Viral load responses to HAART is an independent predictor of a new AIDS event in late stage HIV infected patients: prospective cohort study
© Kazanjian et al; licensee BioMed Central Ltd. 2005
Received: 12 July 2005
Accepted: 31 October 2005
Published: 31 October 2005
A sizeable number of HIV-infected patients receiving HAART do not maintain prolonged virologic suppression. We evaluated long-term HIV viral load (VL) responses to HAART as a risk factor for AIDS events (AE) that is independent of CD4 responses.
A cohort of patients with pre-therapy CD4 < 200/mm3 who had CD4 and VL measurements for > one year after receiving HAART at a university clinic were prospectively enrolled. Cox proportional multivariate regression model was used to determine whether CD4 and VL responses were independently associated with new AE.
The patient (N = 214) mean baseline CD4 = 92/mm3, VL = 219,000 c/mL and follow-up duration 42.3 months (range 13–72 months). A new AE occurred in 56 patients; CD4 cell count response to HAART that remained < 200/mm3 throughout the study period was a significant risk factor for new AE (RR = 9.7–12.5; p < 0.001). Similarly, VL responses that remained > 5,000 c/mL during this period was also a significant risk factor (RR = 6.7–12.8; p = 0.001) that was independent of CD4 response adjusted for <> 200/mm3.
Maintaining adequate long-term virologic responses to HAART provides a clinical benefit independent of CD4 responses.
Despite an overall decline in AIDS-associated illnesses since the introduction of HAART [1, 2], some patients receiving combination antiretrovirals remain at risk for developing new AIDS events (AE) [3, 4]. Those who have a pre-treatment CD4 cell count < 200/mm3 [5–9] or have insufficient CD4 cell responses to HAART  are at risk for progressing to AIDS or dying while receiving therapy. Persistent viremia , independent of insufficient CD4 cell count responses [13–16], has also been shown to be a predictor of disease progression. These studies, however, have differed in regards to the VL that they identify as being predictive of disease progression-- > 1,000 c/mL , > 7,000 c/mL , or > 20,000 c/mL . Furthermore, although these studies have linked short virologic markers in response to HAART with a successful clinical outcome [11–14], the importance of sustaining immunologic and virologic responses to HAART over a more prolonged time period remains to be addressed.
It is important to evaluate whether maintaining adequate responses to long-term HAART treatment has clinical significance for several reasons. First, evaluating whether long-term virologic response is an independent predictor of new AE has relevance for the sizeable number of patients receiving HAART who do not maintain prolonged virologic suppression [1–4, 15]. Also, the meaning of persistent viremia is especially pertinent to those who have high-level resistance to antivirals, but nonetheless sustain an immunologic benefit while continuing their failing HIV regimen . Furthermore, evaluating whether immunologic responses to HIV treatment is an independent risk factor for disease progression is applicable to those patients who are unable to mount significant CD4 cell elevations, regardless of whether they have undetectable VL . Thus, our aim was to determine whether long term virologic, as well as immunologic responses to HAART are important and independent predictors of developing a new AE as well as to identify a level of persistent viremia that is predictive of disease progression. We chose patients with significant immune depletion because they are at highest risk for disease progression [3, 5].
We identified HIV-infected patients treated at the University of Michigan HIV/AIDS Treatment Clinic (UMHATC) who had a baseline CD4 cell count < 200 cells/mm3 before initiating a HAART regimen. The study began on January 31, 1996; patients were enrolled into this prospective cohort from the time therapy was initiated if they received at least twelve months of HIV therapy and were followed to January 1, 2005. The definition of HAART for this study was guided by DHHS guidelines . Patients were included if they were antiviral naïve before starting HAART or if they had previously received one or two nucleoside agents. Patient information was extracted from an electronic database (Solutions™ Patient Management System). The research was carried out according to the principles of the Declaration of Helsinki and the study was approved by the Institutional Review Board of the University of Michigan Health System.
Data on markers of response to HAART (i.e., CD4 and VL levels that were obtained during the course of routine UMHATC practice on an every three month basis), as well as occurrence of a new AE (1993 CDC AIDS Surveillance definition) , were extracted from the electronic database system. CD4 cell counts were determined using flow cytometry, and HIV RNA levels were performed using a reverse transcriptase polymerase chain reaction assay (Amplicor; Roche Diagnostics, Branchburg, New Jersey, USA). For purposes of consistency, a VL value of < 400 copies/mL was considered undetectable for the study period, as lower limit changed to < 50 c/mL in 1999 in the middle of our study period. We also recorded assessment of adherence that was entered into our electronic database by nurses, social workers, and physicians, as was our standardized practice during this study period.
To compare baseline characteristics between patients who developed an AE and those who did not, we used histograms to examine distributions of variables that were collected. Appropriate transformations were applied on variables found asymmetric in the following analysis. Baseline quantitative characteristics of patients (e.g. age, CD4, VL) were described using mean and standard deviation, and qualitative characteristics (e.g. gender, presence of AE at baseline) using proportion. A t test was used to compare quantitative characteristics between patients who developed an AE and those who did not. A Chi-square test was used to compare qualitative characteristics between groups.
To determine probabilities of occurrence of a new AE from HAART initiation, we measured time to first AE and used Kaplan-Meier curves. To determine averaged VL trends after initiating HAART we used lowess smooth curves fitted to all patients, as well as to those who developed an AE and those who did not. We then determined VL values for the entire study period, as well as for four specified periods: 1–12 months, 13–24 months, 25–36 months and 37–48 months after initiating HAART.
To determine risk factors for developing a new AE, we used a Cox proportional hazard multivariate regression model. The proportional hazard assumption was then examined by plotting the complementary log-log transformation of the survival functions separately within subgroups. Pre-treatment factors that were examined as predictors for a new AE included demographic factors (e.g. age and gender), prior NRTI use, and presence of AE at baseline. For a given VL measurement at a particular time point after initiation of HAART, persons were considered at risk for newly diagnosed AE for 1–6 months after the time the measurement was made. For each time period, we classified VL responses to HAART (adjusted for a CD4 response of <> 200 cells/mm3) into separate categories to determine whether a particular VL category was independently associated with presence of a new AE. The relative risks of new AE for patients in specified VL categories were based on the fitted proportional hazard model and 95% confidence intervals were estimated. All statistical tests were two-sided, and a p-value less than 0.05 was considered statistically significant. Statistical software R 1.7.0 was used for all statistical analysis.
Patient Population and AE Description
Demographic, Clinical Features, Mean CD4* and Viral Load** According to Whether an AE Occurred.
Patients with AE (n = 56)
Patients without AE (n = 158)
Age ± SD (range)
37.9 ± 8.28 (16 – 65.4)
39.4 ± 8.79 (15.2 – 67.8)
HIV risk (no.)
OI (no.) at Baseline (%)
Prior NRTI (no.)
Baseline Laboratory Values
Follow Up Laboratory Values
CD4 (Change from Baseline)
VL (Change from Baseline)
29,743 (167, 089)
Number (%) with Undetectable VL at any point during follow up
Number (%) with Undetectable VL
On Treatment CD4 and VL Responses and Association with a New AIDS Event
Risk of Developing New AIDS Event according to CD4 and VL Response Categories.
Time After HAART
Relative Risk and CI
Absolute CD4 < 200
RR = 4.45 (95% CI: 1.8, 10.9);
p < 0.001
RR = 3.95 (95% CI: 1.62, 9.63);
p = 0.002
RR = 5.19 (95% CI: 1.99, 13.5);
p = 0.001
RR = 3.21 (95% CI: 1.61, 10.8);
p = 0.02
Absolute VL > 1,000
RR = 2.34 (95% CI: 0.87, 6.26);
p = 0.09
RR = 2.77 (95% CI: 0.81, 9.48);
p = 0.10
RR = 4.41 (95% CI: 0.99, 19.7);
p = 0.052
RR = 2.74 (95% CI: 0.71, 9.1);
p = 0.14
Absolute VL > 5,000
RR = 4.66 (95% CI: 2.02, 10.7);
p = 0.001
RR = 3.24 (95% CI: 1.06, 9.88);
p = 0.03
RR = 6.41 (95% CI: 1.8, 22.8);
p = 0.004
RR = 4.23 (95% CI: 1.05, 6.0);
p = 0.05
Absolute VL > 50,000
RR = 3.77 (95% CI: 1.68, 8.43);
p = 0.001
RR = 2.78 (95% CI: 1.04, 7.39);
p = 0.021
RR = 2.89 (95% CI: 1.80, 5.51);
p = 0.04
The present study demonstrates that failure to maintain low levels of viremia is a significant risk factor for HIV disease progression that is independent of CD4 responses. Our results are consistent with studies showing the clinical importance of achieving a VL response at 6–12 months that is independent of CD4 responses [10–14], and stress the benefit of maintaining a durable VL response throughout a 48 month period. The VL level identified as carrying an increased risk for an AIDS event in our study, > 5,000 c/mL, is within the range that some reports identify as independently predictive of disease progression, i.e., > 1,000 c/mL , > 7,000 c/mL , but lower than another report (20,000 c/mL) . Regardless of the precise value, a reduction in VL that cannot be maintained at low or undetectable levels, a predictor of new AIDS events in our study, is not uncommon for patients receiving HAART . It is unlikely that the difference in VL response between the groups in our study can be attributed to different adherence rates between the two groups. However, because adherence measurement tools were not used as part of routine clinical care, we could not calculate the percentage of each patient's adherence over a long-term basis and then compare this variable between groups.
In fact, several published studies suggest that virologic failure rates are especially high, varying from 20% to 70%, in those who are either treatment-experienced or have low baseline pre-treatment CD4 cell counts and high VL values [21–23]. In one study, for example, lower CD4 counts and higher VL at baseline predicted virologic failure, but there was no clear cut-off value at which the risk started to increase . In another study, baseline CD4 count < 25 cells/mm3 was associated with a significantly higher risk of virologic failure, as was a baseline VL ≥ 100,000 copies/mL .
Identifying that long-term virologic response is an independent predictor of new AE has relevance for the sizeable number of patients receiving HAART who cannot maintain prolonged virologic suppression . Also, the meaning of persistent viremia is pertinent to those who have high-level resistance to antivirals, but nonetheless sustain an immunologic benefit while continuing their failing HIV regimen . At present, the outcome and optimal management of these patients remains undefined . Switching to an optimized regimen may not be an available option, since the durability of a salvage regimen selected for patients with high level resistance to many antivirals may be brief. Consequently, some physicians continue a failing regimen despite persistent viremia for patients who maintain an immune benefit until other active treatment options become available. By identifying VL as a risk factor for disease progression that is independent of CD4 response, our study suggests that discordant responders may be at risk for disease progression while being maintained on a virologically failing regimen.
It is not rare for patients, such as those followed in the present study, to first receive HAART after they develop an AIDS event or their CD4 cell count falls below 200/mm3. In one report, "late testers" – those whose diagnosis of HIV occurred within 12 months of their AIDS diagnosis, accounted for 39% of persons diagnosed with AIDS in San Francisco between 2001 and 2002 . Also, in a New York study, 27% of persons with newly diagnosed HIV infection were concurrently diagnosed with AIDS . Furthermore, in clinical practice, some patients, despite having been tested earlier, may not have received HAART until a later stage because they may not have wished to take HIV therapy until becoming symptomatic. These reasons, along with sole availability of mono or dual nucleoside agents prior to 1996, accounted for why 21% of patients followed in our practice did not receive HAART until their CD4 count fell below 200/mm3 and therefore met criteria for our study.
Our finding that PCP and MAC remain the most frequent AE in the HAART era is consistent with prior studies . Similarly, our finding that chemoprophylaxis failure among those with the most advanced immunosuppression was the most significant source of new PCP cases is also consistent with a HOPS cohort study , along with studies showing that KS and NHL remain the major AIDS-associated malignancies in the HAART era [32–34]. In contrast, our study showed that CMV and cerebral toxoplasmosis occur less often than in the Pre-HAART era. One possible explanation for this observation is that memory CD4 cells, expanded by HAART, may recognize antigens on previously acquired pathogens that persist in a latent state, thereby preventing clinical reactivation. This possibility, if confirmed, suggests that there may be benefits of receiving HAART that are independent of immunologic and virologic responses currently used to monitor therapy.
In summary, our study emphasizes the clinical benefit of maintaining virologic responses as well as immunologic throughout long-term HAART treatment. Incomplete responses to HAART are not uncommon in clinical practice, and identifying them as risk factors for disease progression emphasizes several treatment options that are available. For example, changing regimens in patients who have persistent virologic replication who have other antiviral options, as recommended by treatment guidelines , is supported by identifying VL as an independent predictor of disease progression. Nevertheless, for those with no option other than to remain on their current regimen for prolonged periods of time because of high level drug resistant HIV, clinical progression may eventually occur, even in those who have a CD4 benefit despite persistent viremia. Our findings stress the importance of developing new potent antiretroviral agents in order to sustain the overall decline in AIDS-associated illnesses that has been witnessed since the introduction of HAART.
Financial Support: There was no financial support.
- Palella F, Delaney K, Moorman A: Declining morbidity and mortality among patients with advanced HIV infection. N Engl J Med. 1998, 338: 853-60. 10.1056/NEJM199803263381301.View ArticlePubMedGoogle Scholar
- Ledergerber B, Egger B, Erard V: AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy. JAMA. 1999, 282: 2220-6. 10.1001/jama.282.23.2220.View ArticlePubMedGoogle Scholar
- Ferrer E, Santamarina E, Santin M: Progression to AIDS/Death is higher in patients initiating ARV with CD4 counts < 350 [abstract]. 10th Conference on Retroviruses and Opportunistic Infections; Boston, MA. 2003, 910-Google Scholar
- Van Sighem AI, van de Wiel MA, Ghani A, on behalf of the ATHENA cohort study group: Mortality and progression to AIDS after starting highly active antiretroviral therapy. AIDS. 2003, 17: 2227-36. 10.1097/00002030-200310170-00011.View ArticlePubMedGoogle Scholar
- Egger M, May M, Chene G, ART Cohort Collaboration: Prognosis of HIV-1 Infected patients starting Highly Active Antiretroviral Therapy: A Collaborative Analysis of Prospective Studies. Lancet. 2002, 360: 119-29. 10.1016/S0140-6736(02)09411-4.View ArticlePubMedGoogle Scholar
- Sterling TR, Chaisson RE, Keruly J: Improved Outcomes with Earlier Initiation of Highly Active Antiretroviral Therapy among Human Immunodeficiency Virus-Infected Patients who Achieve Durable Virologic Suppression: Longer Follow-Up of an Observational Cohort Study. J Infect Dis. 2003, 188: 1659-65. 10.1086/379741.View ArticlePubMedGoogle Scholar
- Tierney C, Lathey JL, Christopherson C: Prognostic value of baseline HIB type 1 DNA measurement for disease progression in patients receiving nucleoside therapy. J Infect Dis. 2003, 187: 144-8. 10.1086/345870.View ArticlePubMedGoogle Scholar
- Wood E, Hogg R, Yip B: Higher baseline levels of plasma HIV type 1 RNA are associated with increased mortality after initiation of triple drug antiretroviral therapy. J Infect Dis. 2003, 188: 1421-5. 10.1086/379201.View ArticlePubMedGoogle Scholar
- de Wolfe R, Spijkerman I, Schellekens P: AIDS prognosis based on HIV-1 RNA, CD4+ T cell count and function: markers with reciprocal predictive value over time after seroconversion. AIDS. 1997, 11: 1799-806.View ArticleGoogle Scholar
- Tarwater PM, Gallant JE, Mellors JW: Prognostic value of plasma HIV RNA among highly active antiretroviral users. AIDS. 2004, 18: 2419-23.PubMedGoogle Scholar
- Raffanti SP, Fusco JS, Sherrill BH: Effect of Persistent Moderate Viremia on Disease Progression During HIV Therapy. J Acquir Immune Defic Syndr. 2004, 37: 1147-54.View ArticlePubMedGoogle Scholar
- ART Cohort Collaboration: Prognostic importance of initial response in HIV-1 infected patients starting patent antiretroviral therapy: analysis of prospective studies. Lancet. 2003, 362: 679-86. 10.1016/S0140-6736(03)14229-8.View ArticleGoogle Scholar
- Kaplan JE, Hanson DL, Jones JL: Viral load as an independent risk factor for opportunistic infections in HIV-infected adults and adolescents. AIDS. 2001, 15: 1831-36. 10.1097/00002030-200109280-00012.View ArticlePubMedGoogle Scholar
- Olsen CH, Gatel J, Ledergerber B: Risk of AIDS and death at given HIV-RNA and CD4 cell counts, in relation to specific antiretroviral drugs in the regimen. AIDS. 2005, 19: 319-30.PubMedGoogle Scholar
- Ledergerber B, Egger M, Opravil M: Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Lancet. 1999, 353: 863-68. 10.1016/S0140-6736(99)01122-8.View ArticlePubMedGoogle Scholar
- Belec L, Piketty C, Si-Mohammed A: High levels of drug resistant HIV in patients exhibiting increasing CD4 T cell counts despite virologic failure of protease inhibitor-containing ARV. J Inf Dis. 2000, 181: 1808-12. 10.1086/315429.View ArticleGoogle Scholar
- Kaufmann G, Furrer H, Perrin L: Low Baseline CD4 T-cell Count and Higher Age Predict Poor CD4 T-cell Recovery in Treated HIV-1 Infected Individuals Suppressing HIV-1 RNA to Levels <1000 copies/mL for 5 Years [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 557-Google Scholar
- Department of Health and Human Services [DHHS] Guidelines. [http://www.aidsinfo.nih.gov]
- Centers for Disease Control and Prevention: 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep. 1992, 41 (RR-17): 1-19.Google Scholar
- Deeks SG, Hecht FM, Swanson M: HIV RNA and CD4 cell count response to protease inhibitor therapy in an urban AIDS clinic: response to both initial and salvage therapy. AIDS. 1999, 13: F35-F43. 10.1097/00002030-199904160-00001.View ArticlePubMedGoogle Scholar
- King MS, Bernstein BM, Walmsley SL: Baseline HIV-1 RNA level and CD4 count predict time to loss of virologic response to nelfinavir but not lopinavir/ritonavir in antiretroviral therapy-naïve patients. J Inf Dis. 2004, 190: 280-4. 10.1086/422037.View ArticleGoogle Scholar
- Ribaudo H, Downey G, Fischl M: Factors Associated with Virologic Failure and Their Impact on Treatment Outcomes: An Analysis of Virologic Failure in ACTG 388 [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 553-Google Scholar
- Paredes R, Mocroft A, Kirk O: Predictors of virologic success and ensuing failure in HIV-positive patients starting highly active antiretroviral therapy in Europe. Arch Intern Med. 2000, 160: 1123-32. 10.1001/archinte.160.8.1123.View ArticlePubMedGoogle Scholar
- Mocroft A, Ledergerber B, Viard JP: Time to Triple Drug Class Failure after Initiation of HAART [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 554-Google Scholar
- van Leth F, Andrews S, Grinsztjen B: Virologic Failure in Antiretroviral Therapy Naive Patients Is Only Determined by Extreme Low Values of CD4+ Cells or High Values of HIV-1 RNA Concentration, Not by Choice of Treatment with Nevirapine or Efavirenz [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 550-Google Scholar
- PLATO Collaboration: Predictors of trend in CD4-positive T cell count and mortality among HIV-1 infected individuals with biological failure to all three antiretroviral drug classes. Lancet. 2004, 364: 51-62. 10.1016/S0140-6736(04)16589-6.View ArticleGoogle Scholar
- Kuritzkes DR: Effect of Antiretroviral Resistance on Response to Treatment-Experienced Patients. The AIDS Reader. 2003, 13: S5-11.PubMedGoogle Scholar
- Schwarcz S, Hsu L: AIDS in the HAART Era: the Extent and Characteristics of Late HIV Testers [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 864-Google Scholar
- Nash D, Manning SE, Ramaswamy C: Descriptive Epidemiology of HIV/AIDS in New York City: Incorporation of Newly Available Population-based Surveillance Data on HIV (non-AIDS), 2001 [abstract]. 11th Conference on Retroviruses and Opportunistic Infections; SF, CA. 2004, 87-Google Scholar
- Centers for Disease Control and Prevention: HIV/AIDS Surveillance Report. 1998, 9: 28-Google Scholar
- Moorman AC, Von Bargen JC, Palella FJ: Pneumocystis carinii pneumonia incidence and chemoprophylaxis failure in ambulatory HIV-infected patients. HIV Outpatient Study (HOPS) Investigators. J Acquir Immune Defic Syndr Hum Retrovirol. 1998, 19: 182-8.View ArticlePubMedGoogle Scholar
- Mocroft A, Lederberger B, Katlama C, EuroSIDA study group: Decline in AIDS and Death rates in the EuroSIDA study: an observational study. Lancet. 2003, 362: 22-39. 10.1016/S0140-6736(03)13802-0.View ArticlePubMedGoogle Scholar
- Ives N, Gazzard BG, Easterbrook PJ: Changing pattern of AIDS defining illnesses with introduction of Highly active antiretroviral therapy in a London clinic. J Infect. 2001, 42: 134-9. 10.1053/jinf.2001.0810.View ArticlePubMedGoogle Scholar
- Sabin C, D'Arminio Monforte A, May M: The Changing incidence of Clinical AIDS events in 12,574 treatment naive patients starting HAART [abstract]. 10th Conference on Retroviruses and Opportunistic Infections; Boston, MA. 2003, 569-Google Scholar
- Yeni PG, Hammer SM, Hirsch MS: Treatment for adult HIV infection: 2004 recommendations of the International AIDS Society-USA Panel. JAMA. 2004, 292: 251-65. 10.1001/jama.292.2.251.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.