Abstract
242420
Introduction: Dosimetry is of increasing clinical interest for radiopharmaceutical treatment personalization and as a potential quantitative descriptor of patient-specific treatment response and side-effects. However, the degree to which absorbed doses to organs-at-risk and tumor targets are associated with imaging metrics from the pre-therapy PET or post-therapy SPECT scans remains largely unknown. The objectives of this work are twofold: (1) to calculate patient-specific 3D dose distributions for patients treated with 177Lu-PSMA-617 and (2) to then investigate the correlation between PET and SPECT imaging parameters with absorbed dose to identify potential predictors of dose delivery.
Methods: Seven patients treated with 177Lu-PSMA-617 with accompanying dosimetry performed after cycle 1 are included. Patients were referred for dosimetry due to a disease pattern or clinical factors that posed concern for organ toxicity (e.g. impaired renal function), or at the onset of retreatment with 177Lu-PSMA-617 therapy. All patients underwent pre-therapy PSMA PET imaging for treatment eligibility and serial 177Lu SPECT imaging for dosimetry at 4 time points following the first cycle of treatment: 4h, 24h, 48h, and 72h. SPECT images were acquired from the vertex to mid-thighs, followed by iterative, quantitative reconstruction. In addition to the metastatic lesions, the following organs were included for dosimetry analysis: salivary glands, lungs, liver, kidneys, and bone marrow. Voxelwise dosimetry calculations were performed using commercial software (Sureplan MRT, MIM Software). Pre-therapy PET and +24h SPECT SUV metrics for bilateral kidneys and salivary glands were extracted from CT-based volumes using a deep learning algorithm (Contour Protégé AI, MIM Software). Metastatic lesions were segmented using a global SUV 3 threshold to generate total tumor burden volumes. Spearman correlation coefficients were calculated between these image-based total tumor volumes, SUV metrics, and absorbed doses for kidneys, parotid glands, and metastatic lesions.
Results: The mean (±SD) absorbed doses to the kidneys, salivary glands, and bone marrow were 3.03 ± 1.56 Gy, 1.42 ± 0.88 Gy, and 0.24 ± 0.10 Gy, respectively (for a single cycle with mean injected activity of 195.8 mCi). In total, 120 metastatic lesions were included (range 7 – 42 per patient) with a mean absorbed dose of 7.12 ± 6.23 Gy (range 0.19 – 23.58 Gy). For kidneys, a moderate positive correlation was observed between the SPECT SUVmean and mean kidney dose (r=0.54; p=n.s.). However, for the parotid glands, the mean SUV by PET was weakly correlated with absorbed dose (r=0.37) and no linear relationship between SPECT SUVmean and absorbed dose was observed (r=0.03). For tumors, there was a weak correlation between increased TTV PET SUVmean and higher dose (r=0.21; p=n.s.).
Conclusions: Individualized dosimetry for patients treated with 177Lu-PSMA-617 revealed tolerable doses to normal organs accompanied by a high degree of heterogeneity in tumor-absorbed doses (mean dose 7.12 ± 6.23 Gy; n=120 metastatic lesions). Investigation into the correlation between imaging metrics from the pre-therapy PET or post-therapy SPECT and the absorbed doses to organs-at-risk showed several positive correlations between mean uptake and dose. No significant correlations were identified between imaging variables and absorbed tumor dose, suggesting that single time points of imaging without additional data may be insufficient to predict dose delivery. Further studies will include an expanded study cohort.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.