In cutaneous T-cell lymphoma (CTCL), skin scoring and lymph node (LN) assessment are essential for scoring disease response in both clinical practice and in trials. Current assessment methods rely on calculating relative changes in the Modified Severity-Weighted Assessment Tool (mSWAT) and target LN size during a patient’s treatment course. However, measuring relative changes fails to account for mixed treatment effects comprising the simultaneous regression of treatment-responsive cells and growth of treatment-resistant cells. Additionally, the clinical significance of LN size on imaging scans remains uncertain in CTCL. LNs are currently measured using linear dimensions that may not capture the extent of nodal involvement or be sensitive enough to track clinically meaningful changes over time. The aim of this study was to perform kinetic analysis using mSWAT data and computed tomography LN measurements from the MAVORIC trial to determine if dynamic changes over time correlate with overall survival (OS) in CTCL. Kinetic modeling evaluated simultaneous rates of growth (g) and regression (d) using a two-phase equation. Unidimensional and volumetric LN sizes were measured using a novel radiomics imaging platform. Tumor doubling times can be estimated based on g, using the equation dt = 0.693/g. An initial cohort of 119 patients were included, and in most patients, g and d values could be estimated. Patients were divided into two groups based on g, and correlation with OS was evaluated. The median g was 0.0013d−1 in the half with slowest rate of increase in SWAT scores (doubling time = 533 days); with g of 0.0057d1 in those with fastest growth (doubling time = 122 days). For unidimensional LN size, volumetric LN size, and mSWAT, g was highly correlated with OS; faster rates of nodal growth or increase in SWAT scores were associated with shorter OS. Median OS for patients with faster unidimensional LN, volumetric LN, and mSWAT growth rates were 29, 27.7, and 17 months, respectively, whereas median OS were not reached for patients with slower growth rates (p = 0.012, 0.0035, and 0.0012, respectively). Our interim analysis demonstrates that skin scoring and lymph node growth kinetics predict survival in CTCL. Thus, g has potential as a novel biomarker for prognostication and treatment response assessment. It may also serve as a clinical trial endpoint that allows for earlier readout of trial results than conventional survival endpoints. Further validation in a larger cohort, and ultimately with prospective studies, is needed. The research was funded by: Kyowa Kirin Keywords: Diagnostic and Prognostic Biomarkers, Cutaneous non-Hodgkin lymphoma, Imaging and Early Detection - Other Conflicts of interests pertinent to the abstract L. J. Geskin Consultant or advisory role Kyowa Kirin Research funding: Kyowa Kirin
Read full abstract