Real-World Lesion and Renal Dosimetry for Peptide Receptor Radionuclide Therapy (PRRT)

Introduction: To demonstrate the feasibility of comprehensive lesion and organ post-therapy SPECT/CT dosimetry in routine follow-up of ¹⁷⁷Lu-DOTA-Octreotate (LuTATE) PRRT and evaluate the impact of lesional radiation dose on change in target volume over multiple cycles. Additionally, renal dosimetry over the therapy course is evaluated to enable a theoretical range of patient-optimized activities to maximize tumor dose while adhering to organ exposure thresholds.

Methods: Dosimetry for kidneys and index lesions were assessed on post-treatment quantitative SPECT/CT imaging 24 hrs after each PRRT cycle in 97 patients. Renal retention was evaluated using an AI autocontouring algorithm based on CT margins. In patients with non-diffuse disease (n=91), uptake to an index tumor lesion with highest avidity at cycle 1 was measured and tracked at each treatment with semi-automated volumetric adaptation. Dosimetry estimates were based on biological half-lives for tumor and kidneys from a previous population study employing multi-timepoint imaging. Lesional response was quantified based on change in localized tumor volume-intensity product (VIP) over multiple cycles.

Results: All patients had neuroendocrine neoplasia (G1 15, G2 51, G3 16 and unknown 9). Prescribed activities were 7.8± 1.4 GBq per treatment. Sixty-five patients received 4 treatment cycles while 19 and 13 received 3 or only 2 cycles, respectively. The mean index lesion absorbed dose at cycle 1 was 5.9±5.6 Gy/GBq (median 4.4 Gy/GBq). Dose values were comparable among G2 and G3 cohorts at 6.9±6.2 and 6.5±7.8 Gy/GBq, but significantly lower in G1 index lesions at 3.7±2.3 Gy/GBq (p=0.003 compared to G2/3 group). In patients receiving all planned cycles (n=65), the index lesion dose reduced by -25.5%, -38.7%, and -45.3% at cycles 2-4, respectively. In terms of index lesion VIP, response was observed in 90% of patients with an overall reduction of -51±42% from cycle 1 to last treatment. Response was inversely related to grade (G3 -72±29%, G2 -50±47%, G1 -26±35%, and unknown grade -62±30%). Radiation dose at cycle 1 correlated with VIP response on treatment cessation (r=-0.39, P<0.001). Notably, all lesions receiving 10 Gy/GBq or greater at the first cycle (n=8) had a reduction in local VIP of at least 70% (-88±8%) regardless of grade and patients with no VIP response (n=9) all received less than 4 Gy/GBq.

Renal dose per GBq administered across 4 treatment cycles increased from baseline (4±19% cycle 2, 13±24% cycle 3 &amp; 12±24% cycle 4) with cycle 1 dose being highly predictive of uptake in subsequent therapies (Pearson r values of 0.81, 0.75, and 0.80 for cycles 2-4). No correlation was observed between index lesion and renal absorbed doses across patients. Theoretical personalized activities to adhere to the nominal 23 Gy renal dose threshold (5.75 Gy/cycle) were 13.4±7.1GBq per treatment. The majority of patients (79%) would be suitable to receive >10 GBq per cycle; only 5% of patients would exceed the dose threshold using a standard activity of 7.4 GBq over four treatments.

Conclusions: Post-treatment LuTATE PRRT dosimetry suggests lesion dose was typically highest in cycle 1, while renal dose was lowest, allowing an opportunity for increasing administered activity and then adjusting subsequent cycles based on renal dosimetry to remain within recommended limits. In most cases, this would allow a significant increase in prescribed activity with potential improvement in treatment efficacy and preserved safety.

• Download figure
• Open in new tab
• Download powerpoint