Dosing ²²⁵Ac-DOTATOC in patients with somatostatin-receptor-positive solid tumors: 5-year follow-up of hematological and renal toxicity

The treatment of metastatic neuroendocrine neoplasms over the past decade has undergone many changes. Conventional platin-/etoposide-based chemotherapies are only active in aggressive G3 but not G1/2 neuroendocrine tumors (NETs). The combination of temozolomide and capecitabine was suggested for NETs of pancreatic origin in 2011 [1]; the first prospective trial was recruited until 2016 (NCT01824875), but a full-text of this study is still not published. Initial phase-3 data evaluating sunitinib and everolimus in well-differentiated pancreatic NETs were published in 2011 [2, 3]; but the extended approval of everolimus for all grade 1/2 NETs of pancreatic, gastrointestinal, and lung origin only occurred in 2016 [4]. Somatostatin analogs were first approved for symptomatic treatment of carcinoid syndrome, but anti-proliferative efficacy for low proliferative (Ki-67 < 10%) enteropancreatic NETs was finally demonstrated in 2014 [5].

In contrast, peptide receptor radiotherapy (PRRT) with radiolabeled somatostatin analogs has a long history of treatment for neuroendocrine neoplasms. Promising initial results of ¹¹¹In-Dota-lanreotid was initially reported in 1998 [6], results of ⁹⁰Y-DOTATOC in 2001 [7], and ¹⁷⁷Lu-Dotatate in 2003 [8]. Nevertheless, pivotal phase-3 evidence for well-differentiated mid-gut NETs and formal approval of ¹⁷⁷Lu-Dotatate only occurred in 2017 [9]; a phase-3 study of ¹⁷⁷Lu-DOTATOC in pancreatic G1/2-NETs is still recruiting (NCT03049189). Also other tumor-entities, such as paraganglioma, Merkel-cell carcinoma, small and large cell neuroendocrine carcinomas, and pheochromocytomas, can express sufficient amounts of somatostatin-receptors. However, these tumors have not yet been systematically evaluated for PRRT. Eventually, physicians considered the anti-tumor activity of β-labeled somatostatin-analogs insufficient for these more aggressive tumors.

Targeted α-radiation therapy (TAT) offers theoretical radiobiological advantages compared to β-radiation [10], but clinical long-term experience is still lacking [11]. In our hospital, ²²⁵Ac-DOTATOC therapies of somatostatin-receptor-positive tumors were conducted on an individual patient basis between July 2011 and March 2015. Pioneering experience with ²¹³Bi-DOTATOC, which demonstrated anti-tumor activity in some patients not responding to β-PRRT [12], provided the rationale to consider ²²⁵Ac-TAT as an appropriate option to escalate systemic PRRT. According to then-available knowledge, we selected patients with well-differentiated gastro-entero-pancreatic (GEP)-NETs (most of our patients were diagnosed before the Ki-67-based WHO classification 2010 was introduced; today these cases would likely be classified as G1/G2 with Ki-67 <10%) for ⁹⁰Y- or ¹⁷⁷Lu-DOTATOC. ²²⁵Ac-DOTATOC was offered as an experimental therapy to patients with additional extra-hepatic involvement, histopathological or clinically more aggressive tumors, more advanced tumor stage, or after insufficient response to previous β-PRRT.

In this work, we evaluate the clinical follow-up data of these patients to retrospectively assess the maximum single cycle and cumulative treatment doses of ²²⁵Ac-DOTATOC.

Dose-dependent thrombocytopenia and leucopenia were the leading acute toxicities of ²²⁵Ac-DOTATOC. Chronic kidney disease (CKD) was the most relevant late effect in this cohort. Other adverse effects included grade 2 hair loss which was reported once. Non-TAT-specific adverse effects included nausea as a typical side effect of the kidney protective solution. Secondary myeloproliferative disease was not observed. The patient population was selected for aggressive histological sub-types and was heavily pretreated (supplement table-1). Known risk factors for CKD were found in 36/39 pts. (supplement table-2). Survival equals the follow-up period provided as the last column of supplement table-3; the median overall survival of the total cohort was 20 months.

In this work, we elaborate on the acute hematological toxicity and chronic renal toxicity of ²²⁵Ac-DOTATOC therapy in patients treated between 2011 and 2015. This time frame enables consideration of 5-year follow-up in the long-term survivors of this cohort.

We found that treatment activities up to 29 MBq were hematologically tolerable (only CTCAE grade 0/1 in 26/27 patients), and severe grade 3/4 toxicities were not observed until single doses of 44 MBq were given. Repeated therapy at 4-month intervals demonstrated no additive toxicity of succeeding cycles if activities were < 25 MBq; in contrast, additive toxicity was observed in a patient treated at 2-month interval. This experience is in line with an empirical dose escalation of ²²⁵Ac-PSMA-617 (a PSMA-targeted agent) that was discontinued due to xerostomia, but without reaching dose-limiting hematological toxicity up to 20 MBq (200kBq/kgBW) [21]. In a recent study, ²²⁵Ac-DOTATATE was given in a 100kBq/kgBW dose to refractory patients or those nearing the maximum cumulative dose of ¹⁷⁷Lu-DOTATATE therapy. The results were in line with our data as no grade 3/4 hematological toxicities were observed. However, the median follow-up of this study was only 8 months, and the rationale of the used dosing regimen was not fully explained [22]. The therapeutic range of a therapy is defined by the ratio of its tolerability to its anti-tumor activity. Thus, determination of a tolerable treatment without concurrent evidence for anti-tumor-activity at that dose level is inconclusive regarding the superiority of α-PRRT over β-PRRT. However, such efficacy analysis is out of the scope of this manuscript, and further efficacy studies will be required.

The mechanism of renal toxicity with either α-PRRT or β-PRRT is incompletely understood and more complex than might be thought at first consideration. Following glomerular filtration, radiolabeled peptides are re-absorbed in the tubular cells. Depending on the tissue penetration range of the emitted particle, β-emitters may reach the glomeruli, which are highly radiosensitive [23]. Some authors have speculated that the intra-renal dose distribution might be responsible for an improved tolerability of ¹⁷⁷Lu-DOTATOC (max 2 mm tissue range) over ⁹⁰Y-DOTATOC (max 12 mm) [17, 24, 25]. If this assumption about tissue range and nephrotoxicity proves true, the short range α particles emissions from ²²⁵Ac (approx. 0.1 mm tissue range) should result in less CKD. However, in recent studies when the biological effective dose was calculated, no inherent benefit of ¹⁷⁷Lu- over ⁹⁰Y-PRRT could be demonstrated [26, 27]. In one preclinical model, glomerulopathy but no tubular degeneration was observed with β-emitting ¹⁷⁷Lu-DOTA-PESIN; in contrast, tubular degeneration was the leading pathology observed for α-emitting ²¹³Bi-DOTA-PESIN [28]. In another preclinical work, α-camera images revealed “hotspots” of ²¹¹At-PSMA only in the proximal convoluted tubules of the kidney, but this was leading to a generally reduced function of the whole nephron. [29]. Thus, the benefit of the “range effect” may be overestimated in importance when considering nephrotoxicity. Renal toxicity may be further complicated by the fate of nuclide-specific daughter isotopes that may cause additional renal damage [16]. For instance, when ²¹³Bi (a daughter isotope of ²²⁵Ac) becomes unbound from the targeting moiety while it is still in circulation, it can accumulate in the kidneys [30]. However, after ²²⁵Ac-TAT internalizes within a cell, the daughter nuclides will be generated intra-cellularly. The latter mechanism is favored with ²²⁵Ac DOTATOC because somatostatin-receptor agonists induce receptor internalization [31] and have rapid blood clearance [32]. Nevertheless, daughter nuclide translocation remains an uncertainty factor regarding accurate dosimetry estimates for the kidney. Indeed, due to the lack of practical α-dosimetry software tools, reliable organ-specific RBE values (relative biological efficacy), and established models for micro-dosimetry and translocation effects, personalized dosimetry estimates have not been reasonable for the presented patients.

As demonstrated in Figure 4, we did not observe a correlation between treatment activity and nephrotoxicity, with or without prior exposure to β therapies. The relatively lower fraction of patients who developed < 5% eGFR-loss per year compared to ¹⁷⁷Lu-Dotatate (Figure 3) can probably be related to the fact that the historical controls were typically radiation-naïve while the majority of our patients had already had β-PRRT and therefore pre-existing a reduced tolerance reserve. Most importantly in both subgroups, we rarely observed more than 20% eGFR-loss per year. At this rate of decline, approximately 5 years must elapse between treatment and the requirement for dialysis, assuming normal kidney function at baseline. Indeed, the kidney failure that was finally observed in two of our patients was a result of moderate but persistent annual eGFR-loss over more than 4 years. Taking into account the limited duration of response to other treatments, α-PRRT probably contributed to a remarkable prolonged survival of these patients. Elsewise they would have likely succumbed to their disease before incurring renal toxicity (Figure 5). Renal scintigraphy was not helpful in identifying such patients in advance. This agrees with observations with β-PRRT [33]. However, renal scintigraphy is still useful to rule out urinary obstruction in advance of radionuclide therapy in selected patients.

Several years ago, a small molecule targeting prostate-specific membrane antigen (PSMA) with tumor uptake and clearance kinetics similar to DOTATOC was developed and labeled with ²²⁵Ac. Recently, a first patient treated with that agent survived 5 years after therapy with 24 MBq ²²⁵Ac-PSMA-617, accompanied by a slow but chronic increase of serum creatinine [34]. Another two patients that were treated with activities of 20–24 MBq ²²⁵Ac-PSMA-617 developed chronic kidney disease, and biopsy revealed both glomerular and tubular pathologies in one of them [35]. All three patients had more than one additional risk factor for occurrence of CKD (age, diabetes, hypertension, etc.). However, the dose difference between the reports of renal toxicity after 24 MBq ²²⁵Ac-PSMA-617 and the moderate renal toxicity we observed after up to cumulative 95 MBq ²²⁵Ac-DOTATOC is remarkable. One explanation could be that despite comparable dose at the whole-organ level, ligand-specific micro-dosimetry might be a significant issue when α-emitters are used. In addition to its nonspecific excretory function, the kidneys also have physiological expression of somatostatin-receptors and PSMA [29, 36]. While PSMA is expressed in neoangiogenesis, renal somatostatin-receptors may play a role in a pro-inflammatory cascade since overexpression has been found in IgE-nephropathy [37]. Highly focused α-radiation to somatostatin-receptor-positive cells may induce a different renal remodeling response than targeting PSMA-positive cells.

There are various reasons for the absence of a dose response relationship for radiotoxicity of the kidney. Even in healthy individuals, there is a physiological decline of kidney function of 10ml/min per decade. In addition, risk factors for CKD are common, and the incidence of kidney failure normally affects patients who were never treated with radioactive drugs. In well-differentiated NETs, additional tumor-related mechanisms are frequent, e.g., hepatorenal syndrome caused by liver metastases, pre-renal azotemia (e.g., diarrhea as part of the carcinoid syndrome), or post-renal (e.g. right heart failure associated with carcinoid heart syndrome); in aggressive tumors, often potentially nephrotoxic chemotherapies are used.

Some centers strictly following the recommended kidney tolerance thresholds and not exceeding 4 × 7.4 GBq ¹⁷⁷Lu-Dotatate reported either no grade 3 or 4 subacute nephrotoxicity in 323 patients [38] or only 1.5% grade 3 or 4 nephrotoxicity in 807 patients [39]. In line with our observations, in these studies, the kidney radiation absorbed dose did not correlate with renal function loss over time. These results suggest higher doses with better efficacy may be possible. Thus, in our center, the general concept is to administer relatively conservative doses to early stage patients and accept higher risks in later-stage or more aggressive tumors. However, if ²²⁵Ac-DOTATOC would be considered for first-line therapy of well-differentiated NETs with a prognosis > 5 years, a 10–15% decline of GFR per year would be considered unacceptable, and lower treatment activities should be administered. We suggest that a 4 × 10 MBq schema would be more appropriate for such patients.

This study has several inherent limitations. The study was retrospective, and treatment doses were defined on a case by case basis—explicitly considering recent lab tests and the patient’s individual prognosis, influenced by total tumor volume, site of metastases, and growth velocity. This naturally leads to selection biases. Patients who were considered to be in a more vulnerable general condition received lower treatment activities than patients who presented in better physical condition. The so-called tumor sink effect, i.e., the observation that normal organ uptake is inversely correlated to the total tumor volume [40, 41], is probably also a relevant confounder. In our cohort, the highest activities were selectively prescribed to patients with high tumor burden. A relevant fraction of the injected radiopharmaceutical was likely trapped in well perfused high-volume tumor lesions, thus accelerating blood-pool clearance and lowering red-marrow dose and the fraction of drug that has to be cleared by the kidneys. These two factors tend to overestimate the risk of lower treatment doses and the tolerability of higher treatment doses. The heterogeneous nature of the patient population affects standardized evaluation. As an example, inclusion of several different tumor entities increases the diversity of previous therapies, which each having diverging nephrotoxic potential making conclusions even more difficult. However, for ethical reasons, potentially irreversible, sometimes lethal and often late occurring side effects such as chronic kidney disease cannot systematically be studied in homogeneous cohorts of early stage patients. Thus, in our opinion, any new information that might positively contribute to patient safety in the future with regard to the design of clinical trials is worth sharing with the community.

Although α-emitters offer potential advantages as radiopharmaceutical therapies and have been used in small studies, the availability of long-term toxicity data is lacking. We found treatment activities of approx. 20 MBq per cycle (4 monthly repetition) and cumulative doses up to 60–80 MBq ²²⁵Ac-DOTATOC reasonable for patients with advanced stage malignancies. This kind of data may be useful when planning prospective clinical trials using this agent or other related therapies based on α-emitters.