Background:  About half of HIV-infected women will develop cervical squamous intraepithelial lesions (SIL).  However, little is known about the natural history of, or the best way to manage, SIL in these women.  

Methods:  A total of 122 HIV-infected and 257 HIV-uninfected women with biopsy-confirmed SIL were enrolled in a multi-arm clinical trial.  Women with satisfactory colposcopic examinations and low-grade SIL (LoSIL) were randomized to either observation or cryotherapy, while those with high-grade SIL (HiSIL) were randomized to cryotherapy or a loop electrosurgical excision procedure (LEEP).  All women with unsatisfactory colposcopic exams had either a cold-knife or LEEP diagnostic conization.  Follow-up exams that included colposcopy, cervical cytology, and, if indicated, cervical biopsy, were performed at 4-month intervals for up to 12 months after enrollment.  

Results:  Mean length of follow-up was 10 and 11 months for HIV-infected and uninfected women, respectively.  Cervical disease status at the last visit is presented below.

 

HIV-uninfected                                       HIV-infected          

 

 Cinical Group                        No.          Normal    LoSIL      HiSIL      No.          Normal    LoSIL      HiSIL

LoSIL - observation                70            61%         30%         9%           25            24%         72%         4%         

LoSIL - cryotherapy               64            95%         5%           -               25            56%         40%         4%  
HiSIL - cryotherapy                41            93%         5%           2%           7             29%         43%         29%
HiSIL - LEEP                          46            85%         11%         4%           12            42%         33%         25%
LEEP - cone                            36            86%         8%           6%           53            45%         30%         25%  

 

Among women with LoSIL randomized to undergo observation there was a significantly lower rate of spontaneous regression during follow-up in HIV-infected (24%) compared to uninfected women (61%) (p <0.001).  However, the rate of progression to biopsy-confirmed HiSIL during observation was low in both HIV-infected and uninfected women (4% vs 9%, p = 0.41).  All treatment modalities were highly effective in HIV-uninfected women, but were significantly less effective in HIV-infected women (all p <0.01).  Failure rates of cryotherapy, LEEP, and cone biopsy were similar in HIV-infected women (p = 0.63).  

Conclusions:  All standard treatment modalities for SIL are significantly less effective in HIV-infected, compared to uninfected, women, and no single treatment modality appears to be more effective in HIV-infected women.  The high failure rate of treatment for LoSIL and the low rate of progression to HiSIL in the absence of treatment suggests that conservative observation should be considered in managing biopsy-confirmed LoSIL in HIV-infected women.