bims-tumhet Biomed News
on Tumor heterogeneity
Issue of 2026–11–08
four papers selected by
Sergio Marchini, Humanitas Research



  1. ESMO Open. 2026 Nov 03. pii: S2059-7029(26)02556-1. [Epub ahead of print]11(11): 108611
       BACKGROUND: The prognostic significance of programmed death-ligand 1 (PD-L1) in high-grade serous ovarian cancer (HGSOC) remains uncertain, because of heterogeneity in assays, scoring systems, and assessment of tumor and immune compartments across studies. We therefore evaluated its biological and prognostic relevance using a standardized pathological approach in a large, well-annotated phase III trial cohort.
    MATERIALS AND METHODS: In this ancillary retrospective biomarker analysis of PAOLA-1/ENGOT-ov25, PD-L1 was assessed by 22C3 pharmDx immunohistochemistry on whole-slide sections from 324 International Federation of Gynecology and Obstetrics stage III-IV HGSOCs. Three predefined scoring systems commonly used in PD-L1 assessment were applied: tumor proportion score (TPS ≥1%), immune cell score (IC ≥1%), and combined positive score (CPS ≥1). Associations with homologous recombination deficiency (HRD), BRCA1/2 status, and survival were examined using multivariable Cox models adjusted for major clinicopathological factors.
    RESULTS: PD-L1 positivity was observed in 38.2% of tumors by TPS, 62.4% by IC, and 66.1% by CPS. Positivity did not significantly vary by sampling site or timing of tissue collection relative to chemotherapy. PD-L1 positivity was enriched in HRD-positive tumors (TPS: 42.3% versus 28.0%, P = 0.017; IC: 65.1% versus 52.0%, P = 0.030; CPS: 70.4% versus 53.0%, P = 0.003) and remained independently associated with improved overall survival in adjusted analyses [TPS: adjusted hazard ratio (aHR) 0.70 (95% confidence interval 0.49-1.00), P = 0.046; IC: aHR 0.68 (0.49-0.95), P = 0.025; CPS: aHR 0.70 (0.50-0.97), P = 0.037]. Associations with progression-free survival were not statistically significant after adjustment.
    CONCLUSIONS: In advanced HGSOC, standardized whole-slide PD-L1 assessment identifies an HRD-associated PD-L1-positive immune context independently associated with improved overall survival. Immune-integrated scoring approaches capture this signal more effectively than tumor-cell-restricted assessment and provide a robust framework for future ovarian cancer studies.
    Keywords:  PD-L1; biomarkers; gynecological cancer (ovarian); homologous recombination deficiency; precision medicine
    DOI:  https://doi.org/10.1016/j.esmoop.2026.108611
  2. Clin Cancer Res. 2026 Nov 03.
      Esophageal cancer sheds little circulating tumor DNA (ctDNA), leading first-generation MRD assays to miss occult disease. Applying ultrasensitive MAESTRO assays to 331 plasma samples from 36 patients treated with curative intent, a recent article found 60% of ctDNA-positive samples were <100 ppm, and every recurrence preceded by postoperative ctDNA detection.
    DOI:  https://doi.org/10.1158/1078-0432.CCR-26-3052
  3. Proc Natl Acad Sci U S A. 2026 Nov 10. 123(45): e2629041123
      Cancer patients carry more circulating cell-free DNA (ccfDNA)-small DNA fragments shed into the blood-than healthy people. Most of this excess comes not from the tumor itself but from immune cells, suggesting that it would simply dilute the much smaller amount of tumor-derived DNA in a blood sample. Using a study of 1,480 previously untreated participants across 43 hospitals-including patients with 14 cancer types spanning all stages, patients with benign growths, and healthy volunteers-we show the opposite: as total ccfDNA rises, it becomes enriched, not diluted, in fragments carrying cancer-associated DNA methylation signatures, independent of cancer type, disease stage, and a broad panel of demographic, clinical, and technical covariates. In fact, total ccfDNA explains more variance in cancer-associated methylation signal than the disease stage itself. This basic biological finding suggests a coupling between immune cell activity and tumor methylation. It also has a practical implication: we show that rising ccfDNA concentration is associated with higher cancer detection sensitivity, even within disease stages and particularly in early-stage disease-stage I sensitivity varied from 0.55 in the lowest ccfDNA concentration tertile to 0.85 in the highest, and stage II from 0.59 to 0.86. This suggests that ccfDNA concentration should be considered as a stage-independent stratification variable for benchmarking cancer-detection tests. Interestingly, similar behavior was observed in patients with symptomatic benign lesions, suggesting a broader biological response to abnormal tissue growth and not something unique to malignancy.
    Keywords:  MCED; cancer; cell-free DNA; methylation
    DOI:  https://doi.org/10.1073/pnas.2629041123
  4. Nat Med. 2026 Nov 02.
      Preoperative immune checkpoint inhibitors (ICIs) are active in patients with deficient mismatch repair (dMMR) colon cancer (CC), whereas proficient mismatch repair (pMMR) CC has been consistently considered resistant to first-generation ICIs. In the phase 2 UNICORN trial, we evaluated short-course preoperative botensilimab (BOT; multifunctional Fc-enhanced anti-CTLA-4 monoclonal antibody) alone or in combination with the anti-PD-1 monoclonal antibody balstilimab (BOT + BAL) in 56 patients with resectable, locally advanced CC. Twenty-eight patients with pMMR CC were sequentially enrolled in BOT (n = 14) and BOT + BAL (n = 14) cohorts. Twenty-eight patients with dMMR CC followed the same sequential allocation strategy. The primary endpoint of centrally assessed pathological major response (pMR) rate (≤10% of residual viable tumor) was met in all cohorts, except pMMR BOT. Specifically, the pMR rate was 0% (95% confidence interval (CI): 0-23.2%) in pMMR BOT, 36% (95% CI: 12.8-64.9%) in pMMR BOT + BAL, 36% (95% CI: 12.8-64.9%) in dMMR BOT and 100% (95% CI: 76.8-100%) in dMMR BOT + BAL. Secondary endpoints included safety, quality of life, describing treatment-induced transcriptional dynamics, peripheral blood mononuclear cell dynamics, radiomics and radiogenomics features, fecal microbiota composition and circulating tumor DNA (ctDNA) dynamics. Eight (29%) and 15 (54%) patients treated with BOT and BOT + BAL, respectively, experienced immune-related adverse events, most commonly flu-like syndrome and diarrhea. Single-cell spatial transcriptomic analysis of matched baseline and post-treatment samples highlighted high baseline PD-L1 stromal expression and low post-treatment regulatory T cell signatures in pMMR responders, whereas dMMR responders displayed high baseline FCGR2A and FCGR3A expression on intratumoral macrophages. Clearance of ctDNA before surgery correlated with pathological response and lack of recurrence at data cutoff. Preoperative BOT + BAL elicits pathological responses in patients with pMMR or dMMR CC, supporting the combination as a promising strategy to expand ICI benefits to the early-stage setting. ClinicalTrials.gov identifier: NCT05845450 .
    DOI:  https://doi.org/10.1038/s41591-026-04710-1