Treatment with protease inhibitors and coinfection with hepatitis C virus are independent predictors of preterm delivery in HIV-infected pregnant women.

In a recent study, Cotter et al. [1] analyzed a cohort of HIV-infected pregnant women followed at a single site, to examine the risk of preterm delivery and other pregnancy outcomes. Their results indicated that combination therapy with protease inhibitors (PIs), compared with monotherapy and combination therapy without a PI, is independently associated with preterm delivery. Such results are in accordance with those reported in a common analysis of 2 European cohorts [2], but a previous US study found no association between combination therapy and preterm delivery [3]. The role played by PIs in preterm delivery remains uncertain, because none of the studies controlled for indication for antiretroviral therapy, thereby not ruling out the possibility that treatment with PIs represents a marker for more-advanced disease [4]. We analyzed data from the largest surveillance study currently being conducted in Italy of the use of antiretroviral drugs in pregnancy [5], to establish the role played by PIs in preterm delivery after controlling for important prognostic cofactors. For the purpose of this analysis, we considered all pregnancies that resulted in the delivery of a live newborn (n=417) or in neonatal death (n=2) from all the HIV-positive women enrolled in our observational study between December 2001 (the date of the start of the study) and March 2006. Nonsingleton pregnancies and pregnancies that ended in spontaneous or voluntary abortion or in intrauterine death were excluded. Information and measurements were collected at routine visits performed during the 3 trimesters of pregnancy (with no restrictions in gestational age at entry into prenatal care). Gestational age at birth was determined on the basis of the last menstrual period, ultrasound biometry, or both. Preterm delivery was defined as delivery before 37 completed weeks of gestation. Unadjusted and adjusted odds ratios and 95% confidence intervals were used to estimate in univariate and multivariate logistic regression models the association of different variables with preterm delivery. All the analyses were performed using SPSS software (version 13.0.1; SPSS). Preterm delivery occurred in 96 (22.9%) of 419 women. Antiretroviral therapy (combination therapy in all cases) was present in 309 women at second trimester (81.1%) and in 366 women (91.3%) at third trimester. PI use occurred in 31.3% and 39.8% at second and third trimester, respectively. Sixty-five percent of the women had an indication for antiretroviral treatment for their own health. In univariate analysis, the following baseline variables were significantly associated with preterm delivery: (1) higher age at conception (P=.023); (2) hepatitis C virus (HCV) coinfection (P<.001); and (3) prior pregnancies (P=.017). No association was found for hepatitis B virus coinfection, a prior AIDS diagnosis, sexually transmitted infections, alcohol or substance abuse, and cigarette smoking (P≥.10, for all). The multivariate logistic regression models included the 3 significantly related variables (1–3) indicated above plus the following: (4) indication for antiretroviral treatment during pregnancy (maternal health or only for prophylaxis of vertical transmission); (5) prior preterm delivery; (6) treatment with PIs; (7) HIV RNA load (log10 transformed); (8) CD4 cell count; and (9) any antiretroviral treatment. The status with respect to variables 6–9 was defined separately at second and third trimester, and 2 separate logistic models were designed using the same baseline data (variables 1–5) plus data from the relevant trimester for variables 6–9. The results of the multivariate analysis are shown in table 1. Use of PI and HCV coinfection remained consistently associated with preterm delivery after age, prior pregnancies, prior preterm deliveries, any antiretroviral treatment, indication for antiretroviral treatment, HIV RNA load, and CD4 cell count were controlled for. View larger version: In this page In a new window Download as PowerPoint Slide Table 1. Association between preterm delivery and prognostic factors in a multivariate analysis. Our study provides important new information: first, we were able to confirm the effect of PIs on preterm delivery after controlling for several potential confounding variables, including indication for antiretroviral treatment. The observed rate of preterm delivery among women receiving PIs at second trimester was 32.2%, compared with 18.2% in women not receiving PIs. These figures and the observed size of the effect—a 2-fold adjusted increase in risk attributable to PI use—confirms the results reported by others. Our study, however, has the additional advantage of being entirely based on women enrolled in recent years, with combination therapy as standard treatment and a high proportion of women receiving PIs. The risk assessment, therefore, is based on current standards of care. We also describe an independent role for HCV coinfection in preterm delivery. Rates of preterm delivery in women with and without HCV coinfection were 38.6% and 17.1%, respectively, with a 2-fold increase in risk. Further studies should explore the mechanisms responsible for such an association.