点击显示 收起
Research Service and Department of Internal Medicine, Saint Louis University Hospital and St. Louis VA Medical Center
Van der Bij et al. retrospectively studied the influence that GB virus C (GBV-C) infection has on HIV-1 disease progression in subjects from the Amsterdam Cohort Study (ACS) [1] and concluded that "GBV-C infection does not seem to protect against either CD4+ cell loss or HIV-1 disease progression" (p. 684). However, there are multiple reasons that this conclusion is not supported by the data presented in their study. First, Van der Bij et al. did not know the time of HIV-1 seroconversion for fully 62% of the subjects studied; they knew only that the subjects were HIV-1 infected at the time they were enrolled in the cohort. Thus, Van der Bij et al. were unable to determine the duration of HIV-1 exposure for the majority of their subjects. Van der Bij et al. model the date of presumed HIV-1 seroconversion using an estimate based on the CD4+ cell count [2], but this is not reliable, because GBV-C infection has been associated with higher CD4+ cell counts in 9 published, peer-reviewed studies (reviewed in [3]); thus, Van der Bij et al.'s model likely underestimates the duration of HIV-1 exposure in GBV-Cinfected subjects. By comparison, Williams et al. documented the time of HIV-1 seroconversion within 18 months in 100% of the subjects in the Multicenter AIDS Cohort Study (MACS) [4].
Second, the database used in Van der Bij et al.'s study is incomplete relative to that used in Williams et al.'s study; in the ACS, HIV-1 loads were tested before June 1996 only in seroconverters [1], and both CD4+ cell counts and HIV-1 loads obtained up to 3 years after HIV-1 seroconversion (documented or modeled) were used as baseline values [1], which they are not. In comparison, in the MACS, HIV-1 loads and CD4+ cell counts were determined within 1218 months of documented HIV-1 seroconversion [4].
In table 2, Van der Bij et al. illustrate the relationship between baseline GBV-C infection and HIV-1 disease progression, and the data show a higher rate of progression to first CD4+ cell count <200 cells/L, syncytium-inducing conversion, AIDS, death, and progression from AIDS to death in GBV-C RNApositive subjects. However, the baseline CD4+ cell counts of the GBV-C RNApositive subjects were significantly lower than those of the GBV-C RNAnegative subjects. This indicates that the GBV-C RNApositive and GBV-C RNAnegative subjects were likely not matched for duration of HIV-1 infection, unless GBV-C infection led to a more rapid decline in CD4+ cell counts (which has not been found in any prior studies of GBV-C) [3, 4].
Contrary to Van der Bij et al.'s interpretation, table 3 does show a statistically significant beneficial effect (P < .0001) of persistent GBV-C infection on survival (hazard ratio [HR] of death, 0.52, 0.57, and 0.57 in models 1, 2, and 3, respectively), as well as a decreased HR of progression from AIDS to death (HR, 0.66; P < .0012) in model 1, as has been shown by others (reviewed in [3]). The benefit from persistent GBV-C infection (HR, 0.74) is not statistically significant only in model 4, which adjusted for updated square root CD4+ cell count. Because GBV-C infection and, particularly, its persistence are associated with higher CD4+ cell counts [4], model 4 is controlling for a variable that is related to GBV-C infection, and thus the statistical model is invalid [5].
Van der Bij et al. also confirm the deleterious effect that the loss of GBV-C infection without the accompanying development of GBV-C envelope protein2 antibodies has on survival, which was observed by others [4, 6]. This again confirms that persistent GBV-C infection is beneficial.
Van der Bij et al. are to be commended for examining the effect that GBV-C acquisition has after HIV-1 seroconversion; previous studies have not included enough patients who had acquired the infection after seroconversion to allow for evaluation of this issue. Again, however, Van der Bij et al. show in table 3 that there is a statistically significant beneficial effect of GBV-C acquisition on the HR of first CD4+ cell count <200 cells/L (HR, 0.34, 0.33, and 0.52), AIDS (HR, 0.56, 0.74, and 0.79), and death (HR, 0.35, 0.53, and 0.44) in models 1, 2, and 3, respectively, and progression from AIDS to death in model 1 (HR, 0.56).
In summary, contrary to Van der Bij et al.'s interpretation, the ACS does show that persistent GBV-C infection has a beneficial effect on HIV-1 disease progressionin accordance with the results of other studieswhen the effect is not adjusted for causal variables associated with GBV-C infection, which is statistically invalid [2]. The results of Van der Bij et al.'s study also confirm previous findings regarding the deleterious effect that the loss of GBV-C infection has, and it is the first study to show that GBV-C acquisition has a statistically significant benefit after HIV-1 seroconversion has occurred.
References
1. Van der Bij AK, Kloosterboer N, Prins M, et al. GB virus C coinfection and HIV-1 disease progression: the Amsterdam Cohort Study. J Infect Dis 2005; 191:67885. First citation in article
2. Geskus RB. On the inclusion of prevalent cases in HIV/AIDS natural history studies through a marker-based estimate of time since seroconversion. Stat Med 2000; 19:175369. First citation in article
3. George SL, Varmaz D. What you need to know about GB virus C. Curr Gastroenterol Rep 2005; 7:5462. First citation in article
4. Williams CF, Klinzman D, Yamashita TE, et al. Persistent GB virus C infection and survival in HIV-infected men. N Engl J Med 2004; 350:98190. First citation in article
5. Kalbfleish JD, Prentice RL. The statistical analysis of failure time data. 2nd ed. New York: Wiley, 2002. First citation in article
6. Bjorkman P, Flamholc L, Naucler A, Molnegren V, Wallmark E, Widell A. GB virus C during the natural course of HIV-1 infection: viremia at diagnosis does not predict mortality. AIDS 2004; 18:87786. First citation in article