1508MO - Rationalizing the use of off-label drugs (OLD) within a special medication (ME) program for cancer patients (pt): The Catalan Institute of Oncology (ICO) prognostic score (ICO MEscore)

Background

Molecularly targeted agents (MTA) and immunotherapies (IO) are approved for cancer pt. OLD can be used before the upcoming approval and in different scenarios or conditions from those reflected in regulatory summary approvals. ME programs allow pt to receive OLD. We aimed to improve ME pt selection by developing a prognostic score (ICO MEscore) that included clinic-analytical (CA) variables readily available in real practice.

Methods

Retrospective cohort of metastatic solid tumor pt treated with ME MTA/IO between Jan’11-Dec’19. A multivariate Cox regression model was fitted to predict the risk of death. The Akaike information criterion was considered in a backward selection method. ICO MEscore was derived from a final prognostic model and was categorized using the nomogram and prognostic scores resulting in similar Kaplan-Meier curves.

Results

In total, 441 out of 541 (82%) cases treated with MTA (67%) or IO (33%) had available CA data. Most pt presented ECOG <2 (87%) and had received <3 previous lines (83%). Most frequent subtypes were thoracic (38%) and breast (19%) cancers, and 39% of all tumors harbored actionable oncogenic drivers. Overall, median overall survival (mOS) was 17.7 months (m) (95% CI 14-23). The final multivariate prognostic model included 8 risk factors: ECOG >=1, Body Mass Index <25, non-breast/thoracic tumors, hemoglobin <10.5 g/dL, neutrophil-to-lymphocyte ratio >3, platelet-to-lymphocyte ratio <150, >=2 sites of metastases and >=3 previous lines. Four prognostic groups were defined: low risk (LR) mOS 41.5 m (95% CI 35-NR), intermediate-low risk (ILR) mOS 18.1 m (95% CI 13.3-24), intermediate-high risk (IHR) mOS 7 m (95% CI 5.6-9.4) and high risk (HR) mOS 1.5 m (95% CI 1.2-4.2)(log rank test, p<0.001). 1-year OS rates were LR 80%, ILR 60% and IHR 36%, whereas there were no alive pt at 1-year in the HR group.

Conclusions

ICO MEscore is the first prognostic index under development for ME pt selection and our results suggest that could serve to identify HR pt with dismal prognosis. Further validation is warranted to confirm its applicability to rationalize ME decision-making.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.