Several conflicting options regarding management of adult ALL are currently discussed. One major issue is the indication for SCT depending on risk factors and age, and the other is the recommendation to use unmodified pediatric protocols for “young” adults. Decision making on SCT is generally based on conventional risk factors – mainly disease characteristics – available at diagnosis, and decision making for “pediatric” chemotherapy on age. It is essential to develop more sophisticated criteria – also to reduce the risk of selection in clinical trials. In order to enhance prognostic models and to better address individual patient characteristics and the course of disease we reanalysed conventional prognostic factors together with new patient specific factors in a large cohort of adult ALL patients. A total of 1657 well characterised pts (15–55 yrs) included in the risk stratified protocols of the German Multicenter Study Group (GMALL) 06/99 and 07/03 was analysed. Treatment and risk stratification have been described (

). Age remained a highly significant factor for CR, survival (OS) and disease free survival (DFS). OS ranged from 58% for 15–25 yrs, 52% for 26–35 yrs, 43% for 36–45 yrs to 32% for 45–55 yrs (p=.0001). Poorer outcome with increasing age was mainly due to early death (ED) and death in CR. CR, OS (45% c/pre-B, 45% pro-B, 38% early T, 47% mature T and 64% thymic T;p<0.0001) and DFS were sign. different for immunophenotypes. WBC above 30.000 was associated with poorer CR, OS and DFS in c/pre-B-ALL but not in T-ALL. Female vs male gender was associated with poorer CR and OS (45% vs 49%;p=.01), but not DFS. Infections, bleeding or cytopenia at diagnosis had no prognostic impact, but ECOG status 0 was associated with a lower ED compared to ECOG 1–4 (3% vs 7%;p=.01) and a better OS (53% vs 45%;p=.04). Low body mass index (BMI) (<=19) was associated with higher failure whereas high BMI (>30) was associated with higher ED (p=.04). BMI (low vs normal vs high) had a significant impact on OS (43% vs 48% vs 39%;p=.003). Additional prognostic factors were identified during therapy. OS was sign.inferior for pts with infections in induction phase II compared to others (39% vs 52%;p<.0001). Bilirubine increases (grade III–IV) were associated with poorer OS (35% vs 51%;p<.0001). Treatment delays lead to inferior OS – particularly an extended break between phase I and II of induction (p=.03) and time from start of induction to first consolidation (p=.02). Overall the GMALL prognostic model (Hoelzer, Blood 1984) was confirmed in a large patient cohort: WBC > 30.000/μl, late CR, Ph/BCR-ABL, proB in B-precursor and early or mature T-ALL in T-lineage ALL are poor prognostic factors. These factors identify pts for SCT. Age has a high prognostic relevance and is not used for this purpose but for selection of age adapted treatment approaches. For the first time other new prognostic factors were described. Poor general condition identifies pts who need specific attention e.g. stabilisation before treatment start. Furthermore complications during induction – particularly infections and liver toxicity – as well as treatment delays in general affect outcome adversely. These findings support the hypothesis that a) results in adult ALL can be improved by better supportive care, protocol compliance and subgroup adjusted treatment and b) these and additional individual factors e.g. comorbidity should be utilised to refine prognostic models and decision making on intensive chemotherapy and/or SCT including dose-reduced SCT.

Supported by supported by Deutsche Krebshilfe 70-2657-Ho2 and partly BMBF 01GI 9971

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author