A checklist for assessing planned phase III immunotherapy combination trials
| Checklist item | Explanation | Priority level (I, II, or III, with I being the highest) |
| Mechanism and biology | ||
| Additive or synergistic biology | In preclinical studies, mechanisms of the combination are well understood and address mechanisms of non-response to either agent (most importantly of the most active agent), which are documented or highly plausible in a substantial proportion of the intended patient population | III |
| Validity of Preclinical Models | Evidence of antitumor activity in multipl preclinical models, which is clearly superior to eaither single agent alone (immune competent, HLA matched, humanized if possible or veterinary models) | III (depending on validity of model, extent of activity) |
| Single-agent activity (preclinical) | The combination partner for anti-PD-(L)1 shows single-agent activity (and if not, an immunotherapy is demonstrated to actually enhance an immune response) | II |
| Phase I and II data | ||
| Single-agent activity (early clinical) | Pharmacodynamic data or neoadjuvant and correlative studies show that the combination (or the single agent being added to anti-PD-(L)1) has the intended immune or biological effect, and that the combination is working as an immunotherapy as the mechanism for antitumor efficacy | I (if IO/IO) II (if other combination) |
| Randomization in phase II | The combination demonstrates increased activity (ORR/PFS/OS) in a well-designed randomized phase II trial, and clearly identifies the contribution of the anti-PD-(L)1 combination partner | I (if IO/IO) II (if other combination) |
| Activity in standard of care/anti-PD-(L)1-resistant tumors | The combination demonstrates convincing antitumor activity (ie, a combination of CR/PR/prolonged SD that exceeds 25% or clinically substantial increases in PFS or OS compared with very well-matched historical controls from large databases) in patients with clearly documented SOC-resistant/refractory tumors in which continuing or introducing anti-PD-(L)1 alone would produce no more than 0%–5% clinical activity (and the combination partner also has minimal activity) | II |
| Activity in tumor types that typically do not respond to PD-1 pathway blockade | The combination displays antitumor effects in tumor types unresponsive to IO therapy (eg, prostate cancer, MSS colon cancer, ER+ breast cancer, hematologic malignancies) and the properties of the response indicate that the activity is immune-mediated (ie, survival curves plateau) | II |
| Responses are durable (including off treatment) | A substantial proportion of the responses observed with the combination in phase I and II trials are prolonged (exceed 1 year) and/or are complete or near complete, and treatment can be discontinued without relapse in most of the responders | II (depending on setting: may be higher for anti-PD-(L)1-resistant tumors) |
| Agents have non-overlapping toxicities | The combination has tolerable and manageable adverse effects, avoiding overlapping/synergistic toxicities—benefit to risk is acceptable | II |
| Trial design | ||
| Translational studies | The clinical trial includes tissue collection (eg, baseline tumor and blood, and on-study tissue sampling) which can be used (within the trial or for future investigation) for informative correlative studies to understand tumor biology and mechanisms of response and resistance | II |
| Integrated biomarker program | Predictive biomarkers can be used that allow selection of patients most likely to benefit from the combination | II |
| Stop therapy | The trial mandates a stop in therapy (perhaps based on a predetermined biomarker such as ctDNA) to determine if durable responses persist | II |
CR, complete response; ctDNA, circulating tumor DNA; ER, estrogen receptor; HLA, human leukocyte antigen; IO, immuno-oncology; MSS, microsatellite stable; ORR, overall response rate; OS, overall survival; PD-(L)1, programmed cell death 1 and its ligand; PFS, progression-free survival; PR, partial response; SD, stable disease; SOC, standard of care.