Correspondence (reply): In Reply

Professor Lippert rightly points out the clinical relevance of tumor classification. He requests specifically that the diagnostic evaluation should provide information regarding the behavior of a tumor in response to systemic treatment. The studies he cites give an interesting insight into currently available methods. They allow conclusions about the possible resistance of a tumor to certain chemotherapeutic regimens. By comparison, predicting a positive therapeutic response—that is, the sensitivity to a particular cytotoxic agent—is subject to far greater uncertainty.

Pathologists have recognized this problem. The relevant subspecialty is known as “predictive pathology,” and in molecular tumor diagnosis, differentiation is made between diagnostic, prognostic, and predictive biomarkers (1, 2). However, the boundaries between these markers are fluid. A confirmed activating epidermal growth factor receptor (EGFR) mutation, for example, is associated not only with sensitivity to treatment with an EGFR tyrokinase inhibitor, but also with a better prognosis for the patient. The mere morphological distinction of small cell and non-small cell lung cancers is already predictive for initial sensitivity or resistance to conventional chemotherapy. Equally, the new, morphology-based classification of adenocarcinoma of the lung yields prognostic as well as predictive information (3, 4). It confirms in an exemplary way that a tumor's histomorphology and cytomorphology contains characteristics whose clinical relevance has hitherto remained unidentified. It may therefore be assumed that differentiated morphological diagnostic evaluation will continue to have a key position in tumor classification. It will be complemented by selective, targeted use of biomarkers to identify clinically relevant subgroups. The challenge for pathologists is therefore in possessing in-depth knowledge of molecular mechanisms in tumor biology and tumor therapy, in addition to having a good eye for morphology.