Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Apr;30(4):421-432.
doi: 10.1016/j.jtct.2024.01.079. Epub 2024 Feb 4.

A Day 14 Endpoint for Acute GVHD Clinical Trials

Nikolaos Spyrou  1 , Yu Akahoshi  1 , Steven Kowalyk  1 , George Morales  1 , Rahnuma Beheshti  1 , Paibel Aguayo-Hiraldo  2 , Monzr M Al Malki  3 , Francis Ayuk  4 , Peter Bader  5 , Janna Baez  1 , Alexandra Capellini  1 , Hannah Choe  6 , Zachariah DeFilipp  7 , Matthias Eder  8 , Gilbert Eng  1 , Aaron Etra  1 , Sigrun Gleich  9 , Stephan A Grupp  10 , Elizabeth Hexner  11 , Matthias Hoepting  9 , William J Hogan  12 , Stelios Kasikis  1 , Nikolaos Katsivelos  1 , Alina Khan  1 , Carrie L Kitko  13 , Sabrina Kraus  14 , Deukwoo Kwon  1 , Pietro Merli  15 , Joseph Portelli  1 , Muna Qayed  16 , Ran Reshef  17 , Tal Schechter  18 , Ingrid Vasova  19 , Matthias Wölfl  20 , Kitsada Wudhikarn  21 , Rachel Young  1 , Ernst Holler  9 , Yi-Bin Chen  7 , Ryotaro Nakamura  3 , John E Levine  1 , James L M Ferrara  22
Affiliations

A Day 14 Endpoint for Acute GVHD Clinical Trials

Nikolaos Spyrou et al. Transplant Cell Ther. 2024 Apr.

Abstract

The overall response rate (ORR) 28 days after treatment has been adopted as the primary endpoint for clinical trials of acute graft versus host disease (GVHD). However, physicians often need to modify immunosuppression earlier than day (D) 28, and non-relapse mortality (NRM) does not always correlate with ORR at D28. We studied 1144 patients that received systemic treatment for GVHD in the Mount Sinai Acute GVHD International Consortium (MAGIC) and divided them into a training set (n=764) and a validation set (n=380). We used a recursive partitioning algorithm to create a Mount Sinai model that classifies patients into favorable or unfavorable groups that predicted 12 month NRM according to overall GVHD grade at both onset and D14. In the Mount Sinai model grade II GVHD at D14 was unfavorable for grade III/IV GVHD at onset and predicted NRM as well as the D28 standard response model. The MAGIC algorithm probability (MAP) is a validated score that combines the serum concentrations of suppression of tumorigenicity 2 (ST2) and regenerating islet-derived 3-alpha (REG3α) to predict NRM. Inclusion of the D14 MAP biomarker score with the D14 Mount Sinai model created three distinct groups (good, intermediate, poor) with strikingly different NRM (8%, 35%, 76% respectively). This D14 MAGIC model displayed better AUC, sensitivity, positive and negative predictive value, and net benefit in decision curve analysis compared to the D28 standard response model. We conclude that this D14 MAGIC model could be useful in therapeutic decisions and may offer an improved endpoint for clinical trials of acute GVHD treatment.

Keywords: Acute GVHD; Biomarkers; Classification and regression tree; Competing risk; Competing risks; Composite; Decision curve analysis; Endpoints; Immunosuppression; Machine learning; Nonrelapse mortality; Surrogate; Time-dependent AUC; Treatment response.

PubMed Disclaimer

Conflict of interest statement

M.A.M. reports consulting from NexImmune and Tscan, research funding from NexImmune; Gilead and Incyte, and participates in the speakers bureau in Sanofi and the board of advisory committees in Hasna Biopharma, Stemline Therapeutics, CarDx, Incyte; P.B. reports research grants from Neovii, Riemser, Medac (to Institution), advisory board for Novartis, Celgene, Amgen, Medac, Servier (personal and to Institution); Speakers Bureau for Novartis, Amgen (to Institution) and Patent and Royalties from Medac; H.C. reports consulting or advisory role with Incyte and research funding from Opna; Z.D. reports research support from Incyte, Corp., Regimmune, Corp and Taiho Oncology, Inc. and has received consulting fees from Sanofi, Omeros Corp., Incyte, Corp., MorphoSys AG, Inhibrx, PharmaBiome AG, and Ono Pharmaceutical; S.GR. reports research and/or clinical trial support from Novartis, Servier, Vertex, and Kite; study steering committees, consulting, or scientific advisory boards: Novartis, Allogene, Cabaletta, Juno, CBMG, GlaxoSmithKline, J&J/Janssen, CRISPR/Vertex, Roche, Jazz, Cellectis, and Kyttaro. All are outside the submitted work; P.M. served on Scientific Advisory Board for Sobi and Jazz and have received fees for speaker’s bureau from Miltenyi and Amgen outside the submitted work; C.L.K. served on an advisory board for Horizon Therapeutics; T.S. serves on an advisory board for Moderna; R.R. reports consulting or advisory role with Atara Biotherapeutics, Allogene, Gilead Sciences, Takeda, Incyte, Instil Bio, TScan, Synthekine, Orca, Quell Biotherapeutics, Capstan and Jasper, expert witness role with Bayer, and research funding from Atara Biotherapeutics, Incyte, Sanofi, Immatics, Abbvie, TCR2, Takeda, Gilead Sciences, CareDx, TScan, Synthekine, BMS, J&J, Genentech and Precision Biosciences; Y.B.C. reports consultancy for Takeda, Incyte, Pharmacosmos, Celularity, Vor, Novo Nordisk; J.E.L. reports consulting fees from Bluebird Bio, Editas, Equillium, Incyte, Inhibrx, Kamada, Mesoblast, Sanofi, and X4 Pharmaceuticals; research support from Genentech, Incyte, and Mesoblast; and royalties from a GVHD biomarker patent;.J.L.M.F. reports royalties from a GVHD biomarker patent. The other authors declare no conflict of interest.

Figures

Figure 1.
Figure 1.. D14 Mount Sinai clinical model in the validation set (n=380).
Favorable MAGIC clinical status is depicted in blue and unfavorable status in red as determined by recursive partitioning according to 12-month NRM. Patients that died before D14 (n=8) were considered as having unfavorable status and are not included in the schematic.
Figure 2.
Figure 2.. NRM and OS of D28 and D14 Mount Sinai clinical models in the validation set (n=380).
Cumulative incidence of NRM (A-B) and Kaplan-Meier plot of OS (C-D) for the D28 standard and the D14 MAGIC clinical models in the validation set (n=380) (non-responders/unfavorable status in red and responders/favorable status in blue). Pie charts depict the proportion of patients correctly classified as non-responders/unfavorable status who experience non-relapse deaths (A-B) or all deaths (C-D) (red, positive predictive value), responders/favorable status and do not experience non-relapse deaths (A-B) or all deaths (C-D) (blue, negative predictive value), and misclassified patients of each response/status group (grey).
Figure 3.
Figure 3.. NRM of D14 MAGIC integrated model in the validation set (n=380).
(A) Sankey diagram of patients classified at D14 by Mount Sinai clinical status and MAP score in the validation set (n=380). (B) Cumulative incidence of NRM for the four patient populations of the MAGIC integrated model in A.
Figure 4.
Figure 4.. NRM of D28 ORR (CR vs PR vs NR) and MAGIC integrated model (good vs intermediate vs poor status).
Cumulative incidence of NRM (A-B) and Kaplan-Meier plot of OS (C-D) for the D28 standard response (A, C) and the D14 MAGIC integrated model (B, D) (good: blue, intermediate: purple, poor: red) in the validation set (n=380). D14 MAGIC integrated model categories: good integrated status, D14 Mount Sinai favorable clinical status with low MAP; intermediate integrated status, either D14 Mount Sinai favorable clinical status with high MAP or D14 Mount Sinai unfavorable clinical status with low MAP; and poor integrated status, D14 MAGIC unfavorable clinical status with high MAP. Pie charts depict the proportion of patients in each category.
Figure 5.
Figure 5.. Quantitative comparisons of D28 and D14 models.
(A) Time dependent AUC for competing risks with NRM as outcome. P-values correspond to the 12-month time point. AUC values at the 12-month time point: D28 standard: 70.57 (95% CI: 64.39–76.75); D14 Mount Sinai clinical: 73.2 (95% CI: 67.1–79.3); D14 MAGIC integrated 77.8 (95% CI: 72.0–83.6) (B) Decision curve analysis (DCA) of the response models for 12 month NRM. Threshold probabilities express preferences of either the patients or the physician.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Westin JR, Saliba RM, De Lima M, Alousi A, Hosing C, Qazilbash MH, et al. Steroid-Refractory Acute GVHD: Predictors and Outcomes. Advances in Hematology. 2011;2011:601953. - PMC - PubMed
    1. Hahn T, Sucheston-Campbell LE, Preus L, Zhu X, Hansen JA, Martin PJ, et al. Establishment of Definitions and Review Process for Consistent Adjudication of Cause-specific Mortality after Allogeneic Unrelated-donor Hematopoietic Cell Transplantation. Biol Blood Marrow Transplant. 2015;21(9):1679–86. - PMC - PubMed
    1. MacMillan ML, DeFor TE, Weisdorf DJ. The best endpoint for acute GVHD treatment trials. Blood. 2010;115(26):5412–7. - PubMed
    1. Martin PJ, Bachier CR, Klingemann HG, McCarthy PL, Szabolcs P, Uberti JP, et al. Endpoints for clinical trials testing treatment of acute graft-versus-host disease: a joint statement. Biol Blood Marrow Transplant. 2009;15(7):777–84. - PMC - PubMed
    1. Levine JE, Logan B, Wu J, Alousi AM, Ho V, Bolaños-Meade J, et al. Graft-versus-host disease treatment: predictors of survival. Biol Blood Marrow Transplant. 2010;16(12):1693–9. - PMC - PubMed

Publication types

LinkOut - more resources