A Prognostic Scoring System for Patients ≤ 60 Years of Age with Acute Lymphocytic Leukemia in First Relapse.

The prognosis of adult pts with ALL in first relapse is poor. We analyzed pt characteristics at the time of relapse to define a prognostic score for pts ≤ 60 yrs of age.

Methods: Pts 18–60 years of age with ALL (WHO criteria) in first relapse treated at the Cleveland Clinic between the years 1990–2007 were evaluated. Pts were identified through the Health Data Service’s Hospital Abstract and BMT databases using the ICD-9 code for ALL. Pts with residual leukemia after their initial induction or those receiving allogeneic bone marrow transplant (alloBMT) in first CR were not included. Relapsed disease was defined as the reappearance of leukemic blasts in the blood or bone marrow (BM) (> 5%) or in any extramedullary site. Cox proportional hazards analysis was used to identify univariate and multivariate factors significantly affecting overall survival (OS) in all pts including those proceeding to alloBMT. Factors evaluated were: age, cytogenetic (CG) risk group, disease status, time from diagnosis to relapse, lactate dehydrogenase (LDH) at the time of relapse, gender, and HLA match (for related or unrelated donors). An HLA match was defined as a 6/6 allele match based on 2 A, B, and DRB1 markers. Factors significant in univariate analysis were used in the multivariate model.

Results: Seventy-seven pts were analyzed. The median age at relapse was 32 yrs [range 18–60] and median time from diagnosis to first relapse 14.8 months [0.8–41.6]. Twenty-six percent of pts had poor risk cytogenetics (CG), 47% normal, 14% miscellaneous abnormalities, and 13% unknown CG as defined by CALGB criteria. Seventy-eight percent of pts had ALL derived from B-cell, 18% T-cell, 1% mixed, and 3% unknown lineage. Ninety-one percent of pts relapsed with blood/BM disease and 21% with central nervous system (CNS) involvement. Nine percent relapsed with leukemia isolated to an extramedullary site. Thirty-seven percent of pts achieved a CR following salvage therapy. Forty-six pts (60%) received an alloBMT, with 24 pts undergoing transplant from a matched-related donor. Forty-eight percent of pts were in CR at the time of alloBMT. The estimated OS for all pts was 41%, 23%, and 6% at 1, 2, and 5 yrs post-relapse. However, the OS was significantly better for those receiving BMT (58%, 33%, and 9% at 1, 2, and 5 yrs) compared to those not receiving an alloBMT (9%, 5%, and 0%) [p< 0.001]. Factors adversely affecting OS in pts proceeding to BMT on both univariate and multivariate analysis were: older age, HLA mismatch, leukemia present at the time of BMT, and higher LDH at the time of relapse [per 200 unit increase (HR 1.77, CI 1.30–2.38, p< 0.001)]. Higher LDH at the time of relapse and older age at the time of diagnosis were predictors of worse OS for all pts, in addition to being predictive for those proceeding to BMT. For pts < age 45 with an LDH at the time of relapse < 1.4 times the upper limits of normal (ULN), 79% proceeded to BMT, and the CR rate to salvage therapy was 67%. Based on this data, we were able to develop a prognostic model. For all pts, a scoring system of 0–3 was used with 1 point for each of the following: age at diagnosis ³45, LDH ³ 1.4 times ULN, not proceeding to BMT (Table 1). Among pts who proceeded to BMT, a similar scoring system was used with 1 point for each of the following: age at diagnosis > 45 years, LDH at the time of relapse > 1.4 times ULN, leukemia present at the time of BMT (Table 1). For pts proceeding to BMT with a score of 2 (based on LDH and age), the median OS was only 4.9 months. The only long-term survivor was a pt who had achieved a CR to salvage therapy.

Table 1 Prognostic Score for Overall Survival

Conclusions: LDH > 1.4 ULN at the time of relapse predicts poor OS and may be incorporated, with age, into a prognostic score that also predicts a pt’s response to salvage therapy and likelihood of proceeding to BMT. Pts with a score of 2 have a poor prognosis, even with BMT, and should be considered for innovative approaches such as Phase 1 clinical trials. This prognostic scoring system will require prospective validation with an independent data set.