Repeated ¹⁷⁷Lu-Labeled PSMA-617 Radioligand Therapy Using Treatment Activities of Up to 9.3 GBq

DISCUSSION

Here, we retrospectively report our clinical experience with various treatment activities of ¹⁷⁷Lu-PSMA-617 during salvage therapy of 40 patients with metastasized castration-resistant prostate cancer.

The first clinical application of PSMA-RLT was done in 2011–2012 using a ¹³¹I-labeled PSMA ligand. The treatment activity was chosen after ¹²⁴I PET-based dosimetry, taking into account a 1-Gy red marrow tolerance dose; hematologic toxicities were mild (10). During repeated treatments, grade 3/4 hematologic toxicities remained rare (16). Any PSA response was demonstrated in 84% (21/25) of the patients (10); a PSA decline of greater than 50% after the first treatment cycle was achieved in 70.6% of 34 patients (17). PSMA-RLT based on ¹⁷⁷Lu has practical advantages, and the PSMA-617 ligand even has refined pharmacokinetics and radiation dosimetry (5,8). Thus, theoretically, ¹⁷⁷Lu-PSMA-617 should improve the therapeutic range of PSMA-RLT. Nevertheless, until now, the reported PSA response rates commonly have been remarkably lower; the largest multicenter investigation of ¹⁷⁷Lu-PSMA-617 reported only a 40% biochemical response rate after the first treatment cycle (17). Therefore, a critical discussion about currently used treatment protocols seems warranted.

Dosimetry studies of ¹⁷⁷Lu-PSMA-617 have been performed with variable methodologies at different centers, but all investigators have reported similar results (6–8,18,19). The essential organs red marrow (∼0.03 Gy/GBq) and kidneys (∼0.6 Gy/GBq) should be considered dose-limiting organs (6–8,18,19). Given a red marrow tolerance dose of 1 Gy, single-cycle activities of up to 30 GBq of ¹⁷⁷Lu could be proposed (7). Taking into account the concept of a biologic effective dose during low-dose-rate radionuclide therapy, a kidney tolerance dose of 28–40 Gy was suggested for ¹⁷⁷Lu-radiopharmaceuticals (20), theoretically making a cumulative treatment activity of 50 GBq of ¹⁷⁷Lu-PSMA-617 reasonable. No grade 3/4 renal toxicity was observed in 55 patients treated with 3 doses of 6 GBq of ¹⁷⁷Lu-PSMA-617 (21). However, the high prevalence of bone metastases, the use of approved chemotherapeutic options before salvage therapy, and the fact that the patients were elderly introduced some doubt about whether the tolerance doses of red marrow and other organs reported in the literature were still valid for the patient cohort. Thus, it was reasonable that ¹⁷⁷Lu-PSMA-617 RLT treatment regimens were initially introduced with caution.

Recent publications predominantly focused on a treatment activity of 6 GBq administered every 2 mo (22–28), and the results were in agreement with each other. All authors reported few cases of grade 3/4 toxicity—that is, in the same dimension as the incidence of tumor-related adverse events observed in the placebo arm of the ALSYMPCA trial (14)—and, despite moderate xerostomia, there was no evidence of relevant treatment-related toxicity. However, the treatment activities used were remarkably lower than the projected maximum tolerance dose, according to dosimetry estimates: a red marrow dose of only 0.2 Gy (6 GBq × 0.03 Gy/GBq). Simultaneously, the PSA response rates were lower than older ¹³¹I-PSMA RLT data obtained with the full 1-Gy red marrow tolerance limit (10,16). Surprisingly, none of the authors (22–28) discussed the possibility that the escalation of ¹⁷⁷Lu treatment activity to an estimated red marrow absorbed dose of 0.2–1.0 Gy should still be tolerated well and should offer the chance to further improve antitumor activity because a positive dose–response relationship is normally expected in radiotherapy. Therefore, after the clinical introduction of ¹⁷⁷Lu-PSMA-617, for us it seemed ethically mandatory to increase treatment activity until either tolerable grade 1/2 toxicity appeared or patients achieved enduring remission.

Because of limited numbers of patients, it is statistically not reasonable and not in the scope of the present report to draw a final conclusion about which dosing regimen provides the optimal therapeutic range. It was already reported that PSA and objective radiologic responses to PSMA-RLT correlated poorly with individual tumor absorbed doses (29). For individual patients, the respective radiosensitivity of the particular tumor and other clinical factors (30) seem to be relevant for determining response probability. It is possible that even high doses of ¹⁷⁷Lu-PSMA-617 cannot achieve the same response rates as ¹³¹I-MIP1095 because this older ligand was used in the preabiraterone/preenzalutamide era and current patients have more previous therapies than these historical controls. Thus, an efficacy analysis cannot be based on a case series of serially treated patients but requires a group comparison after random assignment. For such a purpose, a prospective phase 2 study would be needed.

Up to the highest treatment activity (9.3 GBq), we did not observe increasing numbers of dose-limiting grade 3/4 toxicities. However, in contrast to a formal clinical trial, salvage therapy must be conservative. Thus, we made the clinical decision to stop dose escalation once we observed an incomplete recovery of blood cell counts to the baseline, even before critical absolute numbers were reached. Because there was no randomization (which would define the study as “medical research” and which is not possible during “unproven interventions in clinical practice”), simply by chance the 9.3-GBq group contained the most patients with previous chemotherapy and the highest alkaline phosphatase and PSA levels—which might be a sufficient explanation for the reduced red marrow reserve. Thus, the dose-limiting effects attributed to the 9.3-GBq activity might have been an accidental observation caused by patient selection, and even higher activities might be possible for well-selected patients. However, our aim was to establish a reasonable standard operating procedure appropriate for patients currently scheduled to receive PSMA-RLT and without the routine need for sophisticated pretherapeutic dosimetry to identify statistical outliers in advance. The good tolerability observed in our patient series is well in line with the dosimetry-based expectation. The estimated average red marrow dose of 0.3 Gy/cycle (9.3 GBq × 0.03 Gy/GBq) is far below the accepted red marrow tolerance limit of 1 Gy and, with a ratio of 0.3:1 Gy, there are enough safety margins for individual patient variability to warrant the application of standard doses.

It should be noted that for patients with diffuse red marrow infiltration, we use ²²⁵Ac-PSMA-617 targeted α-therapy (Fig. 1) whenever this radionuclide is available. These types of challenging patients might have been underrepresented in our patient cohort. Modeling of red marrow absorbed doses normally neglects the contribution of β-particles that are emitted from bone metastases to the surrounding healthy red marrow. For ¹⁷⁷Lu, the 1.5-mm maximum β-range in water corresponds to approximately 30 cell layers; this routinely neglected factor might be relevant in patient with very advanced disease. Thus, individual dose reductions should be considered for patients with diffuse red marrow infiltration.