Research Signal – Source date 2022-01-01
A review of clinical trials found that adding another anticancer drug to PD-1 or PD-L1 immunotherapy was associated with more tumor responses overall than using the immune drug alone. But the added drug’s type mattered: combinations containing drugs that interfere with DNA synthesis or microtubules showed statistically significant response improvements in the review’s subgroup analysis, while the examined kinase and angiogenesis inhibitors did not.
That finding is a comparison of tumor responses across groups, not a prediction of what will happen to an individual patient. It also does not tell us whether the combinations extended life, improved quality of life, or offered a better balance of benefits and harms.
The outcome was tumor shrinkage, not time lived
The review measured objective response rate, or ORR: the share of patients whose tumors had a confirmed complete or partial response. A complete response means the measured tumor has disappeared; a partial response means it has shrunk by a defined amount. ORR is one way to assess whether a treatment is affecting tumors, but it does not capture every outcome that matters to patients.
In the 16 studies that compared a combination with a PD-1 or PD-L1 drug alone, the pooled relative risk of response was 1.79 (95% confidence interval 1.46–2.20). In plain terms, responses were about 1.8 times as common in the combination groups as in the checkpoint-drug-only groups, on average across those studies. That is a relative comparison, not a 79 percentage-point increase, and it cannot by itself tell a patient’s chance of responding. That would also depend on the starting response rate for a comparable tumor and treatment setting.
The estimate combines studies of different cancers and drug regimens. It summarizes an average across the included comparisons rather than a result that can be assigned to every pairing.
The narrower comparison asks what the added drug contributes
The authors first identified 36 studies comparing a PD-1/PD-L1 combination with some other treatment. Those control treatments varied, so the pooled result could not isolate the contribution of adding a drug to checkpoint immunotherapy. Across those studies, the combination groups had a higher pooled response rate (relative risk 1.45), but the results varied substantially from study to study (I² = 74%).
To focus on the partner-drug question, the authors then analyzed the 16 studies that included a checkpoint-drug-only arm or cohort. Together, these comparisons involved 3,793 patients: 1,863 in combination groups and 1,930 receiving PD-1/PD-L1 therapy alone. Because both sides of each comparison included checkpoint therapy, this set more directly addressed whether adding another anticancer drug was associated with a different response rate.
The 16-study estimate still came from a mix of randomized and nonrandomized trials. Its variation was lower than in the broader analysis, but not absent: I² was 44%. The narrower comparison helps answer a more specific question; it does not make all the included patients, cancers, or treatment plans interchangeable.
The partner-drug pattern had a biological hypothesis
PD-1 and PD-L1 are parts of a signaling system that can dampen immune activity against tumors. Antibodies that block this interaction can release that immune brake. Researchers have also proposed that some chemotherapy drugs may make tumor cells more visible to immune attack, including by causing a form of cell death called immunogenic cell death.
That idea offered a reason to examine combinations by the added drug’s action. In the review’s subgroup analysis, combinations with drugs that inhibit DNA synthesis (the copying of genetic material) or microtubule formation (needed for cell division) showed statistically significant response improvements compared with checkpoint therapy alone. The authors grouped several chemotherapy medicines into these categories. By contrast, the examined combinations with drugs targeting new blood-vessel growth (angiogenesis inhibitors) or certain cell-signaling enzymes (kinase inhibitors) did not show a statistically significant response improvement.
These are clinical response comparisons; they do not show that immunogenic cell death caused the differences. Nor does a nonsignificant subgroup result mean that a drug class cannot help. Subgroups contain fewer comparisons, and their estimates can be imprecise. The pattern points to a question for more targeted testing, not a ranking of specific medicines for every cancer.
Cancer type and trial design change the reading
The review included several solid tumor types and treatment regimens. Its subgroup estimates varied: for lung cancer, the pooled relative risk was 1.57 (95% CI 1.15–2.14); for gastric or gastroesophageal junction cancers, it was 2.73 (1.86–4.02); and for other tumors, 1.70 (1.25–2.32). These estimates describe different collections of studies. They should not be read as a head-to-head comparison showing that one cancer type responds better than another.
The authors also cautioned that some included trials were not randomized. Differences in tumor type, treatment line, metastatic disease, or the indication being treated could influence comparisons. Random-effects modeling accounts statistically for variation among studies, but it cannot remove unmeasured differences between the people or regimens studied.
The authors extracted eligible trials through December 31, 2020. The review was published in 2022, so its findings should be read as a historical synthesis of the included studies rather than an account of subsequent combination trials.
A pooled response advantage is one useful piece of evidence, but treatment decisions also depend on outcomes and risks this response analysis does not settle. It cannot establish longer survival, better quality of life, or whether added harms are outweighed by benefit. The same broad category of cancer can also involve distinct treatment regimens and patient circumstances.
What this could change
For a person discussing treatment for a defined solid tumor, dependable evidence about a specific checkpoint-drug pairing could make the conversation more precise: how often did tumors respond, what harms accompanied treatment, and how do those outcomes compare with checkpoint therapy alone? Combination treatment already exists. The potential advance is more dependable, tumor-specific selection rather than assuming that adding any anticancer drug improves the result.
This review offers an early bridge toward that possibility: its checkpoint-only comparison favored combinations overall, while its subgroup results differed by partner-drug class. To support a real treatment choice, that pattern would need to hold up in stronger, tumor-specific evidence that also addresses outcomes beyond response.
One nearer milestone would be randomized comparisons of defined drug pairings against checkpoint therapy alone in specific tumor types, reporting tumor response alongside survival and adverse effects. Those results could help clarify which pairings offer a meaningful tradeoff for particular settings.
Key Sources
- Primary source: Anti-Tumor Efficacy of Anti-PD-1/PD-L1 Antibodies in Combination With Other Anticancer Drugs in Solid Tumors: A Systematic Review and Meta-Analysis.
DOI or Journal Page · Review or Meta-Analysis · 2022-01-01
Keep exploring