Pitfalls and major issues in the histologic diagnosis of peripheral T-cell lymphomas: results of the central review of 573 cases from the T-Cell Project, an international, cooperative study

Peripheral T-cell lymphomas (PTCLs) comprise a heterogeneous group of neoplasms that are derived from post-thymic lymphoid cells at different stages of differentiation with different morphological patterns, phenotypes and clinical presentations. PTCLs are highly diverse, reflecting the diverse cells from which they can originate and are currently sub-classified using World Health Organization (WHO) 2008 criteria. In 2006 the International T-Cell Lymphoma Project launched the T-Cell Project, building on the retrospective study previously carried on by the network, with the aim to prospectively collect accurate data to improve knowledge on this group of lymphomas. Based on previously published reports from International Study Groups it emerged that rendering a correct classification of PTCLs is quite difficult because the relatively low prevalence of these diseases results in a lack of confidence by most pathologists. This is the reason why the T-Cell Project requested the availability of diagnostic material from the initial biopsy of each patient registered in the study in order to have the initial diagnosis centrally reviewed by expert hematopathologists. In the present report the results of the review process performed on 573 cases are presented. Overall, an incorrect diagnosis was centrally recorded in 13.1% cases, including 8.5% cases centrally reclassified with a subtype eligible for the project and 4.6% cases misclassified and found to be disorders other than T-cell lymphomas; 2.1% cases were centrally classified as T-Cell disorders not included in the study population. Thus, the T-Cell Project confirmed the difficulties in providing an accurate classification when a diagnosis of PTCLs is suspected, singled out the major pitfalls that can bias a correct histologic categorization and confirmed that a centralized expert review with the application of adequate diagnostic algorithms is mandatory when dealing with these tumours.