Highly Active Antiretroviral Therapy and Cervical Squamous Intraepithelial Lesions in Human Immunodeficiency Virus–Positive Women
Journal of the National Cancer Institute
Highly Active Antiretroviral Therapy and Cervical Squamous Intraepithelial Lesions in Human Immunodeficiency Virus-Positive Women
Linda Ahdieh-Grant 0
Rui Li 0
Alexandra M. Levine 0
L. Stewart Massad 0
Howard D. Strickler 0
Howard Minkoff 0
Michael Moxley 0
Joel Palefsky 0
Henry Sacks 0
Robert D. Burk 0
Stephen J. Gange 0
0 Affiliations of authors: Johns Hopkins Bloomberg School of Public Health , Baltimore, MD (LAG, RL , SJG); University of Southern California School of Medicine, Los Angeles (AML); Southern Illinois University School of Medicine, Springfield (LSM); Albert Einstein College of Medicine , Bronx, NY (HDS , RDB); Maimonides Medical Center/State University of New York Health Sciences at Brooklyn, Brooklyn, NY (HM); Georgetown University Medical Center , Washington , DC (MM); University of California at San Francisco, San Francisco (JP); Mount Sinai School of Medicine , New York, NY (HS). School of Public Health, Department of Epidemiology , Rm. E-7014, Baltimore, MD 21205 , USA
Background: Women infected with human immunodeficiency virus (HIV) have an increased risk of persistent squamous intraepithelial lesions (SILs) of the cervix. We assessed the association between use of highly active antiretroviral therapy (HAART) and regression of SIL in HIV-infected women enrolled in the Women's Interagency HIV Study, a large, multicenter, prospective cohort study. Methods: Of 2059 HIV-infected participants, 312 HIV-infected women had normal cervical cytology at baseline and were subsequently diagnosed during 7 years of follow-up with incident SIL. Pap smears, CD4 T-cell counts, and information regarding use of HAART were obtained every 6 months. The outcome of interest was lesion regression, defined as two consecutive normal Pap smears 6 months apart. Incidence rates of SIL regression were computed among person-years at risk, both before and after HAART initiation. All statistical tests were two-sided. Results: Of 312 women, 141 had lesions that regressed to normal cytology, with a median time to regression of 2.7 years. Overall, the incidence of regression increased (Ptrend .002) over time after HAART was introduced. At incident SIL, median CD4 T-cell counts were lower in women whose lesions did not regress than in women whose lesions regressed (230 versus 336 cells/ L; P<.01). Before HAART was introduced, the rate of lesion regression was 0.0% (95% confidence interval [CI] 0.0% to 2.4%). After HAART was introduced, the rate was 12.5% (95% CI 9.9% to 15.1%) and was related to post-HAART CD4 T-cell counts (Ptrend .002). Conclusions: HAART use was associated with increased regression of SIL among HIV-infected women, and among women who used HAART, increased CD4 T-cell counts were associated with a greater likelihood of regression. However, the majority of cervical lesions among HIV-infected women, even among individuals who used HAART, did not regress to normal. [J Natl Cancer Inst 2004;96:1070 - 6]
Infection with the human immunodeficiency virus (HIV) and
its related immunosuppression are associated with an increased
risk of prevalent, incident, and persistent squamous
intraepithelial lesions (SILs) of the cervix (
). Cervical infection and
persistence of human papillomavirus (HPV) infection, the major
etiologic agent in most cervical neoplasms (3), is also increased
in HIV-infected women (
). It is well established that the use of
highly active antiretroviral therapy (HAART) in HIV-infected
individuals has a profound effect on immune status and greatly
reduces the risk of acquired immunodeficiency syndrome
(AIDS) and death (
). However, only a few small studies
with short follow-up times have examined the effects of
HAART on regression of SIL in HIV-positive women, with the
majority primarily evaluating the effects of HAART on
prevalent cervical lesions. Because the date of onset of a prevalent
lesion could not be known, it was not possible to adequately
estimate the duration of the lesion. Thus, it is not surprising that
the results reported to date have been equivocal regarding
whether HAART increases the probability of regression of SIL.
To more comprehensively study the long-term effects of
HAART on the regression of incident cervical SIL, we analyzed
data from the Women’s Interagency HIV Study (WIHS), a large,
multicenter, prospective cohort of HIV-infected women who, at
the time of this analysis, had been followed every 6 months for
more than 7 years. Two important strengths of this study are the
analysis of only incident SIL and the availability of data from
before and during the HAART era.
SUBJECTS AND METHODS
Participants of the WIHS were recruited from October 1994
through November 1995, during which time 2059
HIVseropositive and 569 HIV-seronegative women were enrolled at
six sites (Bronx/Manhattan, NY; Brooklyn, NY; Chicago, IL;
Los Angeles, CA; San Francisco Bay Area, CA; Washington,
DC, metropolitan area). The study protocol has been described
). All participants gave written informed consent,
and the study was approved by institutional review boards at all
sites. In brief, after an initial baseline visit, WIHS participants
were evaluated every 6 months. Each evaluation consisted of
administering a structured questionnaire and performing a
physical and gynecologic examination. Blood specimens were
collected to determine levels of CD4 T cells and plasma HIV
RNA. T-cell subsets were determined by flow cytometry in
laboratories participating in the AIDS Clinical Trials Quality
Assurance Program. Plasma HIV RNA levels were measured
using a nucleic acid sequence-based amplification technique
(Organon Teknika, Durham, NC) according to the
manufacturer’s recommended protocol.
Assessment of Cervical SIL and HPV Infection
Pap smears were collected using a plastic Ayre spatula and an
endocervical brush, and the cells were spread onto a single slide.
Cytology samples from all six participating sites were centrally
evaluated at Kyto Meridien Laboratories (New York, NY) using
the 1994 Bethesda System criteria for cytologic diagnosis (
Two cytotechnologists, who were blinded to participants’ HIV
infection status, read all smears. All abnormal smears and a
randomly selected sample of 10% of all negative smears were
confirmed by a cytopathologist. The cytopathologist’s diagnosis
was considered final. Diagnoses were reported as normal or
benign, atypical squamous cells of undetermined significance
(ASCUS), low-grade SIL (LSIL), moderate or high-grade SIL
(HSIL), or cancer.
HPV infection was determined from cells collected by
cervicovaginal lavage. This protocol involved rinsing the cervical
os and the exocervix with 10 mL of sterile normal saline and
then aspirating the fluid from the posterior vaginal fornix. A
comparison of cervicovaginal lavage and cytobrush methods for
the detection of HPV found that both methods provided similar
estimates of the prevalence of HPV (
). All specimens were
stored in aliquots at –70 °C. Two laboratories performed HPV
testing on samples collected at the baseline visit by using a
polymerase chain reaction with MY09/MY11/HB01 L1
consensus primers, which are generic but not HPV-type–specific, and a
control primer set that simultaneously amplifies a beta-globin
DNA fragment, as described (
). Samples that were positive
using the consensus probe were re-probed to identify the
following specific HPV types: HPV6, HPV11, HPV13, HPV16,
HPV18, HPV26, HPV31, HPV35, HPV39, HPV40, HPV42,
HPV45, HPV51, HPV59, HPV61, HPV62, HPV64, HPV66,
HPV73, HPV81, HPV85, HPV89, AE9, and AE10.
Use of HAART
At each visit, participants were asked for the names of the
antiretroviral medications they had used since their previous
visit. Photo-medication cards were provided to assist the
participants in their recall. The primary exposure of interest for this
analysis was the use of HAART. Women reporting use of one of
the following combinations since the time of their last visit were
considered as being on HAART: (a) two or more nucleoside
reverse transcriptase inhibitors (NRTIs) in combination with at
least one protease inhibitor or one non-nucleoside reverse
transcriptase inhibitor (NNRTI); (b) one NRTI in combination with
at least one protease inhibitor and at least one NNRTI; (c) a
regimen containing ritonavir and saquinavir in combination with
one NRTI and no NNRTIs; or (d) an abacavir-containing
regimen of three or more NRTIs in the absence of both protease
inhibitors and NNRTIs. Combinations of zidovudine and
stavudine with either a protease inhibitor or an NNRTI were not
considered HAART because of contraindications. The definition
of HAART was based on the U.S. Department of Health and
Human Services–Kaiser Panel guidelines (
) and the
International AIDS Society–USA Panel (
). The date of HAART
initiation was determined to be the midpoint between the first
visit at which a participant reported using HAART and the
preceding visit. Participants who began HAART were classified
as HAART users from the date they first started HAART.
Study Population and Outcome of Interest
For this analysis, the study population consisted of all
HIVinfected women enrolled in the WIHS who had a normal Pap
smear at study enrollment (i.e., baseline) and subsequently
developed a cytologic lesion during at least 7 years of follow-up.
A cytologic lesion was defined as LSIL or worse. Our outcome
of interest was cytologic regression, which was defined as two
consecutive normal Pap smears 6 months apart. Pap smears were
classified as normal when both squamous and endocervical cells
were within normal limits. Participants were excluded from the
study if they had a diagnosis of ASCUS at baseline. A diagnosis
of ASCUS during follow-up after the incident SIL event was not
considered an endpoint of interest. Because gynecologic
treatment has an impact on the course of cervical SIL, individuals
who reported treatment for cervical neoplasia were censored at
the time of treatment. The following procedures were considered
cervical disease treatments: interferon; 5-fluorouracil;
cryotherapy; laser ablation; conization by diathermy loop, laser or
scalpel; radiation; and hysterectomy. A report of cervical biopsy
alone was not considered a treatment, and participants who
reported such were included in the analysis.
We compared participants whose cervical lesions regressed
with participants whose lesions did not regress to a normal Pap
smear in terms of their demographic, immunologic, and
virologic characteristics at the time of the incident cytologic lesion.
Continuous variables were compared using Wilcoxon tests, and
discrete variables were assessed using chi-squared tests. All
statistical tests were two-sided.
We characterized the rate of cytologic regression using a
person-time approach. Specifically, follow-up time began with
the first visit after the incidence of a cytologic lesion and
continued until the event of interest (cytologic regression) or
time of censoring. Three analyses of the incidence of regression
were conducted using Poisson regression methods, and
confidence intervals (CIs) were calculated using exact methods
(StatXact; Cytel Software, Cambridge, MA). First, we examined
time trends in regression rates of cytologic lesions among all
HIV-infected women in sequential annual calendar periods.
Second, we evaluated the trends around the time an individual began
HAART. The usual method of investigating this factor would be
to partition person-time by whether the individual began
HAART. Thus, “no HAART” time would be a combination of
pre-HAART visits among women who later began HAART and
visits from women who never began HAART. Because HAART
was administered by primary care providers rather than as part of
this observational study, there are a number of key differences
between women who began HAART and women who did not.
Specifically, women who initiate HAART generally have more
advanced disease than those who do not begin therapy (
Because of the challenges in capturing factors relating to
initiating therapy and to minimize the bias introduced by such
selection by indication, we therefore restricted the primary
analysis to a comparison of incidence rates pre- and post-HAART
initiation among those women who began HAART at some time
during follow-up. Third, we conducted analyses in which
person-time was partitioned by quartiles of the overall CD4
T-cell count distribution. CD4 T-cell count measurements were
incorporated as a time-dependent variable and, as such, are
reflective of the ongoing heterogeneity in the effectiveness of HAART.
At study enrollment there were 1243 individuals with a
normal Pap smear. Of these women, 343 had an incident
cytologic lesion, of which 312 had subsequent follow-up during
which regression could be examined. This group of 312 women
constituted the study population for this analysis. There were no
differences in characteristics between the 312 women with
subsequent follow-up and the remaining 31 women (data not
shown). Among the 312 women, 299 had LSILs and 13 had
HSILs. Cytologic regression was observed in 141 (45.2%)
women, and the median time to regression was 2.7 years
(interquartile range 1.0 year to more than 4.6 years). A total of 33
women were censored because they received a gynecologic
A comparison of the characteristics of women whose lesions
regressed and women whose lesions did not regress is shown in
Table 1. At the time of detection of incident SIL there was no
difference between groups by race/ethnicity, but women whose
lesions regressed were younger than women whose lesions did
not (36.8 versus 39.2 years; P .02). Compared with women
whose lesions regressed, women whose lesions did not regress
had more advanced HIV disease, including statistically
significantly lower CD4 T-cell counts and higher HIV RNA levels at
the time of their incident SIL (both P .001). Furthermore, a
greater proportion of women whose lesions did not regress had
prior ASCUS findings than women whose lesions did regress
(64.9% versus 51.1%, respectively; P .02). Because the
women included in our analysis had normal Pap smears at study
enrollment and developed incident LSIL or worse during
followup, the women with prior ASCUS had this diagnosis made
subsequent to baseline. There were no statistically significant
differences between the two groups of women in the prevalence
of baseline HPV infection (P .15). The median time from
baseline to incident cytologic lesion was 2.9 years, with no
statistically significant difference between women whose lesions
regressed and women whose lesions did not (P .65; data not
Among the 312 women in our study population, 221 began
HAART during study follow-up. At the time of their incident
cytologic lesion, 193 women had never used HAART. Of these
women, 91 never began HAART and 102 began HAART
subsequent to their incident lesion.
The component classes of initial HAART regimens, which
were similar to those reported by the WIHS population overall
(data not shown), did not differ among women whose lesions
regressed and women whose lesions did not. Among women
who began HAART, 94.6% of all regimens consisted of two or
more NRTIs in combination with at least one protease inhibitor
or one NNRTI, 3.2% of all regimens consisted of one NRTI in
combination with at least one protease inhibitor and at least one
NNRTI, and 2.3% of all regimens consisted of ritonavir and
saquinavir in combination with one NRTI and no NNRTIs.
Changes in the incidence of regression among the
HIVinfected women over calendar time are shown in Fig. 1. Using
this population-level approach, we found that, before July 1,
1996, the incidence of cytologic regression was 3.8% per
6-month period (five events out of 133 person-visits). After July
1, 1996, the incidence of cytologic regression increased over
time, concomitant with the increased use of HAART among the
cohort (Fig. 1). An unadjusted test for linear trend of these rates
was statistically significant (Ptrend .002). Moreover, after
adjustment for CD4 T-cell counts at each visit, the linear trend
remained statistically significant (Ptrend .002). However, the
rates of regression in HIV-positive women, even in the latter
time periods, remained lower than the overall rate of regression
among the 27 HIV-negative women (i.e., 16 events out of 84
person-visits, 19.0% per 6-month period).
Incidence rates of cytologic regression per year before and
after beginning HAART are shown in Table 2. The denominator
for these rates is the respective number of person-years
contributed before and after beginning HAART. Overall, there were no
instances of cytologic regression in the pre-HAART period (i.e.,
before HAART initiation), although more than 160 person-years
were contributed. However, the rate of cytologic regression
among women in the post-HAART period (i.e., after HAART
initiation) was 12.5% (95% CI 9.9% to 15.1%), and the
likelihood of regression was dose-dependently associated with
CD4 T-cell count (Ptrend .002). This statistically significant
trend for increasing regression rates with higher CD4 T-cell
counts persisted after adjusting for age at incident SIL. For
example, the regression rate was 7.9% (95% CI 5.3% to
11.7%) among women with CD4 T-cell counts of less than
200 cells/ L and 17.0% (95% CI 10.9% to 26.5%) among
women with CD4 T-cell counts of greater than or equal to 500
cells/ L. Similar results were noted when we limited the
analysis to individuals with low-grade lesions and excluded the 13
individuals whose incident lesion was considered high-grade or
above. Furthermore, there was also an inverse dose–response
association between HIV RNA levels and the likelihood of
regression; specifically, regression rates associated with viral
loads of less than 400 copies/mL, 400 –20 000 copies/mL, and
greater than or equal to 20 000 copies/mL were 17.2% (95% CI
11.9% to 24.6%), 13.3% (95% CI 9.6% to 18.5%), and
8.3% (5.4% to 12.6%), respectively.
Only a small subset of women did not begin HAART during
study follow-up. It is not surprising that this number is relatively
low because cytologic abnormalities are often associated with
levels of immunosuppression that fall within the guidelines of
HAART initiation. The median CD4 T-cell count was 204
cells/ L for women who began HAART and 326 cells/ L for
women who did not (data not shown). Among the latter group,
the rate of cytologic regression per year was 23.2% (95% CI
17.6% to 30.6%). There was an association between CD4
T-cell count and regression, with regression rates ranging from
11.1% (95% CI 4.7% to 26.0%) for women with CD4 T-cell
counts of less than 200 cells/ L to 33.3% (95% CI 21.3% to
52.1%) for women with CD4 T-cell counts of greater than or
equal to 500 cells/ L (Table 3). It is of note that regression rates
for these women were higher than rates for those in the
postHAART era, supporting our previous observations (
) of a
major selection by indication for HAART initiation, i.e.,
individuals who were selected to remain HAART-naive were the
healthiest, with the least advanced HIV disease in our study
population. Moreover, when we collapsed person-years
contributed by women who never began HAART with pre-HAART
time contributed by women who began HAART, regression
rates were similar to those of women in the post-HAART period.
This result is likely due to the strong association between CD4
T-cell count and both HAART initiation and the incidence of an
abnormal Pap smear.
HIV-related immunosuppression has been shown to alter the
clinical course of HPV and SIL by increasing the likelihood of
viral persistence and lesion progression, respectively (
Although HAART has lowered the incidence of various
opportunistic diseases associated with immunosuppression, its
possible impact on cytologic lesions is unclear and, to date, has been
addressed in only a few studies (
In this study, which was based on a large cohort of
HIVinfected women followed prospectively since 1994, we found
that, after controlling for level of immunosuppression, the
probability of cytologic regression was greater after HAART
initiation than before HAART initiation. In fact, whereas there were
no instances of cytologic regression before HAART initiation,
regression rates were higher than 10% after beginning HAART.
These data suggest that the duration of a cytologic lesion is
shorter for women who take HAART than for women who do
not, although it is essential to note that, even in the post-HAART
period, the majority of lesions did not regress. Among those
lesions that did regress to normal cytology, the average duration
of cytologic abnormalities was nearly 3 years. By comparison,
the likelihood of regression among HIV-negative women was
high, with 16 (approximately 59%) of the 27 women with
incident LSIL or worse regressing to normal.
Because the WIHS continues to ascertain longitudinal CD4
T-cell count measurements, we were able to incorporate the
heterogeneity of the effect of HAART by modeling CD4
T-cell counts in a time-dependent fashion and examining
sequential person-visits contributed at different levels of
immunosuppression. We applied an “intent to remain on treatment”
assumption; thus, our estimate of the effect of HAART may be
an underestimate because some women discontinued therapy or
were noncompliant with their HAART regimen. Indeed, it
should be noted that between October 1996 and September
1999, only a minority of women switched to a non-HAART
regimen (10.0%) or discontinued HAART (6.7%) (
frequency of switching among HAART regimens, however, was
very common (
). Models that use post-HAART initiation data
and allow for switching or discontinuing HAART regimens are
complicated by biases inherent in the fact that individuals who
alter therapies after initiation are different from those who
remain on prescribed regimens (
). Because such selection
biases are an inherent challenge to using observational data to
understand treatment effects (27), one active field of
investigation is focused on using more comprehensive models, such as
causal models (
). We took a different approach in which we
examined trends at the population level and linked such trends to
changes in exposures (Fig. 1). Although this approach is subject
to ecologic distortion, the observation that SIL regression rates
decreased as the use of HAART increased is highly suggestive
of an association between HAART and the course of cytologic
Our findings confirm and extend our previous observations
that linked HAART use to the likelihood of short-term, 6-month
cytologic regression (
). In our previous study, the odds of
regression to any lower grade lesion were 40% greater among
individuals using HAART than among individuals not using
HAART. This pattern of higher regression rates for women
using HAART was also observed when high-grade lesions were
examined separately. Since the publication of these findings, two
additional studies (
) have examined longitudinal patterns
of HPV infection and cytologic lesions in the context of women
who use HAART. Heard et al. (23) found that 67 (39.9%) of 168
women had low-grade SIL that regressed to a lower grade
abnormality and that the likelihood of regression was doubled
for women receiving HAART. Similar to the results reported by
Minkoff et al. (
), regression rates increased in individuals who
used HAART, regardless of lesion grade. A major strength of the
analysis by Heard et al. (
) was the availability of histologic
specimens to ascertain cervical disease status.
Unlike our results and those of Minkoff et al. (
) and Heard
et al. (
), a recent Italian study of 163 women found no
association between HAART and the regression of cytologic
). Furthermore, HAART affected neither the
persistence of high-risk HPV infection nor the incidence of higher
grade lesions, defined as either a new lesion or progression.
Participants in the study were followed for 15.4 months, and the
outcomes of interest were assessed at baseline and at a single
follow-up visit. The sample size for the analysis of cytologic
regression was, however, relatively small. Only 43 subjects had
a cytologic lesion at baseline and could subsequently be
evaluated for regression, and only 74 subjects used HAART among
all 163 participants. Additional studies are needed to fully
resolve questions regarding the role of HAART on cervical disease.
HAART has dramatically changed the outcome of HIV
infection and the clinical course of many of its commonly
associated conditions. Analyses of data from the WIHS (
) and a
large international collaborative study involving more than
45 000 individuals (
) evaluated the risk of specific cancers
before and after the introduction of HAART and found
statistically significant declines in the risk for Kaposi’s sarcoma (P
.03) and, to a lesser extent, for non-Hodgkin lymphoma (P
.06), although not for cervical cancer. Much attention is
currently focused on gaining a better understanding of the
mechanisms leading to these reduced incidence rates, with
immunosuppression believed to play an important role.
Our study has several limitations. We used longitudinal
cytologic specimens to assess cervical disease rather than
colposcopy and histologic samples. We cannot exclude the possibility
that negative cytology may mask persistent histologic disease or
that abnormal cytology may reflect non-neoplastic changes. It
should be emphasized, however, that all smears were read by
two trained cytotechnologists with expertise in cervical cytology
at a well-established central laboratory. The cytotechnologists
were blinded to any identifying information from participants.
Therefore, there is no reason to believe that any misclassification
that occurred was differential according to HAART use.
Furthermore, random misclassification of cervical disease status
would only dilute estimates of the effect of HAART. To address
potential non-differential misclassification, we applied a
conservative outcome measure, i.e., regression was defined by two
consecutive normal Pap smears rather than only one. Despite the
use of incident SIL to measure duration of lesions, there is
evidence that even these estimates are imprecise, as suggested
by the very high incidence of ASCUS before diagnosis of SIL,
indicating that at least some lesions may have occurred earlier
than estimated using a diagnosis of SIL, given the limitations of
cytopathology. Finally, because the majority of abnormal smears
in our analysis were low-grade and we had no concomitant
biopsy results, it is possible that our results reflect an effect of
HAART on HPV expression rather than on oncogenic or
premalignant processes. However, an effect of HAART on viral
persistence would be meaningful in itself because the
persistence of HPV is generally recognized as the first and most
crucial step in cervical oncogenesis.
Notwithstanding these challenges in using cytologic
specimens to study questions of cervical disease, a core advantage to
their use is that Pap smears are routinely obtained from all
subjects at all visits and on the same day that other biologic
specimens and patient data are collected. This temporality is an
important concern in the WIHS. Cervical lesions are often
transient and, especially in the era of HAART, the immune
status of a patient may change, making it essential that HIV
status, CD4 T-cell count, HIV RNA level, and HAART data
be collected concurrently with diagnoses of cervical pathology.
The use of cytologic endpoints for epidemiologic research
purposes is consistent with another large, well-established,
prospective cohort study of HPV and cervical disease (
histology is the clinical gold standard from a research
perspective, cytology and histology both present problems regarding
diagnostic accuracy, because it has been repeatedly
demonstrated that inter-rater agreement is only fair to good with either
approach, even among expert pathologists (
). It should be
noted that HPV infection is another important endpoint of
interest with respect to cervical disease. Additional studies to
elucidate the effect of HAART on HPV natural history will
necessitate the inclusion of the larger subset of women who did
not have incident SIL and will also require a different analytic
approach that accounts for the fact that multiple, concurrent
type-specific infections are common, especially among
HIVinfected women. We believed that it was most appropriate to
address the question of the effect of HAART on HPV infection
separately and therefore focused this study on the clinically
relevant consideration of the effect of HAART on cervical
disease because it has a unique epidemiology that is of interest.
Because HAART was introduced relatively soon after WIHS
recruitment, there was limited follow-up time from the
preHAART era. Consequently, it is possible that there was
insufficient observed follow-up during which an incident lesion could
resolve and that, with extended follow-up, additional
preHAART regressions would have been noted. To investigate
whether this was the case, we examined regression rates among
the 275 women with prevalent LSIL or worse at baseline (data
not shown). Consistent with the observations for incident cases,
there were no regression events before the introduction of HAART.
In summary, we found that individuals who began HAART
during follow-up had a higher risk of cytologic regression than
individuals who did not. These findings underscore the
importance of ensuring that women who are immunosuppressed have
full access to antiretroviral therapy. It should be emphasized,
however, that HIV-infected women on HAART must still
receive careful gynecologic follow-up and close routine
monitoring. Furthermore, additional studies are needed to
comprehensively examine how HAART impacts the incidence of cervical
abnormalities and the persistence of high-risk HPV infections.
Data in this manuscript were collected by the Women’s Interagency HIV
Study (WIHS) Collaborative Study Group with the following centers (principal
investigators): New York City/Bronx Consortium (Kathryn Anastos); Brooklyn,
NY (Howard Minkoff); Washington DC Metropolitan Consortium (Mary
Young); The Connie Wofsy Study Consortium of Northern California (Ruth
Greenblatt, Phyllis Tien); Los Angeles County/Southern California Consortium
(1) Massad LS , Riester KA , Anastos KM , Fruchter RG , Palefsky JM , Burk RD , et al. Prevalence and predictors of squamous cell abnormalities in Papanicolaou smears from women infected with HIV-1. Women's Interagency HIV Study Group . J Acquir Immune Defic Syndr 1999 ; 21 : 33 - 41 .
(2) Massad LS , Ahdieh L , Benning L , Minkoff H , Greenblatt RM , Watts H , et al. Evolution of cervical abnormalities among women with HIV-1: evidence from surveillance cytology in the Women's Interagency HIV Study . J Acquir Immune Defic Syndr 2001 ; 27 : 432 - 42 .
(3) Bosch FX , Lorincz A , Munoz N , Meijer CJ , Shah KV . The causal relation between human papillomavirus and cervical cancer . J Clin Pathol 2002 ; 55 : 244 - 65 .
(4) Ahdieh L , Klein RS , Burk R , Cu-Uvin S , Schuman P , Duerr A , et al. Prevalence, incidence, and type-specific persistence of human papillomavirus in human immunodeficiency virus (HIV)-positive and HIV-negative women . J Infect Dis 2001 ; 184 : 682 - 90 .
(5) Autran B , Carcelain G , Li TS , Blanc C , Mathez D , Tubiana R , et al. Positive effects of combined antiretroviral therapy on CD4 T cell homeostasis and function in advanced HIV disease . Science 1997 ; 277 : 112 - 6 .
(6) Hammer SM , Squires KE , Hughes MD , Grimes JM , Demeter LM , Currier JS , et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less . N Engl J Med 1997 ; 337 : 725 - 33 .
(7) Schapiro JM , Winters MA , Stewart F , Efron B , Norris J , Kozal MJ , et al. The effect of high-dose saquinavir on viral load and CD4 T-cell counts in HIV-infected patients . Ann Intern Med 1996 ; 124 : 1039 - 50 .
(8) Yamashita TE , Phair JP , Munoz A , Margolick JB , Detels R , O'Brien SJ , et al. Immunologic and virologic response to highly active antiretroviral therapy in the Multicenter AIDS Cohort Study . AIDS 2001 ; 15 : 735 - 46 .
(9) Palella FJ Jr, Delaney KM , Moorman AC , Loveless MO , Fuhrer J , Satten GA , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection . HIV Outpatient Study Investigators. N Engl J Med 1998 ; 338 : 853 - 60 .
(10) Detels R , Munoz A , McFarlane G , Kingsley LA , Margolick JB , Giorgi J , et al. Effectiveness of potent antiretroviral therapy on times to AIDS and death in men with known HIV infection duration . JAMA 1998 ; 280 : 1497 - 503 .
(11) Gange SJ , Barron Y , Greenblatt R , Anastos K , Minkoff H , Young M , et al. Effectiveness of highly active antiretroviral therapy among HIV-1 infected women . J Epidemiol Community Health 2002 ; 56 : 153 - 9 .
(12) Barkan SE , Melnick SL , Preston-Martin S , Weber K , Kalish LA , Miotti P , et al. The Women's Interagency HIV Study (WIHS)- design, methods, sample, cohort characteristics and comparison with reported AIDS cases in U.S. women . Epidemiology 1998 ; 9 : 117 - 25 .
(13) Kurman RJ , Solomon D. The Bethesda System for reporting cervical/ vaginal cytologic diagnoses . New York (NY): Springer-Verlag; 1994 .
(14) Goldberg GL , Vermund SH , Schiffman MH , Ritter DB , Spitzer C , Burk RD . Comparison of cytobrush and cervicovaginal lavage sampling methods for the detection of genital human papillomavirus . Am J Obstet Gynecol 1989 ; 161 : 1669 - 72 .
(15) Palefsky JM , Minkoff H , Kalish LA , Levine A , Sacks HS , Garcia P , et al. Cervicovaginal human papillomavirus infection in human immunodeficiency virus-1 (HIV)-positive and high-risk HIV-negative women . J Natl Cancer Inst 1999 ; 91 : 226 - 36 .
(16) Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Developed by the Panel on Clinical Practices for the Treatment of HIV Infection, convened by the U.S. Department of Health and Human Services . March 23 , 2004 . Available at: http://hivatis.org. [Last accessed: June 9 , 2004 .]
(17) Carpenter CC , Cooper DA , Fischl MA , Gatell JM , Gazzard BG , Hammer SM , et al. Antiretroviral therapy in adults. Updated recommendations of the International AIDS Society-USA Panel . JAMA 2000 ; 283 : 381 - 90 .
(18) Ahdieh L , Gange S , Greenblatt R , Minkoff H , Anastos K , Young M , et al. Selection by severity of potent antiretroviral therapy usage in a large cohort of HIV-infected women . Am J Epidemiol 2000 ; 152 : 923 - 33 .
(19) Sun XW , Kuhn L , Ellerbrock TV , Chiasson MA , Bush TJ , Wright TC . Human papillomavirus infection in women infected with the human immunodeficiency virus . N Engl J Med 1997 ; 337 : 1343 - 9 .
(20) Wallin KL , Wiklund F , Angstrom T , Bergman F , Stendahl U , Wadell G , et al. Type-specific persistence of human papillomavirus DNA before the development of invasive cervical cancer . N Engl J Med 1999 ; 341 : 1633 - 8 .
(21) Ellerbrock TV , Chiasson MA , Bush TJ , Sun XW , Sawo D , Brudney K , et al. Incidence of cervical squamous intraepithelial lesions in HIV-infected women . JAMA 2000 ; 283 : 1031 - 7 .
(22) 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 - 64 .
(23) Heard I , Tassie JM , Kazatchkine MD , Orth G . Highly active antiretroviral therapy enhances regression of cervical intraepithelial neoplasia in HIVseropositive women . AIDS 2002 ; 16 : 1799 - 802 .
(24) Lillo FB , Ferrari D , Veglia F , Origoni M , Grasso MA , Lodini S , et al. Human papillomavirus infection and associated cervical disease in human immunodeficiency virus-infected women: effect of highly active antiretroviral therapy . J Infect Dis 2001 ; 184 : 547 - 51 .
(25) Ahdieh-Grant L , Silverberg MJ , Palacio H , Minkoff H , Anastos K , Young MA , et al. Discontinuation of potent antiretroviral therapy: predictive value of and impact on CD4 cell counts and HIV RNA levels . AIDS 2001 ; 15 : 2101 - 8 .
(26) Kirstein LM , Greenblatt RM , Anastos K , Levine A , French AL , Minkoff H , et al. Prevalence and correlates of highly active antiretroviral therapy switching in the Women's Interagency HIV Study . J Acquir Immune Defic Syndr 2002 ; 29 : 495 - 503 .
(27) Munoz A , Gange SJ , Jacobson LP . Distinguishing efficacy, individual effectiveness and population effectiveness of therapy . AIDS 2000 ; 14 : 754 - 6 .
(28) Robins JM , Hernan MA , Brumback B . Marginal structural models and causal inference in epidemiology . Epidemiology 2000 ; 11 : 550 - 60 .
(29) Hessol NA , Seaberg EC , Massad S , Melnick S , Sacks HS , Levine AM , et al. Increased incidence of cancer among participants in the Women's Interagency HIV Study (WIHS) . XIV International AIDS Conference, July 7-12 , 2002 , Barcelona, Spain.
(30) International Collaboration on HIV and Cancer. Highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virus-infected adults . J Natl Cancer Inst 2000 ; 92 : 1823 - 30 .
(31) Schlecht NF , Kulaga S , Robitaille J , Ferreira S , Santos M , Miyamura RA , et al. Persistent human papillomavirus infection as a predictor of cervical intraepithelial neoplasia . JAMA 2001 ; 286 : 3106 - 14 .
(32) Stoler MH , Schiffman M. Interobserver reproducibility of cervical cytologic and histologic interpretations: realistic estimates from the ASCUSLSIL Triage Study . JAMA 2001 ; 285 : 1500 - 5 .
(Alexandra Levine); Chicago Consortium (Mardge Cohen); Data Coordinating Center (Stephen J. Gange , Alvaro Mu n˜oz) . The WIHS is funded by Public Health Service Grants U01 -AI- 35004 , U01 -AI- 31834 , U01 -AI- 34994 , U01 -AI- 34989 , U01 -AI- 34993 , and U01-AI42590 from the National Institute of Allergy and Infectious Diseases; U01- HD-32632 from the National Institute of Child Health and Human Development; and M01-RR00079 and M01-RR00083 from the National Center for Research Resources; by supplemental funding from the National Cancer Institute, the National Institute on Drug Abuse, and the National Institute of Dental and Craniofacial Research, National Institutes of Health; and by the Centers for Disease Control and Prevention; all from the Department of Health and Human Services . Manuscript received May 1 , 2003; revised May 5 , 2004 ; accepted May 18, 2004 .