Development Of Theranostic Response Criteria In Solid Tumors (Thercist) And Tumor Burden Quantification Methods For Ga-68-Pet/Ct And '"Lu-Qspect/Ct

626 Objectives: Theranostics based on 68Ga for PET/CT imaging and 177Lu for therapy are growing fast, owing to the tremendous success of PRRT of neuroendocrine tumors and PSMA RLT of prostate cancer. As 177Lu can also be imaged with quantitative SPECT (QSPECT)/CT, there is an opportunity to use both molecular imaging modalities for objective response assessment during and after multi-cycle radionuclide therapies, and this could overcome some limitations of radiological imaging for this purpose (e.g. RECIST). Furthermore, significant lesion uptake is required for eligibility to these therapies, which facilitates tumor segmentation. Our aims were to evaluate practical methods to quantify the tumor burden with both 68Ga-PET/CT and 177Lu-QSPECT/CT, and to propose adapted molecular response criteria: the Theranostic Response Criteria in Solid Tumors (THERCIST). Methods: 17 consecutive patients with neuroendocrine tumor who underwent 68Ga-octreotate-PET/CTs within 4 months prior to and after a PRRT induction course of 3 to 4 cycles were included. The 177Lu-QSPECT/CTs performed 3 days after the first and the last of these 177Lu-octreotate administrations were analyzed. Tumor was segmented using relative thresholds of 30% and 15% of SUVmax for PET and QSPECT, respectively, with exclusion of organs and urine activity, to derive the total lesion fraction (TLF: total lesion activity divided by body weight, expressed in %IA) and molecular tumor volume (MTV). Correlations were made between 68Ga and 177Lu tumor burden indices. The molecular imaging response was assessed using these categories: CMR (complete molecular response): disappearance of all lesions; PMR (partial molecular response): ≥50% decrease in tumor burden; MMR (minor molecular response): ≥30% but <50% decrease in tumor burden; SMD (stable molecular disease): <30% decrease or <50% increase in tumor burden; PMD (progressive molecular disease): ≥50% increase in tumor burden, or appearance/progression of avid or non-avid lesion(s). Results: Baseline median (range) TLF was 1.8 (0.2 - 32.2) %IA on QSPECT and 2.3 (0.5 - 33.9) %IA on PET, and baseline median MTV was 104 (24 - 2210) cc on SPECT and 101 (22 - 1480) cc on PET. BothTLF and MTV were highly and linearly correlated between PET and QSPECT (Pearson r = 0.90 and 0.96, respectively; P < 0.0001). Molecular response assessed with MTV better corresponded to that perceived visually, and was thus preferred over TLF-based response. Using MTV, there was no CMR or PMD in this cohort. There was an agreement between QSPECT and PET in 9 cases (2 PMR, 2 MMR, 5 SMD), while one MMR on QSPECT converted to PMR on PET, and of 7 SMD on QSPECT, 2 converted to PMR and 5 to MMR on PET. Therefore, in all patients, the QSPECT response category was maintained or did improved on later PET, which is consistent with a further delayed response that is often seen after PRRT. Conclusion: Using a simple tumor segmentation approach with distinct thresholds, both 68Ga-PET/CT and 177Lu-QSPECT/CT have a fairly comparable capability to quantify the tumor burden. Molecular theranostic imaging is an appealing alternative to anatomical imaging for real-time objective response monitoring of PRRT, PSMA RLT, as well as future theranostic applications. Research Support: Scholarship from Fonds de recherche du Quebec - Sante to JMB.