Purpose <p>Extranodal natural killer/T-cell lymphoma (ENKTL) is an aggressive malignancy and usually involves the nasal cavity. Although the Lugano criteria is widely used for lymphoma response assessment, its applicability in ENKTL is limited by reliance on a single pre-treatment reference and a positive reference threshold with high false-positive rate. This study proposed and evaluated the ability of a novel evaluation system derived from the Lugano criteria to predict ENKTL treatment response.</p> Methods <p>A total of 259 patients underwent baseline, interim, and end-of-treatment [¹⁸F]FDG PET/CT. Treatment responses were evaluated on PET/CT using the Lugano, IPET (with interim PET/CT as the reference time point), NPU (with nasal mucosal uptake as the positive reference), and IPN (with integration of interim PET/CT as the reference time point and nasal mucosal uptake as the positive reference) criteria. Endpoints were progression-free survival (PFS) and overall survival (OS). Cox regression analysis was used to assess prognostic factors, and model discrimination was quantified using Harrell’s concordance index (C-index).</p> Results <p>The three novel systems reclassified patients initially evaluated by Lugano. Reclassification from partial response or stable disease to a worse response category by IPET and IPN correlated with worse outcomes, whereas reclassification to complete response by NPU or IPN predicted improved outcomes. The IPN achieved the highest C-index for prediction of PFS and OS and remained independently associated with survival after multivariable adjustment.</p> Conclusion <p>The IPN, IPET, and NPU criteria were superior to the Lugano classification for prognostic stratification. The IPN criteria had the best prediction performance and allowed early identification of high-risk patients.</p>