Abstract
<jats:p>Background: TNM staging is essential for cancer registries but is often embedded in unstructured pathology reports, requiring manual extraction. Transformer-based models have recently been proposed for automated TNM classification from English pathology texts. Objectives: To assess the robustness and cross-lingual transferability of pre-trained English-language TNM classifiers applied without fine-tuning to German pathology reports. Methods: Three publicly available transformer-based TNM classifiers (T, N, M) were applied to a synthetic German pathology dataset of 109 breast, lung, and prostate cancer reports, using expert-assigned TNM labels as the gold standard. Results: The models achieved high precision and specificity but moderate sensitivity for tumor and nodal staging and low sensitivity for metastasis, frequently abstaining by predicting “Unknown.” Conclusion: English-trained TNM classifiers can extract staging information from German pathology reports in a zero-shot setting with high reliability when predictions are made, but reduced recall. This conservative behavior supports their use as high-precision screening tools in registry workflows and could be improved through limited domain adaptation.</jats:p>