Presented here are some highlights of preliminary studies presented at the 2023 American Association for Cancer Research (AACR) Annual Meeting. These presentations provide signals for the utility of an off-the-shelf chimeric antigen receptor (CAR) T-cell product in renal cell carcinoma, an artificial intelligence model called HECTOR that may predict recurrence in endometrial cancer, and a potential new combination of olaparib and ceralasertib that may overcome resistance to PARP inhibition. Further studies are needed to confirm the findings of these three studies.
Off-the-Shelf CAR T-Cell Therapy for Kidney Cancer
The allogeneic anti-CD70 CAR T-cell therapy ALLO-316 achieved signals of antitumor activity with a tolerable safety profile in patients with advanced or metastatic clear cell renal cell carcinoma, according to dose-escalation findings from the phase I TRAVERSE trial.1
In 18 patients evaluable for efficacy, the objective response rate to ALLO-316 was 17%, and the disease control rate was 89%, with a median follow-up of 7.8 months. In 10 patients with CD70-positive renal cell carcinoma, 3 patients (30%) experienced a partial response, and the disease control rate was 100%. The median progression-free survival was 5 months in CD70-positive patients.
Greater tumor reduction was correlated with higher baseline tumor CD70 immunohistochemistry H-Score, said lead study author Samer A. Srour, MB ChB, MS, of The University of Texas MD Anderson Cancer Center, Houston. (The H-score is a semiquantitative score system that calculates a score ranging from 0 to 300 based on both the intensity of tumor cytoplasmic staining and the percentage of cells stained, with 0 meaning no signal.)
Samer A. Srour, MB ChB, MS
“With this off-the-shelf CAR T-cell product, we have seen encouraging antitumor activity, with no unexpected safety signals,” said Dr. Srour. “The treatment was initiated within a median of 5 days from enrollment, which is a proof of concept for this off-the-shelf allogeneic product [that] can be given for patients with AN unmet need [and/or in need of] urgent treatment.”
Regarding safety, one dose-limiting toxicity, grade 3 type 2 autoimmune hepatitis, was reported at the second dose level with conditioning therapy and ALLO-316. There were two cases of grade 3 neurotoxicity, and one patient died of respiratory failure from COVID-19 infection, which was unrelated to the study treatment. Cytokine-release syndrome was found to be manageable, and there were no reports of immune effector cell–associated neurotoxicity syndrome or graft-vs-host disease. The safety data were similar to those reported with autologous CAR T-cell therapies. Infections were managed with prophylaxis.
CD70 is a promising target for CAR T-cell therapy, which is expressed in up to 80% of patients with renal cell carcinoma and is detected in other malignances, with little expression in normal tissue. TRAVERSE is a first-in-human, dose-escalation, multicenter trial of ALLO-316. “ALLO-316 has some unique features that might make it more attractive to treat [patients with] kidney cancer,” Dr. Srour noted.
Two conditioning regimens are being explored in this ongoing study: fludarabine at 30 mg/m2 and cyclophosphamide at 300 mg/m2, both daily for three cycles, with and without ALLO-316 at 10 mg daily for three cycles. Four cell dose levels of ALLO-316 were also evaluated. Safety and response were assessed on day 28. Follow-up continued through 12 months, and long-term follow-up is planned for 3 years.
The 19 patients included in the safety analysis set had advanced or metastatic (stage IV) clear cell renal cell carcinoma and had received prior treatment with an immune checkpoint inhibitor and tyrosine kinase inhibitor. The median patient age was 62 years, and 16% were female. Most patients had an Eastern Cooperative Oncology Group performance status of 0, and 79% had undergone a nephrectomy. The median number of prior lines of therapy was three.
High CAR T-cell expansion was seen following both conditioning regimens and at relatively low doses. Results suggest that ALLO-316 may eliminate CD70-positive host T cells, which in turn may prevent allorejection and support persistence.
Expansion cohorts are planned to enroll by the end of 2023, and they will be enriched with CD70-positive subtypes.
Olaparib Plus Ceralasertib in Pediatric Advanced Cancers
AcSé-ESMART is an international European proof-of-concept platform trial intended to match pediatric, adolescent, and young adult patients with relapsed or treatment-refractory cancers with a treatment regimen targeted to their cancer’s mutational profile. Part of the inclusion criteria is availability of comprehensive molecular profiling of the tumor (whole-exome/genome sequencing ± RNA sequencing). Susanne Gatz, MD, PhD, of the Institute of Cancer and Genomic Sciences of the University of Birmingham, United Kingdom, and colleagues, including Birgit Geoerger, MD, PhD, Head of the AcSé-ESMART trial, have so far evaluated 15 different treatments, mostly combination strategies, in more than 220 children.
Susanne Gatz, MD, PhD
Birgit Geoerger, MD, PhD
Three trial arms within this platform trial evaluated agents targeted toward the DNA repair pathway in this platform trial: WEE1 and carboplatin (arm C; CT087), PARP and irinotecan (arm D; CT088), and PARP/ATR (arm N presented here2). Arm N is s a first-in-child phase I/II combination study of a PARP inhibitor (olaparib) and an ATR inhibitor (ceralasertib). The phase I component of this trial was reported at AACR 2023. Of the 18 patients enrolled in phase I part in arm N across three dose levels, 100% were enriched for defects in DNA replication and damage repair. Phase II is ongoing.
The combination was well tolerated, and signals of activity were observed. Patients received a median of 3.5 cycles of treatment. There were two confirmed partial responses (one patient with pineoblastoma and one with neuroblastoma). Best response was stable disease in nine patients, including one unconfirmed partial response (in Ewing sarcoma) and three prolonged stable diseases for more than four treatment cycles (neuroblastoma, papillary spinal cord tumor). The remaining seven patients had progressive disease.
“To our knowledge, the combination of PARP inhibitors and ATR inhibitors has not been widely investigated in adult tumor types,” said Dr. Gatz. Patients who had a diagnosis of a relapsed or refractory pediatric malignancy were enrolled. Patients were enrolled to the dose-escalation phase between February 2021 and September 2022, with a data cutoff of January 2023. Of the 18 patients 8 had sarcomas, 5 central nervous system tumors, 4 neuroblastomas, and 1 a carcinoma. The median age at enrollment was 16 years (range, 4–24 years). Most patients had metastatic disease at study entry (89%), and all had progressive disease and prior exposure to chemotherapy. The median number of prior lines of therapy was three.
Sufficient data for the recommended phase II dose were available in the older-age cohort of patients between 12 and 18+; the optimal dose going forward was defined as 80 mg of ceralasertib on days 1 to 14 plus 150 mg of olaparib on days 1 to 28. Both agents were administered twice daily. Identification of the recommended phase II dose in younger patients and pharmacokinetic analysis for olaparib and ceralasertib in all patients are ongoing.
Dr. Gatz noted that none of the patients who experienced clinical benefit in the study had BRCA mutations. Regarding safety, a total of five patients experienced dose-limiting toxicities across three dose levels. Grade 3 or higher thrombocytopenia and neutropenia were the most common dose-limiting toxicities. “The main toxicities were hematologic and gastrointestinal,” she added.
“So far, it is unclear whether the molecular alterations on which the patients were enrolled in this trial are the sole reasons for response,” Dr. Gatz commented. “Further, it may be difficult to identify patterns of response in specific tumor types because of the tumor-agnostic nature of the study. Nevertheless, this study design may offer preliminary indications of signals in specific alterations and tumor types and may provide the basis for future clinical trials.”
Predicting Risk of Distant Recurrence in Endometrial Cancer
A deep learning, risk prediction model accurately identified patients with endometrial cancer at low and high risk of distant recurrence.3 The model—called HECTOR (histopathology-based; endometrial; cancer; tailored treatment and distant recurrence; outcome; risk prediction deep learning model)—outperformed baseline Cox models based on typical prognostic factors such as tumor type, stage, grade, and molecular class.
HECTOR was able to identify 81 patients at low risk of distant recurrence, 176 at intermediate risk, and 96 at high risk of distant recurrence among a data set of 353 patients that was not used to train the model. Subsequently, 3.4%, 15.4%, and 36% of patients categorized by the model as being at low, intermediate, and high risk of recurrence, respectively, experienced a distant recurrence.
“Endometrial cancer is the most common gynecologic cancer, and the primary treatment is surgery,” said Sarah Fremond, a PhD candidate in the Department of Pathology at Leiden University Medical Center in the Netherlands. “Most patients have a good prognosis and do not require any adjuvant treatment. [However], there is a proportion who will develop distant recurrence, and for those, you want to recommend adjuvant chemotherapy. This is the only treatment in the adjuvant setting known to lower the risk of distant recurrence, but it also causes morbidity. We wanted to accurately identify patients at low and high risk of distant recurrence to reduce under- and overtreatment.”
Sarah Fremond
The artificial intelligence model used long-term follow-up data from patients with endometrial cancer who had not previously received adjuvant chemotherapy and were enrolled in clinical trials as well as three cohorts of patients treated at clinics (n = 1,761). HECTOR used a single histopathologic slide from each of the 1,408 individual patients to train and optimize itself. HECTOR is a collaborative effort between the AIRMEC Consortium and the TransPORTEC Consortium.
The data set had a median follow-up of 8.5 years. Patients with stage IV disease and those who had previously received chemotherapy were excluded from the model. Prognostic factors currently used, such as histologic type and grade, stage, and molecular class, are challenging to assess correctly because of variability, limited use of visual information in tumor slides, costs, turnaround time, and interpretation. HECTOR was developed to overcome this limitation.
“HECTOR uses only two types of input data: histopathologic tumor slide and stage. We then integrate the molecular class information of endometrial cancer without the need for molecular testing because we use the histopathologic tumor slide–based molecular class from a previously published model. In the end, HECTOR combines [information from] the histopathologic tumor slide–based molecular class, the pure morphologic information from the tumor slide, and the stage category to predict distant recurrence–free survival. HECTOR can be put into a standard workflow using only a histopathologic tumor slide and stage [without the need for molecular testing],” she said.
DISCLOSURE: Dr. Srour has served as a consultant and speaker for Novartis Pharmaceuticals. Dr. Gatz has served as a consultant to EMB Serono and received a grant from Bayer. Ms. Fremond reported no conflicts of interest.
REFERENCES
1. Srour SA, Kotecha R, Curti B, et al: A phase 1 multicenter study (TRAVERSE) evaluating the safety and efficacy of ALLO-316 following conditioning regimen in patients with advanced or metastatic clear cell renal cell carcinoma. 2023 AACR Annual Meeting. Abstract CT011. Presented April 17, 2023.
2. Gatz S, Harttrampf AC, Brard C, et al: Phase I/II study of the PARP inhibitor olaparib and ATR inhibitor ceralasertib in children with advanced malignancies. 2023 AACR Annual Meeting. Abstract CT019. Presented April 17, 2023.
3. Fremond S, Adani S, Wolf JB, et al: Deep learning risk prediction model of distant recurrence from H&E endometrial cancer slides. 2023 AACR Annual Meeting. Abstract 5695. Presented April 18, 2023.
