NETTER-1 trial demonstrated high efficacy and low toxicity of four cycles of Peptide Receptor Radionuclide Therapy (PRRT) in patients with metastasized NET. The present study evaluates the outcome of further PRRT cycles in the so called salvage setting in patients after initial response to four therapy cycles and later progression. Thirty five patients (pat.) (25 male, 10 female, 63 ± 9 years) with progressive, metastasized NET (23 small intestinal, 5 lung, 4 CUP, 1 rectal, 1 gastric and 1 paraganglioma) were included. All patients previously received 4 PRRT cycles with 177Lu-DOTATATE and showed initial response. SPECT based dosimetry was applied to determine kidney and tumor doses. Therapy response was evaluated using 68Ga-DOTATATE PET/CT (with high dose CT), CT alone or MRI (RECIST 1.1), toxicity was defined using CTCAE 5.0 criteria. 99mTc99-MAG3 scintigraphy was used to assess potential renal tubular damage. Progression free survival (PFS) and Overall survival (OS) analysis was performed with the Kaplan-Meier-method. The median PFS after initial PRRT was 33 months (95% CI: 30–36). The mean cumulative dose for including salvage PRRT was 44 GBq (range 33.5–47). One pat. (2.9%) showed grade 3 hematotoxicity. Kidney dosimetry revealed a mean cumulative kidney dose after a median of 6 PRRT cycles of 23.8 Gy. No grade 3 / 4 nephrotoxicity or relevant decrease in renal function was observed. Follow-up imaging was available in 32 patients after salvage therapy. Best response according to RECIST 1.1. was PR in one patient (3.1%), SD in 26 patients (81.3%) and PD in 5 patients (15.6%). PFS after salvage therapy was 6 months (95% CI: 0–16; 8 patients censored). Mean OS after initial PRRT was 105 months (95% CI: 92–119) and 51 months (95% CI: 41–61) after start of salvage therapy. Median OS was not reached within a follow-up of 71 months after initial PRRT and 25 months after start of salvage PRRT, respectively. Salvage therapy with 177Lu-DOTATATE is safe and effective even in patients with extensive previous multimodal therapies during disease progression and represents a feasible and valuable therapy option for progressive NET.

The bone marrow (BM) is a main risk organ during Lu-177-PSMA ligand therapy of metastasized castration-resistant prostate cancer (mCRPC) patients. So far, BM dosimetry relies on S values, which are pre-computed for reference anatomies, simplified activity distributions, and a physiological BM distribution. However, mCRPC patients may show a considerable bone lesion load, which leads to a heterogeneous and patient-specific activity accumulation close to BM-bearing sites. Furthermore, the patient-specific BM distribution might be significantly altered in the presence of bone lesions. The aim was to perform BM absorbed dose calculations through Monte Carlo (MC) simulations and to investigate the potential value of image-based BM localization. This study is based on 11 Lu-177-PSMA-617 therapy cycles of 10 patients (10 first cycles), who obtained a pre-therapeutic Ga-68-PSMA-11 PET/CT; quantitative Lu-177 SPECT acquisitions of the abdomen 24 (+CT), 48, and 72 h p.i.; and a Lu-177 whole-body planar acquisition at 24 h post-therapy. Patient-specific 3D volumes of interest were segmented from the Ga-68-PSMA-11 PET/CT, filled with activity information from the Lu-177 data, and imported into the FLUKA MC code together with the patient CT. MC simulations of the BM absorbed dose were performed assuming a physiological BM distribution according to the ICRP 110 reference male (MC1) or a displacement of active BM from the direct location of bone lesions (MC2). Results were compared with those from S values (SMIRD). BM absorbed doses were correlated with the decrease of lymphocytes, total white blood cells, hemoglobin level, and platelets. For two patients, an additional pre-therapeutic Tc-99m-anti-granulocyte antibody SPECT/CT was performed for BM localization. Median BM absorbed doses were 130, 37, and 11 mGy/GBq for MC1, MC2, and SMIRD, respectively. Significant strong correlation with the decrease of platelet counts was found, with highest correlation for MC2 (MC1: r = − 0.63, p = 0.04; MC2: r = − 0.71, p = 0.01; SMIRD: r = − 0.62, p = 0.04). For both investigated patients, BM localization via Tc-99m-anti-granulocyte antibody SPECT/CT indicated a displacement of active BM from the direct location of lesions similar to model MC2 and led to a reduction in the BM absorbed dose of 40 and 41% compared to MC1. Higher BM absorbed doses were observed for MC-based models; however, for MC2, all absorbed doses were still below 2 Gy. MC1 resulted in critical values for some patients, but is suspected to yield strongly exaggerated absorbed doses by neglecting bone marrow displacement. Image-based BM localization might be beneficial, and future studies are recommended to support an improvement for the prediction of hematoxicities.

Background: We performed an initial evaluation of non-invasive ECG-gated [¹⁸ F]FDG-positron emission tomography (FDG-PET) for serial measurements of left ventricular volumes and function in murine models of dilated (DCM) and ischemic cardiomyopathy (ICM), and then tested the effect of erythropoietin (EPO) treatment on DCM mice in a preliminary FDG-PET therapy monitoring study. Methods: Mice developed DCM 8 weeks after injection with Coxsackievirus B3 (CVB3), whereas ICM was induced by ligation of the left anterior descending artery. LV volumes (EDV and ESV) and the ejection fraction (LVEF) of DCM, ICM and healthy control mice were measured by FDG-PET and compared with reference standard results obtained with 1.5 T magnetic resonance imaging (MRI). In the subsequent monitoring study, LVEF of DCM mice was evaluated by FDG-PET at baseline, and after 4 weeks of treatment, with EPO or saline. Results: LV volumes and the LVEF as measured by FDG-PET correlated significantly with the MRI results. These correlations were higher in healthy and DCM mice than in ICM mice, in which LVEF measurements were somewhat compromised by absence of FDG uptake in the area of infarction. LV volumes (EDV and ESV) were systematically underestimated by FDG-PET, with net bias such that LVEF measurements in both models of heart disease exceeded by 15% to 20% results obtained by MRI. In our subsequent monitoring study of DCM mice, we found a significant decrease of LVEF in the EPO group, but not in the saline-treated mice. Moreover, LVEF in the EPO and saline mice significantly correlated with histological scores of fibrosis. Conclusions: LVEF estimated by ECG-gated FDG-PET significantly correlated with the reference standard MRI, most notably in healthy mice and mice with DCM. FDG-PET served for longitudinal monitoring of effects of EPO treatment in DCM mice.

The bone marrow (BM) is a main organ at risk in Lu-177-PSMA-617 therapy of prostate cancer and Lu-177-Octreotate therapy of neuroendocrine tumours. BM dosimetry is challenging and time-consuming, as different sequential quantitative measurements must be combined. The BM absorbed dose from the remainder of the body (ROB) can be determined from sequential whole-body planar (WB-P) imaging, while quantitative Lu-177-SPECT allows for more robust tumour and organ absorbed doses. The aim was to investigate a time-efficient and patient-friendly hybrid protocol (HP) for the ROB absorbed dose to the BM. It combines three abdominal quantitative SPECT (QSPECT) scans with a single WB-P acquisition and was compared with a reference protocol (RP) using sequential WB-P in combination with sequential QSPECT images. We investigated five patients receiving 7.4 GBq Lu-177-Octreotate and five patients treated with 3.7 GBq Lu-177-PSMA-617. Each patient had WB-P and abdominal SPECT acquisitions 24 (+ CT), 48, and 72 h post-injection. Blood samples were drawn 30 min, 80 min, 24 h, 48 h, and 72 h post-injection. BM absorbed doses from the ROB were estimated from sequential WB-P images (RP), via a mono-exponential fit and mass-scaled organ-level S values. For the HP, a mono-exponential fit on the QSPECT data was scaled with the activity of one WB-P image acquired either 24, 48, or 72 h post-injection (HP24, HP48, HP72). Total BM absorbed doses were determined as a sum of ROB, blood, major organ, and tumour contributions. Compared with the RP and for Lu-177-Octreotate therapy, median differences of the total BM absorbed doses were 13% (9-17%), 8% (4-15%), and 1% (0-5%) for the HP24, HP48, and HP72, respectively. For Lu-177-PSMA-617 therapy, total BM absorbed doses deviated 10% (2-20%), 3% (0-6%), and 2% (0-6%). For both Lu-177-Octreotate and Lu-177-PSMA-617 therapy, BM dosimetry via sequential QSPECT imaging and a single WB-P acquisition is feasible, if this WB-P image is acquired at a late time point (48 or 72 h post-injection). The reliability of the HP can be well accepted considering the uncertainties of quantitative Lu-177 imaging and BM dosimetry using standardised organ-level S values.