B7-H3-targeted Radioimmunotherapy of Human Cancer

Abstract

Background:

Targeted radioimmunotherapy (RIT) is an attractive approach to selectively localize therapeutic radionuclides to malignant cells within primary and metastatic tumors while sparing normal tissues from the effects of radiation. Many human malignancies express B7-H3 on the tumor cell surface, while expression on the majority of normal tissues is limited, presenting B7-H3 as a candidate target for RIT. This review provides an overview of the general principles of targeted RIT and discusses publications that have used radiolabeled B7-H3-targeted antibodies for RIT of cancer in preclinical or clinical studies.

Methods:

Databases including PubMed, Scopus, and Google Scholar were searched for publications through June 2018 using a combination of terms including “B7-H3”, “radioimmunotherapy”, “targeted”, “radiotherapy”, and “cancer.” After screening search results for relevancy, ten publications were included for discussion.

Results:

B7-H3-targeted RIT studies to date range from antibody development and assessment of novel radioimmunoconjugates (RICs) in animal models of human cancer to phase II/III trials in humans. The majority of clinical studies have used B7-H3-targeted RICs for intra-compartment RIT of central nervous system malignancies. The results of these studies have indicated high tolerability and favorable efficacy outcomes, supporting further assessment of B7-H3-targeted RIT in larger trials. Preclinical B7-H3-targeted RIT studies have also shown encouraging therapeutic outcomes in a variety of solid malignancies.

Conclusion:

B7-H3-targeted RIT studies over the last 15 years have demonstrated feasibility for clinical development and support future assessment in a broader array of human malignancies. Future directions worthy of exploration include strategies that combine B7-H3-targeted RIT with chemotherapy or immunotherapy.

1. INTRODUCTION

1.1. Introduction to Targeted Radioimmunotherapy (RIT) of Cancer

Targeted RIT is an attractive approach to selectively localize therapeutic radionuclides to malignant cells while sparing the majority of normal tissues from the effects of radiation [1]. This approach uses a particle-emitting radionuclide that is linked to a carrier molecule, typically an antibody or an antibody fragment, which binds to an antigen or a receptor that is overexpressed on the cancer cells relative to normal cells. After administration into a subject and distribution through accessible tissues, the radiolabeled conjugate will be selectively retained in tissues that express the respective antigen or receptor and will be cleared from other tissues that lack target expression. This approach is desirable for treating patients who are not candidates for localized brachytherapy due to disseminated disease or who have malignancies that are resistant to gamma radiation. Although there are many physical, chemical, biological, and pharmacological variables that influence therapeutic outcomes of radiotherapy strategies [1–8], three main components that must be considered when designing conjugates for targeted radiotherapy will be discussed here: a) selection of the target, b) selection of the targeting agent and c) selection of the radionuclide. The following general introduction will discuss these considerations in relation to B7-H3-targeted radiotherapy of human cancer.

1.1.1. Desirable Features of Antigens or Receptors for Targeted RIT

The main factor required for effective targeted RIT is selecting a target that is overexpressed on malignant cells relative to normal cells. Many studies have shown that B7-H3 protein is highly expressed on the surface of cancer or stromal cells in the tumor microenvironment in a wide variety of solid tumors [9–17] but has a low expression or is not detectable in most normal tissues [13, 15, 18]. Although the percentage of B7-H3-positive malignant cells can vary among individuals that have a particular type of cancer, B7-H3 expression increases with the stage of malignant progression. Another variable to consider is the stability of antigen or receptor expression during the course of RIT. Mutations or isoform variability can alter the conformation of an antigen or receptor during the course of RIT, such that cancer cells become resistant to a particular therapeutic agent (e.g., patients with EGFR-positive [epidermal growth factor receptor] cancer developing resistance to cetuximab [19]). Heterogeneous expression of an antigen or receptor in an individual patient is also a potential concern [20–21], as targeted agents can select against the antigen- or receptor-positive cells without affecting the antigen- or receptor-negative cells, resulting in the resurgence of disease through the proliferation of target-negative tumor cell populations [22]. As targeting B7-H3 is still in relatively early stages of development, it is not yet known how expression changes with targeted RIT. Preclinical studies indicate that tumor xenografts maintain B7-H3 expression during targeted therapy with antibody-drug conjugates [23]. Future studies have to determine if cancer cells maintain B7-H3 expression during treatment with radioimmunoconjugates (RICs) in preclinical and clinical settings. Internalization after binding of the targeting ligand to the antigen or receptor is another factor to consider for targeted RIT. Relative to targeting non-internalizing antigens or receptors on cancer cells, RICs that internalize have been shown to result in greater therapeutic efficacy in preclinical models of human cancer [24]. Preclinical studies using human breast, pancreatic, or ovarian cancer cell lines have indicated moderate levels of B7-H3-targeted RICs are internalized after binding to these cells in vitro [25–26].

1.1.2. Monoclonal Antibodies (mAbs) as Targeting Agents for RIT

Selection of appropriate targeting agents is also essential for designing effective targeted RIT approaches. RIT uses a mAb, which binds with high affinity and specificity to a defined antigen or receptor present on malignant cells, as a carrier molecule to deliver a therapeutic radionuclide to target cells. As discussed in greater detail below, several mAbs that bind to extracellular B7-H3 epitopes have been developed and used for targeted RIT in preclinical and clinical studies. mAb fragments or smaller constructs (e.g., nanobodies) derived from intact mAbs are also used in targeted RIT but have not yet been used in B7-H3-targeted RIT. mAbs are advantageous as delivery agents for targeted RIT due to their high serum stability, slow clearance from target tumors, and amenability for direct chemical conjugation with halogen radionuclides or with chelates that stably coordinate metal radionuclides [2, 5, 27]. As opposed to small molecules, where structural modifications can significantly alter the biological properties of the resulting conjugate compared to the original targeting agent, antibodies can be conjugated under controlled conditions while maintaining their original affinity, stability, and pharmacokinetics [28–29]. However, it is well known that chemical modifications within the antigen binding regions or addition of too many chelates to the other regions of a mAb for radiometal labeling can negatively influence antibody immunoreactivity, cause aggregation, or induce recognition of the RICs by cells in the reticuloendothelial system (RES) in vivo [30–32]. All these conditions would impair the specific accumulation of a RIC in a tumor.

The immunogenicity of the antibody can also influence the tolerability and success of targeted RIT regimens, particularly during repeated or fractionated therapy. The majority of the available mAbs recognizing human tumor cells are first developed from murine hosts. The B7-H3-targeted RIT studies to date have all used murine mAbs as carriers for localizing the radionuclide to malignant cells. Humoral responses against non-human or humanized chimeric antibodies can preclude multiple administrations, and have been known to induce toxicity during initial infusion in some patients [33]. Therefore, fully human antibodies are often preferred for use in human patients, although they can induce an anti-idiotypic immune response. Several humanized or fully human mAbs that bind to B7-H3 have been developed [13, 23, 34], although none has yet been employed for targeted RIT applications.

Using antibodies as carriers for targeted RIT is particularly attractive for intra-cavity therapy, such as treating peritoneal malignancies. In this approach, malignant cells within the cavity are directly accessible to the RIC. In contrast, exposure to normal tissues outside of the cavity is limited until the RIC slowly clears from the cavity through lymphatic or other processes [7–8]. These variables favorably contribute to localizing the therapeutic dose to cancer cells. Despite these favorable properties, the slow extravasation rates into solid tumors and prolonged circulation times of mAbs present challenges to eradicate all malignant cells at tolerable dose levels of regionally or systemically administered RICs, while also contributing to toxicity from absorbed doses to normal tissues [5–6, 35–40]. These variables influence the selection of radionuclides to generate RICs for the target of interest.

1.1.3. Selecting a Radionuclide for Targeted RIT

A variety of particle-emitting radionuclides have been used in targeted RIT strategies against cancer (Table 1). The half-life of the radionuclide, the type of particle emitted, and the energy of particle emission are important variables to consider when designing RICs for targeted RIT. Additionally, gamma rays emitted during radionuclide decay also contribute to selecting a particular radionuclide for targeted RIT. The majority of radionuclides used to treat cancer in clinical settings have half-lives in the range of several days, although radionuclides with half-lives shorter than 1 h have also been used in human trials [41–42]. With longer-lived radionuclides, a single administration can deliver a cumulative dose of several Gy at moderate dose rates over the decay period, which promotes tolerability [1–2, 43–45]. In many applications, targeted RIT using radionuclides that have half-lives spanning several days is desirable due to the pharmacokinetic half-life of the antibody to which the radionuclide is bound. Using a radionuclide with a physical half-life that is significantly shorter than the blood clearance half-life of the RIC will result in delivery of the majority of the administered dose outside of the intended tumor target. Alternatively, using a short-lived radionuclide for RIT is desirable in cases where the target antigen or receptor is highly expressed and immediately accessible, such as in blood-borne malignancies (e.g., leukemia) where the malignant cells essentially “scavenge” the administered RIC rapidly after systemic administration [41, 46–48].

Table 1.

Commonly used particle-emitting radionuclides for targeted RIT of cancer.

Radionuclide	Half-life	Major particle(s) emitted	Maximum energy of emitted particle(s) (keV)
64Cu	12.7 h	β− and β+	653 (β+), 579 (β−)
90Y	2.7 d	β−	2280
124I	4.2 d	β+	2138
125I	59.4 d	Auger electron	35
131I	8.0 d	β−	807
177Lu	6.6 d	β−	498
186Re	3.7 d	β−	1070
188Re	17.0 h	β−	2120
211At	7.2 h	α	5870 (7450 α from daughter 211Po)
212Pb	10.6 h	β− (α and β− in daughter radionuclides)	570 (8785 α from 212Po in decay chain)
213Bi	45.6 min	α	5875 (8376 α from daughter 213Po)
225Ac	10.0 d	α (α and β− in daughter radionuclides)	5830 (8376 α from 213Po in decay chain)

Radionuclides that emit α-particles, β⁻-particles, β⁺-particles (positrons), or Auger electrons have been used for preclinical or clinical B7-H3-targeted RIT studies, as discussed further below. α-particles are emitted with defined energy (0.5–10 MeV depending on radionuclide), travel a short range in tissues (50–80 μm) and are considered high linear energy transfer (LET) radiation (80–100 keV/μm). α-particles are advantageous in targeted RIT of cancer due to their high relative biological effectiveness and ability to halt the proliferation of malignant cells that are relatively resistant to X-ray radiotherapy or chemotherapy [4, 8, 44, 49–50]. While these features present α-particle RICs as attractive candidates to kill recalcitrant cancer cells, the therapeutic benefits must be carefully weighed against potential toxicity from α-particles to normal tissues while the RICs are in the circulation or are metabolized and excreted [40, 51–52]. Delivering sufficient therapeutic doses of α-particle RICs to eliminate malignancies, particularly solid tumors, prior to de-bulking surgery or chemotherapy has been challenging in clinical studies [47, 53]. Because of these concerns, coupled with the limited range of α-particles, targeted α-particle RIT is often considered as best suited to treat blood-borne malignancies (e.g., leukemia) or metastatic disease that is confined within a body cavity (e.g., peritoneal malignancies) [4, 49, 54].

As opposed to α-particles, β⁻-particles and positrons are emitted over a spectrum of energies (0.1–5 MeV), travel a moderate range in tissues (0.1–5 mm), and are considered low LET radiation (0.1–10 keV/μm). The main advantage of these particles for RIT of cancer is their ability to cause ionization over several hundred cell diameters [55–56]. This factor has been exploited for treating solid tumors, which often have dense tumor parenchyma that limits homogeneous penetration of RICs throughout the tumor volume [2, 37, 57]. The extended “cross-fire” range of β⁻-particles and positrons can offset this limitation of RIC penetration and allow irradiation of the entire tumor during targeted RIT. The efficacy of this approach is limited by the low LET of β⁻-particles and positrons, which requires high doses to halt malignant cell proliferation. Achieving high enough doses to control tumor growth, particularly in radiation-resistant or hypoxic solid tumors, can damage adjacent normal tissues due to the cross-fire effect [7, 58–59]. The limited ability of low LET particle radiation to control cancer cell growth at tolerable doses is a particular challenge associated with targeted RIT with β⁻-particles and positrons.

Auger electrons are low-energy electrons (<0.2 MeV) that have the shortest range (<1 μm) of all particles used for therapy and are considered high LET radiation (5–50 keV/μm). These particles are most effective at inducing cell death or proliferation arrest when the radionuclides are present in the nucleus; they are much less effective when localized to the cell membrane or cytoplasm due to the limited penetration range of the Auger electrons [60–62]. While several Auger electron-emitting radionuclides are used to treat patients via brachytherapy in the clinical setting [63–64], none have been approved for targeted RIT of human malignancies. No Auger electron RICs that target B7-H3 have yet been developed for therapy in patients with cancer.

1.2. A Brief Overview of the Current Landscape in Targeted RIT of Cancer

To date, ¹³¹I-tositumomab and ⁹⁰Y-ibritumomab tiuxetan are the only two RICs that have received approval from the U.S. Food and Drug Administration (FDA) for use in human patients [65–66], although production of ¹³¹I-tositumomab was halted by the manufacturer (GlaxoSmithKline) in 2014. Both of these RICs are generated from murine mAbs that bind to CD20 and received approval for B cell malignancies (non-Hodgkin’s lymphoma). As reviewed elsewhere [1, 4, 48–49, 67–72], many other RICs have been developed and evaluated in preclinical and clinical trials against various types of cancer, although none have yet received regulatory approval. Many RICs are typically given as a single therapeutic administration during initial studies to evaluate tolerability in patients with advanced stages of cancer [7, 35, 39, 48, 73]. Most clinical trials have indicated that systemically administered RIT as a sole form of treatment against cancer does not eradicate disease, particularly in patients with large tumor burden. Similar to external radiation therapy, many RICs in the research setting are evaluated in an adjuvant or combinatorial therapeutic approach. Adjuvant RIT after debulking tumor burden through surgery, chemotherapy, or external radiation therapy has resulted in improved survival outcomes as opposed to treatment without RIT in several studies [74–75]. Treating early metastatic disease with RIT has also produced favorable outcomes in preclinical studies [76–79]. The cumulative evidence from RICs that bind to antigens or receptors overexpressed on solid cancer cells indicates the opportunity and continued need for the development of effective RIT agents that can be utilized across a wide variety of human malignancies. The existing literature on B7-H3-targeted RIT reviewed below supports continued exploration of this therapeutic strategy to improve the outcomes of patients with cancer.

2. REVIEW OF THE FUNCTION AND EXPRESSION OF B7-H3 IN HUMAN MALIGNANCIES

B7-H3 (CD276) is a type I transmembrane protein in the B7 protein superfamily, with nearly 30% sequence homology to other proteins in this superfamily. The extracellular isoform expressed in humans (4IgB7-H3) is comprised of identical pairs of immunoglobulin variable domain and constant domain segments [80]. The intracellular tail has no known signaling functions. The binding receptor for B7-H3 has not been identified. B7-H3 mRNA is widely expressed in normal human tissues including liver, heart, intestines, pancreas, spleen, and reproductive organs, although expression at the protein level in normal tissues is much more restricted [18]. B7-H3 protein is expressed on endothelial cells and activated cells of the immune system including antigen presenting cells, dendritic cells, and other monocyte lineages. The roles of B7-H3 in innate and adaptive immune signaling have not yet been clearly defined. Several studies have shown that B7-H3 has co-stimulatory effects on immune cell function, while other studies have indicated co-inhibitory functions [17, 81–82].

While ongoing studies are attempting to elucidate the molecular roles that B7-H3 plays in normal and pathologic immune regulation, consensus from a growing body of literature has shown that B7-H3 is overexpressed in malignant tissues relative to normal tissues [9–17], thus presenting this protein as an attractive candidate for targeted therapy applications. As summarized in previous reviews [17, 83], malignancies including colorectal cancer, breast cancer, renal cell carcinoma, lung cancer, pancreatic cancer, cervical cancer, esophageal cancer, ovarian cancer, prostate cancer, neuroblastoma, hepatocellular carcinoma, squamous cell carcinoma, melanoma, glioma, acute leukemia, and multiple myeloma aberrantly express B7-H3 in 17–93% of tested clinical specimens. Tumor cells, stromal cells, and endothelial cells in these specimens have shown membrane and cytoplasmic expression of B7-H3. For the majority of malignancies examined to date, increased B7-H3 expression in tissues and soluble B7-H3 present in the circulation correlates with worse prognostic outcomes [9–10, 84]. These results indicate that the development of B7-H3-targeted therapeutic strategies, including targeted RIT, could be applicable across an array of human malignancies [14–15, 17, 85–86].

3. REVIEW OF PUBLISHED B7-H3-TARGETED RIT STUDIES

Several mAbs that bind to human B7-H3 have been developed and used for targeted RIT of human cancer in preclinical or clinical research studies. The articles discussed below were initially identified by searching scientific databases (PubMed, Scopus, Google Scholar) using various combinations of the terms “B7-H3”, “radioimmunotherapy”, “targeted”, “radiotherapy”, and “cancer.” All publications through June 2018 were considered and filtered to exclude studies that a) did not use a radiolabeled antibody that binds to B7-H3, and b) did not perform a therapeutic efficacy study against cancer (whether in vitro, in vivo, or clinical trial). Ten publications met the selection criteria and were included for discussion (Table 2). The sections below summarize the biodistribution, dosimetry, toxicology, and therapeutic results (when available) from each independent study. Relevant unpublished results from the authors’ own studies are included in the sections below.

Table 2.

List of publications meeting the selection criteria for discussion of B7-H3-targeted RIT.

Type of study	Tumor type in study	RIC used	Major result	Refs.
Preclinical	Human rhabdomyosarcoma	131I-8H9, 125I-8H9	Targeted RIT with 131I-8H9 significantly inhibited growth of s.c. human rhabdomyosarcoma xenografts in mice relative to controls (no treatment or isotype control RIC) with no adverse toxicological effects	[87]
Preclinical	N/A	131I-8H9, 124I-8H9	CED of RICs in brainstem of healthy rats and non-human primates showed no adverse toxicology	[88]
Clinical phase I (18 patients)	Metastatic CNS neuroblastoma	131I-8H9, 124I-8H9	Intrathecal RIT with 124I-8H9 (up to 2220 MBq) in combination with alternative therapy was tolerable and resulted in durable complete responses	[12]
Clinical (retrospective, 96 patients)	Metastatic CNS neuroblastoma or medulloblastoma	131I-8H9	Intrathecal RIT after external radiation therapy resulted in low rate (1–2%) of radiation necrosis	[89]
Clinical (retrospective, 3 patients)	CNS relapsed rhabdomyosarcoma	131I-8H9	Intrathecal RIT after resection and external radiation therapy resulted in longer survival compared to no RIT	[90]
Clinical phase I (28 patients)	DIPG	124I-8H9	No grade 4 adverse effects after CED of RIT (up to 148 MBq) with high retention of activity in the brainstem	[91–92]
Preclinical	Human ovarian cancer	212Pb-376.96	RIT with 212Pb-376.96 significantly prolonged survival of mice with i.p. tumor xenografts relative to controls (no treatment or isotype control RIC)	[25]
Preclinical	Human PDAC	212Pb-376.96	Targeted RIT with 212Pb-376.96 significantly inhibited growth of s.c. PDX PDAC tumors in mice relative to controls (no treatment or isotype control RIC)	[26]
Preclinical	Human RCC	131I-4H7	Targeted RIT with 131I-4H7 significantly inhibited growth of s.c. RCC xenografts in mice relative to controls (no treatment or isotype control RIC)	[93]

3.1. Studies Using β-particle RICs of mAb 8H9

The murine mAb 8H9 has been the most frequently used carrier in B7-H3-targeted RIT studies. This IgG₁ mAb was generated by hyperimmunizing Balb/c mice with human neuroblastoma cells [11]. mAb 8H9 is currently the only B7-H3-targeted mAb that has been used for RIT of malignancies in both preclinical and clinical research studies, which will be discussed separately in the sections below. High-affinity chimeric and humanized versions of mAb 8H9 have been generated and tested in preclinical biodistribution studies [34], although no studies have yet been published regarding their use in targeted RIT of human cancer.

3.1.1. Preclinical RIT Studies

An early report used B7-H3-targeted RIT with mAb 8H9 in a preclinical model of human rhabdomyosarcoma [87]. Rhabdomyosarcoma is the most frequent form of sarcoma in pediatric patients and is associated with a high rate of metastasis. Although it is usually sensitive to chemotherapy and external radiotherapy, treatments are rarely curative in high-risk patients. Immunohistochemical staining of human rhabdomyosarcoma with mAb 8H9 indicated epitope expression in 94% of tested clinical specimens (31 of 33), suggesting that B7-H3-targeted RIT may be a viable approach to treat this malignancy. In vitro studies with ¹²⁵I-8H9 showed that human HTB82 rhabdomyosarcoma cells expressed approximately 10⁵ binding sites per cell and minimal internalization of the RIC. Biodistribution studies with ¹²⁵I-8H9 in mice bearing subcutaneous (s.c.) HTB82 xenograft tumors showed higher tumor uptake compared to an isotype control RIC at 120 h post injection. Toxicology studies of 4.4 or 18.5 MBq ¹³¹I-8H9 in tumor-bearing mice showed no long-term adverse effects on tissues or mouse weights at days 15 or 30 post injection. During RIT studies, tumors in mice given 18.5 MBq ¹³¹I-8H9 partially regressed during the first 3 weeks after treatment and subsequently re-grew starting 50–60 days after treatment, while a control RIC did not inhibit tumor growth. Treated tumors showed sustained expression of the epitope recognized by mAb 8H9, suggesting that fractionated dosing with the B7-H3-targeted RIC could be performed to enhance anti-tumor efficacy.

Preclinical studies investigated the distribution, dosimetry, and toxicology of ¹²⁴I- or ¹³¹I-8H9 delivered via convection-enhanced delivery (CED) into the brainstem of rats and non-human primates [88]. CED is used to enhance the distribution of an administered solution throughout a tissue or compartment that has restricted the flow of the injected solution [88, 94]. In the study, rats received a brainstem infusion of 3.7–37 MBq ¹³¹I-8H9 and were monitored for toxicology. One of five rats given 37 MBq ¹³¹I-8H9 showed severe hemiparesis that was due to necrosis at the infusion site. No other rats showed behavioral abnormalities, weight loss, or histopathological damage throughout the three-month study period. MicroPET imaging studies in rats injected with ¹²⁴I-8H9 into the pons showed variable concentration of activity (18–75 %ID/g) and absorbed doses (0.43–1.5 Gy/MBq) at the injection site, with <5% of the infused activity remaining at the injection site or in the brain by 72 h. One male cynomolgus monkey was injected in the pons with 37 MBq ¹²⁴I-8H9. PET/CT and MRI imaging at 2 and 40 h showed a high concentration of activity in the pons and midbrain, while activity was also apparent over the brain surface, in the spinal cerebrospinal fluid (CSF), and the thyroid due to systemic clearance of the RIC through the disrupted BBB at the infusion site. Dosimetry analyses indicated a biological absorbed dose of 0.10 cGy/MBq, with the thyroid showing the highest absorbed dose (2 Gy) outside of the central nervous system (CNS). No adverse neurological, behavioral, or hematologic effects were observed throughout the nine-month study period. Histopathological analyses after terminating the primate showed mild neuronal loss and gliosis at the infusion site. These preclinical studies demonstrated the feasibility of using CED-delivered ¹²⁴I-8H9 in the brainstem to monitor the volume of distribution and calculate dosimetry levels of the RIC in vivo. The results from these studies paved the way for translation of CED-delivered ¹²⁴I-8H9 into clinical studies in human patients with CNS-confined malignancies, as discussed further below.

3.1.2. Clinical RIT Studies

Results from the first clinical study that included B7-H3-targeted RIT was reported in 2010 [12]. The study used intrathecal targeted RIT for curative treatment of recurrent, metastatic neuroblastoma in the CNS. Metastatic CNS neuroblastoma is associated with a poor prognosis, where conventional therapies fail to improve survival beyond 6 months in the majority of patients [12, 95]. Patients were enrolled in trials using either ¹³¹I-8H9 or the anti-GD2 mAb ¹³¹I-3F8 as therapies against CNS or leptomeningeal cancers (NCT00445965, NCT00089245). As mAb 8H9 was generated using human neuroblastoma cells as the antigen [11], using RICs of mAb 8H9 represents a rational approach for targeted RIT of neuroblastoma. In the reported study, eighteen pediatric patients with CNS neuroblastoma were treated with a salvage combinatorial regimen starting with resection of bulk CNS disease followed by chemotherapy, craniospinal irradiation, peripheral blood stem cell rescue (if necessary due to myelosuppression), intrathecal compartment RIT (1–2 doses administered monthly, 370–2220 MBq/injection) with ¹³¹I-8H9 or the anti-GD2 mAb ¹³¹I-3F8, and subsequent systemic immunotherapy and chemotherapy. Patients in the study were given ¹³¹I- or ¹²⁴I-8H9 before RIT to estimate dosimetry to CSF, spinal cord, normal brain, ventricles, and blood as determined by single photon emission computed tomography or PET imaging. Dosimetry estimates indicated that individual patients received 0.46–3.8 cGy/MBq ¹³¹I-8H9, yielding 10.4–88.2 Gy absorbed doses to the CSF due to RIT. Fourteen of the treated patients achieved complete responses that lasted throughout the study period (7–74 months depending on the patient), and one patient showed evidence of progressive disease in the CNS compartment (occurred 14 months after the initial CNS event). The other patients died from either non-CNS metastatic disease (2 patients) or infection without the appearance of residual neuroblastoma at autopsy (1 patient). RIT was well tolerated, with transient and acceptable effects on serum chemistry at single-administration dose levels up to 1110 MBq ¹³¹I-8H9. Five patients required prolonged thyroid hormone replacement intervention due to treatment-related hypothyroidism. These results supported the feasibility of managing CNS-restricted neuroblastoma with the combinatorial regimen incorporating RIT. While the study did not have multiple arms to evaluate which components of the therapeutic regimen were sufficient or necessary to halt disease progression, the observed durable responses were noteworthy compared to outcomes from alternative treatments previously tested in patients with CNS neuroblastoma. It was noted that the selection criteria for the study could have biased the favorable responses observed in the treated patients compared to historical controls with rapidly progressing neuroblastoma.

A retrospective analysis examined the rate of radiation necrosis in patients treated with either intrathecal compartment ¹³¹I-8H9 (370–2590 MBq/injection, up to two injections) or ¹³¹I-3F8 (370 MBq/injection, up to four injections) after external radiation therapy. Patients were from the single-institution clinical trials noted above and had CNS metastatic disease of either medulloblastoma (38 patients) or neuroblastoma (58 patients). All except two patients had been treated with radiation therapy directed to the CNS prior to RIT (mean 28 or 21.7 Gy cerebrospinal dose in medulloblastoma or neuroblastoma patients, respectively); also received post-RIT XRT (mean 27.5 or 24 Gy cerebrospinal dose in medulloblastoma or neuroblastoma patients, respectively) in subsets of patients. The mean cerebrospinal dose from RIT in medulloblastoma or neuroblastoma patients was 18.6 or 32.1 Gy, respectively. Two cases of radiation necrosis were noted during the median 41.5-month follow-up period in this cohort of patients, which is lower than that reported in larger, historical studies (>4% incidence) [89].

In a retrospective review of 23 patients with CNS relapse of rhabdomyosarcoma [90], patients with B7-H3-positive biopsy samples (3 patients) were treated with one or two intrathecal doses of ¹³¹I-8H9 RIT (dose not reported). Prior to B7-H3-targeted RIT, treatment of these patients included surgical excision of parenchymal disease and CNS radiation therapy (cerebrospinal or whole brain) with or without chemotherapy; prior CNS-directed radiotherapy had been completed 1.1–2.4 months prior to treatment with RIT. These three patients survived 13–50 months after CNS relapse of disease, which was significantly longer than patients who did not receive ¹³¹I-8H9 RIT (p=0.003). Patients treated with CNS-directed radiotherapy also survived significantly longer than patients without radiotherapy (p<0.001). It was noted that due to the limited sample size and retrospective nature of the study, patients that received radiotherapy or RIT could have had less aggressive disease or greater inherent sensitivity to radiation, which would have biased the favorable outcomes seen in these patients compared to the patients in alternative groups. The percentage of B7-H3-positive rhabdomyosarcoma biopsy samples was lower in this study compared to prior studies [11]. This result was attributed to differences in biopsy techniques between studies, and to potential differences in expression of B7-H3 in primary vs. metastatic disease. The authors pointed out the potential benefits to explore these questions (e.g., B7-H3 expression in initial vs. relapsed disease) in more detail for future investigations.

Recently, results have been reported from a phase I clinical trial (NCT01502917) testing the maximum tolerated dose of CED of ¹²⁴I-8H9 in 28 pediatric patients with diffuse intrinsic pontine glioma (DIPG) [92, 94]. DIPG is the most frequent brainstem malignancy in pediatric patients and has a median survival of less than 12 months, indicating the dire need for effective therapies against DIPG [96]. The phase I study with ¹²⁴I-8H9 was the first clinical trial to evaluate the safety of CED in DIPG pediatric patients (primary outcome), and the first to determine both dosimetry and therapeutic outcomes from a single RIC in humans with brain cancer (secondary outcomes). Enrolled patients had previously been treated with external beam radiation therapy and showed no evidence of clinical or radiographic disease progression at time of enrollment. Targeted RIT at this stage of disease was expected to maximize tolerability and potential therapeutic outcome. Cohorts of 3–6 patients received intra-tumor, single infusions of ¹²⁴I-8H9 (9.25, 18.5, 27.75, 37, 92.5, 120.25, or 148 MBq per patient per cohort) via CED. Patients were given saturated potassium iodide and liothyronine to prevent uptake of ¹²⁴I in the thyroid. All 28 patients were assessed for adverse events for a 30-day period following CED of ¹²⁴I-8H9 to determine the tolerability of treatment [92]. There were no treatment-related grade 4 adverse events or deaths in the evaluated patients. The most common grade 1 or 2 side effects included headache, cranial neuropathy, pain, and motor weakness, or raised aspartate aminotransferase levels. Grade 3 events were limited to transient hemiparesis or skin infection. Minor motor abnormalities typically resolved spontaneously or with steroid treatment. The maximum tolerated dose was not reached in the evaluated patients, indicating the tolerability of treatment in the reported study. No patients tested positive for human anti-mouse antibodies during the study period.

Secondary outcomes from the phase I study with ¹²⁴I-8H9 included estimations of volume of distribution after CED, dosimetry analyses, and survival after initial diagnosis. MRI and PET scans to determine volume of distribution and dosimetry indicated increased necrosis within the tumor at the end of the 30-day study period compared to baseline in five patients among the cohorts treated with either 92.5, 120.25, or 148 MBq ¹²⁴I-8H9. The volume of injectate distribution correlated linearly with the volume injected across all evaluable patients (r²=0.75). A case report of one patient from the study suggested that adequate delivery to the tumor was achieved despite the presence of an intra-tumoral cyst [91]. PET/CT scans indicated 61±22% of the injected activity remained in the brainstem within 1–6 h of infusion, with high retention in brainstem up to day 14 after infusion. Clearance of activity from the target lesion followed a biexponential clearance profile. Dosimetry analyses indicated an average absorbed dose of 0.39±0.20 Gy/MBq in the lesion, with a lesion/whole-body absorbed dose ratio of 1285±1019. These results indicate that intratumor CED of ¹²⁴I-8H9 to DIPG tumors provides high tumor-to-normal tissue dosimetry and is tolerated in patients. The median survival of the evaluable patients was 15.3 months after diagnosis of disease. At time of interim analysis, three of the evaluable patients remained alive with a median follow-up time of 36.1 months. All other patients had died from disease progression. Preliminary statistical comparison of median survival with dose level did not reveal a significant correlation (p=0.053), although the study was not powered to determine overall survival. Many of the enrolled patients were treated with repeat infusions or alternative therapies after exiting the trial, further complicating the evaluation of treatment efficacy. Despite these limitations in secondary outcome measures, the primary outcome of the study indicated the safety and feasibility of CED for RIT of pediatric patients with DIPG. The information from the study will be useful in designing a phase II trial using this therapeutic approach.

The favorable outcomes of the collective clinical studies led to breakthrough therapy designation from the FDA for a multi-site phase II/III trial with ¹³¹I-8H9 (burtomab, Y-mAbs Therapeutics, Inc., New York, NY) in pediatric patients with neuroblastoma CNS or leptomeningeal metastases (NCT03275402) [97]. A separate phase I clinical trial was designed to test the safety of intraperitoneal (i.p.) RIT using ¹³¹I-8H9 in patients with i.p. desmoplastic small round cell tumors or peritoneal metastases of other B7-H3-positive solid tumors (NCT01099644). While final results from the study are not yet available, preliminary results indicated dose levels of 1110–2220 MBq/m² were tolerated in patients. Dosimetry estimates were performed with a pre-treatment dose of ¹²⁴I-8H9 [98]. This study will be important in demonstrating whether i.p. treatments with mAb 8H9 result in human anti-murine immune responses, which have not been frequently observed during the other clinical trials using mAb 8H9 for targeted RIT within the CNS compartment.

3.2. Studies Using α-particle RICs of mAb 376.96

The murine mAb 376.96 was generated from Balb/c mice hyperimmunized with human melanoma cells [99]. This IgG2_a mAb binds to an extracellular epitope of 2Ig and 4Ig human B7-H3 that is expressed on both differentiated cancer cells and cancer initiating cells (CICs) from various malignancies [15]. CICs, also referred to as cancer stem cells, are implicated in generating recurrent, metastatic, and chemotherapy-resistant disease [100–101]. Therefore, destroying these cells as well as bulk cancer cells is thought to be essential to effectively eliminate the malignancy from a subject [102]. Immunohistochemical analyses have indicated that the mAb 376.96-defined B7-H3 epitope has restricted expression in most normal tissues [15]. These properties prompted the selection of mAb 376.96 as a carrier to prove the concept for targeted α-particle RIT of in vitro and in vivo preclinical models of human cancer.

3.2.1. Review of Preclinical Studies Using ²¹²Pb-376.96 for i.p. RIT of Peritoneal Malignancy

Targeted RIT in the peritoneal compartment is relevant for ovarian cancer, as the peritoneal compartment is the main location of treatment failure in patients. Effective therapies against ovarian cancer are needed, as many patients treated with conventional chemotherapy eventually present with recurrent, chemotherapy-resistant disease [103]. Previous clinical trials have shown that i.p. targeted RIT with β-particles is only modestly effective in patients with ovarian cancer [7], possibly due to the inability of low LET radiation to kill CICs or microscopic cell clusters at tolerable doses [8]. More recent phase I trials using i.p. targeted RIT with α-particles have shown this therapy strategy to be tolerable in patients [39, 53, 104–105]. In a dose escalation study of ²¹²Pb-trastuzumab injected i.p. in 18 patients with HER2-positive (human epidermal growth factor receptor 2) peritoneal malignancies that failed prior conventional therapies, patients in the highest 3 dose cohorts (16.3–27.4 MBq/m²) showed stable disease and reduced tumor growth at 6 weeks post treatment, with decreased blood levels of tumor antigens (e.g, TAG-72), as compared with patients in the 3 lowest dose cohorts (7.4–12.6 MBq/m²) [39]. Therefore, continued preclinical and clinical investigations are warranted to evaluate the efficacy of α-particle RIT using mAb carriers that bind to antigens overexpressed in the majority of ovarian cancer patients. Immunohistochemical studies have shown that 93% of tested human ovarian cancer specimens (96 of 103) express B7-H3, where expression levels in tumor endothelium correlate with worse prognosis [16]. Immunohistochemical and flow cytometry staining experiments have shown that mAb 376.96 binds to solid ovarian cancer lesions and ascites cells collected from the peritoneal compartment of patients with ovarian cancer [106]. These factors prompted preclinical evaluations with ²¹²Pb-labeled mAb 376.96 (²¹²Pb-376.96) against in vitro and in vivo models of human ovarian cancer [25].

²¹²Pb-376.96 inhibited the clonogenic survival of human ovarian cancer cells grown as adherent monolayers or as non-adherent CICs. The sensitivity of the ovarian cancer differentiated cells or CICs to ²¹²Pb-376.96 paralleled the relative number of binding sites present on the cells as determined by Scatchard assays with ²¹²Pb-376.96. Notably, ²¹²Pb-376.96 halted the proliferation of ovarian cancer cells and CICs in vitro at dose levels that are clinically achievable with alternative α-particle RICs in the peritoneum of human patients without causing toxicity [8, 53, 104]. Combining ²¹²Pb-376.96 with the radiosensitizing chemotherapy drug carboplatin was significantly more effective than the individual agents at inhibiting the in vitro clonogenic survival of carboplatin-resistant ovarian cancer cells and CICs. The inhibitory growth effects due to targeting specificity was confirmed by the significantly higher clonogenic survival of ovarian cancer cells and CICs treated with the isotype control ²¹²Pb-F3-C25 (three- to forty-fold lower sensitivity relative to ²¹²Pb-376.96).

Biodistribution studies in mice bearing i.p. xenografts of human ovarian cancer showed that ²¹²Pb-376.96 resulted in higher uptake of the injected ²¹²Pb activity in ascites cells or tumor nodules compared to any other tissue at 24 h post injection. The ²¹²Pb activity in tumors of mice given the control RIC ²¹²Pb-F3-C25 was lower compared to the mice given ²¹²Pb-376.96, while the uptake of activity in normal tissues was comparable between the two RICs, confirming the ability of ²¹²Pb-376.96 to result in specific retention of activity in ovarian tumors in vivo.

Therapeutic efficacy studies showed that mice bearing ES-2 i.p. ovarian cancer xenografts survived significantly longer after i.p. treatment with 0.17–0.51 MBq ²¹²Pb-376.96 compared to non-treated control mice. Similarly, mice bearing A2780cp20 i.p. ovarian cancer xenografts showed a dose-dependent increase in survival time after i.p. treatment with 0.35–0.70 MBq ²¹²Pb-376.96; these mice survived significantly longer than non-treated mice, mice treated with carboplatin, or mice treated with a comparable dose of the control RIC ²¹²Pb-F3-C25 regardless of the tumor model (ES-2 or A2780cp20). Transient toxicity was noted due to drops in body weight in all mice treated with ²¹²Pb-376.96 starting one week after dosing, although all mice recovered to their initial weights within 2–4 weeks after RIT treatment.

3.2.2. Review of Preclinical Studies Using ²¹²Pb-376.96 for Systemic RIT of Solid Tumors

Due to the relatively short path length of α-particles in tissues, targeted α-particle RIT after systemic administration is typically regarded as most effective for treating myeloid malignancies (e.g., leukemia) or microscopic metastases of solid malignancies (e.g., breast cancer). Several preclinical studies have also demonstrated that established solid tumors can also be effectively treated with α-particle RICs administered systemically. A recent study tested the efficacy of ²¹²Pb-376.96 to halt the progression of preclinical models of solid human pancreatic ductal adenocarcinoma (PDAC) tumors [26]. PDAC is the fourth most common cause of death due to cancer and is associated with a 5-year survival rate of less than 10% [107], indicating a need for effective therapies to improve the prognosis for patients with PDAC. PDAC is known to overexpress B7-H3, where aggressive and invasive phenotypes of PDAC have higher expression compared to low-grade PDAC, thus presenting B7-H3-targeted RIT as an attractive approach against this malignancy.

In vitro clonogenic survival assays indicated that human PDAC3 cells, which were derived from ascites fluid of a patient with metastatic PDAC, grown as adherent monolayers and grown as non-adherent tumorspheres with CIC characteristics were 3–6 times more sensitive to ²¹²Pb-376.96 than to ²¹²Pb-F3-C25. Similar to the results seen with ovarian cancer cells, the PDAC3 cells grown under CIC conditions showed more binding sites per cell and were more sensitive than adherent PDAC3 cells to ²¹²Pb-376.96.

Biodistribution studies at 24 h after i.v. injection of ²¹²Pb-376.96 or ²¹²Pb-F3-C25 in mice bearing s.c. patient derived xenograft (PDX) PDAC tumors showed significantly greater uptake of ²¹²Pb activity in tumors from the ²¹²Pb-376.96 group as opposed to the isotype control group. Except for the spleen, all other normal tissues showed lower uptake of the ²¹²Pb activity compared to the tumor at 24 h after injection of ²¹²Pb-376.96. The pattern of activity localization in the spleen and liver after i.v. injection of ²¹²Pb-376.96 in mice suggests the RIC may be recognized and cleared from circulation by cells in the reticuloendothelial system. Biodistribution studies in mice bearing orthotopic PDAC3 xenografts showed significantly more uptake of ^99mTc relative to ¹²⁵I activity in the tumors at 24 h after i.v. co-injection of ^99mTc-376.96 and ¹²⁵I-F3-C25. The higher uptake of activity in the tumors observed with the mAb 376.96 RICs relative to the isotype control RICs in two distinct PDAC models confirmed that the mAb 376.96 RICs specifically targeted human PDAC tumors in mice.

Tumor growth inhibition studies in mice bearing s.c. PDX PDAC tumors showed that 0.2–0.73 MBq ²¹²Pb-376.96 slowed tumor growth relative to non-treated control tumors; 0.54 MBq and 0.73 MBq dose levels of ²¹²Pb-376.96 were equally effective at inhibiting tumor growth. In support of the B7-H3 targeting specificity, 0.36 MBq ²¹²Pb-376.96 showed similar anti-tumor growth efficacy as a higher dose (0.46 MBq) of the control ²¹²Pb-F3-C25. All mice given RIT showed transient drops in body weight during the first week after injection of the RICs, although all mice quickly recovered (within 2 weeks), indicating that acute toxicity due to RIT likely occurred in the studies. Other analyses of toxicology due to systemic RIT with ²¹²Pb-376.9 were not performed in the studies. Similar to the results in the ovarian cancer models, no PDAC tumor regressions were observed in these studies, suggesting that fractionated doses of ²¹²Pb-376.96 alone or in combination with chemotherapy may be required to effectively halt proliferation of malignant cells.

3.2.3. Unpublished In Vitro Studies Using ²¹²Pb-376.96 in Combinatorial Therapy Approaches

Triple negative breast cancer (TNBC) is another solid malignancy with a poor prognosis that has been shown to express B7-H3 [108]. Biodistribution studies in athymic nude mice bearing orthotopic xenografts of human SUM159 TNBC cells showed significantly greater uptake of ²¹²Pb relative to ¹²⁵I activity in the tumors (11.2±2.0% and 6.5±1.1% ID/g, respectively) at 24 h after co-administration of ²¹²Pb-376.96 and ¹²⁵I-F3-C25 via tail vein injection (Fig. (1). The specific uptake of ²¹²Pb in SUM159 xenografts (4–7 mm diameter) at 24 h after injection of ²¹²Pb-376.96 was higher than that observed in SUM159 xenografts targeted with the chondroitin sulfate proteoglycan 4 (CSPG4) specific mAb 225.28 labeled with ²¹²Pb (8.3 ± 5.7% ID/g at 24 h) [109]. mAb 225.28 binds to an epitope of CSPG4 that is expressed on 73% of primary human TNBC specimens (32 of 44) and pleural effusions from 12 patients with metastatic TNBC. Similar to the mAb 376.96-defined B7-H3 epitope, the mAb 225.28-defined CSPG4 epitope has restricted expression in normal tissues [110]. The higher uptake of injected activity in SUM159 tumors mediated by ²¹²Pb-376.96 relative to ²¹²Pb-225.28 is consistent with the respective in vitro expression of the proteins on differentiated SUM159 cells. Specifically, SUM159 cells express approximately 10⁵ binding sites/cell of the mAb 376.96 defined B7-H3 epitope compared to 2.9×10⁴ binding sites/cell of the mAb 225.28 defined CSPG4 epitope [109]. The results from these studies in mice with solid tumors support the concept for B7-H3-targeted α-particle RIT against multiple types of human cancer. Future preclinical studies to determine tissue dosimetry and toxicology during RIT with ²¹²Pb-376.96 or other candidate RICs would be needed prior to translational pilot studies with this RIT strategy in human patients.

Fig. (1).

²¹²Pb-376.96 shows specific localization in human TNBC xenografts in mice. Female athymic nude mice (n=5) were injected with 4 million SUM159 human TNBC cells, 1:1 with Matrigel, in the mammary fat pad 7 days prior to the experiment. Tumors were 3–7 mm in diameter at the time of the experiment. On the day of the experiment, mice received an i.v. injection containing 1.47 MBq (10.5 μg) ²¹²Pb-376.96 and 16.7 kBq (10.7 μg) ¹²⁵I-F3-C25 (isotype control) in 200 μL PBS. Mice were euthanized at 24 h after injection, when tumors and normal organs were resected, weighed, and counted using a gamma counter to determine the %ID/g of each radionuclide by comparing the tissue activity to solutions with known activity of the radionuclide of interest. ¹²⁵I was counted after the ²¹²Pb activity had fully decayed. Data are presented as mean±standard deviation. **p<0.01 (two-tailed student’s t test)

3.2.4. Unpublished In Vitro Studies Using ²¹²Pb-376.96 in Combinatorial Therapy Approaches

The above studies demonstrate that B7-H3-targeted RIT when used alone results in significant therapeutic advantages compared to control radiolabeled mAb, although it fails to eradicate malignant progression in the majority of cases. Combining targeted RIT with chemotherapy is a common and established method to enhance therapeutic outcome in preclinical and clinical settings [1, 12, 25, 69–70]. Using selective chemotherapy agents that halt the proliferation of differentiated cancer cells and CICs is a particularly attractive strategy to prevent disease recurrence or resistance to conventional chemotherapies, which typically enrich the CIC population [111–113]. The sonic hedgehog signaling pathway is aberrantly regulated in various malignancies, resulting in enhanced metastatic potential of malignant cells [114–119]. LDE225 (Novartis, Basel, Switzerland) is a Smoothened antagonist that has been shown in preclinical studies to inhibit sonic hedgehog signaling, resulting in reduced growth of chemotherapy-resistant tumors and CICs derived from various tumor cell lines [114, 120]. In vitro studies have shown that combining LDE225 with ²¹²Pb-376.96 is more effective than either agent alone at inhibiting the proliferation of CICs derived from human PDAC or ovarian cancer cell lines (Fig. (2). The canonical Wnt/β-catenin signaling pathway is also aberrantly activated in differentiated cancer cells and CICs [121–124]. Increased Wnt/β-catenin signaling has been shown to lead to chemotherapy resistance and expression of genes that are characteristic of CICs [111, 125–128]. SRI32529 (Southern Research, Birmingham, AL, USA) is a novel small molecule inhibitor of the Wnt/β-catenin signaling pathway that has been shown to inhibit the proliferation of human TNBC cells, although the mechanism of action of SRI32529 is not exclusively limited to the Wnt/β-catenin pathway [129]. As cancer cells and CICs often exploit multiple signaling pathways to evade cell death when a single pathway is blocked, initial experiments were performed to determine if SRI32529 in combination with B7-H3-targeted RIT could reduce the survival of malignant cells compared to the individual agents. In vitro clonogenic survival assays showed that human PDAC3 cells and CICs treated with both ²¹²Pb-376.96 and SRI32529 had lower clonogenic survival compared to cells treated with SRI32529 alone (Fig. 3).

Fig. (2).

Combinatorial treatment with ²¹²Pb-376.96 and Smoothened inhibitor LDE225 reduces the in vitro clonogenic survival of human ovarian cancer cells (ES-2; A,B) or PDAC cells (PDAC3; C,D) grown as adherent differentiated cells or as non-adherent tumorspheres with enriched CIC characteristics (CICs). Adherent cells (A,C) or CICs (B,D) were treated for 1 d with vehicle or LDE225 (10 μM). Treated adherent cells or CICs were incubated for 2 h with the indicated concentration of ²¹²Pb-376.96, washed, and returned to incubation with LDE225 or vehicle for 2 d. Adherent cells and CIC tumorspheres were collected, dissociated into single cells, and plated in 6-well plates (60–100 cells/well) for 10 days. Colonies were fixed with 10% formalin and stained with 0.01% crystal violet, and colonies with >50 cells were counted. Data are presented as mean % clonogenic survival (± standard deviation) relative to vehicle-treated controls (no drug and no radiation). n=3 to 6 replicate wells per condition. ***p<0.001.

Fig. (3).

Combinatorial treatment with ²¹²Pb-376.96 and Wnt/β-catenin pathway inhibitor SRI32529 reduces the in vitro clonogenic survival of PDAC3 human cancer cells grown as adherent differentiated cells or as non-adherent tumorspheres with enriched CIC characteristics (CICs). PDAC3 adherent cells (A) or CICs (B) were treated for 1 d with vehicle or SRI32529 (0.2 μM). Treated cells or CIC tumorspheres were incubated for 2 h with the indicated concentration of ²¹²Pb-376.96, washed, and returned to incubation with SRI32529 or vehicle for 2 d. Adherent cells and CICs were collected, dissociated into single cells, and plated in 6-well plates (100 cells/well) for 10 days. Colonies were fixed with 10% formalin and stained with 0.01% crystal violet, and colonies with >50 cells were counted. Data are presented as mean % clonogenic survival (± standard deviation) relative to vehicle-treated controls (no drug and no radiation). n=3 to 6 replicate wells per condition. *p<0.05

In addition to aberrantly activated intracellular signaling pathways in cancer, epigenetic DNA modifications are now widely recognized to fundamentally contribute to aberrant gene expression patterns associated with many human malignancies [130–131]. Several types of cancer show increased levels of class I histone deacetylases (HDACs), which cause aberrant gene expression leading to increased malignancy and immune cell deactivation in the tumor microenvironment. Targeted inhibition of HDACs has thus been explored as a therapeutic strategy against these malignancies [132–136]. Entinostat is a selective class I HDAC inhibitor that has been given breakthrough approval by the FDA for recurrent, aromatase-resistant breast cancer [137]. In vitro studies have shown that entinostat leads to significantly increased protein expression of B7-H3 in several human ovarian cancer cell lines (Table 3). While HDAC inhibitors in ovarian cancer have not been effective in clinical trials [136], these results raise the possibility to develop combinatorial targeted therapy strategies directed against downstream products that are induced in ovarian cancer cells but not in normal cells during HDAC inhibition. Such a strategy could potentially widen the therapeutic window to more effectively combat ovarian cancer cells and CICs through B7-H3-targeted RIT or other approaches.

Table 3.

HDAC inhibition of human ovarian cancer cells leads to increased expression of the mAb 376.96 defined B7-H3 epitope.^a

	Number of B7-H3 binding sites/cell (×104)		Entinostat/control ratio
	Control	Entinostat
Adherent Hey-A8	6.54	30.8b	4.71
CIC Hey-A8	6.07	22.0b	3.62
Adherent ES-2	9.48	55.3c	5.83
CIC ES-2	10.4	27.7c	2.66
Adherent SKOV3.ip1	16.7	25.1c	1.50
CIC SKOV3.ip1	19.7	38.3c	1.94
Adherent SKOV3-TR	24.4	40.0c	1.64
CIC SKOV3-TR	13.7	27.7c	2.02

^aScatchard results from binding assays with ^99mTc-376.96 and ovarian cancer cell lines cultured with or without entinostat for 3 days. At 24 h after seeding ovarian cancer cell lines in culture as adherent monolayers (Adherent) or as non-adherent tumorspheres with enriched CIC characteristics (CIC), vehicle (control) or entinostat at the indicated concentration was added to the cells for 3 days. After the 3-day treatment period, adherent monolayers or dissociated CICs in suspension were incubated with dilutions of ^99mTc-376.96 with or without unlabeled mAb 376.96 as a blocking agent for 1 h at 37 °C in Scatchard cell binding assays. The data were analyzed as previously described to determine the number of binding sites per cell [25].

^b1 μM entinostat

^c2.5 μM entinostat.

3.3. Review of Studies Using β-particle RICs of mAb 4H7

A murine anti-human 2IgB7-H3 mAb, 4H7, was generated by hyperimmunizing Balb/c mice with L929 murine fibroblast cells transfected with human 2Ig B7-H3 [138]. mAb 4H7 has since been used as a carrier for biodistribution and RIT studies in mice bearing s.c. xenografts of 786–0 human clear cell renal cell carcinoma (RCC) cells [93]. RCC is the most common type of kidney cancer in adults and fails to respond to conventional chemotherapy or radiation therapy in the majority of patients, resulting in a poor prognosis for the disease [139]. IHC staining has shown expression of B7-H3 in RCC tumor cells and vasculature [140–141], indicating the potential to use B7-H3-targeted RICs for therapy of RCC. Biodistribution studies showed that ¹³¹I-4H7 had higher uptake than the isotype control ¹³¹I–mIgG in the tumors from 2–24 h after injection, while the uptake in the tumors at 48 and 72 h was comparable for both RICs. Tumor-to-normal tissue ratios of 1.29–5.02 for the lung, kidneys, liver heart, and muscle were observed at 24 h after injection of ¹³¹I-4H7. Tumor growth inhibition studies indicated that 7.4 MBq ¹³¹I-4H7 was significantly more effective than an equivalent dose of ¹³¹I-mIgG, non-targeted ¹³¹I, or saline at inhibiting tumor growth by day 28 after injection. Immunohistochemical staining of tumors from mice treated with ¹³¹I-4H7 showed larger areas of necrosis compared to tumors from the other treatment or control groups. There were no tumor regressions observed in the therapy studies, although a dose-escalation experiment indicated 3.7–29.6 MBq ¹³¹I-4H7 significantly inhibited tumor growth in a dose-dependent manner.

3.4. Additional B7-H3-targeted mAbs that could be used in Future RIT Studies

Two additional B7-H3-targeted mAbs have been developed for therapy of human cancer, but have not yet been employed in targeted RIT studies. Enoblituzumab (MGA271) is a humanized mAb containing an engineered Fc domain to enhance antibody-dependent cellular cytotoxicity [13]. Enoblituzumab is currently being evaluated in several clinical trials to determine its safety and efficacy for immunotherapy of malignancies that express B7-H3 (NCT02982941, NCT02381314, NCT02923180, NCT02475213, NCT01391143). m276 is a fully human mAb obtained through affinity maturation studies with the B7-H3 protein [23]. m276 cross-reacts with human, murine, and non-human primate B7-H3. Preclinical studies using antibody drug conjugates of m276 have shown therapeutic efficacy against B7-H3-expressing human and murine cancer cell lines and tumors grown in mice [23]. The ability of both enoblituzumab and m276 to bind to a wide array of malignancies with little cross-reactivity to normal tissues suggests that these mAbs would be worthy of consideration as carriers for B7-H3-targeted RIT in future studies.

CONCLUDING REMARKS AND FUTURE OUTLOOK

The field of B7-H3-targeted RIT has shown favorable progress since the first results of this approach were published nearly 15 years ago. Considering the number of human malignancies that express B7-H3, it is likely that many more RIT applications can be explored using the above RICs or using novel mAbs that bind to B7-H3. The availability of humanized and fully human mAbs that target B7-H3 is particularly advantageous for systemic RIT of solid tumors in human patients, as RICs developed from these mAbs could be administered in fractionated doses without inducing immunogenic responses against xenobiotic mAbs. Having multiple mAbs and therapeutic radionuclides to select from when designing future B7-H3-targeted RICs is anticipated to allow researchers to optimize RIT efficacy and tolerability for advancement through clinical studies in the future.

Despite the benefits of B7-H3-targeted RIT observed to date, it is possible that this RIT strategy when employed as a single therapeutic regimen may not eradicate all malignant cells, as has been frequently observed with alternative RIT approaches [1, 4–5, 7, 26, 38, 44, 51, 67, 109, 142]. Combinatorial therapy strategies that synergistically inhibit tumor cell proliferation are desirable to maximize therapeutic efficacy and expand the therapeutic window. B7-H3-targeted RIT in combination with novel therapeutic agents that are active against CICs represents an attractive approach to halt recurrent or recalcitrant malignancies. An increasing number of small molecules are being developed against various pathways that are active in CICs [143–148]. Combining B7-H3-targeted RIT with inhibitors of CIC proliferation would be worthy of future investigation. An alternative combinatorial approach against advanced malignancies would be to utilize both targeted RIT and immune checkpoint blockade immunotherapy that blocks inhibitory immune signals (e.g., PD-1, CTLA-4) on effector immune cells [149–152]. While the role of B7-H3 in immune regulation is not yet fully understood, studies have shown that blocking B7-H3 enhances therapeutic outcomes at least partially through pathways of the adaptive immune response [17, 86, 153]. Radiotherapy is also known to stimulate immune responses in solid tumors [154–157], although the effects of targeted RIT on adaptive immune responses have not yet been explored. Combining B7-H3-targeted RIT with immunotherapy thus represents a novel area of research in the field of cancer therapy. As the field of targeted RIT of cancer continues to progress, additional B7-H3-targeted RICs and therapeutic strategies are likely to be developed to enhance the therapeutic efficacy against B7-H3-expressing malignancies. Such advancements will be important steps to improve the prognosis for patients with cancer.

LIST OF ABBREVIATIONS

CED

Convection-enhanced Delivery

CIC

Cancer Initiating Cell

CNS

Central Nervous System

CSF

Cerebrospinal Fluid

CSPG4

Chondroitin Sulfate Proteoglycan 4

CT

Computed Tomography

DIPG

Diffuse Intrinsic Pontine Glioma

EGFR

Epidermal Growth Factor Receptor

FDA

Food and Drug Administration

HDAC

Histone Deacetylase

HER2

Human Epidermal Growth Factor Receptor 2

%ID/g

Percent Injected Dose Per Gram

LET

Linear Energy Transfer

mAb

Monoclonal Antibody

MRI

Magnetic Resonance Imaging

PET

Positron Emission Tomography

PDAC

Pancreatic Ductal Adenocarcinoma

PDX

Patient Derived Xenograft

RCC

Renal Cell Carcinoma

RES

Reticuloendothelial System

RIC

Radioimmunoconjugate

RIT

Radioimmunotherapy

TNBC

Triple Negative Breast Cancer

XRT

X-ray Radiotherapy