B176
The Royal Marsden Hospital (RMH) Prognostic Score for Phase I Studies
Andre Brunetto, Joo Ern Ang, David Olmos, Hendrik-Tobias Arkenau, Barriuso Jorge, Bee Ayite, David Eaton, Johann de Bono, Ian Judson, Stan Kaye
The Royal Marsden NHS Foundation Trust, Sutton, Surrey, UK
Background
A major challenge of Phase-I trials is that of identifying those patients with very poor prognosis for whom trial participation is inappropriate. In view of this, we previously performed 2 retrospective studies on patients treated in the Drug Development Unit (DDU), RMH, between 2003 and 2007, and an objective prognostic score (range 0-3) comprising low albumin, high LDH and > 2 metastatic sites was obtained. We have further examined the validity of this score prospectively and assessed its impact on the pattern of inpatient admission in DDU.
Method
In order to validate the score with 90% power, we applied the prognostic score prospectively to 78 consecutively treated patients over a 4-month period and correlated the presence of high or low risk with overall survival (OS). Additionally, we applied the score (at baseline) to 310 DDU inpatient admissions and examined its utility in predicting unplanned admissions.
Results
In the prospective cohort, 43 patients (55%) with a risk score of 0-1 had an OS of 33 weeks (95%CI: 24.3-41.6), and the remaining 35 patients with a score of 2-3 had an OS of 15.7 weeks (95%CI: 10.6-20.8), p=0.036. Additionally, multivariate analysis confirmed that the risk score was independent in predicting OS with a hazard ratio of 1.4 (95%CI: 1.02-1.88). In the unplanned inpatient admissions, we note that 78% of patients had a risk score of 2-3.
Conclusion
Our simple, objective prognostic score has been validated in a prospective analysis of patients entering Phase I trials and could guide future patient selection.
