A first-in-human study investigating biodistribution, safety and recommended dose of a new radiolabeled MAb targeting FZD10 in metastatic synovial sarcoma patients

Synovial Sarcomas (SS) account for 2.5–5% of all soft tissue sarcoma and affect all ages, but is most prevalent among adolescents and young adults (15–40 years of age) [1]. Therapy for patients with localized disease is based on surgery followed by external radiotherapy. Five-year overall survival rates range from 36 to 76%. Recurrences may be local (30–50%) or distant (40%), with lung being the most common site of distant metastases. Survival is better in children, possibly because of major biological differences between SS of adults and children [2]. In the setting of advanced disease, systemic chemotherapy with doxorubicin and/or ifosfamide is the standard of care. However, median overall survival becomes close to 12–18 months [3, 4]. Most targeted agents have failed to show significant activity except the VEGFR inhibitors [5], pazopanib [6], and more recently regorafenib [7]. Thus, therapeutic options in advanced SS remain an unmet clinical need.

In a study by Nagayama et al., 26 genes were found to be commonly upregulated in SS based on the genome-wide gene expression profiles of 13 SS cases by cDNA microarray, including Frizzled homologue 10 (FZD10) [8]. FZD10 belongs to the Frizzled family of seven-pass transmembrane receptors for molecules in the Wnt signaling pathway. It was found to be overexpressed in SS samples and was almost absent in remaining normal adult tissues except the placenta [9]. Based on these findings, a specific monoclonal-antibody that targets the N-terminal extracellular domain of FZD10 (MAb 92–13) was developed as a first step toward the development of an antibody-based therapy for SS [10]. In vitro, MAb 92–13 has only a weak antagonistic activity on cell growth and no/little antibody­dependent cell­mediated cytotoxicity and complement-dependent cytotoxicity. In vivo, when radiolabelled with Indium-111 (¹¹¹In), it bound and accumulated in up to 5 days after IV injection in SS tumor cells overexpressing FZD10 (SYO-1) implanted in nude mice and not in FZD10 negative SS tumor cells (LoVo). This provided evidence for a specific binding. MAb 92–13 was proved to get internalized into tumor cells by confocal microscopy and flow cytometric (FACS) analyses. When it was radiolabeled with Yttrium-90 (⁹⁰Y), a highly energetic beta emitter radioisotope, tumor shrinkage was observed in immunocompromised Balb-c mice bearing established FZD10-positive SS tumor subcutaneous xenografts (SYO-1 cell line), without significant toxicity [10]. Indeed, tumor volumes were markedly reduced immediately after treatment after a single administration of 3,7 MBq. Median time to tumor progression was 58 days in treated mice and 9 days in the control group.

In the present report, we describe the results of a FIH, first in class phase I study evaluating the use of ¹¹¹In-OTSA-101 and ⁹⁰Y-OTSA-101 in patients with advanced synovial sarcoma following a theranostic approach.

We reported results of a FIH study with a radio-immuno-conjugate targeting FZD10 in patients with advanced synovial sarcoma. The concept of vectorising cytotoxic and radionuclide has emerged 20 years ago when the first monoclonal antibodies as anti-cancer agents where approved, with some successes like radiolabelled anti-CD20 (Zevalin®) applied to non-Hodgkin-lymphoma as reviewed by Rizzieri [12] or more recently radiolabelled anti-PSMA applied to metastatic prostate cancer [13]. The overall aim of these anti-body drug conjugates (ADC) and radio-immunotherapy (RIT) is to use the specificity of monoclonal antibodies for their target to deliver a highly toxic payload (cytotoxic drug or radionuclide) to tumor cells, thus avoiding systemic toxicity of this payload. Key aspects in developing ADC/RIT are: 1) the target of the monoclonal antibody and its specificity, 2) payload and its individual cytotoxic potency, 3) the linker technology which will vary depending on the type of payload.

The linker technology was selected for maximum stability of the loaded antibody. In the present study we chose to have an assessment of target expression using FZD10-imaging with an ¹¹¹In-labeled version of OTSA-101. Our study aimed to evaluate the in vivo ¹¹¹In-OTSA-101 biodistribution and tumors biding rather than evaluating the maximal tolerated dose. As this was our main aim, we did not organize our study as a real 3 + 3 dose escalating study. Once biodistribution and tumors uptake would have been considered compatible with treatment, patients would be randomized to receive different doses of ⁹⁰Y-OTSA-101, to give them more chance to receive an efficient dose. In standard dose escalation studies, as most responses occur between 80 and 120% of the Maximum Tolerated Dose (MTD), the first cohorts of patients are often treated with low sub-therapeutic dose and only few patients actually receive doses at or near the recommended therapeutic dose. Patients would rather beneficiate from a randomized study, giving more chance to patients to receive the effective dose treatment.

In the imaging part, we observed significant intra-patient and inter-patient differences in uptake intensity and uptake rates. Indeed, some patients had very low or no significant uptake of ¹¹¹InOTSA-101 while in some patients we observed rapid uptake of target lesions. Although this heterogeneity was anticipated based on the preclinical data used as a basis for this study (8 of 13 SS patients had overexpression of FZD10, 4/13 had detectable expression and one patient had no detectable FZD10 in the initial study by Nagayama et al. [3]), its importance was underestimated. The precise molecular or pharmacological mechanisms for these differences have not been explored in the current study. Another hurdle encountered during the preparation and conduct of this study, was the lack of validated immunohistochemistry (IHC) assay to assess FZD10 expression on tumor cells. This lead us to propose the imaging part of the study as a biomarker, with the aim of reducing the number of patients exposed to Yttrium 90 to those most likely to benefit due to high in vivo expression. Due to the lack of available data at the time the study was initiated, we chose an arbitrary threshold of uptake based on a visual assessment of tumor uptake compared to background following Krenning’s grades applied for peptide receptors radionuclide therapy (PRRT) of endocrine tumors [14]. This visual scale allows for patients selection for PRRT as higher objective responses leading to longer survival and improved quality of life have been observed in case of higher grade tumors uptake [15]. Molecular imaging allows for a non-invasive whole body in vivo characterisation of the heterogeneity of tumors antigen or receptors expression between primary and metastases and as well as between metastases, supplementing the traditional role of using imaging for localizing and measuring disease. For example, it is currently used to conduct treatment of endocrine tumors as somatostatin receptors scintigraphy and ¹⁸F-FDG PET/CT are able to quantify sites of well and poorly differentiated disease, respectively, and therefore to treat more aggressively ¹⁸F-FDG PET/CT positive endocrine tumors [16]. Indeed, these whole body images are less influenced by biopsy bias. The in vivo expression of FZD10 was not assessed on tumor biopsy as this procedure is exposed to sampling error and may not reflect the entire tumor cells heterogeneity. This is bypassed by whole body scintigraphies that allow tracking the radiotracer in the entire body and in every tumor. This is how we observed the difference of radiotracer uptake from one lesion to another and how we can proceed to dosimetry studies. Changes in radiotracers compounds can influence and change the biodistribution. This is the way we can improve the way they target tumors rather than normal tissues and thereafter, based on dosimetry studies, select the best molecule. This is the basis of theranostic in nuclear medicine [17, 18].

Kinetic of the lesions uptake reached maximum intensity as soon as 1 h post-injection in some patients while it would become obvious only on the latest acquisitions in others, consistent with specific tumor suptake. The tumors radiotracer retention observed confirmed radiotracer internalization. Some bulky tumors demonstrated heterogeneous tracer uptake as a result of the various tumors vascularizations, tumor necrosis depending on tumor size, the capacity of the antibody to diffuse into the tumors, the expression of FZD10, the antibody internalization. Thus, bystander effect or even the abscopal effect of radiopharmaceutical may occur only on a portion of the viable cells that would not show tracer uptake.

Tracer kinetic depends on tumors vascularization but also on tracer size. Antibodies are large molecules known to have poor tissues penetration and long circulating half-life. For this reason, more recent radionuclide therapy trials favour smaller radiolabelled vectors than antibody, such as fragments of antibodies (FAb), or mostly peptides, displaying more rapid tissue distribution and faster blood clearance, leading to almost no toxicity [19, 20]. Therefore, the development of FAb targeting FZD10 may improve radiotracer kinetic and lesions uptake and reduce toxicity. It could be also possible to radiolabel it with a long physical half-life positron emitter such as Zirconium-89 (78.4 h) rather than Indium-111 to increase radiotracer detection in tumors by using Positron Emission Tomography (PET) camera, known for a better resolution than SPECT.

Overall loaded ⁹⁰Y-OTSA-101 was well tolerated at all dose levels and no formal MTD was determined. No significant drug-related AE was observed with ¹¹¹In-OTSA-101 but the injected doses were low in part 1 of the study to limit receptor occupancy by non-therapeutic antibody. In the second part of the study, the most common AEs were cytopenia and fatigue, with thrombocytopenia being the more problematic. Bone marrow suppression, which results in cytopenia, is most likely due to the exposure of the bone marrow to ⁹⁰Y either bound to OTSA-101 during blood perfusion of bone marrow, or free ⁹⁰Y which tends to accumulate in the bone marrow. Other isotopes such as Lutetium 177 (¹⁷⁷Lu) for example are associated with reduced bone marrow suppression compared to ⁹⁰Y [21, 22], due to a shorter mean pathway of less energetic electrons emitted (12 mm mean pathway in the water and Maximum 2270 Kev for 90Y vs 1.5 mm and 497 Kev for ¹⁷⁷Lu).

We performed 2D dosimetry on ¹¹¹In-OTSA-101 whole body acquisitions to estimate the absorbed dose of normal organs (liver, spleen, kidneys) when the patient would be treated with either 370 or 1110 MBq of ⁹⁰Y-OTSA-101. As usually observed in RIT, liver was the most targeted normal organ by the radiolabeled antibody. However, the mean absorbed liver dose was calculated after Indium 111 in order to prevent reaching the 20 Gy Maximal Tolerated Dose (MTD) limit applied to external beam radiotherapy. When looking at liver toxicity, only one grade 1 bilirubin increase was reported in one patient treated by the lowest activity, contrasting with systematic bone marrow toxicity despite low visual bone marrow uptake and no diffuse medullary metastases in any patient. We recently reported the results of a 3D dosimetry based on Montecarlo simulations performed on these patients [23]. The calculated absorbed doses in the liver would range from 4.3 to 13 Gy, consistent with the 2D dosimetry results and the absence of liver toxicity. On the other hand, the absorbed dose in the bone marrow would be greater than 1.8 Gy for 5 patients, which is close to the MTD estimated around 2 Gy, partially explaining the observed bone marrow toxicity.

Efficacy of ⁹⁰Y-OTSA-101 on tumors was mostly transient stabilized disease until disease progression. Most of the patients included in this phase I trial presented with bulky lung metastases. For this reason, a highly energetic electrons emitter such as ⁹⁰Y would be more suitable to treat these lesions. However, these lesions were rapidly growing while radionuclide therapies demonstrate delayed response to treatment. This is well known for slowly progressive disease such as neuroendocrine tumors or thyroid cancers, with non-bulky disease. Moreover, the highly heterogeneous and mostly low ¹¹¹In-OTSA-101 uptake observed in these tumors would lead to low absorbed doses in the tumors partially explaining the lack of treatment response especially in bulky tumors.

Altogether, when using ¹⁷⁷Lu radiolabelled small molecules radionuclide therapy tend to be less toxic allowing increased injected doses in order to increase absorbed tumors doses. As a recent example, PSMA was first targeted with antibodies (J591) for diagnosis and therapeutic purpose of prostate adenocarcinoma. The maximum tolerated dose (MTD) due to marrow toxicity was 4 times lower when using ⁹⁰Y-J591 than with ¹⁷⁷Lu-J591 [22, 24]. For the past few years, high radiolabeled PSMA avid small molecules have been developed and used with great success for diagnosis and treatment. Different types of ¹⁷⁷Lu-PSMA appeared to be highly tolerated with administered activities being twice the MTD of ¹⁷⁷Lu-J591 with almost no grade 3–4 hematotoxicity in these populations of heavily pre-treated patients with bone metastases [19, 25, 26]. Moreover, dosimetry studies evaluate the absorbed tumors dose sometimes exceeding 100 Gy due to intense homogenous target lesions tracer uptake, which is far above the 10 Gy maximal absorbed dose in tumors we estimated on 3D [23].

We confirmed the in vivo capacity of a new radiolabeled antigen targeting FZD10 to get trapped in at least one metastasis in 20 patients presenting with metastatic synovial sarcoma. However, the ¹¹¹In-OTSA-101 tumors uptake appeared highly heterogeneous and was considered intense enough to select patients for ⁹⁰Y-OTSA-101 treatment in only half of the 20 patients included. A theranostic approach appeared feasible in a first in human phase I study in a very rare cancer using a national networking. Transient stable disease was observed in 5 of the 8 treated patients. However, as most of the patients presented significant hematotoxicity related to treatment, the use of small fragments of antibodies targeting FZD10 radiolabeled with lower energetic electron emitter such as Lutetium 177 would be a favored option for further development.