Prediction of Progression-Free Survival in Patients With Multiple Myeloma Following Anthracycline-Based Chemotherapy Based on Dynamic FDG-PET

Methods:Dynamic positron emission tomography (PET) studies with F-18deoxyglucose were performed in patients with multiple myeloma who received anthracycline-based chemotherapy to evaluate the impact of full kinetic analysis and assess its value with regard to progression-free survival (PFS). The evaluation included 19 patients (56 metastatic lesions) with multiple myelomas. All patients received combined anthracycline-based chemotherapy. PFS served as a reference for the PET data. All patients were examined prior to the onset of chemotherapy and on days 23 to 28 after the onset of the first cycle (prior to the second cycle). The following parameters were retrieved from the dynamic PET studies: Standardized Uptake Value (SUV), fractal dimension (FD), 2 compartment model with computation of K1, k2, k3, k4 (unit: 1/min), the fractional blood volume (vB), and the FDG-influx according to Patlak were calculated. Results:The observed PFS varied from!1 month to 64.1 months with a median PFS of 26 months. Most kinetic parameters demonstrated only small changes, primarily declining after 1 cycle. We compared the kinetic data of each study using a Wilcoxon matched-pairs signed rank test. The results were considered significant forP!0.05. The test revealed a significant change for the SUV (z"4.954,P!0.0000), the FD (z"5.036,P! 0.0000), the fractional blood volume vB (z"4.116,P!0.0000) and influx (z"2.614,P!0.0090) when the absolute values of the first and the second study were compared. We dichotomized the patients according to the PFS of 18 months and defined 2 survival groups. The data demonstrate that the correct classification rate (CCR) of group 2 (survival:#18 months) was generally higher (exceeding 94%) than for group 1. The use of the baseline SUV led to a CCR of 82% for the group 2 with the longer survival. The CCR of group 1 with the short survival varied between 55% and 70% depending on the parameter and the study used for prediction. Furthermore, the CCR for both groups based only on the data of the second study was somewhat lower (74%‐75%) as compared with the baseline FDG study (75%‐82%). Finally, the combined use of the 6 predictor variables, namely SUV, k3, and FD (selected by the Wilcoxon rank sum test) of each study led to the highest CCR of 85% for both groups. This combination was in particular useful for the prediction of group 2 with the longer survival with a CCR of 94%. Best cutoff-values for the differentiation between short and long PFS were SUV of 4.0 and a k3 of 0.07 of the baseline study. Conclusions:The results demonstrate, that a full kinetic analysis of the FDG studies prior and after 1 chemotherapeutic cycle in patients with multiple myeloma is helpful for the prediction of PFS and may be used to identify those patients who benefit from this chemotherapeutic protocol. A high SUV (#4.0 SUV) as well as a high k3 (#0.07) of the baseline study were bad prognostic parameters and related to a short PFS.