Abstract P008

Brentuximab Vedotin (BV) Exposure and Long-Term Efficacy Analysis in Patients With Classical Hodgkin’s Lymphoma (cHL): Analysis of Phase 3 ECHELON-1 Study

Background: In the phase 3 E1 (NCT01712490) study, BV vs bleomycin in combination with doxorubicin, vinblastine, and dacarbazine (A+AVD vs ABVD) showed superior overall survival (OS; HR, 0.59; 95% CI, 0.40-0.88; P=.009) in previously untreated stage III or IV cHL. BV dose adjustments, including dose modifications (eg, reduction) and discontinuations, were recommended for managing adverse events (AEs), including peripheral neuropathy (PN). We evaluated the impact of dose adjustments on efficacy by exploring the exposure-response (ER) relationships between BV and OS and progression-free survival (PFS).

Methods: In E1, pts were randomized 1:1 to receive A+AVD or ABVD for six 28-day cycles. Included pts had received ≥1 BV dose and had evaluable BV pharmacokinetic (PK) data (n=661). Average BV concentrations (Cavg) were estimated via a validated population PK model and used for ER analyses. Incidences of dose adjustments and grade ≥2 (G≥2) PN and duration of OS and PFS were compared across exposure quartiles and with the comparator ABVD arm (n=659). Univariate Cox regression analysis was used to assess ER relationships.

Results: Of 661 pts, 60.5% had BV dose modifications and 11.0% discontinued BV (Table). Median treatment duration was similar in pts with vs without BV dose modifications (25 vs 24 wk), suggesting manageable AEs. Lower BV exposure quartiles had higher dose modification rates, but discontinuation rates were relatively similar across the quartiles. Higher G≥2 PN incidences were observed in higher BV exposure quartiles; Cavg was predictive of G≥2 PN (P=8 x 10-6). A+AVD provided OS benefit in pts with (n=199) and without (n=462) G≥2 PN events, with estimated 6-year OS (95% CI) of 95% (91-98) and 93% (90-95), respectively, compared with 89% (87-92) with ABVD. OS and PFS benefits over ABVD were observed in all BV exposure quartiles, inclusive of dose adjustments. Average on-treatment BV exposure was not a statistically significant predictor for OS (P=.091), while higher early exposure (Cavg up to the end of cycle 2) was predictive of higher OS (P=.003).

Conclusion: At 6 years of follow-up, A+AVD provided benefit over ABVD for all BV exposure ranges, inclusive of dose adjustments for AE management. This showed that recommended dose adjustments effectively managed AEs, including G≥2 PN, while keeping most pts on treatment and subsequently maintaining survival benefits in E1. High initial BV exposure was associated with high probability of response.