AUTHORS
Saubia Fathima, MD, Abirame Gopinath, MBBS2, Amina Akram, MBBS3, Ali Alsugair, MBBS4, Priyansh Faldu5, Rania Abdelaziz6, James Foran, FRCPC7, Abhishek Avinash Mangaonkar, MBBS8, Mark R. Litzow, MD9, Hemant Murthy, MD10, Lisa Sproat, M.D.11, Jeanne Palmer, MD12, Aasiya Matin, MBBS13, Ernesto Ayala, MD14, Madiha Iqbal, MBBS, MD15, Nandita Khera, MD16, Mehrdad Hefazi Torghabeh, MD17, Saurabh Chhabra, MD, MBBS18, Kebede Begna, M.D.19, Animesh Pardanani, PhD, MBBS20, William Hogan, M.B., B.Ch.21, Mithun Shah, MD, PhD22, Naseema Gangat, MBBS23, Hassan B Alkhateeb, M.D.24, Ayalew Tefferi, MD25
AFFILIATIONS
1 Mayo Clinic, Rochester, United States
2 Mayo Clinic, Rochester, United States
3 Mayo Clinic, Rochester, United States
4 Mayo Clinic, Rochester, United States
5 Mayo Clinic, Rochester, United States
6 Mayo Clinic, Rochester, United States
7 Mayo Clinic, Jacksonville, Florida, United States
8 Mayo Clinic, Rochester, United States
9 Mayo Clinic, Rochester, United States
10 Mayo Clinic, Jacksonville, Florida, United States
11 Mayo Clinic, Phoenix, Arizona, United States
12 Mayo Clinic, Phoenix, Arizona, United States
13 Mayo Clinic, Rochester, United States
14 Mayo Clinic, Jacksonville, Florida, United States
15 Mayo Clinic, Jacksonville, Florida, United States
16 Mayo Clinic, Phoenix, Arizona, United States
17 Mayo Clinic, Rochester, United States
18 Mayo Clinic, Phoenix, Arizona, United States
19 Mayo Clinic, Rochester, United States
20 Mayo Clinic, Rochester, United States
21 Mayo Clinic, Rochester, United States
22 Mayo Clinic, Rochester, United States
23 Mayo Clinic, Rochester, United States
24 Mayo Clinic, Rochester, United States
25 Mayo Clinic, Rochester, United States
ABSTRACT
Background
Allogeneic hematopoietic stem cell transplantation (ASCT) remains the only curative treatment modality for primary (PMF) and secondary (SMF) myelofibrosis. Limited studies have evaluated predictors of post-transplant outcomes in MF but accurate interpretation of the results was confounded by inclusion of patients with both PMF and SMF as well as chronic- (PMF-CP) and accelerated-phase (PMF-AP) disease. The objective of the current study was to identify clinical and genetic predictors of post-transplant overall survival (PT-OS), non-relapse mortality (NRM), and relapse, in a carefully curated cohort of PMF-CP, in whom ASCT was performed before disease transformation into PMF-AP or blast phase PMF (PMF-BP).
Methods
The study cohort was selected from an institutional database of 1,036 patients with PMF based on documentation of ASCT prior to disease transformation into PMF-AP or PMF-BP. Clinical, genetic and transplant-related variables were collected at the time of ASCT. PT-OS and relapse-free survival (RFS) were estimated by Kaplan-Meier analysis with Cox proportional hazards modeling for multivariable analysis (MVA). Receiver operator curves at 2 and 5 years were used to identify relevant cut offs. Competing-risk methods and Fine-Gray regression were used to evaluate relapse and NRM.
Results
The core study cohort included 176 patients with PMF-CP (median age 63 years, range 30-77, 65% male). Driver mutation distribution included JAK2 (60%), CALR (21%), MPL (9%) and triple-negative (10%). Median time from diagnosis to ASCT was 17 months (IQR 8-50). At a median post-transplant follow-up of 61 months, 50 (28%) deaths and 28 (17%) relapses were documented. Day-100 and one-year mortality rates were 6% and 14%, respectively.
Survival analysis
Median PT-OS was 143 months (95% CI 100- not reached) and RFS 135 months (95% CI 79-159). In univariate analysis, age ≥60 years (HR 2.9, p<0.01), Hematopoietic Cell Transplantation specific Comorbidity Index (HCT-CI) ≥4 (HR 2.3, p=0.02), spleen size >20 cm (HR 1.96, p=0.04), non-myeloablative (NMA), compared to myeloablative or reduced intensity conditioning (HR 2.5, p=0.02), haploidentical/mismatched unrelated donor (HR 1.2, p=0.02), DNMT3AMUT (HR 3.2, p=0.03), TP53MUT (HR 3.6, p=0.01), and bone marrow (BM) blasts ≥5% at ASCT (HR 2.8, p=0.03) were associated with inferior PT-OS. On MVA, TP53MUT (HR 4.4, p<0.01), HCT CI ≥4 (HR 2.3, p=0.04) and NMA conditioning (HR 2.5, p=0.04) remained significant. On the other hand, specific driver mutational status, high risk mutation status, cytogenetics, prior JAK2 inhibitor use, and use of post-transplant cyclophosphamide did not impact PT-OS (p>0.1 in all instances).
Non-relapse mortality
Cumulative incidences of NRM at 1, 2, and 5-years were 13%, 15% and 20% respectively. In univariate analysis, 5-year NRM was higher in patients with age ≥60 years (39% vs. 28%, p <0.01), HCT CI ≥4 years (32% vs. 17%, p =0.03), myeloablative conditioning (23% vs. 6%, p=0.02), platelet count <150 X 109 (24% vs 12%, p=0.03), BM blasts ≥5% ( 51% vs. 18%, p=0.03), DNMT3AMUT (57% vs 21%, p<0.01), and TP53MUT (50% vs 23%, p=0.04). On MVA, TP53MUT (HR 6.4, p<0.01), platelet count <150 X 109 (HR 2.7, p=0.04) and age ≥60 years (HR 4.0, p=0.05) remained significant.
Relapse
Cumulative incidences of relapse at 1, 2, and 5-years were 6%, 12% and 16% respectively. In univariate analysis, 5-year relapse rates were lower in patients with triple negative MF (0% vs. 18%, p =0.02), and higher with haploidentical donor (41% vs. 13%, p =0.03), NMA conditioning (39% vs. 15%, p<0.01), and SF3B1MUT (43% vs 17%, p<0.01). On MVA, triple-negative mutational status (no relapses observed in this subgroup, p<0.01) and SF3B1MUT (p=0.01) remained significant.
TP53MUT outcomes
Pre-transplant, TP53MUT was detected in 6 patients, including 3 with multihit TP53MUT; the latter, compared to the former, displayed shorter median PT-OS (5 months vs. not reached) and 2-year PT-OS (0% vs. 67 %; p=0.4).
Conclusion
Post-transplant survival in PMF-CP is adversely affected by TP53MUT, HCT CI ≥4, and use of NMA conditioning. NRM was also affected by TP53MUT, as well as by older age and thrombocytopenia while SF3B1MUT was associated with higher relapse rate. Conversely, triple-negative mutational status was associated with lower risk of relapse. Prospective confirmation of these findings is warranted and could improve pretransplant risk assessment.
DISCLOSURES
Saubia Fathima, MD: No relevant conflicts of interest to declare.