About
    • AMMDI is an open-notebook hypertext writing experiment, authored by Mike Travers aka mtraven. It's a work in progress and some parts are more polished than others. Comments welcome! More.
Search
Page Contents
Full

PICI/PRINCE

28 Dec 2025 - 04 Feb 2026
Open in Logseq
    • Aka PICI0002 Pancreatic cancer. WTF did I put the paper?

    • Note from ccab 8/2022

      • Hi PICI family,
      • I'm excited to share that yesterday the PRINCE team completed the database lock of the clinical database. This is a HUGE accomplishment that was 5 years in the making: PRINCE was the first in-house clinical trial initiated at PICI and we dosed our first patient in August 2017.
      • Per my count, there were 60,477 CRFs (case report forms) entered into the PRINCE database, which translates roughly to 500,000+ data points! Each data point was entered manually by site personnel, source data verified by a PICI CRA, reviewed by multiple PICI team members, and frozen/locked by a PICI data manager. During the data review process, over 55,000 queries were issued that required action by the site and then re-review by PICI. To put it simply, a mountain of work and effort was put into ensuring the completeness and accuracy of the clinical database.
      • This truly was a team effort where contributions were made by more PICI personnel (past and present) than I could name. However, an extra special shout out is deserved for a few current PICI employees: our Trial Leads (Jaclyn, Sultan), Clinical Scientist (Jingying), Safety team (Janelle, Marko, Stephen), Data Managers (Martin, Vivy), and CRAs (Janie, Sherri, Kit, Natalie, Diana, Jann, Annette, Stephen).
      • Please help me congratulate the entire PRINCE team for achieving this milestone.
      • Best,
      • Chris

    • Claude Summary

      • Summary of the PRINCE Trial

      • The Problem Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest cancers, projected to become the second-leading cause of cancer death in the US by 2030. While immunotherapy drugs called "checkpoint inhibitors" have revolutionized treatment for many cancers (melanoma, lung cancer, etc.), they have been remarkably ineffective in pancreatic cancer, with response rates below 5%. This is thought to be because pancreatic tumors create a highly immunosuppressive environment that prevents the immune system from attacking cancer cells.

      • Study Design The PRINCE trial was a randomized phase 2 clinical trial testing whether combining immunotherapy drugs with standard chemotherapy (gemcitabine + nab-paclitaxel) could improve outcomes for patients with newly diagnosed metastatic pancreatic cancer. 105 patients were randomized into three treatment arms: | | Arm | | Treatment | | Mechanism | | |
        - | | | 1 | | Nivolumab + chemotherapy | | Nivolumab blocks PD-1, releasing the "brakes" on T cells | | | | 2 | | Sotigalimab + chemotherapy | | Sotigalimab activates CD40, stimulating antigen-presenting cells | | | | 3 | | Both drugs + chemotherapy | | Combined immune activation | | The primary endpoint was 1-year overall survival rate, compared against a historical benchmark of 35% for chemotherapy alone.

      • Clinical Results Key survival findings:

      • Nivolumab/chemo: 57.7% 1-year survival (P=0.006) — met the primary endpoint
      • Sotigalimab/chemo: 48.1% 1-year survival (P=0.062) — did not meet endpoint
      • Triple combination: 41.3% 1-year survival (P=0.233) — did not meet endpoint Notably, the triple combination performed worst, suggesting the two immunotherapy approaches may interfere with each other rather than synergize.

      • Biomarker Analysis (The Core Scientific Contribution) The researchers performed extensive "multi-omic" profiling—analyzing blood samples and tumor biopsies using flow cytometry, mass cytometry, protein panels, and RNA sequencing—to identify which patients benefited from each treatment. For nivolumab/chemotherapy responders:

      • Higher baseline levels of activated, antigen-experienced T cells (particularly CD4+ and CD8+ T cells expressing markers like CD38, PD-1, and CD39)
      • Higher T follicular helper cells (Tfh cells)
      • Less immunosuppressive tumor microenvironment (lower TGF-β and TNF-α signaling signatures)
      • Fewer iNOS+ macrophages in tumors For sotigalimab/chemotherapy responders:
      • Higher baseline cross-presenting dendritic cells (CD141+ DCs)
      • More activated B cells (HLA-DR+CCR7+)
      • Higher type-1 helper T cells (Tbet+) in circulation
      • Greater intratumoral CD4+ T cell infiltration
      • Th1 and IFN-γ response gene signatures in tumors For triple combination:
      • Few biomarkers predicted benefit
      • Higher frequencies of "exhausted" T cells (CD38+ non-naive T cells) were associated with worse outcomes
      • Appearance of potentially regulatory B cells (CCR7+CD11b+CD27−) correlated with shorter survival

      • Key Scientific Conclusions

      • Treatment-specific biomarkers exist: Different patients benefit from different immunotherapy approaches, and these can potentially be identified before treatment begins using blood tests.
      • CD4 T cells may be critical: Unlike other cancers where CD8 "killer" T cells drive responses to immunotherapy, pancreatic cancer responses appear more dependent on CD4 "helper" T cells—a potentially important mechanistic insight.
      • More isn't always better: The triple combination may cause immune "hyperactivation" leading to T cell exhaustion rather than effective anti-tumor immunity.
      • Clinical implications: These treatments should not be used for all pancreatic cancer patients, but biomarker selection could identify the ~40-50% who might benefit significantly.

      • Data Collected The study generated an unusually comprehensive dataset:

      • Clinical data: Survival, response rates, adverse events for 105 patients
      • Blood samples: Collected at 5 timepoints (before and during treatment)
        • Flow cytometry (28-color panel for T cell phenotyping)
        • Mass cytometry (CyTOF for broad immune profiling)
        • Serum proteomics (172 proteins via Olink panels)
      • Tumor biopsies: Before and during treatment
        • Multiplex immunofluorescence imaging
        • Whole exome and transcriptome sequencing

      • Limitations

      • No chemotherapy-only control arm (compared to historical data instead)
      • Relatively small sample size (35 patients per arm)
      • Biomarker findings are hypothesis-generating and need prospective validation

      • Significance This study provides a template for how detailed immune profiling in clinical trials can identify patient subsets likely to benefit from specific immunotherapy approaches—moving toward "precision immunotherapy" for a disease where standard immunotherapy has failed.