B290

A Clinical Prognostic Index for Assessing Patients aged >60 Being Considered for High‐Dose Therapy and Autologous Stem‐Cell Transplant in Relapsed or refractory High‐Grade Non‐Hodgkin Lymphoma

Edwards, David; Kirkland, Keiren 1; Pearce, Rachel; Robinson, Stephen 1; Cook, Gordon

1BSBMT

Abstract

Introduction Patients with relapsed high‐grade NHL or disease refractory to first‐line therapy can still be

cured with high‐dose therapy and autologous stem cell transplant if they respond to salvage

chemotherapy. This aggressive algorithm is accepted in younger patients but is less well established in the

elderly. Age >60 has a negative predictive score in the International Prognostic Index (IPI) and there are

concerns that the outcomes of HDT in these patients are significantly worse. Deciding which older patients

will benefit from HDT is challenging and there are no established predictive tools to guide physicians. We

present a clinical prognostic index derived from information readily available at the time a patient is being

assessed for ASCT Patients and Methods The BSBMT audited the outcomes for UK patients aged >60

transplanted between 2004 – 09 (n=371) and benchmarked against the European Bone Marrow Transplant

(EBMT) database for the same period (n= 2695). The primary outcome was progression‐free survival (PFS)

but data was also analysed for overall survival (OS), relapse rate (RR) and non‐relapse mortality (NRM). We

included all patients with a diagnosis of high grade NHL and the following demographic features were also

analysed: Age at diagnosis; age at transplant; M/F; year of transplant; CR/Not CR at transplant; no. of prior

therapies; no. of cells infused; clinical staging; Karnofsky status at transplant; histology; IPI at diagnosis;

mobilising regime and conditioning regime. Candidate prognostic indices were factors achieving

significance in univariate and multivariate analyses of the main outcomes by regression analysis. The best

prognostic index was selected based on the BSBMT dataset and then applied to the rest of the EBMT

dataset (the validation dataset). Results There were no significant differences in patient characteristics

between the UK and non‐UK groups nor in outcomes of PFS, OS, RR or NRM. (Figure 1). In both univariate

and multivariate analysis the following features were associated with a significantly worse outcome for PFS,

OS, RR and NRM : Age>66, Karnofsky Score