Immunotherapy linked to longer survival in advanced cancers with DNA repair gene alterations
In a retrospective MD Anderson analysis of 662 patients, immunotherapy tracked with longer survival than chemotherapy or anti-DDR drugs, with or without DNA repair defects.
Summary
DNA repair defects cause genomic instability, which may make tumors more visible to the immune system. This retrospective analysis of the MD Anderson IMPACT2 study looked at 662 patients with advanced cancers and at least one gene alteration. Of these, 109 had DNA damage response (DDR) alterations and 553 did not. Among treated patients, immunotherapy (IO) was independently associated with longer overall survival than chemotherapy or anti-DDR agents in the DDR-altered group (n=85). IO was also associated with longer survival in the DDR-wild-type group (n=426). Progression-free survival did not differ significantly in the DDR group. The study is observational, non-randomized and small in the DDR subgroup, so it generates hypotheses rather than guiding treatment.
Detailed Summary
Defects in DNA repair let mutations accumulate. Laboratory work suggests they may also leave tumors more visible to the immune system, for example through cytosolic DNA that triggers cGAS-STING signaling, type I interferon release and T-cell recruitment. Whether DDR alterations predict benefit from immuno-oncology (IO) drugs across tumor types has not been systematically studied in patients. This paper takes a first look.
The authors retrospectively analyzed patients from the MD Anderson IMPACT2 study (NCT02152254). Of 829 enrolled, 692 completed molecular testing and 662 had at least one alteration. Of these, 109 (16.5%) had DDR alterations and 85 of them were treated. The other 553 (83.5%) were DDR wild-type, and 426 of them were treated. Treatment groups were IO (alone or combined with non-IO agents), anti-DDR agents such as PARP inhibitors, other targeted therapies, and chemotherapy. Outcomes were response, progression-free survival (PFS) and overall survival (OS), analyzed with Cox models.
In the DDR-altered group, median PFS was 4.1 months and median OS was 11.6 months. The overall OS comparison across treatment groups was not significant (log-rank p=0.056), and PFS did not differ by treatment (p=0.45). In multivariable analysis, IO was associated with longer OS than chemotherapy (HR for chemotherapy vs IO 2.17, p=0.021) and than anti-DDR agents (HR 2.08, p=0.048). Absence of liver metastases predicted longer OS in univariate analysis (p=0.029) but lost significance in the multivariable model. Objective response was 20.6% with IO, 11.8% with chemotherapy, 8.3% with anti-DDR agents and 0% with other targeted therapy. Only 3 of the 34 IO-treated patients had TMB-high or MSI-high tumors, so the signal was not simply a hypermutation effect.
In the DDR-wild-type group, IO also tracked with longer OS than every other treatment group (all p≤0.006). Liver metastases, low albumin and high LDH predicted worse OS, and mTOR pathway alterations predicted markedly worse outcomes (HR about 10.8). Anti-DDR agents did worst in both cohorts, which the authors attribute partly to toxicity. Because IO also performed best in patients without DDR alterations, the data do not show that DDR status specifically predicts IO benefit.
The main caveats are design and size. This is a retrospective, non-randomized analysis from one center in heavily pretreated patients, with treatment chosen by clinicians. The 34 IO-treated DDR patients are a small group, and IO alone and IO combinations were pooled. The mechanism (cGAS-STING-driven immunogenicity) comes from prior preclinical work and was not tested here. Prospective studies are needed before DDR alterations can be treated as an IO biomarker.
Key Findings
- In 85 treated DDR-altered patients, IO was independently linked to longer OS than chemotherapy (HR 2.17, p=0.021) and anti-DDR agents (HR 2.08, p=0.048).
- The DDR group's overall OS comparison across treatments was not significant (p=0.056), and PFS did not differ by treatment (p=0.45).
- IO also had the longest OS among 426 DDR-wild-type patients, so the data do not isolate DDR status as an IO predictor.
- Anti-DDR agents, including PARP inhibitors, had the worst survival in both cohorts, possibly due to toxicity and heavy pretreatment.
- mTOR pathway alterations in DDR-wild-type patients predicted much worse OS (HR about 10.8, p<0.001), alongside liver metastases and low albumin.
Methodology
Retrospective analysis of the MD Anderson IMPACT2 trial (NCT02152254) in 662 patients with advanced cancers and at least one alteration: 109 with DDR alterations (85 treated) and 553 without (426 treated). Outcomes were ORR, PFS and OS by treatment group, analyzed with Kaplan-Meier and Cox univariate/multivariate models, plus gene enrichment and co-occurrence analyses. Patients were not randomized to treatment.
Study Limitations
The study is retrospective and non-randomized, with a small DDR-altered treated cohort (n=85) and only 34 IO-treated patients. Mixed tumor types, heavy pretreatment, pooled IO groups and confounding by indication limit causal conclusions. IO also did well in non-DDR patients, and the immune mechanism was not measured in this study.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe:
