Is it safe to drop CD4+ monitoring among virologically suppressed patients

For over two decades, the measurement of CD4+ cell count has been the principal means of assessing eligibility for initiation of antiretroviral therapy (ART) and monitoring response to treatment [1]. In high-income settings, treatment response is also supported by routine virologic monitoring as the preferred way to assess adherence and detect virologic failure early and accurately [2,3]. In low-income, high HIV burden settings, access to viral load monitoring remains limited due to the complexity and cost of current technology. However, there has been a progressive evolution favouring increased use of viral load monitoring, and the latest guidelines from the WHO in June 2013 recommend that countries use viral load as the preferred approach to monitoring response to ART [4]. In support of this recommendation, major international agencies are working with countries to support scale-up of viral load capacity in resource-limited settings [5]. In settings where viral load is routinely available, the added value of CD4+ monitoring is increasingly being questioned [6]. Several studies from high-income settings have suggested that CD4+ monitoring provides little additional information in situations where viral load is available and patients are virologically suppressed [7,8]. However, data from low-income, high HIV burden settings are lacking. Differences in immune status at baseline, antiretroviral drug regimens, healthcare delivery, and background opportunistic infections mean that data from high-income settings may not be readily applicable to high HIV burden resource-limited settings. We assessed CD4+ changes among virologically suppressed patients in a routine programme setting in South Africa. Starting in 2001, the Khayelitsha programme is one of the largest and longest standing treatment programmes in South Africa, serving a population of around 500 000 people through 11 primary healthcare facilities. The majority of patients on ART are managed by nurses, with adherence clubs for stable patients facilitated by lay counsellors [9,10]. First-line adult ART regimens have varied over time, with zidovudine (ZDV)/lamivudine (3TC)/efavirenz (EFV) being the predominant regimen in the first 3 years, with stavudine replacing ZDV in 2004, which in turn was replaced by tenofovir. By 2010, TDF/3TC/EFV was used as the first-line ART regimen for the majority of adult patients [11]. Prospective patient-level data have been collected since the start of the programme, using electronic capture of clinical records. We identified adult patients (≥16 years old) started on ART between 2001 and 2012, who achieved a viral load below 400 copies/ml and a CD4+ cell count above 200 cells/μl between 9 and 15 months on ART; and who subsequently (from 21 months onwards) had paired (within a month of each other) viral load and CD4+ cell count measurements while remaining virologically suppressed. Patients were censored if duration between viral loads exceeded 15 months. The proportion of patients maintaining CD4+ cell counts above 200 cells/μl were described by duration on ART. The time to a drop in CD4+ cell count to below 200 cells/μl was assessed using Kaplan–Meier survival curves (Fig. 1).Fig. 1: Kaplan–Meier survival estimates of sustained immunologic recovery.A total of 14 792 adults were followed for 2 years or more, of whom 7250 met the inclusion criteria, and 5697 (79%) additionally had at least one eligible set of paired CD4+ cell count and viral load data for evaluation. Median age at ART initiation was 34 years [interquartile range (IQR) 29–40 years]. Women constituted two-thirds of cohort enrolment over time and were younger at ART initiation than men (median age 32 vs. 37 years; P < 0.001). Median baseline CD4+ cell count increased from 63 cells/μl (IQR 22–122 cells/μl) between 2001 and 2003 to 225 cells/μl (IQR 127–300 cells/μl) in 2012, reflecting both increasing trends in enrolment and South African national guideline changes in CD4+ cell count threshold. Of the 17 991 paired results, 230 CD4+ cell counts (1.3%) were below 200 cells/μl. This proportion was higher earlier on ART (1.5, 1.0 and 0.9% before 3 years, 5 years and subsequently). At 2, 5 and 10 years on ART, 99.3, 95.8 and 92.9% of patients with ongoing virologic suppression maintained CD4+ cell counts continuously above 200 cells/μl. This proportion increased in patients with higher 12-month CD4+ cell counts (Fig. 1). A follow-up CD4+ cell count was available in 137 of the 230 patients falling below 200 cells/μl, and had returned above this value in the majority (133, 97%) of cases. In this study, only a very small proportion of virologically suppressed adults with initial immunological recovery had subsequent low CD4+ cell counts, most of which increased again without intervention. These data are consistent with data from randomized trials and observational cohorts which suggest that once HIV-infected patients on ART are virologically suppressed, CD4+ cell count does not decline over time in the vast majority of patients [7,8]. Strengths of this study include the large cohort size, long duration of follow-up, and data which come from a routine programme setting in a high HIV prevalence setting with a high rate of HIV/tuberculosis (TB) co-infection. There is growing appreciation that, from a public health management perspective, CD4+ monitoring adds little additional value to viral load monitoring once patients are stable on ART. Guidelines issued by the Southern African HIV Clinicians Society in 2013 recommend that in patients being monitored with viral loads, CD4+ cell count testing can be stopped once the CD4+ cell count is above 200 cells/μl and the viral load is suppressed; CD4+ cell count should be repeated if virologic or clinical failure occurs [12]. The latest US guidelines for the management of persons infected with HIV recommend that CD4+ measures can be reduced among patients who are virologically suppressed, from 3–4 months to 6–12 months [13–15]. CD4+ cell count at baseline continues to be important for initial clinical management decisions, including screening and prophylaxis for major opportunistic infections, and stratifying risk among patients presenting for care. CD4+ cell count will also continue to play an important role in determining ART eligibility, despite a trend towards an increasing number of CD4+-independent ART initiation scenarios, and will also likely be required for a small minority of patients to support individual clinical decision-making, for example, for those who fail to respond to treatment or patients taking concomitant immunosuppressive therapy. However, once treatment has been initiated, the key focus is ensuring virologic suppression. Research is needed to determine the optimal clinical and virological criteria for stopping CD4+ following ART initiation. In conclusion, these data support considerations to reduce the frequency of CD4+ cell counts, or eliminate them altogether, from routine use once patients are stable on ART in settings where both CD4+ and viral load are routinely available. Acknowledgements Author's contributions: N.F., K.S., and A.B. conceived the study. K.S. and A.B. analysed the data. All authors contributed to writing, critical revision, and finalization of the manuscript. Conflicts of interest There are no conflicts of interest.