Ranking therapeutic recommendations of Molecular Tumor Board with an evidence-based algorithm to deliver optimal care and improve outcomes for cancer patients- continuous study

Background: The Molecular Tumor Board (MTB) has brought about significant advancements in precision oncology, but there is a long-standing challenge in selecting the most effective patient-specific therapeutic strategy due to the molecular rationale, disease relevance, and patient-specific issues. We have developed an algorithm that incorporates both molecular and clinical evidence-based criteria to rank therapeutic strategies to deliver optimal care and improve outcomes in patients with malignancy. This part of our studies was to evaluate the effectiveness and accuracy of our novel algorithm. Methods: History of present illness and comprehensive genomic profiling results of 571 cancer patients were reviewed by Avera MTB from June 2021 to December 2024. Therapeutic recommendations were provided with the Ranking Score (R, -1 to 12). The progression free survival (PFS) and overall survival (OS) of the patients that received recommended treatments (Cohort 1) or not (Cohort 2) were then assessed. Results: The median duration of follow-up was 14.1 months by the data cut-off on July 31, 2025. Of the 571 patients, 542 (94.9%) patients with cancer were evaluable in this study, including 286 (52.8%) that received matched therapeutic plans recommended by our MTB, and 256 (47.2%) did not. No significant differences in PFS/OS were observed between patients in Cohort 1 and in Cohort 2. In Cohort1, patients who received early line (≤ line 3, n=236) therapies had longer PFS (median PFS 8.8 vs 4 months, p <0.0001) and OS (median OS 24.5 vs 13.7 months, p =0.0058) when compared with those that received later lines (> line 3, n=46) of treatment. Among patients in Cohort 1, the median PFS/OS for patients with R≥10 (n=218, 8.6/23.9 months) were longer than patients with R<10 (n=68, 4.9/13.1 months), p =0.036/ 0.0009; the OS of patients that received Standard of Care treatments with R≥9 (n=230, 22.2 months) were longer than R<9 (n=27, 13.1 months), p =0.0053, whereas no significant difference was found in PFS. The PFS/OS of patients who received Off-Label treatments with R≥7 (n=10, 35.6/not reached months) were longer than R<7 (n=5, 3.5/10 months), p =0.0095/0.012; however, no significant differences in PFS/OS were observed among patients treated on Clinical trials. Conclusion: Our novel molecular and clinical evidence-based algorithm may be used to support oncologists' decision-making to utilize the most clinically appropriate and effective therapeutic options to benefit patients. Further validation studies and development of a user-friendly computational ranking platform based on the algorithm are planned in order.