SMART study halted due to safety concerns with significantly more AIDS events in the treatment interruption arm

Simon Collins, HIV i-Base

The Executive Committee for the SMART study, which randomised close to 5,500 patients to either use continuous HAART, or to interrupt HIV treatment after achieving CD4 increases to >350 cells/mm3, and to restart if CD4 count subsequently fell to <250 cells/mm3, have stopped further recruitment due to safety concerns. People on the treatment interruption arm will receive a recommendation that it is prudent to restart treatment.

Study investigators have been contacted with an executive summary of the data. Patients will be notified of the implications of these results by their doctors. A public presentation and more detailed analysis of these data will be the subject of a late breaker session at the 13th Conference of Opportunistic Infections and Retroviruses, to be held in mid-February in Denver.

Although the SMART Data and Safety Monitoring Board (DSMB) agreed in November 2005 that the study was safe to continue, a more recent review of the data showed a statistically significant difference in serious AIDS events between the continuous treatment arm and the treatment interruption arm.

A higher rate of events in the interruption arm was expected, and allowed for in the study design. However, the design also expected a reduced number of serious organ events that might be attributed to continuous ARV treatment. The analysis in January also showed a higher rate of major organ complications including cardiovascular, renal and hepatic events in the interruption arm. Continuous treatment was therefore found to be safer than originally assumed when the study was designed.

The DSMB made a recommendation to stop the study, and the Executive Committee, after reviewing the unblinded dataset, agreed to follow this recommendation.

The steering committee have still to make a formal announcement as we went to press but the investigator letter detailed: ‘The interim data over a mean follow-up of 15 months indicate that there is approximately a 2-fold increased risk of disease progression in the discontinuation group compared to the [continued treatment] group (3.5 versus 1.7 events per 100 person years). Furthermore, there is no evidence of short-term benefit of the discontinuation strategy in reducing the risk of major cardiovascular and metabolic adverse events’.

At present, UK sites, investigators and community representatives have been notified.

Recommendations are that:

  • further recruitment is stopped immediately
  • data will continue to be collected as per protocol
  • patients in the discontinuation arm who are currently off-treatment will be asked to dicsuss these results with their doctor, and that the current analysis of the data show that it is prudent to restart treatment.


This final recommendation is probably the most immediate practical issue, as many patients may be stable and well off-treatment, maintaining a high CD4 count that is well above the trials determined protocol for restarting treatment.

These results show that STIs as defined and applied in SMART are not safe. They do not prevent further research into treatment interruptions however, and an amended protocl for the current study may emerge from further amalysis of the data, and this could help define which patients may be at lower risk from a interruption of treatment.

One of the first messages that is also likely to be widely circulated, and which has a impact for the pharmaceutical industry is that ‘treatment is for life’. This may not be the most useful or relevant conclusion to draw from the decision that has been made to stop the study.

The SMART trial is the largest HIV study, with the longest planned follow-up. That it has already been able to answer the main study question, before recruitment was completed, and after less than two of the planned seven years follow-up, actually supports the importance of this research. This international collaboration has addressed key management questions promptly and carefully. The study would have generated an important long-term data-set, and this may still be possible to some degree.

Any further speculation is not possible without seeing the detailed study results. It appears that the study has already shown that the benefits of continued treatment with currently used treatments, outweigh the risks of treatment interruption, over a short period f fllow-up than was initially anticipated.

This highlights the importance of large-scale, well planned and coordinated international trials, and as such, the investigators should be supported.

This study was strongly supported by the patient community and the success of the investigators in answering important questions on patient management, through such a large international, randomised trial, should also be recognised.

The SMART cohort also initiated several important sub-studies, including body composition, Quality of Life, risk behaviour, and more recently, neurology and anal dysplasia.

Sources: NIH press release and personal communication with MRC investigators

SMART Strategies for Management of Anti-Retroviral Therapy

National Institute of Allergy and Infectious Disease

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