Serious Pulmonary Toxicity with SGN-30 and Gemcitabine, Vinorelbine, and Liposomal Doxorubicin in Patients with Relapsed/Refractory Hodgkin Lymphoma (HL): Cancer and Leukemia Group B (CALGB) 50502

Background: SGN-30 is a chimeric anti-CD30 antibody that is well-tolerated but has modest single agent activity in classical HL. Synergistic cytotoxicity is observed in vitro when SGN-30 is combined with gemcitabine. In CALGB 59804, a phase II study of gemcitabine, vinorelbine, and liposomal doxorubicin (GVD), 70% of patients (pts) with HL responded (

Methods: 30 pts with CD30-positive classical HL relapsed/refractory after „□ 1 therapy were treated with GVD + 12 mg/kg SGN-30 or placebo. Pts were permitted to come off study after „□ 2 cycles for stem cell transplant (SCT). To assess the safety of the combination, 16 pts in part 1 of the trial received open label SGN-30 + GVD, without unexpected grade 3–4 toxicity during cycle 1. In part 2, 14 pts were randomized to SGN-30 + GVD (n=7) or placebo + GVD (n=7). After 2 pts developed late pulmonary events in part 1, part 2 was amended to stop the trial if the risk of grade 2–5 pulmonary events with SGN-30 + GVD compared to placebo + GVD exceeded 20%.

Results: In 30 pts, median age was 35 years (range, 19–78 years), median prior therapies 2 (range 1–4), 11 pts had undergone a previous autologous SCT, 12 pts had received radiation, and 12 pts were refractory to their last regimen. After a median of 3 cycles of therapy, ORR was 63% (32% complete response (CR)) in 30 pts, 65% (35% CR) in 23 pts receiving SGN-30 + GVD, and 58% (29% CR) in 7 pts receiving placebo + GVD. Median EFS was 9 months (7.8 months for those pts with prior SCT and not yet reached for pts without prior SCT), with no statistically significant differences in EFS with the addition of SGN-30. Grade 2–5 pneumonitis occurred in 5 pts, all receiving SGN-30 + GVD, leading to closure of the trial. In part 1, 1 pt (prior ABVD, ICE) developed grade 4 pneumonitis after 3 cycles with fevers, hypoxia, and ground glass infiltrates requiring mechanical ventilation. A 2nd pt (prior ABVD, spine XRT) died from pneumonitis after 2 cycles despite intubation, antibiotics, and steroids. In part 2, 2 pts (prior ABVD and prior ABVD, ICE, SCT, mediastinal XRT) developed grade 3 pneumonitis after 4 and 2 cycles, respectively, that improved with steroids, and 1 pt (prior ABVD) developed grade 5 pneumonitis after 5 cycles. Autopsy in this pt demonstrated mild interstitial fibrosis and Pseudomonas and Enterococcal lobar pneumonia without evidence of HL. In the 4 other pts, bronchoalveolar lavage (n=4) and transbronchial lung biopsy (n=2) demonstrated no evidence of infection or progressive HL. In univariate analysis, age, gender, number of prior therapies, time from last treatment, prior autologous SCT, prior XRT, use of GCSF, and the number of GVD cycles were not significantly associated with the development of grade 2–5 pneumonitis (p-values=0.34–1.00).

Conclusion: With grade 2–5 pneumonitis in 5/23 pts receiving SGN-30 + GVD, no events in the placebo arm (n=7), and historical results from CALGB 59804 demonstrating a 2% incidence of pulmonary events with GVD, the data from CALGB 50502 suggest that SGN-30 cannot safely be administered concurrently with GVD. Compared with historical GVD outcomes, the addition of SGN-30 to GVD does not appear to improve ORR or EFS.