Immunotherapy of Relapsed and Refractory Solid Tumors With Ex Vivo Expanded Multi-Tumor Associated Antigen Specific Cytotoxic T Lymphocytes: A Phase I Study

Abstract

Purpose: Tumor-associated antigen cytotoxic T cells (TAA-Ts) represent a new, potentially effective and nontoxic therapeutic approach for patients with relapsed or refractory solid tumors. In this first-in-human trial, we investigated the safety of administering TAA-Ts that target Wilms tumor gene 1, preferentially expressed antigen of melanoma, and survivin to patients with relapsed/refractory solid tumors.

Materials and methods: TAA-T products were generated from autologous peripheral blood and infused over three dose levels: 1, 2, and 4 × 107 cells/m2. Patients were eligible for up to eight infusions administered 4 to 7 weeks apart. We assessed dose limiting toxicity during the first 45 days after infusion. Disease response was determined within the context of a phase I trial.

Results: There were no dose-limiting toxicities. Of 15 evaluable patients, 11 (73%) with stable disease or better at day 45 postinfusion were defined as responders. Six responders remain without progression at a median of 13.9 months (range, 4.1 to 19.9 months) after initial TAA-Ts. Patients who were treated at the highest dose level showed the best clinical outcomes, with a 6-month progression-free survival of 73% after TAA-T infusion compared with a 38% 6-month progression-free survival with prior therapy. Antigen spreading and a reduction in circulating tumor-associated antigens using digital droplet polymerase chain reaction was observed in patients after TAA-T infusion.

Conclusion: TAA-Ts safely induced disease stabilization, prolonged time to progression, and were associated with antigen spreading and a reduction in circulating tumor-associated antigen DNA levels in patients with relapsed/refractory solid tumors without lymphodepleting chemotherapy before infusion. TAA-Ts are a promising new treatment approach for patients with solid tumors.

Trial registration: ClinicalTrials.gov NCT02789228.

Conflict of interest statement

The contents of this work are solely the responsibility of the authors and do not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health.

Figures

FIG 1.

Treatment summary. Multimodality therapy administered before tumor-associated antigen cytotoxic T cell (TAA-T) infusion. Patients experienced relapsed disease after completion of therapy as well as disease progression while on treatment. (*) Targeted therapy includes the following: denosumab (patient 1 [P1]), dinutuximab (P2 and P3), radiolabeled 131I-MIBG (P2 and P3), lorvotuzumab (P2 and P4). ES, Ewing sarcoma; 131I-MIBG, 131I-meta-iodobenzylguanidine; NB, neuroblastoma; OS, osteosarcoma; SD, stable disease; PD, progressive disease; RMS, rhabdomyosarcoma; STS, soft tissue sarcoma; WT, Wilms tumor.

FIG 2.

Characterization of tumor-associated antigen cytotoxic T cell (TAA-T) products. (A) Flow cytometry demonstrates a variable phenotype of polyclonal, polyfunctional T-cell products in patients in the Responding group. (B) Patients in the Nonresponding group showed a comparatively lower percentage of CD8+CD3+ cells and CD16+CD56+CD3+ cells with high percentages of CD4+CD3+ cells. (C) Luminex assay to measure cytokine secretion by TAA-T products. Interferon gamma (IFNγ), tumor necrosis factor α (TNFα), and MIP1b were the cytokines most commonly detected in response to antigen stimulation. (D) Product TAA specificity as measured by IFNγ enzyme-linked immunospot assay. Number on the x-axis corresponds to patient number and, when applicable, multiple products are numbered accordingly (eg, T4, T4.2, and T4.3 are the first, second, and third products administered to P4). TAA-T products demonstrated variable specificity to the targeted antigens. GM-CSF, granulocyte-macrophage colony-stimulating factor; IL, interleukin; MIP-1b, macrophage inflammatory protein 1β; PRAME, preferentially expressed antigen of melanoma; WT1, Wilms tumor 1.

FIG 3.

Disease response. (A) Outcome for evaluable patients who received at least one tumor-associated antigen cytotoxic T cell (TAA-T) infusion. Many patients were able to receive multiple TAA-T infusions without adverse reactions. Eleven of 15 patients met criteria for response, which was defined as stable disease or better at the day 45 evaluation. (B) Median time to progression for patients enrolled in dose level (DL) 1 and 2 (n = 7) was 2.8 months compared with 9.3 months for patients enrolled in DL 3 (n = 8; P = .034). (C) Progression-free survival of patients after TAA-T therapy treated at the highest DL was 73% at 6 months and 58% at 12 months compared with their immediate prior therapy regimen (38% and 25%, respectively; P = .73). ES, Ewing sarcoma; NB, neuroblastoma; OS, osteosarcoma; P, patient; RMS, rhabdomyosarcoma; STS, soft tissue sarcoma; WT, Wilms tumor.

FIG 4.

Antigen spreading and T-cell persistence postinfusion in responders. (A) Interferon gamma (IFNγ) enzyme-linked immunospot assay was used to evaluate antitumor immunity to the targeted antigens (Wilms tumor 1, preferentially expressed antigen of melanoma, survivin), as well as four nontargeted antigens commonly identified in solid tumors (MAGE A3, MAGE A4, SOX-2, SSX-2). Ten of 11 responders demonstrated evidence of antigen spreading while receiving tumor-associated antigen (TAAs) cytotoxic T cell infusions. Patient 1 (P1) did not show increased specificity for targeted or nontargeted antigens until after disease progression at week 12. (B) Increased specificity to targeted and nontargeted antigens correlated with the expansion of unique T-cell receptor (TCR) clonotypes as detected by TCR sequencing in P10, a responder.

FIG 5.

Patient tumor samples express targeted antigens by immunofluorescence. Tumor sample from patient 14 (P14) demonstrates strong expression of Wilms tumor 1 (WT1; +++; top left), preferentially expressed antigen of melanoma (PRAME; ++; top middle), and survivin (+++; top right). Immunofluorescent staining of tumor samples was graded qualitatively for available patient tumor samples (bottom left); qualitative grading system using survivin as an example (bottom right). (*) Tumor sample was obtained from a calcified pulmonary nodule. NR, nonresponder; R, responder.

FIG A1.

Patient inclusion/exclusion criteria. ANC, absolute neutrophil count; ATG, antithymocyte globulin; G-CSF, granulocyte colony-stimulating factor; GVHD, graft-versus-host disease; LVEF, left ventricular ejection fraction; LVSF, left ventricular shortening fraction; TAAs; tumor-associated antigens; TBI, total-body irradiation.

FIG A2.

Participant flow diagram. Patients with high-risk solid tumors were eligible for enrollment. Eighteen patients were enrolled and 15 were infused with tumor-associated antigen cytotoxic T cells (TAA-T) at the time of manuscript. One patient was removed from treatment before day 45 as a result of disease progression. Fourteen patients remained evaluable for toxicity.

FIG A3.

Tumor-associated antigen cytotoxic T-cell (TAA-T) products express low levels of exhaustion markers. Exhaustion markers T-cell immunoglobulin and mucin domain–containing-3 (TIM3), programmed cell death 1 (PD1), and cytotoxic T lymphocyte antigen 4 (CTLA4) were at uniformly low levels in TAA-T products as detected by flow cytometry. Lymphocyte-associated gene 3 (LAG3) was higher in products generated for nonresponders compared with responders.

FIG A4.

Circulating cytokines in patients receiving tumor-associated antigen cytotoxic T cells (TAA-Ts). Inflammatory cytokines remained low in circulation after TAA-T infusion. Interleukin-8 (IL-8) did decrease in patients P4, P5, and P6 after TAA-T infusion. P5 and P6 had subsequent increases in IL-8 that correlated with clinical disease progression. G-CSF, granulocyte colony-stimulating factor; IFNγ, interferon gamma; MCP-1, monocyte chemoattractant protein 1; MIP-1b, macrophage inflammatory protein 1β; TNFα, tumor necrosis factor α.

FIG A5.

Representative imaging in patients receiving tumor-associated antigen cytotoxic T cells (TAA-Ts). (A) Computed tomography (CT) chest scan shows disease in patient 1 (P1) before TAA-T (pulmonary nodule on top panel; left) and at the time of progression after cycle 2 with new disease in the anterior and posterior right lobe (right). (B) 123I-meta-iodobenzylguanidine (123I-MIBG) imaging for P3 before cells (left) and at progression after cycle 1 (right) with new avid right posterior skull and vertebral lesions. (C) Positron emission tomography imaging from P8 before TAA-T with multifocal disease (left femur, precariat lymph node, spine, humerus), which remained stable through four infusions before eventual progression. (D) CT scan from P9 pre–TAA-T demonstrates calcified hilar lesion, which remained stable through four infusions and in follow up after therapy.

FIG A6.

Digital droplet polymerase chain reaction (ddPCR) results in responding versus nonresponding patients. (A) DNA from targeted antigens Wilms tumor 1 (WT1), preferentially expressed antigen of melanoma (PRAME), and survivin identified by ddPCR in responding patient P9 at baseline and post–TAA-T (tumor-associated antigen cytotoxic T cell) infusion. Threshold levels were determined using the median of healthy individuals for WT1 (left), PRAME (middle), and survivin (right). (B) Circulating TAA DNA levels in P9 and P10 compared with threshold levels. Patients remain clinically well without evidence of disease progression. (C) TAA DNA were measured in two nonresponding patients and compared with healthy controls. Both had elevated levels of at least one TAA at the time of disease progression. P13 had increases in DNA of all three TAAs, which correlated with significant clinical disease progression. BCR/ABL, breakpoint cluster region-Abelson.