIV-positive patients had squamous cell carcinomas, compared to 87% of the HIV-negative patients. No HIV-positive patient presented with early clinical disease (defined as IA and nonbulky IB), compared to one-half of the HIV-negative patients (P < 0.05) (Table 2). Seventy-one percent of HIV-positive patients had clinical stage II or more, compared to 37% of HIV-negative patients. When patients were upstaged based on surgical-pathologic findings, all but one HIV-positive patient had stage II or more, including one patient with grossly metastatic paraaortic nodes, one with metastatic scalene nodes, and one with malignant ascites. Four HIV-positive patients and ten HIV-negative patients had pretreatment T-cell studies (Fig. 1). The mean T4:T8 cell ratio was 0.49 in the HIV-positive group, compared to 1.86 in the HIV-negative group (1.40 in HIV-negative patients with advanced disease and 1.98 in HIV-negative patients with early disease). The mean T4 cell count was 362 cells/mm3 in HIV-positive patients compared to 775 cells/mm3 in HIV-negative patients. All but one HIV-positive patient completed primary
HIV AND CERVICAL
379
NEOPLASIA
TABLE 2 Stage at Diagnosis of Invasive Carcinoma of the Cervix by HIV Status HIV-positive Stage
N
IA/IB IB, bulky II-IV Total
0 2 5 7
Percent
N
HIV-negative
..Percent
A. Clinical stage 0 15 29 4 71 11 100 30
50 13 31 100
IA/IB vs IS bulky: P < .05 in xz test; I dJ IA/IB IB, bulky II-IV Total
0 I 6 7
B. Surgical-pathologic stage 0 II 14 3 86 I6 100 30
Months After Diagnosis FIG. 2. Recurrence-free survival by HIV status and stage. 31 IO 53 100
patients gave a history of intravenous drug abuse, compared to 2% of the HIV-negative group (P < 0.05). IA/lB vs IB bulky-IV: P < 0.10 in ,$ test; I & The cytologic and histologic findings are shown in Table 3. More than half the HIV-positive patients had highgrade cytologic abnormalities (CIN 11 or worse) comtreatment and were followed from 10 to 30 months. All pared to less than one-quarter of HIV-negative patients six treated HIV-positive patients had persistent or re- (P < 0.05). This relationship was confirmed on biopsy, current disease. compared to 58% of HIV-negative pa- when more than half of HIV-positive patients had CIN II-III lesions, compared to less than one-third the HIVtients with advanced cancer and none of the HIV-negative patients with early disease. The median time to negative group. The majority of HIV-positive patients had multifocal recurrence was 1 month in HIV-positive patients and 9 months in the HIV-negative patients with advanced dis- or extensive cervical lesions, in contrast to the HIVease (Fig. 2). The median time to death was 10 months negative patients of whom 63% had focal cervical lesions in HIV-positive and 23 months in HIV-negative patients (Table 4). In addition, the neoplasia involved multiple (Fig. 3). All four HIV-positive patients who died were sites of the lower genital tract significantly more often asymptomatic with regard to specific HIV disease and in the HIV-positive patients than in the HIV-negative patients. Perianal disease was found in almost half the their deaths were recorded as due to cervical cancer. HIV-positive patients with multiple-site involvement, but Among the 77 women with abnormal smears and in none of the HIV-negative patients. known HIV status, 25 were HIV-positive, of whom 16 Overall, there was either cytologic or histologic eviwere asymptomatic, 1 had ARC, and 8 had AIDS. The dence of HPV infection in 97% of HIV-positive patients, mean age of both HIV-positive and HIV-negative women compared to 56% of HIV-negative patients (P < 0.01). was 28.6 years. Thirty-two percent of the HIV-positive In addition, almost half the HIV-positive patients had other gynecologic infections, compared to one-fifth of the HIV-negative group.
T4:T8
1.0 1
STAGE
Advanced
HIV STATUS
Positive
Advanced
Early
Negative
FIG. 1. T4:TS cell ratio by HIV status.
0 0
I,,,,,.,,,,,,,~,,~,~ 4 6
12
16
20
24
28
Months After Diagnosis FIG. 3. Survival by HIV status and stage.
380
MAIMAN
TABLE 3 HIV Infection and Intraepithelial Neoplasia: Cytology and Histology by HIV Status
Cytology HPV only CIN I CIN II-III Invasive carcinoma Total
HIV-positive
HIV-negative
N
Percent
N
Percent
2 9 13 I 25
8 36 52 4 100
5 35 12 0 52
10 67 23 0 loo
HPV vs CIN I vs CIN II +: p < 0.05 in x2 test; 2 df Histology No pathology HPV only CIN I CIN II-III Total
0 7 4 14 25
0 28 16 56 100
4 15 17 16 52
7 29 33 31 loo
ET AL.
TABLE 4 HIV Infection and Intraepithelial Neoplasia: Multifocality and Multiple-Site Involvement by HIV Status HIV-positive
Cervical Lesion Focal Multifocal/extensive Total Genital Sites Single site Multiple sites Total Perianal involvement
HIV-negative
N
Percent
N
Percent Significance
5 20 25
20 80 100
33 19 52
63 37 100
14 11 25 5/l 1
54 46 100 45
47 5 52 O/5
90 10 100 0
P < 0.01
P < 0.01 N.S.
The potential association between HIV-associated immunodeficiency and cervical neoplasia is, however, confounded by the fact that cervical neoplasia and HIV are sCIN I vs CIN II-III: p < 0.05 in x2 test: I df both associated with high-risk sexual behavior and with Note. All categories of neoplasia may or may not include HPV. several genital infections including HPV. Intravenous drug use is associated with both HIV and high-risk sexual Although there was a trend toward higher-grade lebehavior. sions in AIDS/ARC patients compared to asymptomatic Given the multiple common risk factors, our finding HIV-positive patients, there were no statistically signifof a 19% prevalence of asymptomatic HIV infection in icant differences between the two groups (Table 5). young patients with invasive cervical carcinoma was not surprising, but extremely disturbing. We have previously DISCUSSION reported a HIV seropositivity rate of 11% in women Immunodeficient individuals are at increased risk for attending a colposcopy clinic for evaluation of abnormal developing certain malignancies, including Kaposi’s sar- Pap smears in the same high-risk population [8]. These coma, non-Hodgkin’s lymphoma, and squamous cell car- HIV prevalence rates of 19 and 11% in patients with cinomas of the skin and genitalia. This has been dem- neoplasia are significantly higher than those reported in onstrated in congenitally immunodeficient individuals [ 11 obstetrical populations in the same high-risk areas [91. There have been several reports of cytologic abnorand in iatrogenic immunodeficiency, especially organ transplant patients on immunosuppressive therapy [23. Data from the Denver Transplant Tumor Registry [3] TABLE 5 and from Schneider et al. [41 demonstrate the increased HIV Infection and Intraepithelial Neoplasia: Comparison of risk for anogenital neoplasia in women. Halpert et al. Cytology and Histology, in Asymptomatic and Symptomatic HIV[5] showed a 9-fold increased risk of cervical neoplasia Infected Women and a 17-fold increased risk of HPV infection in renal ARC/AIDS Asymptomatic transplant patients on immunosuppressive drugs. Percent More recently, AIDS has been demonstrated to be N Percent N associated with Kaposi’s sarcoma and non-Hodgkin’s lymphomas and there are numerous reports of squamous Cytology 0 HPV only 2 13 0 cell neoplasia developing in HIV-infected persons [6]. It 33 CIN I 38 3 6 is probable that, as in congenitally and iatrogenically 67 CIN II-III 50 6 8 loo Total 9 16 100 immunodeficient populations, the immunodeficiency resulting from HIV infection allows an oncogenic virus to Histology 11 HPV only 6 37 1 flourish [7], and inhibits the body’s natural immunologic 11 CIN I 3 I9 1 defense mechanisms that suppress the development of 78 CIN II-III I 7 44 neoplasia. In anogenital neoplasia in women, there is 100 9 Total 16 100 evidence that the oncogenic virus is one or more type of HPV. Note. No associations are statistically significant.
HIV AND CERVICAL
malities in HIV-positive patients. Provencher et al. [lo] found only 37% “normal” Pap smears in HIV-positive patients, compared to 95% in HIV-negative patients, with a higher rate of abnormalities after seroconversion than before seroconversion. In their study however, “abnormal” smears included hyperkeratosis, parakeratosis, Trichomonas, Herpes, and atypia. Schrager et al. [ 1I] found a 3 1% increase in abnormal smears (defined as atypical or worse) among 35 HIV-positive patients in New York City. Our findings demonstrate higher-grade dysplastic lesions, not only on cytology, but also in the colposcopically directed biopsies of HIV-positive women compared to HIV-negative. We also consistently found more cytologic and histologic evidence of HPV in the HIV-infected women and high rates of other genital infections. A high prevalence of HIV infection in women with cervical neoplasia and even a greater severity of neoplasia in HIV-infected women could be explained by the common risk factors of sexual behavior and intravenous drug use. However, we also found that HIVpositive patients had higher rates of multifocal cervical disease, multiple site involvement, and perianal involvement, characteristic of the lower genital tract field effect in iatrogenically immunosuppressed women [ 121. This suggests that immunodeficiency may make an independent contribution to the development of genital neoplasia in HIV-infected women. In addition, the lack of significant differences between asymptomatic HIV-positive patients and those with AIDS in the severity of neoplasia suggests that overt signs of immunodeficiency are not necessary to manifest the consequences of HIV on lower genital tract neoplasia. In HIV-infected patients with invasive cervical cancer, young age combined with advanced stage of several patients stood out as sentinel indicators of an alarming problem. The 16-year-old patient with stage IIIB disease is the youngest patient ever reported with advanced squamous cell cervical carcinoma known to the authors, but her young age is less anomalous in view of her 8year history of HPV infection. Kling and Buchsbaum [I31 reported eight patients under 21 years of age, the youngest of whom was 18 with stage IA disease. In our 40-year tumor registry, the only invasive cervical carcinomas in women under 20 years have been adenocarcinemas, several associated with maternal diethylstilbestrol use. The more advanced stage of presentation, the higher recurrence and death rates, and the shorter interval to recurrence and death in asymptomatic HIV-positive compared to HIV-negative patients have many implications. The findings suggest that occult HIV-related immunodeficiency may contribute heavily to the natural history of cervical carcinoma, and that HIV-positive patients need not demonstrate other signs of immuno-
381
NEOPLASIA
suppression, such as opportunistic infections, for their neoplasia to be adversely affected by HIV. Our finding that our HIV-infected patients, although asymptomatic with respect to HIV disease, had depressed T-cell function suggests that the first indication of HIV positivity may be the advanced stage and rapid progression of cervical cancer. We, as others [14,1.5], have found poorer survival in younger women with cervical cancer than in older women and believe the contribution of latent HIV infection to this trend in high-risk populations warrants investigation. We certainly recommend that young patients with cervical cancer in areas of high HIV prevalence be tested for the virus. Kaposi’s sarcoma in AIDS patients differs from that seen in transplant [16] and other patients, in that lymph node involvement, systemic disease, and fatalities are much more frequent in the AIDS patients. Our data suggest a similar pattern may exist in squamous cell carcinoma of the cervix in the HIV-infected, and imply that unique treatment approaches are needed to combat the more advanced and aggressive form of disease resulting from the interaction of these two disease processes. A multidisciplinary approach to treatment is essential since conventional radiation therapy and chemotherapy may be inadequate for invasive disease, as may standard therapies for preinvasive disease. Laboratory services for monitoring the immune system and consultation with medical and infectious disease specialists on the treatment of the HIV disease are essential. Psychological and counseling services that address the needs of patients with two life-threatening disease processes are needed. Decisions concerning treatment priorities must be made: Should the HIV infection be treated first? Should it be treated at all? Or, should oncologic therapy be instituted initially? The association between HIV infection and cervical neoplasia is undoubtedly complex. As new drugs that suppress, but do not destroy, HIV are developed with wider indications for use, life will be prolonged and we expect to see cervical neoplasia in HIV-infected women with greater frequency. We believe that careful study of this disease interaction will provide invaluable insight into carcinogenesis, immunology, and neoplasia, and that development of research and treatment protocols should proceed in a scientific and compassionate manner. REFERENCES 1.
Gatti, R. A., and Gard, R. A. Occurrence of malignancy in immunodeficient diseases, Cancer 28, 89 (1971).
2.
Penn,I. Malignanciesassociatedwith immunosuppressive or cy-
3.
Penn,I. Cancersof the anogenitalregionin renaltransplantre-
totoxic therapy, Surgery 83, 492-502 (1978). cipients, Cancer 58, 611 (1986).
382
MAIMAN
4. Schneider, V., Kay, S., and Lee, H. M. Immunosuppression as a high risk factor in the development of condyloma accuminatum and squamous neoplasia of the cervix, Acta Cytol. 97, 220-224 (1983).
5. Halpert, R., Fruchter, R. G., Sedlis, A., Butt, K., Boyce, J. G., and Sillman, F. H. Human papillomavirus and lower genital neoplasia in renal transplant patients, Obstet. Gynecol. 68, 251-258 (1986). 6. Fahey, J. C., and Emmanuel, 0. Acquired immune deficiency syndrome (AIDS) and neoplasia, Amer. J. Pediatr. Hematol./Oncol. 9, 193-195 (1987). 7. Purtilo, D. T., Manolov, G., Manolova, S., Harada, S., and Lip-
scomb, H. Squamous cell carcinoma, Kaposi’s sarcoma and Burkitt’s lymphoma are consequences of impaired immune surveillance of ubiquitous virus in acquired immune deficiency syndrome, allograft recipients and tropical African patients, IARC Pub. 83,749770 (1984).
8. Maiman, M., Fruchter, R. G., Serur, E., and Boyce, J. G. Prevalence of human immunodeficiency virus in a colposcopy clinic, J. Amer. Med. Assoc. 260, 2214-2215 (1988). 9. Shapiro, C. N., Schulz, S. L., Lee, N. C., and Dondero, T. J.
Review of human immunodeficiency virus in women in the United States, Obstet. Gynecol. 74, 800-808 (1989).
ET AL. 10. Provencher, D., Valme, B., Averette, H. E., Ganjei, P., Donata, D., Penalver, M., and Sevin, B. U. HIV status and positive Papanicolau screening: Identification of a high-risk population, Gynecol. Oncol. 31, 184-188 (1988). Il. Schrager, L., and Friedland, G. Increased risk of cervical and/or vaginal squamous atypia in women infected with HIV, in Proceedings, dejiciency
Third International Syndrome (AIDS),
Conference on Acquired ImmunoWashington, DC, p. 86 (1987).
12. Sillman, F. H., Stanek, A., and Sedlis, A. The relationship between human papillomavirus and lower genital neoplasia in immunosuppressed women, Amer. J. Obstet. Gynecol. 150, 300 (1984). 13. Kling, T., and Buchsbaum, H. Cervical carcinoma in women under twenty-one years of age, Obstet. Gynecol. 42, 205-207 (1973). 14. Chapman, G. W., Abreo, F., and Thompson, H. E. Carcinoma of the cervix in young females (35 years and younger), Gyecol. Oncol. 31, 430-434 (1988). 15. Fedorkow,
D. M., Robertson, I., Duggan, M. A., McGregor, S. E., and Stuart, G. C. E. Invasive squamous cell carcinoma of the cervix in women less than 35 years old. Recurrent versus nonrecurrent disease, Amer. .I. Obstet. Gynecol. 158, 307-311 (1988). 16. Penn, I. Cancer is a complication of severe immunosuppression, Surg. Gynecol.
Obstet. 162, 603-609 (1986).
