Studying When to Add Biological Therapy to Cytotoxic Therapy in Advanced Cancer Care Using Real World Data: How to...

Since real-world data lack information on treatment assignment (usually the only information available is treatment prescription/administration) and early outcomes can happen in advanced cancer, some design choices may introduce immortal time due to selection or due to treatment strategy misclassification.

To illustrate design choices that would biassedly introduce immortal time and propose unbiased alternatives, using as a case study the estimation of the effects of different monoclonal antibody (mAb) sequencing strategies on overall survival in colorectal cancer patients.

We specified a target trial to estimate the effect on overall survival of initiating mAb within 8 weeks of starting first line chemotherapy versus initiating mAb with second line chemotherapy. The first biased design choice would be requiring the initiation of second line chemotherapy for eligibility. This was avoided by aligning eligibility with the initiation of first line chemotherapy. The second biased design choice would be excluding from the first line mAb strategy those who die during the grace period without starting mAb. This was avoided separately via cloning, censoring and weighting and by sequential emulation.

There were 1014 eligible patients in GEMCAD 1401. The CCW approach estimated a 4-year survival difference (second-first line mAb) of 4.3% (95% CI) (-3.3%; 10.0%) and a risk ratio (RR) of 0.95 (0.88; 1.04). Sequential trial emulation estimated a survival difference of -2.8% (-7.5%; 1.2%) and a RR of 1.04 (0.98; 1.10). Biased analyses that introduced immortal time estimated an implausible early survival benefit for mAb initiation with second-line (6-month RR of 0.13 (0.03; 0.28)).

Our results are compatible with no effect on survival by mAb sequencing. Misalignment of time zero, treatment assignment, and eligibility introduced immortal time.