A Novel Scoring System Using Geriatric Nutritional Risk Index, C-Reactive Protein, and Disease Status Predicts Survival after Second Allogeneic Stem Cell Transplantation

Introduction: Second allogeneic stem cell transplantation (SCT2) is generally associated with poorer overall survival (OS) and non-relapse mortality (NRM) than first SCT (SCT1). However, some patients survive long after SCT2, and the number of SCT2 cases increases yearly. Thus, a prognostic stratification system is required to select optimally eligible patients for SCT2. We have previously shown that the geriatric nutritional risk index (GNRI) predicts OS and NRM for SCT2 (Ann Haematol. 2020;99:1655). In this study, using a multicenter database, we aimed to validate the importance of the GNRI and establish a scoring system that can more accurately stratify the prognosis after SCT2.

Methods: We retrospectively analyzed patients with hematopoietic malignancies who underwent SCT2 between 2004 and 2020 at transplant centers of the Kanto Study Group for Cell Therapy (KSGCT), using the KSGCT database. The primary endpoint was 3-year OS. The secondary endpoints were disease-free survival (DFS), NRM, and cumulative incidence of relapse (CIR). The GNRI was defined as 1.489×albumin(g/L) + 41.7×(weight/ideal weight) (Am J Clin Nutr. 2005;82:777). We chose a GNRI cutoff value of 92.0 according to our previous report. We defined acute leukemia and malignant lymphoma in the first or second complete remission, CML in the 1st or 2nd chronic phase, and MDS with refractory anemia as standard risk, and other malignant diseases as high risk.

Results: Among 722 patients who underwent SCT2, we were able to calculate the GNRI and perform a detailed analysis for 376. The median patient age was 44 years (range: 16-69). The underlying diseases were AML (n=207, 55.1%), ALL/LBL (n=107, 28.5%), MDS (n=33, 8.8%), CML (n=11, 2.9%), and other (n=18, 4.8%). The median follow-up period for survivors was 1375 days (range: 223-4213). Three years after SCT2, the OS, DFS, NRM, and CIR were 26.6%, 21.7%, 40.4%, and 37.9%, respectively. Based on previous research, we divided patients into two groups: low GNRI ( ≤92) (n=162, 43.1%) and high GNRI (>92) (n=214, 56.9%). In the low GNRI group, OS was better, NRM was lower, and CIR was comparable to that in the high GNRI group (3y-OS 36.0% vs. 14.3%, p<0.001; 3y-NRM 29.5% vs. 55.2%, p<0.001; 3y-CIR 33.4% vs. 41.3%, p=0.146). Through univariate analysis and stepwise multivariate regressions on the dependent variables for OS, low GNRI (hazard ratio [HR] 1.39, 95%CI 1.05-1.85, p=0.024), high C-reactive protein (CRP) (>0.5 mg/dl; HR 1.96, 95%CI 1.16-2.63, p<0.001), and disease status at SCT2 (HR 1.54, 95%CI 1.16-2.04, p=0.003) were associated with poor OS. Likewise, low GNRI (HR 1.71, 95%CI 1.18-2.47, p=0.0045), high CRP (HR 1.69, 95%CI 1.20-2.38, p=0.0029), high HCT-CI (≥3) (HR 1.58, 95%CI 1.12-2.23, p=0.0092) were unfavorable factors for NRM, while the shorter interval from SCT1 to SCT2 (HR 1.53, 95%CI 1.10-2.13 p=0.011) and high-risk disease status at SCT2 (HR 1.62, 95%CI 1.16-2.27, p=0.0046) were associated with high CIR. To predict OS after SCT2, we developed a scoring system consisting of three factors: low GNRI, high CRP level, and high-risk disease status. As the HRs for OS were comparable, a score of 1 was assigned to each factor. OS at 3 years was 50.7%, 29.5%, and 11.1% for patients with low (score 0, n=89), intermediate (score 1, n=106), and high (scores 2 and 3, n=174), respectively (p<0.001, Fig 1). Differences between groups were revealed using the Holm-Bonferroni correction.

Conclusion: Low GNRI, high CRP level, and high-risk disease status were significantly associated with inferior OS after SCT2. The scoring system with these factors stratified the entire cohort into three risk groups including low risk with over 50% OS, and high risk with a very limited chance of survival.