Case Report of AdAPT-001-Mediated Sensitization to a Previously Failed Checkpoint Inhibitor in a Metastatic Chordoma Patient

Chordoma is a rare, locally invasive, and destructive malignancy of the axial skeleton, which only very rarely metastasizes, and which arises from the remnants of the notochord. It usually presents due to mass effect on and displacement of adjacent structures, i.e., brainstem, cranial nerves, nasopharynx, spinal cord and accounts for 1% of intracranial tumors and 4% of all primary bone tumors [1]. In approximately 30–50% of cases, the primary site is the sacrum. In 30–35% of cases, the tumor location is in the spheno-occipital region, and in 15–30% of cases, the tumor is in the vertebral bodies [2, 3]. Historically, chordomas are divided into three histopathologic subtypes: typical (conventional), chondroid, and dedifferentiated. The course for the first 2 forms is more indolent, with a 3-year overall survival rate of 90% in contrast to the more aggressive dedifferentiated subtype.

Most common in male Caucasians, the incidence of chordoma peaks at 50–60 years of age [4]. The prognosis is typically poor due to the locally aggressive and recurrent nature of these tumors as well as their resistance to radiotherapy and standard chemotherapy, with an overall 10-year survival of approximately 40% [5]. The accepted standard of treatment is surgical resection followed by adjuvant radiation therapy [6]. Complete surgical resection is ideal but rarely achieved due to the proximity of these tumors to critical vascular and neural structures [7]. Targeted therapies are lacking primarily due to the “quiet” chordoma genome with few known drivers of disease [8].

AdAPT-001 is a tumor-selective, replication-competent type 5 oncolytic adenovirus, which expresses a receptor-Fc fusion TGF-bata (TGFβ) trap (TβRII-Fc) that is under investigation in a phase 1 clinical trial called BETA PRIME (NCT04673942) [9]. Emerging evidence suggests that (1) aberrant activation of the TGFβ signaling pathway is integral to chordoma genesis [9] and that (2) immune checkpoint inhibitors (ICIs) may provide benefit although due to the small patient population data are scarce. One small 17-patient study reported 1 complete response (6%), 3 partial responses (18%), and 11 cases of stable disease (65%) with pembrolizumab in chordoma. Two patients experienced progressive disease. The overall duration of response was a median of 12 months overall, 13 months in patients who achieved a complete response or partial response, and 6 months in patients with stable disease [10]. This was despite the fact that Duan et al. [11] recently reported low-level expression of MHC class 1 and programmed cell death ligand 1 (PD-L1) genes in chordoma tumor cells, which mechanistically may argue against the use of PD-1 checkpoint inhibitors for chordoma.

Another one of the issues with ICIs, and the probable reason for the lack of universal benefit with them, since anti-PD-L1/anti-PD-1 antibodies are only active in a minority of patients is that to simply boost a “sleeping” immune system is not enough to overcome the immunosuppressive mechanisms such as TGFβ overexpression that are operant in the tumor microenvironment. An important feature of AdAPT-001 is that it is “armed” with a TGFβ trap to sequester TGFβ and, thereby, to decrease TGFβ-induced signaling and, by extension, immunosuppression [12]. Moreover, because AdAPT-001 sensitizes syngeneic mouse tumors previously refractory to the therapeutic effects of ICIs [12], synergy between AdAPT-001 and PD-1/L1 inhibitors is anticipated.

In the ongoing single-arm, open-label, solid tumor BETA PRIME phase 1 study, patients with palpable tumors underwent a single intratumoral injection of AdAPT-001 at escalating doses (ranging 2.5E¹¹, 5.0 E¹¹, 1.0E¹² viral particles [vp]). At the initial dose level, 2.5E¹¹ vp of AdAPT-001, a chordoma patient with aggressively metastatic disease demonstrated control of treated tumor for 4 months, as defined by RECIST criteria, in the absence of any AdAPT-001-related serious adverse events. Following treatment with AdAPT-001 and a brief treatment-free interval, March 2021 to August 2021, the patient initiated pembrolizumab (200 mg q 3 weeks). The patient remained on single agent pembrolizumab for 8 months (August 2021 to April 2022) at which time patient had documented progression. The CARE checklist has been completed by the authors for this case report, attached as online supplementary material (see www.karger.com/doi/10.1159/000529503) [13].

A heavily pretreated, post-surgical, and post-radiotherapy 66 years old dedifferentiated chordoma patient, RMF, with recurrent, metastatic sacral chordoma, enrolled in March 2021 on the BETA PRIME study. He previously received – and failed – nivolumab as well as chemotherapy. Prior to the start of therapy, RMF was symptomatic with chest wall pain and shortness of breath due to lung metastases. Approximately 2 months after direct injection of one dose of AdAPT-001 at a dose of 2.5 × 10¹¹ vp in a metastatic right extrapleural extending lesion, measurements of his injected lesions were stable; i.e., the lesions were non-progressive per RECIST criteria. No treatment-related adverse events were noted. Symptomatically, the patient was improved as chest wall pain and shortness of breath were no longer listed as adverse events. Five months later, the patient initiated pembrolizumab 200 mg IV every 3 weeks in August 2021 to which he responded with tumor shrinkage that lasted 7 months as subsequent imaging from April 2022 showed progressive disease. The comparison of the two scans, prior to enrollment and after treatment with AdAPT-001 and subsequent immunotherapy (Keytruda), is shown in Figure 1. A timeline is shown in Figure 2.

Chordoma is a rare tumor, occurring with a prevalence of approximately 2,000 in the USA, with few treatment options other than surgical resection and radiation therapy, and recurrence rate is high [14]. No targeted therapies have been approved for chordoma, due in large part to the difficulty of conducting adequately powered randomized controlled clinical trials.

This case report is the first ever to report on the use of AdAPT-001, a new biological entity, in chordoma, a rare, TGFβ-driven spinal cord tumor, which is a strength. A limitation of this case report, given that it is an n-1, is an inability to generalize. AdAPT-001 is potentially active directly as an oncolytic virus and can sensitize tumors to immunotherapies due to expression of a TGFβ-receptor trap. In preclinical studies, AdAPT-001 expressed the TGFβ-receptor trap which was detectable systemically for up to 2 weeks following treatment [15]. Synergism with anti-PD-L1 against distant, uninfected tumor areas was demonstrated in a mouse model where monotherapy with anti-PD-1 was ineffective [9]. Importantly, no virus was present in the contralateral nontreated tumors. Hence, the infection and replication of AdAPT-001 in tumors with concomitant overexpression of the TGFβ trap have the potential to systemically activate the anticancer immune response, thereby converting “cold” or nonresponsive tumors into “hot” or sensitive ones, which more effectively respond to various combination therapies such as ICIs (e.g., PD-1/PD-L1 and CTLA-4), cytokines, tumor vaccines, CAR-T, TLR, and STING agonists. Given the lack of approved treatments in chordoma, outside of en bloc surgical resection and radiotherapy and the logistical difficulties to conduct clinical trials in such a rare tumor type, further studies with AdAPT-001 and checkpoint inhibitors in chordoma patients are warranted.

Ethical approval is not required for this case report in accordance with local and national guidelines. Written informed consent was obtained from the patient in question to publish their medical case and accompanying images.

The authors have no conflicts of interest to declare.

EpicentRx Inc. is the sponsor of the BETA PRIME clinical trial (NCT04673942).

Santosh Kesari contributed to data collection, data analysis, manuscript writing, manuscript editing, and manuscript review. Jeannie Williams, Erica Burbano, Meaghan Stirn, Scott Caroen, Bryan Oronsky, Tony Reid, and Christopher Larson contributed to data collection, data analysis, manuscript writing, manuscript editing, and manuscript review.

All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.