University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Progression and regression of premalignant cervical lesions in HIV-infected women from Soweto: a prospective cohort
Global Health Sciences Literature Digest
Published February 28, 2011
Journal Article

Omar T, Schwartz S, Hanrahan C, Modisenyane T, Tshabangu N, Golub JE, et al. Progression and regression of premalignant cervical lesions in HIV-infected women from Soweto: a prospective cohort. AIDS. 2010; 25(1):87-94.


To evaluate prevalence of and changes in cervical dysplasia among HIV-infected women in Soweto


An HIV wellness clinic in Soweto, South Africa

Study Design

Prospective observational cohort


HIV infected women ≥18 years of age

Main Outcome Measures

Cervical smears were assessed using the 2001 Bethesda reporting system,(1) with the following classifications: ASCUS (atypical cells of uncertain significance); LSIL (low-grade intraepithelial lesion); HSIL (high-grade squamous intaepithelial lesion); ASCH (atypical cells cannot exclude HSIL); AGUS (atypical glandular cells of uncertain significance); invasive carcinoma. Progression of PAP smears was defined as a subsequent smear (at an interval of at least 5.5 months) of HSIL, ASCH or cancer, among women with baseline LSIL; and progression to LSIL or higher grade lesions among women with baseline normal PAP smears. Regression was defined as an initial LSIL smear followed by a subsequent normal cervical smear at least 11.5 months later. Incidence rates, and risk factors for progression/regression using Cox proportional hazard modeling were calculated.


HIV-positive women were enrolled, and received a basic HIV care package including screening for and syndromic treatment of sexually transmitted infections (STI); women eligible for ART were referred to treatment clinics. Annual cervical smears were obtained and were read by two cytologists. Women with LSIL or ASCUS had a repeat smear in 6-12 months; those with HSIL, ASCH or other high grade lesion were referred for colposcopy and further treatment. Women with smears that could not be evaluated because of inflammation or inadequate sampling were requested to have a repeat smear either immediately or following treatment for infection.


There were 2475 women in the cohort of whom 2325 had at least one cervical smear between Aug 2003 and May 2009; a total of 4425 cervical smears were obtained, of which 22 were unsatisfactory and could not be analyzed. Median age of women was 31.7 yrs; median CD4 count was 312 cells/µl (IQR 168-486); 20.8% had STI symptoms; 32.7% were using hormonal contraception; 152 women were on HAART (6%) at baseline, and during followup, 457 were newly initiated on ART. At their first smear, 38% of all women had a premalignant cervical lesion (ASCUS 3.1%, LSIL 20.4%, ASCH 1.0%, HSIL 13.5%); none had cancer. Overall, 18.6% had STI symptoms (17.2% of those with LSIL and 21.0% of those with HSIL).Women with an abnormal smear had lower CD4 cells counts (254 cells/µl vs 351 cells/µl, p<0.001). Factors independently associated with less chance of an abnormal baseline smear included: higher CD4 count (adjusted OR= 0.87 per each 100 CD4 cells/ µl increase, 95% CI [.83-0.92]), greater age (aOR=0.90 per each 5 years, 95% CI [.77-0.87]). During followup, 1193 (51.3%) women had at least one followup smear; the 119 women with baseline ASCH, HSIL or AGUS smears were excluded from progression/regression analyses, leaving 1074 women with smears who were prospectively analyzed. The median followup time from first to last smear was 2.5 years (IQR 1.7-3.4). Overall, 10.5 % of women with baseline normal or LSIL smears progressed to higher grade lesions. Among those with initial normal smears (n=832) 21.8% progressed to LSIL (9.6 per 100 person years), 7.3% to HSIL (3.3 per 100 person years) , and 1.3% to ASCH. Of 242 women with baseline LSIL, 21.5% developed HSIL, and 0.8% ASCH (combined incidence of 9.8 per 100 person years). Among women whose initial smear was LSIL (n=225), the incidence of regression to a normal smear was 21.2 per 100 person years. In multivariable regression analysis, baseline CD4 strata was strongly associated with progression among women with baseline normal or LSIL smears: aHR=1.96 (95% CI [1.33-2.88]) for those with <200 cells/µl vs. >500 cells/µl, and aHR=1.65 (95% CI [1.14-2.39]) for those with 350-500 CD4 cells/µl vs. >500 cells/µl.There was a small reduction in risk of progression among those on HAART (aHR=0.72, 95% CI [0.52-0.99]). CD4 cell count strata was not associated with regression to normal smears.


A very large proportion of these HIV infected women had abnormal cervical smears (38.1%); approximately ¼ progressed to abnormal or higher grade lesions; fewer than ½ with baseline LSIL regressed to normal. The ability to evaluate the association of HAART was limited, because treatment eligibility was at CD4 cell counts <200 cells/µl. Even women with CD4 cells/counts <500 cells/µl were more likely to have an abnormal cervical smear, and/or to progress. to a higher grade lesion.

Quality Rating

This was a high quality observational study. It was not possible to determine the effect of HIV on cervical abnormalities, as there was not an HIV-negative cohort for comparison. Information on HPV was not obtained.

In Context

Cervical cancer is the second leading cancer in women worldwide, and the leading cause of cancer death in women in many developing countries.(2, 3) The long pre-malignant phase from early lesions to invasive cancer allows preventive screening and curative excision, and is the reason for lower rates in more developed countries. In HIV infected women, there are higher rates of human papilloma virus (HPV), more rapid progression of disease, and poorer outcomes. Similar rates of LSIL (35%) and HSIL (13%) have been reported from South Africa,(4, 5) with 68% having high-risk HVP types (16 and 56). In these other studies, progression rates to LSIL and HSIL were lower than in this study. Conflicting findings on the effect of HAART of risk of progression have been reported.(6, 8)

Programmatic Implications

Data from this study indicate that cervical abnormalities may be highly prevalent among HIV infected women, even among those with CD4 counts well above the limit for HAART eligibility. Screening and treatment should be made available as widely as possible. In addition, screening intervals shorter than the once every 10 years (current South Africa recommendations) should be implemented. The fact that there were no cancerous lesions detected in this study could be due to the fact that most women in this clinic were enrolled with comparatively high CD4 cell counts, and were regularly screened; those with high grade lesions were referred for treatment. This study did not have enough data to evaluate the effect of HAART on abnormal cervical lesions, although there did appear to be some protective effect; further studies are needed. It is known that HPV is associated with cervical dysplasia and cancer, although HPV data were not obtained here. Nevertheless, these results suggest that the HPV vaccine could be extremely beneficial in prevention, if cancer-associated HPV subtypes in this region are similar to those in the currently available vaccine.


  1. Solomon D, Davey D, Kurman R, Moriarty A, O'Connor D, Prey M, et al. The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA 2000; 287:2114-2119.
  2. Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: cancer incidence, mortality and prevalence worldwide. IARC Cancer-Base No. 5. version 2.0. Lyon: 2004 IARC Press.
  3. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55:74-108.
  4. Denny L, Boa R, Williamson AL, Allan B, Hardie D, Stan R, Myer L. Human papillomavirus infection and cervical disease in human immunodeficiency virus-1-infected women. Obstet Gynecol 2008; 111:1380-1387.
  5. Firnhaber C, Van Le H, Pettifor A, Schulze D, Michelow P, Sanne IM, et al. Association between cervical dysplasia and human papillomavirus in HIV seropositive women from Johannesburg South Africa. Cancer Causes Control 2010; 21:433-443.
  6. Minkoff H, Ahdieh L, Massad LS, Anastos K, Watts DH, Melnick S, et al. The effect of highly active antiretroviral therapy on cervical cytologic changes associated with oncogenic HPV among HIV-infected women. AIDS 2001; 15:2157-2164.
  7. Heard I, Schmitz V, Costagliola D, Orth G, Kazatchkine MD. Early regression of cervical lesions in HIV-seropositive women receiving highly active antiretroviral therapy. AIDS 1998;12:1459-1464.
  8. Paramsothy P, Jamieson DJ, Heilig CM, Schuman PC, Klein RS, Shah KV, et al. The effect of highly active antiretroviral therapy on human papillomavirus clearance and cervical cytology. Obstet Gynecol 2009; 113:26-31.