AI pathology links TROP2 location to drug outcomes
A Nature Medicine study links TROP2’s location within tumor cells to outcomes on datopotamab deruxtecan, but its biomarker threshold still needs evaluation in an independent dataset.
A November 11 Nature Medicine study uses deep-learning pathology to measure TROP2 staining separately in tumor-cell membranes and cytoplasm.
Its normalized membrane ratio is membrane signal divided by membrane plus cytoplasmic signal. Lower ratios mean relatively more signal inside the cell. [1]
Researchers classified a tumor sample as biomarker-positive when at least 75% of its tumor cells had a ratio at or below 0.56. “Positive” therefore means a predominance of lower-ratio cells, not more membrane staining. The threshold was selected using the same trial dataset in which outcomes were assessed. [1]
The analysis covered 221 patients with biomarker-evaluable nonsquamous lung cancer and no actionable genomic alterations in TROPION-Lung01. Among 108 receiving datopotamab deruxtecan (Dato-DXd), median progression-free survival was 7.2 months in the positive group and 4.0 months in the negative group. Among 113 receiving docetaxel, the corresponding medians were 4.1 and 4.4 months. These compare biomarker groups within each treatment, not the treatment effect of one drug versus the other. [1][2]
The treatment-dependent pattern warrants further study. Resampling tested the threshold’s stability, but cannot replace evaluation on independent patients. The next decisive test is performance after fixing the threshold in advance of a new dataset. [1][3]
Sources & details 4 sources
Discuss in your voice app
Take the reporting, visual explanations and original sources into a conversation. Copy the brief, paste it into your preferred assistant, then start its voice mode.
Paste the brief into your voice app to discuss this reporting.Revised 12 November 2026. Revised visual explanations and copy; added Nvidia investigation context and clarified tool eligibility.
Reporting sources
- Nature Medicine · TROP2 membrane-ratio study
November 11 original paper and full methods. Threshold optimized within TROPION-Lung01; independent threshold validation remains a limitation. Publication date is distinct from the March 29, 2023 efficacy-data cutoff.
- Nature Medicine · Figure 4
Original figure opened and visually inspected. Dato-DXd: NMR-positive n=68, negative n=40; docetaxel: positive n=72, negative n=41. This is a second artifact from the same study, not independent replication.
- Journal of Clinical Oncology · parent randomized trial
Separate original trial report confirms randomized Dato-DXd versus docetaxel design and efficacy endpoint. Does not independently validate the later TROP2 threshold.
- Roche · diagnostic development status
April 29, 2025 announcement of Breakthrough Device designation and the membrane/cytoplasm measurement approach. Development partner statement; designation is not approval.
Images & figures
The candidate biomarker separates outcome groups within Dato-DXd more strongly than within docetaxel. Its threshold was selected using the same trial.
Euangelion · Source · Usage terms · Download figure