Why START would make an excellent long-term cohort

Nathan Geffen, CSSR

Results from the international START study were so important that rapidly changed national and international treatment guidelines for routine management of HIV; antiretroviral treatment (ART) is now routinely recommended even at high CD4 counts.

These results were 18 months earlier than expected, and although everyone in the study now has been recommended to start ART, follow-up with continue for all participants until the end of 2017.

But as a unique cohort – never to be available again in the future – there might be important reasons to follow up these people for longer still in a cohort study.

Long-term cohort groups of people living with HIV have provided useful information on life-expectancy, morbidity and the role of factors such as earlier HIV treatment, smoking and differences in health indicators based on mode of infection. (1–6) Long-term cohort groups do not have the evidential value of clinical trials, but they can provide insight into questions that are too expensive, impractical or unethical to answer in clinical trials.

The Strategic Timing of Antiretroviral Treatment (START) study is a randomized controlled clinical trial that has compared early versus deferred antiretroviral treatment. In May 2015, participants on the deferred arm began to be offered immediate treatment, after the study question was answered, about a year and-a-half before the trial was scheduled to complete. The primary end-point – any serious AIDS related event, serious non–AIDS-related event, or death from any cause – was reached in 42 patients on the immediate arm versus 96 on the deferred one (HR: 0.43; 95%CI: 0.3-0.62; p<0.001). (7)

START has features that collectively would make it a unique long-term follow-up cohort. [7]

  • It is large; START has 4,685 participants.
  • It is geographically widespread; there are 215 sites in 35 countries.
  • The participants are diverse by measures of sex (27% female), sexual orientation (38% reported acquiring HIV from a person of the opposite sex), by self-reported race or ethnic group (9% Asian, 30% Black, 14% Latino or Hispanic, 44% white), reported income (46% high vs 54% moderate or low), smoking status (32% smokers) and mode of acquisition (64 participants acquired HIV through IDU and 5% acquired HIV through blood products or other means, including unknown).
  • It has comprehensive patient baseline and monitoring data and excellent follow-up capacity, better than most of the large cohorts.
  • Because of the study design the participants were relatively healthy and early in HIV infection when they enrolled, with a median time of one year since infection, a median baseline CD4 count of 651 cells/mm3 [IQR: 584-765] and a median 10 year cardiovascular event risk of just under 2%.
  • Participants have been taking ART regimens combinations with negligible risk of lipoatrophy, lipohypertrophy, neuropathy, pancreatitis and lactic acidosis. This was mostly tenofovir/emtricitabine/efavirenz (75% in the immediate arm, with less use over time of efavirenz (approximately 50% of people in the deferred arm),
  • Many participants have been placed on newer drug regimens. This provides an opportunity to study possible long-term toxicities.
  • The study has already collected an large number of urine and blood samples. Typically five blood vials and a urine sample have been taken from each participant every four months. Use of these as part of a long-term follow-up cohort would require further participant consent.

Potential to learn from START as an observational cohort

The most compelling reason for follow up would be continued comparison of morbidity and mortality between the two arms. This is a randomised cohort that due to the study findings will not be unethical to ever run again.

Rates of drug resistance and treatment failure, as well as the rate at which mutations occur, can be measured. There is also potential to examine long-term outcomes broken down by sex, mode of acquisition, geographical area, income, self-identified race and smoking status. Comparisons of the age/disease profiles that affect the participants could potentially be compared to background populations, albeit that such analyses should be treated with caution.

START would make an ideal long term follow up observational cohort with the potential for important insights in the same tradition as other cohorts like EuroSIDA, IeDEA, Multicenter AIDS Cohort Study, Danish HIV Cohort Study and Swiss HIV Cohort Study. (8–12)

The study is due to terminate at the end of 2016, and participants are expected to move to other treatment institutions in 2017. It would be unfortunate if in 2017 the opportunity was lost to turn START into a long-term cohort.

Many of the participants might prefer the opportunity to continue getting a high standard of care in a clinical trial environment as well as the opportunity to continue contributing to scientific knowledge, so it is unlikely there will be resistance to turning START into a long-term cohort from most participants.

No doubt there will be some complex logistical and funding requirements to make this work. Nevertheless these challenges should be manageable. The scientific benefits and consequent benefits to people with HIV of making START work as a long-term observational cohort could be immense.

Nathan Geffen is a member of the Community Advisory Board for the START Study.


  1. Sterne JAC et al. Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort study. Lancet Lond Engl. 2005 Aug 30;366(9483):378–84.
  2. Lohse N et al. Survival of persons with and without HIV infection in Denmark, 1995-2005. Ann Intern Med. 2007 Jan 16;146(2):87–95.
  3. Sewankambo NK et al. Mortality associated with HIV infection in rural Rakai District, Uganda. AIDS Lond Engl. 2000 Oct 20;14(15):2391–400.
  4. Glynn JR et al. Survival from HIV-1 seroconversion in Southern Africa: a retrospective cohort study in nearly 2000 gold-miners over 10 years of follow-up. AIDS Lond Engl. 2007 Mar 12;21(5):625–32.
  5. Helleberg M et al. Risk of cancer among HIV-infected individuals compared to the background population: impact of smoking and HIV. AIDS Lond Engl. 2014 Jun 19;28(10):1499–508.
  6. Reekie J et al. Regional differences in AIDS and non-AIDS related mortality in HIV-positive individuals across Europe and Argentina: the EuroSIDA study. PloS One. 2012;7(7):e41673.
  7. INSIGHT START Study Group. Initiation of Antiretroviral Therapy in Early Asymptomatic HIV Infection. N Engl J Med. 2015 Aug 27;373(9):795–807.
  8. Council FD. EuroSIDA [Internet]. 2011 [cited 2015 Aug 27].
  9. IeDEA [Internet]. IeDEA. [cited 2015 Aug 27].
  10. Multicenter AIDS Cohort Study (MACS) | UCLA AIDS Institute [Internet]. [cited 2015 Aug 27].
  11. Omland LH et al. Cohort profile update: the Danish HIV Cohort Study (DHCS). Int J Epidemiol. 2014 Dec;43(6):1769–1769e.
  12. Introduction – Swiss HIV Cohort Study [Internet]. [cited 2015 Aug 27].

Links to other websites are current at date of posting but not maintained.