LAUSR

Primary cutaneous CD30-positive T-cell lymphoproliferative disorders (PCLPDs) comprise the second most common group of cutaneous T-cell lymphomas (CTCLs) and include lymphomatoid papulosis, primary cutaneous anaplastic large cell lymphoma (ALCL), and borderline cases (Willemze et al, 2017). Despite a variety of clinicopathological manifestations, PCLPDs are considered to belong to a single disease spectrum characterized by indolent clinical behaviour with relatively uncommon extracutaneous spread (Bekkenk et al, 2000). Diagnostically, PCLPDs must be distinguished from reactive processes, other CTCLs, such as mycosis fungoides (MF), and secondary cutaneous involvement by systemic lymphoma. Clinical outcomes in patients with PCLPDs are generally excellent but a minority of tumours demonstrates more aggressive behaviour; 5-year overall survival rates range from 70% to 100% depending on the series. However, few prognostic factors have been identified. Specifically, the impact of ethnicity or race on outcomes has been difficult to evaluate due to the overall low incidence of the disease and low event rates (Yu et al, 2008). The age-adjusted incidence rate of T cell lymphomas is lower for Whites than for Blacks, and improved outcomes in Whites have been reported both for T cell lymphomas overall and in MF (Abouyabis et al, 2008; Nath et al, 2014). However, racial and/or ethnic differences in the incidence and outcomes of PCLPDs remain unknown. In this issue of the British Journal of Haematology, Su et al (2018) sought to identify prognostic factors in 1496 PCLPD patients in the National Cancer Database (NCDB), with special emphasis on outcomes among ethnic groups. Using multivariate analysis, they found, among other significant associations, that Blacks had significantly poorer overall survival than Whites. This difference persisted after adjusting for interaction between ethnicity and income, particularly in younger patients. The reason for poorer outcomes in Blacks with PCLPDs is probably multifactorial, including biological/ genetic factors, more advanced disease stage at presentation, access to care and associations with other risk factors, such as socioeconomic status; unfortunately, a comprehensive dissection of the interplay among these factors is not possible given the study design and limitations inherent in the source database. A multi-level model for examining these factors and their interactions has been proposed for diffuse large B cell lymphoma and other diseases with health disparities (Warnecke et al, 2008; Flowers & Nastoupil, 2014). Similar approaches could be forwarded for PCLPD once appropriate supporting data sources become available. In addition, as the authors acknowledge, the NCDB lacked data on specific disease subtype and cause of death data and might be subject to diagnostic coding errors. Nevertheless, while the immediate clinical utility of the findings for diagnosis and management of PCLPD remains unclear, the finding that ethnicity or race is a significant prognostic factor has important implications for future research. Data from registry-based outcome studies reflecting primarily White populations, as the authors postulate may be the case for much of the published experience in PCLPD (Su et al, 2018), might not be equally applicable to more racially mixed populations. Establishment of prospective multi-ethnic cohorts with adequate patient outreach is critical to understanding the diversity of outcomes in such populations (Anderson et al, 2015). In addition, the establishment of multi-ethnic cohort studies with sufficient availability of biospecimens for genetic and other molecular studies is needed to identify candidate biomarkers or therapeutic targets, the applicability of which may differ among racial and ethnic groups (Rodriguez et al, 2016). Therefore, biomarker studies should be adequately powered to identify high-risk signatures in these groups. Outcome differences also may reflect disparities in access to state-of-the-art screening, Correspondence: Andrew L. Feldman, Department of Laboratory Medicine and Pathology, Mayo Clinic, Hilton Building, Room 8-00F, 200 First Street Southwest, Rochester, MN 55905, USA. E-mail: [email protected] editorial comment