Increasing patient compliance in SPRING-1 phase IIb HIV clinical trial in ART-naïve patients with dolutegravir (DTG, S/GSK1349572)

Background: To maximise generalisability of large-scale clinical trials, recruitment and retention of a diverse patient population is key. With commonly used algorithms (Snapshot, TLOVR, missing=failure), all dropouts in HIV studies, regardless of reason, are classed as failures in efficacy analyses. Good patient compliance therefore improves statistical power and the quality of trial results.
Methods: In four recent GSK/ViiV-sponsored phase IIIB/IV HIV ART-naïve studies, 24% of subjects withdrew and approximately 2/3 of these withdrawals (16%) were potentially avoidable (i.e. not treatment related). With the aim of increasing future subject retention, new tools were developed and trialled in the DTG Phase IIb treatment-naïve study, SPRING-1. Retention tools developed included opt-in text-messaging for study visit reminders, patient compliance support materials and presentations on prior withdrawal rates and associated risk factors at investigator meetings.
Results: Week 48 withdrawals, both overall and for potentially avoidable reasons, were considerably lower in SPRING-1 than historical HIV trials. 27% of subjects in 13 out of 35 sites in SPRING-1 opted to receive text-messaging visit reminders.

StudyNWeekWithdrawals from study n (%)Withdrawals for unavoidable reasonsWithdrawals for potentially avoidable reasons
HEAT/KLEAN/APV109141/ARIES~229036-96552 (24%)179 (8%)373 (16%)
SPRING-12052411 (5%)7 (3%)4 (2%)
SPRING-12054813 (6%)8 (4%)5 (2%)
[Comparing Withdrawal Rates ]

~ Pooled data from EPZ104057 (HEAT), ESS100732 (KLEAN), APV109141 and EPZ108859 (ARIES)
Conclusion: Incorporating specific patient compliance initiatives in SPRING-1 contributed to lower percentages of withdrawals than historical HIV studies. Additional patient compliance tools are in use in Phase III studies of DTG, where patient compliance may be more challenging due to increased numbers of sites and subjects. Identifying the specific impact of patient compliance tools is acknowledged as difficult since patient compliance can be influenced by overall study design, investigational product tolerability profile, current standard of care and treatment access for the disease under study.

J. Harris1, S. Hughes2, C. Brothers3
1GSK Research & Development, Stockley Park, United Kingdom, 2ViiV Healthcare, Brentford, United Kingdom, 3GSK Research & Development, RTP, United States