CONVERT

Background

Phase 3 open-label RCT (CONVERT). 547 patients with limited-stage SCLC. Enrolled at 73 centers in 8 countries (April 2008–November 2013). Designed to test whether QD 66 Gy was superior to BID 45 Gy (the INT-0096 standard). Chemotherapy: cisplatin 75 mg/m² day 1 + etoposide 100 mg/m² days 1–3 × 4–6 cycles (q21 days). RT started with cycle 1 after day 22.

Interventions and follow up

Arm A: 45 Gy in 30 fractions of 1.5 Gy BID over 19 days + cisplatin/etoposide × 4–6 cycles (N=274)
Arm B: 66 Gy in 33 fractions of 2 Gy QD over 45 days + cisplatin/etoposide × 4–6 cycles (N=273)
Primary endpoint: OS (superiority of QD over BID; 12% higher 2-yr OS required)
mFollow up: Median 45 month

Results

Median OS: 30 months (BID) vs 25 months (QD), HR 1.18 (95% CI 0.95–1.45), P=.14 — not significant
2-yr OS: 56% (BID) vs 51% (QD), absolute difference 5.3% (95% CI −3.2% to 13.7%)
Grade ≥3 esophagitis: 19% (BID) vs 19% (QD) — identical
Grade 4 neutropenia: 49% (BID) vs 38% (QD), P=.05

Adverse events

Main adverse events: Grade 3–4 esophagitis: 19% in both arms — identical, significantly less than INT-0096 BID arm (27%), reflecting improved supportive care. Grade 4 neutropenia higher with BID (49% vs 38%). Treatment-related deaths: 3 (BID) vs 8 (QD). Most toxicities were comparable between arms with modern RT planning and supportive care.

Conclusions

Twice-daily RT (45 Gy/30 fx BID) was not inferior to once-daily RT (66 Gy/33 fx QD) for limited SCLC — since the trial was designed to show superiority of QD and failed, BID remains the reference standard. However, both regimens showed similar esophagitis rates (~19%) and comparable survival, suggesting QD 66 Gy is a valid alternative when BID scheduling is not feasible.

Key Limitations

Key Limitations: Powered for superiority of QD — NOT an equivalence or non-inferiority trial; failure to show QD superiority does not prove equivalence. The numerically better OS in the BID arm (30 vs 25 months) suggests BID may actually be more efficacious, though the difference is not statistically significant. Modern supportive care reduced esophagitis rates vs INT-0096 for both arms. Lack of consistent staging (CT vs PET) across international sites. The BID arm's numerical OS advantage is consistent with the BID hypothesis from INT-0096.

Clinical Context

CONVERT confirmed that BID 45 Gy CRT remains the reference standard for limited SCLC — the QD 66 Gy approach did not demonstrate superiority. However, QD 66 Gy is now considered acceptable for patients for whom BID scheduling is impractical (e.g., long travel distance, elderly patients). Current standard of care: cisplatin/etoposide × 4–6 cycles + 45 Gy BID starting cycle 1, followed by PCI if CR. Ongoing trials are evaluating higher RT doses with proton therapy.

References

References: Faivre-Finn C et al, Lancet Oncol 2017 (CONVERT)