PRO1107, a PTK7 antibody-drug conjugate, ended at 33 patients on benefit-risk

What was tried

Genmab ran NCT06171789, a Phase 1/2 open-label, single-group study of PRO1107 (also called GEN1107), an antibody-drug conjugate that binds protein tyrosine kinase 7 (PTK7) and carries a proprietary hydrophilic MMAE-based linker-drug at a drug-to-antibody ratio of 8 (ProfoundBio first-patient-dosed release, 2024, and ClinicalTrials.gov, NCT06171789). Dose escalation used a modified toxicity probability interval design, with the primary endpoints set as adverse events and dose-limiting toxicities (ClinicalTrials.gov, NCT06171789). The trial opened to six tumor types: ovarian, endometrial, triple-negative breast, non-small cell lung, gastroesophageal, and urothelial cancers (ClinicalTrials.gov, NCT06171789). It started on 2024-01-03 and reached an actual enrollment of 33 participants before termination, with a primary completion date of 2025-08-18 and no posted results (ClinicalTrials.gov, NCT06171789). Genmab recorded the reason as a decision to discontinue development because the overall benefit-risk profile no longer supported continuation (ClinicalTrials.gov, NCT06171789).

The biological hypothesis

PTK7 is a catalytically dead Wnt co-receptor, a pseudokinase that is over-expressed across several epithelial cancers and is linked to tumor-initiating cell biology and metastasis (Therapeutic targeting of PTK7 in cancer, 2025). A first-generation PTK7 ADC reduced tumor-initiating cell frequency and produced durable regressions in patient-derived xenografts, which established the antigen as a delivery handle for cytotoxic payloads (doi:10.1126/scitranslmed.aag2611). Because PTK7 has no kinase activity, the therapeutic logic was not signal blockade but antigen-directed payload delivery: bind PTK7 on tumor cells, internalize, and release an antimitotic auristatin inside the cell. PRO1107 was built to widen the therapeutic index of that approach with a hydrophilic linker and homogeneous DAR-8 loading (ProfoundBio first-patient-dosed release, 2024).

What actually happened

The study closed after 33 patients in a dose-finding stage, and no efficacy or pharmacodynamic results were posted (ClinicalTrials.gov, NCT06171789). The openFDA FAERS database held zero reports naming GEN1107 or PRO1107, consistent with a small, early, controlled-exposure population (openFDA FAERS, queried 2026-06-02). Genmab discontinued the program on a benefit-risk basis, one year after acquiring ProfoundBio for $1.8 billion (Genmab acquisition release, 2024). In Open Targets, the PTK7 association with neoplasm scored 0.1175, carried almost entirely by the literature datatype at 0.966, with no human genetic evidence (Open Targets, PTK7 ENSG00000112655). PTK7 sits in a worked landscape: the prior auristatin ADC cofetuzumab pelidotinib reached a first-in-human study without advancing to approval (doi:10.1158/1078-0432.CCR-20-3757), and additional PTK7 ADCs such as MTX-13 remain preclinical (doi:10.1158/1535-7163.MCT-23-0164).

Failure mechanism, best guess

The posted reason points to a therapeutic-index problem rather than a single safety event or a clean efficacy miss. A DAR-8 MMAE conjugate delivers a high payload load per antibody, which raises potency but also raises off-tumor toxicity risk when antigen expression is not tumor-selective enough to spare normal tissue. The basket design spread 33 patients across six histologies, which limited the depth of evidence in any single tumor type and made an early benefit-risk call more likely. PTK7 carries weak direct genetic support (Open Targets 0.1175, literature-only, no genetic datatype), so the program leaned on expression and preclinical regression data rather than human causal evidence (doi:10.1126/scitranslmed.aag2611, doi:10.1093/oncolo/oyae290). The most likely chain is that early safety and preliminary activity together did not define a dose with an acceptable window, and Genmab stopped rather than push a marginal index forward.

How to prevent this next time

Two quantitative tools would have sharpened the go-or-stop decision.

First, a historical base-rate adjustment anchored to the PTK7 ADC track record. Across oncology, the probability of moving from Phase 1 to approval sits near 5 to 10 percent, and antigen-delivery programs built on expression-only targets cluster at the low end (Clinical Development Success Rates, BIO, Informa, and QLS, 2021). A prior that explicitly debited the failure of cofetuzumab pelidotinib and the absence of PTK7 genetic evidence would have set a demanding bar for the dose-escalation readout before expansion.

Second, a Bayesian predictive-probability framework with explicit power for a target-engagement and therapeutic-index readout. A DAR-8 auristatin conjugate needs a pre-specified exposure-toxicity and exposure-response model, with the minimum tumor-selective dose defined as a gate.

PP(success)=∫θ​P(final success∣θ,interim data)⋅π(θ∣interim data)dθ

Pairing a predictive-probability gate with a payload-tolerability power calculation converts a six-histology basket into a decision-grade index experiment instead of a broad exposure screen.

The single highest leverage change would have been to pre-specify a therapeutic-index gate, a tumor-selective minimum effective dose validated against a base-rate-adjusted Bayesian predictive-probability threshold, before opening expansion across six tumor types.

What this means for similar programs

PTK7 remains a plausible delivery antigen with thin genetic support and at least one prior clinical ADC that did not reach approval. Programs that target expression-only antigens with high-DAR auristatin payloads should treat the therapeutic window as the central scientific question, not a tuning detail. Narrowing first-in-human entry to the one or two histologies with the strongest PTK7 expression and the clearest unmet need would buy deeper per-indication evidence than a wide basket.

Open questions

Did PRO1107 reach a dose with measurable PTK7 target engagement, and will any pharmacokinetic or response data be released? Was the limiting problem on-target normal-tissue toxicity, off-target payload toxicity, or insufficient activity at tolerable doses? Does DAR-8 loading help or hurt the PTK7 window relative to lower-DAR designs?

Sources

1. • ClinicalTrials.gov, NCT06171789 (Phase 1/2 single-group mTPI escalation, six tumor types, enrollment 33, primary endpoints adverse events and dose-limiting toxicities, start 2024-01-03, primary completion 2025-08-18, termination reason, no posted results). - ProfoundBio first-patient-dosed release, 2024 (PRO1107 binds PTK7, hydrophilic MMAE-based linker-drug, drug-to-antibody ratio 8). - Genmab acquisition of ProfoundBio, 2024 (transaction value $1.8 billion). - Open Targets Platform: PTK7 (ENSG00000112655) association with neoplasm 0.1175, literature datatype 0.966, no genetic datatype. - openFDA FAERS, queried 2026-06-02 (0 reports for GEN1107 or PRO1107). - PTK7-targeted ADC reduces tumor-initiating cells, 2017,- First-in-human cofetuzumab pelidotinib (PF-06647020), 2021,- MTX-13, a PTK7-directed ADC with widened therapeutic index, 2023,- Therapeutic targeting of PTK7 in cancer: an overview, 2025,. Available from: https://clinicaltrials.gov/study/NCT06171789.