Background

The MAVORIC phase 3, randomized trial, compared safety and efficacy of mogamulizumab (Moga) with vorinostat (Vori) in patients (pts) with relapsed/refractory mycosis fungoides (MF) or Sézary syndrome (SS). Among Moga-treated pts, global overall response rate (ORR) was 28% (52/186), with median duration of response (DOR) of 14.1 months (mo). This analysis assessed the clinical and demographic characteristics of MAVORIC cohorts with varying ORR durations.

Methods

This post hoc analysis, divided pts into 4 cohorts based on minimum DOR (≥4, ≥6, ≥8, and ≥12 mo; ORR4, 6, 8, and 12, respectively). Linear regression analysis (odds ratios [ORs] for ORR12 response) and stepwise multivariate analysis were performed for gender, baseline ECOG Performance Status, disease type, disease stage (IB-IV), blood involvement, baseline CCR4 expression, age, time from diagnosis, mSWAT, and LDH. Blood samples at multiple time points after blood CR were collected from two SS ORR12 patients; malignant TCR frequency was monitored with the clonoSEQ next generation sequencing platform (Adaptive Biotechnologies; Weng 2013). Samples were assessed at baseline and every 3-6 mo to monitor molecular minimal residual disease (MRD).

Results

Response rates by disease compartment for pts treated for ≥12 mo with Vori and Moga are shown in the Table. Among Moga-randomized pts (n=186), ORRs lasting ≥4, ≥6, ≥8, and ≥12 mo were seen in 25.3%, 21.0%, 16.1%, and 10.8%, respectively. When baseline characteristics of ORR12 pts were compared with all other Moga-treated pts, ORR12 pts were more likely to have SS (P=0.016, OR 0.29), stage IVA1 disease (P=0.0002, OR 11.13), and blood involvement (P=0.03, OR 0.19). There was no correlation between baseline skin CCR4 expression and ORR12. When ORR12 pts were compared to those with shorter response, long-term responders were more likely to have stage IVA1 disease (P=0.006, OR 7.3). Stepwise multivariate analyses confirmed SS as a significant long-term response predictor. In 2 ORR12 SS pts from a single center (best response global CR), molecular MRD was monitored after clearance of blood Sézary cells by routine flow cytometry. One patient showed blood CR after 1 Moga cycle and in skin at cycle 3. At last follow-up, malignant TCR sequences in blood remained barely detectable at <1 copy/million nucleated cells, confirming deep remission in blood lasting ≥47 mo. Another patient achieved blood CR at cycle 1 and skin CR at cycle 3. Moga was discontinued after 16 cycles due to surgery but global CR was maintained at last follow-up with mostly non-detectable MRD (<1 copy/million nucleated cells) in blood 63 mo from first dose—50 mo since discontinuing treatment.

Conclusion

Compared with patients with shorter responses, MAVORIC trial pts who achieved long-term (≥12 mo) responses had SS (stage IVA1) or blood involvement. Available MRD analyses on ORR12 pts indicate the ability of Moga to produce lasting and deep responses in some pts.