EBMT 2018 | The Minnesota refined acute GvHD score is useful to identify high risk patients after haploidentical transplantation with post-transplant cyclophosphamide

During the 44^th Annual Meeting of the European Society for Blood and Marrow Transplantation (EBMT), Lisbon, Portugal, on Wednesday 21^st March 2018, Luca Castagna from Humanitas Cancer Center, Bone Marrow Transplant Unit, Rozzano, Italy, presented data on behalf his colleagues from their retrospective analysis.

Castagna began his talk by highlighting the importance of risk stratification after allogenic stem cell transplantation (allo-SCT) to accurately identify patients at higher risk of acute graft-versus-host disease (aGvHD). In a recent publication, McMillan and colleagues established a new clinical Minnesota refined risk score that was better than the IBMTR scale to predict overall response rate (ORR) to treatment and 6-month OS.¹ For the reason that this score was primarily applied in patients receiving either a MRD or a MUD/MMUD donor transplant, this analysis was set to identify whether the use of this scoring system may be also beneficial in patients undergoing haploidentical transplantation (haplo-SCT) and receiving post-transplant cyclophosphamide (PT-Cy). The analysis included 318 patients who underwent haplo-SCT with PT-Cy between 2009 and 2016.²

Patient characteristics:

• Median age = 56
• Graft sources:
  • Peripheral blood stem cells (PBSC) in 64 (73%) patients
  • Bone marrow (BM) in 23 (37%) patients
  • 56% of the patients received a reduced intensity (RIC) conditioning
  • 38% a non-myeloablative (NMA) conditioning
• Acute GvHD characteristics:
  • 87 patients (27%) had grade 2-4 aGvHD
  • 63 (72%) patients had only skin GvHD
  • 24 (28%) patients had skin and visceral aGvHD
  • 65 (75%) patients were classified as standard risk (SR)
  • 22 (28%) patients as high-risk according
  • Front line therapy was steroid (59%) or steroid plus extracorporeal photochemotherapy (41%)

Key findings: 

• CI of aGVHD = 21%
• 3-year OS = 55%
• 1-year NRM = 23%
• Day 28 CI of NR varied between MN SR and high-risk patients: 15% vs 36%, P = 0.09
• Glucksberg, IMBTR and visceral/non visceral grading had a similar capacity to identify patients with different chance of response
• On multivariate analysis no variable was independently associated with day 28 probability of NR
• Patients with high risk MN score were at higher risk of death: 50% vs 14%, P = 0.0002
• 3-year OS was 42% vs 60%, P = 0.04
• Stage III-IV aGVHD associated with inferior NRM

In conclusion, the refined MN aGVHD risk score is beneficial to detect patients receiving haplo-SCT with PT-CY at higher risk of non-response and mortality. Castagna further added that "for patients with higher risk of to not respond, different treatment should be tested." The key limitations of the study included the low number of patients, and the heterogenous treatment because 41% of them were treated with steroids and ECP.