The authors thank Barreiro and colleagues for their interest in our study. 1,1a In this report we demonstrated that patients to whom nonnucleoside reverse transcriptase inhibitor (NNRTI)-based highly active antiretroviral therapy (HAART) was initially prescribed had shorter time to a CD4 cell count increase of 50 cells in comparison to patients initially prescribed protease inhibitor (PI)-based HAART, among participants enrolled in an observational cohort study. Since we had recently published a demonstration of the major selection factors operating in comparisons of various HAART regimens in observational databases, 2 we cautiously concluded that our results “indicate the rather comparable effects of NNRTIs and PIs on CD4 cell count responses.” 1 Nevertheless, Barreiro and colleagues suggest some additional analyses and argue that PI-based HAART should be provided universally for patients with low CD4 cell counts. We will address the points they raise below.
We should note that Barreiro and colleagues draw incorrect assumptions from a 2002 report in which we concluded that “virtually all CD4 increases can be attributed to transient or partial pVL suppression.” 3 Contrary to their suggestion, the analyses presented in this paper did include patients to whom NNRTIs were prescribed. The findings of this report indicated that CD4 benefits are most closely tied to the degree and duration of virologic suppression. These conclusions are consistent with studies suggesting that virologic suppression is the main determinant of CD4 responses. 4–6
In terms of the suggested subanalyses, when the patient population that was assessed in our earlier report is restricted to the 1343 patients (88.2%) who did not switch their HAART regimen in the first 12 months, we find evidence that that the time to the first CD4 cell count response ≥50 cells/mm3 was faster for the 402 patients (29.9%) to whom NNRTIs were initially prescribed (log rank P = 0.005), when compared with those who started on PIs, although there was no statistical difference when we looked at the time to the first of 2 CD4 measures ≥ 50 cells/mm3 from baseline (log rank P = 0.161). Among these patients, the CD4 gain from baseline at 12 months was 120 (interquartile range [IQR]: 20–220) for patients on PIs and was 115 (IQR: 20–210) for patients on NNRTIs (P = 0.465). Note that this was restricted to 1005 (74.8%) of the above patients who had a CD4 measure available between 6–12 months from baseline.
We also repeated our earlier analyses among the 874 patients (57.4%) who did not switch therapy in the first 12 months and who had a plasma HIV RNA measure <500 copies/mL in the first 6 months of HAART. Here, the time to the first CD4 cell count response ≥50 cells/mm3 was faster for the 295 patients (33.8%) to whom NNRTIs were initially prescribed (log rank P = 0.006), though there was no statistical difference when we looked at the time to the first of 2 CD4 measures ≥50 cells/mm3 from baseline (log rank P = 0.311). Among these patients, the CD4 gain from baseline was 140 (IQR: 60–250) for patients on PIs and was 150 (IQR: 60–240) for patients on NNRTIs (P = 0.635). Note that this was restricted to 705 of the above patients (80.7%) who had a CD4 measure available between 6–12 months from baseline.
While the stratifications suggested by Barreiro and colleagues may help to address some sources of potential confounding, they introduce the concern that we are forced to examine a highly selected population that may in itself introduce confounding and limit generalizability. As such, we must reiterate that the data presented here, and in our earlier report, 1 are from an observational study and should thus be interpreted with substantial caution. 1,2 Nevertheless, all analyses conducted support our original conclusion that CD4 responses appear to be similar.
Barreiro and colleagues also argue that NNRTI-based HAART has a lower genetic barrier to resistance than PI-based HAART and that the extent of virologic rebound among patients in whom PI-based HAART fails may be lower than patients in whom NNRTI-based HAART fails. While this may be true, if we assume that the lower genetic barrier hypothesis is relevant to our analyses, this would have served to lead to superior CD4 count responses among patients to whom PIs were prescribed, yet this was not observed.
We believe that there is presently insufficient evidence to support Barreiro and colleagues' recommendation that PI-based HAART should be universally prescribed to patients with low CD4 cell counts. 7–9 On the contrary, patients who present for therapy with severe immunodeficiency would be best served by a randomized controlled trial comparing the impact of the various initial HAART regimens on plasma HIV RNA and CD4 cell count responses, and more importantly survival, among this patient population.
1. Barreiro P, González-Lahoz J, Soriano V. CD4+
T-Cell Gain With Nonnucleoside or Protease Inhibitors: Convenience May Not Always Be the Most Convenient. J Acquir Immune Defic Syndr.
1a. Wood E, Hogg RS, Yip B, et al. CD4 cell count response to nonnucleoside reverse transcriptase inhibitor- or protease inhibitor-based HAART in an observational cohort study. J Acquir Immune Defic Syndr
2. Wood E, Hogg RS, Heath KV, et al. Provider bias in the selection of non-nucleoside reverse transcriptase inhibitor and protease inhibitor-based highly active antiretroviral therapy and HIV treatment outcomes in observational studies. AIDS
3. Wood E, Hogg RS, Yip B, et al. “Discordant” increases in CD4 cell count relative to plasma viral load in a closely followed cohort of patients initiating antiretroviral therapy. J Acquir Immune Defic Syndr
4. The AVANTI and INCAS Study Groups. Highly active antiretroviral therapy including protease inhibitors does not confer a unique CD4 cell benefit. AIDS
5. Staszewski S, Miller V, Sabin C, et al. Determinants of sustainable CD4 lymphocyte count increases in response to antiretroviral therapy. AIDS
6. Wood E, Yip B, Hogg RS, et al. Full suppression of viral load is needed to achieve an optimal CD4 cell count response among patients on triple drug antiretroviral therapy. AIDS
7. Walmsley S, Bernstein B, King M, et al. Lopinavir-ritonavir versus nelfinavir for the initial treatment of HIV infection. N Engl J Med
8. Staszewski S, Morales-Ramirez J, Tashima KT, et al. Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med
9. van Leth F, Hassink E, Phanuphak P, et al. Results of the 2NN Study: a randomized comparative trial of first-line antiretroviral therapy with regimens containing either nevirapine alone, efavirenz alone or both drugs combined together with stavudine and lamivudine. Paper presented at: 10th Conference on Retroviruses and Opportunistic Infections; February 10–14, 2003; Boston, MA.