Pure Pathologic Response Is Associated with Improved Overall Survival in Patients with Advanced Systemic Mastocytosis Receiving Avapritinib in the Phase I EXPLORER Study

Introduction: Advanced systemic mastocytosis (AdvSM) comprises a heterogeneous group of clonal mast cell neoplasms, primarily driven by KIT D816V. Measures of AdvSM response, including the International Working-Group for Myeloproliferative Neoplasms Research and Treatment and European Competence Network on Mastocytosis (IWG-MRT-ECNM) criteria, are based on improvements in mast cell-related organ damage (C-findings), and further sub-classified by the extent of reduction in measures of mast cell disease (e.g. serum tryptase level, bone marrow mast cell burden). However, assessment of some C-findings lacks precision (such as splenomegaly and its resolution). Normalization of C-findings may not be an adequate surrogate for important clinical outcomes such as overall survival (OS). We evaluated whether pure pathologic response (PPR) criteria based on changes in bone marrow mast cells, serum tryptase, and complete blood count was more closely correlated with OS compared to the modified IWG-MRT-ECNM (mIWG-MRT-ECNM) criteria.

Methods: As an exploratory post-hoc analysis of the phase 1 EXPLORER study of avapritinib in AdvSM, we evaluated responses lasting ≥12 weeks by both mIWG-MRT-ECNM and PPR criteria. At baseline, evaluability for mIWG-MRT-ECNM response required ≥1 evaluable C-findings; PPR required presence of bone marrow mast cell aggregates and/or serum tryptase ≥20 ng/mL. Per PPR, morphologic complete remission (mCR) is absence of bone marrow mast cell aggregates, serum tryptase <20 ng/mL and full (or partial [mCRh]) hematologic recovery; morphologic partial remission (mPR) is ≥50% reduction in bone marrow mast cells and serum tryptase level. OS was analyzed by Kaplan–Meier method and was time from first dose to death. OS comparisons were by log-rank test, performed for best response and landmark analyses at various cycles.

Results: As of the data cut-off of August 30, 2019, 80 patients enrolled including 62 with AdvSM (7 with aggressive SM [ASM], 44 SM with an associated hematologic neoplasm [SM-AHN] and 11 with mast cell leukemia [MCL]). Ten (16%) AdvSM patients (7 ASM, 3 SM-AHN) were not response evaluable (RE) per mIWG‑MRT‑ECNM criteria, due to a lack of an evaluable C-finding at baseline, and 4 additional AdvSM patients were recently enrolled and were not yet response evaluable.

Of the 48 RE patients (3 ASM, 35 SM-AHN and 10 MCL) the best overall response rate (ORR) per mIWG-MRT-ECNM was 77% (8% CR, 19% CRh, 42% partial response [PR], and 8% clinical improvement [CI]). Non-responders had stable disease (SD; 21%) or were not evaluable (NE) due to insufficient (<13 weeks) follow-up (2%). Responders (CR/CRh/PR/CI) had 18-month OS of 85% (CR/CRh, 100%; PR, 77%; CI, 75%; Figure 1A); non-responders had 18-month OS of 48% (SD, 53%; NE, 0%) (P=0.042). Per PPR criteria, the best ORR was similar at 79%; however, a greater proportion of patients were assessed as being in a complete remission (15% mCR, 27% mCRh and 38% mPR). Non-responders by PPR all had SD (21%). This demonstrates that elimination of measurable mast cell burden can be discordant with complete C‑finding resolution. Responders (mCR/mCRh/mPR) by PPR had 18-month OS of 88% (mCR/mCRh: 100%; mPR: 72%; Figure 1B); non-responders (all SD) had 18-month OS of 21% (P=0.0001).

Eventually, all 62 AdvSM patients will be evaluable by PPR criteria, including those 10 patients without mIWG evaluable C-findings at baseline; however, 5 patients had insufficient follow-up at the time of analysis. For the 57 AdvSM patients with sufficient follow up, the best ORR per PPR criteria was similar at 77% (14% mCR, 26% mCRh and 37% mPR). Overall, no patients had a best response of progressive disease based on mIWG or PPR criteria.

Landmark analyses of PPR at the end of 6 cycles showed a trend in 18-month OS of mCR/mCRh>mPR>SD in patients with similar starting avapritinib doses of ≥200 mg daily (n=48 of 57 PPR-evaluable patients).

Conclusions: In the phase I EXPLORER study, response assessment in AdvSM using PPR criteria increases the evaluable population, significantly correlates with OS, and should be explored as a potential primary endpoint for future trials.