Persistence of CTL Clones Targeting Melanocyte Differentiation Antigens Was Insufficient to Mediate Significant Melanoma Regression in Humans

Cancer regression in patients with metastatic melanoma can now be achieved with three mechanistically distinct types of immunotherapies that augment naturally existing antitumor T-cell responses: (i) systemic cytokine therapy (1, 2); (ii) checkpoint inhibition (3–6); and (iii) adoptive transfer of autologous tumor-infiltrating lymphocytes (TIL; refs. 7–9). These clinical findings have drawn attention to the significant therapeutic potential of exploiting endogenous T-cell populations for cancer therapy. However, efforts to improve current immunotherapies are hindered by a limited understanding of the specific lymphocyte populations that were responsible for the observed tumor responses. Furthermore, the tumor antigens associated with durable and complete cancer regression remain unclear, thus hindering the development of targeted immunotherapeutics. We hypothesized that a clinical strategy involving the iterative adoptive transfer of highly selected autologous antigen-specific T-cell clones could help systematically define immunologic targets associated with successful cancer therapy, without the interpretative ambiguity of transferring polyclonal T-cell populations. In this approach, T-cell clones could be selected ex vivo based on high avidity recognition of specific tumor antigen epitopes, expanded to large numbers, and reintroduced into the autologous host after a lymphodepleting preparative regimen to eliminate regulatory cells and augment homeostatic expansion.

Here, we report two sequential phase II clinical trials for patients with refractory metastatic melanoma in which the class of melanocyte differentiation antigens (MDA) was targeted with highly avid CD8⁺ T-cell clones specific for either gp100 or MART-1, respectively. The targeting of these MDAs, which are expressed in both normal melanocytes and melanoma tumors, was prompted by the significant natural immunogenicity of these proteins as evident by the high frequency of primed MDA-specific CD8⁺ T cells found within the TIL of melanoma metastases (9–12). Furthermore, there has been a long observed association between the development of vitiligo and uveitis due to melanocyte destruction and melanoma tumor regression in patients undergoing immunotherapy (13–17).

We previously reported a proof-of-concept experience isolating MDA-specific CD8⁺ T-cell clones from peripheral blood using high-throughput in vitro sensitization that enabled rapid clone isolation for clinical therapy (18, 19). In the initial 5 patients, we found that MDA-specific effector clones could target skin melanocytes in an autoimmune fashion, persist long term in peripheral blood, and undergo self-renewal to repopulate the memory pool after adoptive transfer. We now update this experience with the clinical results from 15 patients with metastatic melanoma treated with MDA-specific CD8⁺ T-cell clones. Our findings suggest that despite successful clonal repopulation with autologous MDA-specific CD8⁺ T cells, the targeting of MDAs was insufficient to mediate meaningful cancer regression in patients with metastatic melanoma. These findings raise significant concerns regarding future immunotherapy efforts directed against MDAs.

HLA-A2⁺ patients with metastatic melanoma were treated with either gp100-specific CD8⁺ T-cell clones (n = 10) or MART-1–specific CD8⁺ T-cell clones (n = 5) at the Surgery Branch, National Cancer Institute (NCI; Bethesda, MD), between January 2009 and January 2013 on two consecutive phase II clinical protocols (NCT00665470 and NCT01495572) approved by the Institutional Review Board (IRB) and U.S. Food and Drug Administration. All patients gave informed consent for treatment in accordance with the Declaration of Helsinki. The patients were required to be 18 years of age or older and have measurable metastatic melanoma that expressed gp100 or MART-1 and major histocompatibility complex (MHC) class I by immunohistochemistry. Before clone infusion, patients were transiently lymphoablated with a nonmyeloablative lymphodepleting regimen including intravenous administration of cyclophosphamide (60 mg/kg) for 2 days followed by fludarabine (25 mg/m²) for 5 days as previously described (9). One day after completion of their lymphodepleting regimen, patients received expanded CD8⁺ T-cell clones intravenously, either with or without high-dose interleukin-2 (IL2; 720,000 IU/kg) every 8 hours to tolerance. Patients received baseline computed tomography (CT) and/or magnetic resonance imaging (MRI) and/or positron emission tomography (PET) before treatment. Tumor size was evaluated monthly for 3 months and at regular intervals thereafter by CT, MRI, or documented with photography for cutaneous/subcutaneous lesion. Tumor measurements and patient responses were determined according to Response Evaluation Criteria in Solid Tumor (RECIST).

Human cultured cell lines, including T2 cells (HLA-A2⁺ peptide transporter–associated protein deficient T-B hybrid) and melanoma tumor lines, 526 mel (HLA-A2⁺/MART-1⁺gp100⁺), 624 mel (HLA-A2⁺/MART-1⁺gp100⁺), 888 mel (HLA-A2⁻/MART-1⁺gp100⁺), 938 mel (HLA-A2⁻/MART-1⁺gp100⁺), were routinely cultured in complete medium (CM) as previously described (19). T2 cells and the melanoma cell lines, 526mel, 624mel, 938mel, and 888mel, were obtained from the cell production facility in the Surgery Branch, NCI. The tumor cells had been characterized to confirm tumor morphology, antigen, and HLA expression by immunohistochemistry; they were obtained and used within 6 months of testing. Human peripheral blood mononuclear cells (PBMC) used in this study were obtained by leukapheresis from patients with HLA-A2⁺ metastatic melanoma evaluated on IRB-approved protocols at the Surgery Branch, NCI (NIH, Bethesda, MD). Human PBMC and CD8⁺ T-cell clones were cultured in CM with 10% heat-inactivated human AB serum (Gemini Bio-Products).

PBMCs from patients with HLA-A2⁺ melanoma underwent depletion of CD4⁺ lymphocytes by magnetic bead separation (Miltenyi Biotec) and were plated as individual microcultures in 96-well flat-bottomed plates at approximately 1 × 10⁵ cells per well. The cells underwent in vitro sensitization for 10 to 14 days in the presence of 1 μg/mL of either gp100_154–162 (KTWGQYWQV) or the modified 10-mer MART-1_26–35(27L) (ELAGIGILTV) GMP grade peptide (Multiple Peptide Systems) and IL2 (90 IU/mL) as previously described (19). Individual microcultures that exhibited specific peptide reactivity either by a high-throughput qPCR cytokine screening or a high-throughput flow cytometry tetramer screening assay were selected for further expansion. To derive antigen-specific CD8⁺ T-cell clones, limiting dilution was typically performed by plating between 1 and 3 T cells in each well of a 96-well U-bottomed plate in 0.2 mL of conditioned medium containing anti-CD3 monoclonal antibodies (mAb) Orthoclone OKT3 (50 ng/mL; Ortho-Biotech) and IL2 (300 IU/mL) with 5 × 10⁴ autologous irradiated (40 Gy) PBMCs. On day 5 and every 3 to 4 days thereafter, half of the medium in each well was replaced with fresh medium containing IL2. Growth-positive culture rate was typically approximately 10% to 12%. Characterization of clone function was performed with enzyme-linked immunosorbent assay (ELISA) to quantify IFNγ secretion in response to limiting concentrations of gp100_154–162 or the native MART-1_27–35 (AAGIGILTV) peptide pulsed onto T2 cells and antigen-positive tumor lines. Selected clones were subsequently expanded with anti-CD3 mAb (50 ng/mL), IL2 (300 IU/mL), and 3 × 10⁷ irradiated allogeneic PBMCs in upright 25-cm² flasks for 14 days. Final large-scale expansion of the clones for patient therapy was performed by the Surgery Branch Cell Production Facility using methods previously described (9, 20). The production time for clone generation from in vitro stimulation to the final treatment product was approximately 6 to 7 weeks (18–20).

Confirmation of CD8⁺ T-cell clonality and determination of T-cell receptor (TCR) clonotype was performed by purifying RNA from each clone using Qiagen RNeasy kits. 5′ RACE was performed using BD SmartRace reagents and protocol, using the universal 5′ forward primer, and a 3′ gene-specific reverse primer for the TCR α constant region, or C1 or C2 β constant regions. Results were run on an agarose gel and appropriately sized bands (800–900 bp) were excised, subcloned into pCR2.1 (Invitrogen Life Technologies) vector, and sequenced.

Phycoerythrin-conjugated MART-1_26–35(27L) (ELAGIGILTV) peptide/HLA-A*0201 tetramer complexes were obtained from Immunotech, Beckman Coulter. Phycoerythrin-conjugated gp100_154–162 (KTWGQYWQV) peptide/HLA-A*0201 tetramer complexes were obtained from the NIH Tetramer Facility. Anti-human CD8, CD3, CD45RO, CD62L, and CD95 mAbs were obtained from BD Biosciences. Immunofluorescence, analyzed as the relative log fluorescence of live cells, was measured using a FACSCanto II flow cytometer with FACSDiva software (BD Biosciences) and FlowJo software (TreeStar, Inc.).

Responder cells (1 × 10⁵ T cells) and stimulator cells (1 × 10⁵ peptide pulsed T2 cells or tumor lines) were coincubated in a 0.2-mL volume in individual wells of a 96-well plate. Supernatants were harvested from duplicate wells after 20 to 24 hours and IFNγ secretion was measured in culture supernatants using commercially available IFNγ ELISA kits (Endogen). All data are presented as a mean of duplicate samples. Cultures with IFNγ production greater than 100 pg/mL and twice background were considered as having specific antigen reactivity.

Antigen-specific T cells were cocultured with T2 cells pulsed with the cognate peptide versus a control peptide for 2 hours. Brefeldin A (eBioscience) was added to the coculture for another 4 hours. The cells were then fixed in 2% paraformaldehyde, permeabilized, and stained with anti-CD3 and anti-CD8 (BD Biosciences) along with anti-IFNγ-PE-Cy7 and anti-IL2-APC (eBioscience). Cytokine staining was assessed on CD3⁺CD8⁺-gated cells.

HLA-matched and mismatched target tumor lines were loaded with 15 μmol/L calcein-AM (Invitrogen) for 30 minutes at 37°C, washed three times, then plated at 5,000 cells per well/100 μL CM in a 96-well round-bottomed plate. An equal volume of effectors was added at different concentrations to target cells (E:T ratios) as indicated in the experiment. After a 4-hour coculture, supernatants were harvested and free calcein was quantitated using Glomax UV detection system (Promega). The percentage of specific cytotoxicity was calculated as (experimental release − spontaneous release)/(maximum release − spontaneous release) × 100. Spontaneous release was determined by incubating the targets with 100 μL of CM instead of effector cells, and maximum release was determined by incubating the targets with 100 μL of 0.5% Triton-X. All data are presented as the mean + SEM of triplicate samples.

To evaluate in vivo persistence of T cells in the peripheral blood, PBMCs were prepared from samples drawn on day 0 (pre infusion) and at day 30 (postinfusion) and cryopreserved so that all samples could be analyzed simultaneously. PBMCs (1–2 × 10⁶) were stained for 30 minutes at 4°C to 9°C with peptide–MHC tetramer-PE, anti-CD8-APC, and the cell viability dye, propidium iodide, to exclude dead cells, and analyzed by flow cytometry. The degree of persistence of transferred clones is presented as the frequency of tetramer-positive, CD8⁺ lymphocytes over the total number of CD8⁺ cells. To determine long-term engraftment, we analyzed PBMC samples by flow cytometry over an extended period of time (days 60, 90, 120, and 150) and derived the absolute number of infused clones (CD8⁺TET⁺Vβ⁺) per microliter of blood.

A total of 15 HLA-A*0201⁺ patients with refractory metastatic melanoma, including 5 patients described in a previous report (19), were enrolled upon two consecutive clinical trials (NCT00665470 and NCT01495572) in which they received ex vivo–expanded MDA-specific CD8⁺ T-cell clones in conjunction with a nonmyeloablative lymphodepleting conditioning regimen. The protocol was designed to evaluate the persistence, safety, and therapeutic efficacy of MDA-reactive clones. The characteristics of the patients and their cell therapy are shown in Table 1. At the time of enrollment, all patients had demonstrated progression of their metastatic disease after prior systemic therapy. The first 10 patients were treated with CD8⁺ T-cell clones specific for the gp100_154–162 epitope and the next 5 patients received CD8⁺ T cells specific for the MART_27–35 epitope. The TCR clonotype for each clone was defined by complete molecular sequencing of the β-chain variable region (Vβ). Each patient received a single clonotype except for patient M2 who received two unique clonotypes. The mean number of infused CD8⁺ T-cell clones was 15.8 × 10⁹ (range, 0.1–58.3 × 10⁹). Patients, who were medically eligible, received concomitant high-dose IL2 infusions that were administered to tolerance. All patients in the MART-1 cohort presented with significant clinical comorbidities and were therefore ineligible for IL2 administration.

The functional and phenotypic attributes of the isolated MDA-specific CD8⁺ T-cell clones were assessed immediately before infusion. Each of the gp100- and MART-specific CD8⁺ T-cell clones demonstrated highly specific and avid antigen recognition by secreting significant amounts of IFNγ in response to 100 ng/mL cognate peptide pulsed on T2 target cells (Fig. 1A) and naturally presented peptide on allogeneic HLA-A2⁺ melanoma tumor lines (Fig. 1B). Analysis of the clone reactivity by intracellular FACS for IL2 and IFNγ production after antigen stimulation revealed an effector cytokine profile with 77% ± 3% of the reactive cells producing only IFNγ, 22% ± 4% of cells producing both IFNγ and IL2, and an insignificant population of cells producing only IL2 (1% ± 1%; Fig. 2A). Furthermore, the transferred clones were found to be highly cytolytic and efficiently lysed tumor in an MHC-dependent manner (Fig. 2B). Phenotypic profiling of the clones demonstrated high cell-surface expression of CD45RO and CD95 and minimal expression of the lymph node homing molecule, CD62L, consistent with a differentiated effector status (Fig. 2C). In summary, the adoptively transferred cells represented a highly selected homogenous population of lytic and differentiated effector CD8⁺ T cells that specifically targeted MDAs.

Each of the treated patients experienced transient neutropenia and thrombocytopenia induced by the lymphodepleting preparative chemotherapy regimen. The patients who received postinfusion IL2 were additionally noted to have well-described self-limited toxicities associated with systemic cytokine therapy (7). All of the chemotherapy and IL2-related adverse effects were found to be reversible with clinical symptoms and laboratory test values returning to appropriate levels within 2 weeks. With respect to the MDA-specific CD8⁺ T-cell clones, within 7 days of infusion, there was evidence of immune-mediated targeting of skin melanocytes. Of the 15 treated patients, 11 patients (73%) developed a diffuse erythematous skin rash (Fig. 3, Table 1) with skin biopsies revealing CD8⁺ T-cell infiltration into the melanocytic layer consistent with autoimmune dermatitis. These histologic observations, noted at a time when the endogenous lymphocytes had been depleted, strongly suggested that the epidermal melanocytes were the targets of immune attack by the transferred clones. There were no cases of uveitis or ototoxicity detected in any of the patients, indicating that resident melanocytes found in the eyes and inner ears were not targeted. In all cases, the autoimmune dermatitis resolved without treatment after approximately 10 to 14 days, with the exception of patient M2, who developed progressive patchy skin vitiligo, indicating ongoing melanocyte destruction after therapy.

We next sought to evaluate the antitumor effects of the administered clones. Despite evidence of melanocyte targeting in the majority of the patients, none of the treated patients demonstrated an objective tumor response by standard oncologic RECIST criteria (Table 1). We observed mixed and minor biologic activity related to the T-cell clone transfer with radiographic shrinkage of individual tumors in patients GP1, GP5, M1, and M2 (examples shown in Supplementary Fig. S1); however, these findings did not appear to provide meaningful clinical benefit for these patients.

In prior adoptive transfer clinical trials, the ability of the transferred cells to persist was strongly associated with tumor regression in patients with metastatic melanoma (21, 22). Thus, to help understand the lack of objective tumor responses in the current trials, we sought to determine whether the MDA-specific CD8⁺ T-cell clones had successfully engrafted and repopulated the immune repertoire. To evaluate the in vivo survival of the transferred clones, peripheral blood samples, obtained before and 1 month after cell infusion, were compared by FACS for the percentage of CD8⁺ T cells that were tetramer positive (Fig. 4A). Furthermore, each CD8⁺Tetramer⁺ population was FACS sorted to >99% purity to allow TCR molecular sequencing to determine the antigen-specific clonotypes that were present in the peripheral blood before and after clone infusion. The presence of the infused clonotype after infusion, but not before infusion, was defined as clonal persistence. The percentage of this clonotype among total CD8⁺ T cells was used to define the persistence frequency. Using this stringent criterion, we detected engraftment of transferred clones in 8 of the 15 patients (53%) with 1 month clonal persistence ranging from 0.3% to 7.8% of all circulating CD8⁺ T cells (Table 1 and Fig. 4B). To evaluate the long-term fate of the persisting clones, we obtained extended peripheral blood samples from selected patients and demonstrated the sustained presence of circulating clones beyond 100 days (Fig. 4C). We next assessed the ability of the persisting CD8⁺ T cells from all patients to re-respond to antigenic stimulation. Without culturing or addition of exogenous cytokines, peripheral blood samples obtained 1 month after cell infusion were assayed against T2 cells pulsed with the cognate peptide or a control peptide. Intracellular cytokine FACS for IFNγ revealed that the persisting cells from all of the patients were highly reactive against their respective peptide targets (data not shown). Collectively, these findings demonstrated that the MDA-specific CD8⁺ T-cell clones had engrafted and persisted as a functionally active population in the immune repertoire of half of the patients after adoptive transfer.

The adoptive transfer of autologous TIL in conjunction with lymphodepleting conditioning regimens can mediate durable complete tumor regression in selected patients with metastatic melanoma (7, 8, 23, 24). However, efforts to further improve upon these clinical findings are currently hindered by an incomplete understanding of the specific lymphocyte populations that were responsible for the tumor responses. The cellular composition of administered melanoma TIL is polyclonal, varies across individual patients, and remains largely unknown. As such, the tumor antigens that were instrumental in inducing sustained and complete immune responses are also unclear. One strategy to provide clarity to these issues involves the iterative isolation and adoptive transfer of tumor reactive T-cell clones with single antigen specificity. The transfer of cloned lymphocytes would allow a precise determination of the in vivo fate and function of a genetically trackable population of antigen-specific T cells. Here, we report the results from two sequential clinical trials in which MDAs were targeted with autologous CD8⁺ T-cell clones in patients with metastatic melanoma. The decision to target MART-1 and gp100 stemmed from a number of observations that suggested that these tumor antigens represented favorable therapeutic targets to treat melanoma. First, these antigens have been reported to be commonly and highly expressed in metastases among many patients with melanoma (25, 26). Second, the MDAs are highly immunogenic. Since the original identification of HLA-A2 restricted MART-1 and gp100 epitopes recognized by naturally occurring TIL (10, 11), high frequencies of primed CD8⁺ T cells specific for these antigens have been routinely isolated from peripheral blood and tumor derived lymphocyte populations (11, 27, 28). Finally, there has been a long observed and intriguing association between the development of autoimmunity against normal melanocytes (for example, uveitis and vitiligo) and melanoma tumor regression in patients treated with immune therapies (13–17). Collectively, these findings prompted our prospective evaluation of the effectiveness of MART-1 and gp100 as tumor regression antigens.

We previously reported a high-throughput technique that allowed the rapid isolation and expansion of high-avidity MDA-specific CD8⁺ T-cell clones from peripheral blood for use in adoptive transfer clinical studies (18, 19). The current report further demonstrates that these clones could be routinely isolated and that they possess potent and specific in vitro lytic capability against MDA-expressing melanoma tumor lines. In vivo evidence of clone activity after adoptive transfer was seen in the targeting of normal melanocytes residing in the skin of the majority of treated patients. Furthermore, these clones were observed to engraft in over half of the treated patients and survive long term in the circulating immune repertoire. Despite these findings, we did not observe clinically significant tumor regression. The precise explanation for these seemingly paradoxical findings still remains unclear.

One potential deficiency of our treatment may reside in the intrinsic nature of the infused cell product. The clones generated in this study were derived from the peripheral blood and as such, these cells may lack necessary tumor trafficking and homing abilities that are present in tumor-derived lymphocytes, such as TIL. Furthermore, although our CTL clones were derived using a rapid cloning approach, the cells still underwent massive in vitro proliferation and differentiation. Murine studies have suggested that CD8⁺ T cells that have undergone such extensive in vitro expansion progressively lose in vivo proliferative potential and rapidly undergo apoptosis after adoptive transfer (29). However, this theory is difficult to reconcile with the observation that the clones in the current study could persist for very long periods in the host (Fig. 4C), suggesting that cell survival was not a limiting factor.

Another possibility is that the degree of clonal persistence noted in our studies may not have been sufficient to mount a sustained antitumor response. The persistence in our treated patients was heterogeneous and ranged from 0.3% to 7.8% of the circulating CD8⁺ T-cell population (Table 1). Previous clinical trials studying the adoptive transfer of autologous TIL have reported a significant and positive correlation between tumor response and the capability of the transferred cells to persist. However, the magnitude of persistence was not examined as a correlate of clinical response. In an effort to determine whether a threshold of persistence is critical for tumor regression, we are currently evaluating approaches to enhance in vivo persistence after cell transfer in preclinical murine models. Variability in the administration of high-dose IL2 might have also influenced the level of persistence or the activation state of the transferred clones. Unfortunately, our study was not designed to address the impact of IL2 dosing on clonal persistence and activity.

Finally, perhaps the most compelling hypothesis for the results of the current clinical trials is that MDAs are suboptimal tumor regression antigens. In fact, a comprehensive review of published adoptive transfer studies targeting the MART-1 and gp100 antigens in patients with metastatic melanoma over the last decade has, similarly, found poor therapeutic efficacy using either MDA-specific clones (28, 30–35), MDA-enriched polyclonal bulk infusions (36–38), or MDA TCR gene-modified PBL (Table 2; refs. 17, 39, 40). Although these trials were conducted by various groups and had differences in their clinical design, the characteristics of the infused cell product, and the use of conditioning regimens and cytokines, they shared the exclusive targeting of either MART or gp100. Our review found few, if any, objective clinical responses by standardized oncologic criteria. The most potent in vivo targeting of MDAs was evident in the trials transferring peripheral blood lymphocytes that were genetically engineered to express high-affinity TCRs against MART and gp100 (17). The transfer of these TCR-transduced cells was associated with severe on target-off tumor toxicity in normal tissues in the eye, inner ear, and skin which all harbor populations of normal melanocytes. Although objective tumor responses in patients were demonstrated in these trials, the tumor shrinkage was typically partial and transient in nature. In reviewing the collective experience of published clinical trials targeting MDAs, we found a very low complete tumor response rate of 3.7%, which usually involved regression of small volume lymph node metastases. This complete response rate is significantly lower than the 22% 5-year complete response rate associated with bulk polyclonal TIL therapy in patients with advanced melanoma (9). Thus, the cumulative findings of these published reports, in concert with our current study, raise significant concerns regarding future immunotherapy efforts targeting this class of tumor antigens.

One major biologic limitation of targeting MDAs may be the nonessential role that these proteins play in malignant melanoma transformation and the metastatic phenotype. We recently reported the profiling of over 3,000 biopsies of metastases from 1,514 patients with melanoma using quantitative immunohistochemistry to detect gp100, MART-1, and tyrosinase expression (41). We observed significantly low expression or complete loss of expression of each of these MDAs in approximately 30% of lesions. Furthermore, within metastases with detectable antigen expression, there was significant heterogeneity among the tumor cells within the same metastases. This heterogeneity in MDA antigen expression may explain the partial, transient, and mixed responses observed when MDAs are immunologically targeted.

In an attempt to retrospectively define potentially more immunogenic targets responsible for the durable remissions with TIL transfer, the source tumors from these patients have recently undergone whole-exomic sequencing to identify tumor-specific nonsynonymous mutations. Candidate epitopes containing these mutated amino acids were screened for HLA-class I binding motifs and those predicted by in silico algorithm to bind with high affinity to the patient's pertinent class I alleles were synthesized. In all cases so far, TIL were identified that immunologically recognized one or more of these mutant epitopes (42). These findings provide compelling evidence that somatic mutations in metastatic cutaneous melanoma tumors can generate neo-epitopes that can elicit a robust autologous immunologic response against mutant proteins. The search for more suitable antigenic targets has directed our current efforts toward identifying T cells that target neo-epitopes generated by cancer-specific somatic mutations and, in particular, driver mutations. Classic driver mutations encode for proteins that typically contribute to tumor development and maintenance of the malignant phenotype. Unlike MDAs, the neo-proteins encoded by these mutations may serve as more effective therapeutic targets given their tumor specificity and their essential role in the survival and proliferation of metastatic cells.

No potential conflicts of interest were disclosed.

Conception and design: S.S. Chandran, R.M. Sherry, U.S. Kammula

Development of methodology: S.S. Chandran, B.C. Paria, A.K. Srivastava, M.E. Dudley, U.S. Kammula

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.S. Chandran, B.C. Paria, A.K. Srivastava, L.D. Rothermel, D.J. Stephens, J.R. Wunderlich, J.C. Yang, U.S. Kammula

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S.S. Chandran, B.C. Paria, A.K. Srivastava, L.D. Rothermel, D.J. Stephens, J.C. Yang, S.A. Rosenberg, U.S. Kammula

Writing, review, and/or revision of the manuscript: S.S. Chandran, B.C. Paria, A.K. Srivastava, D.J. Stephens, M.E. Dudley, R.M. Sherry, J.C. Yang, S.A. Rosenberg, U.S. Kammula

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B.C. Paria, D.J. Stephens, R. Somerville, S.A. Rosenberg, U.S. Kammula

Study supervision: S.A. Rosenberg, U.S. Kammula

The authors thank the Surgery Branch cell production facility and the immunotherapy clinical and support staff for their contributions. The authors also thank Arnold Mixon and Shawn Farid for assistance with cell sorting by flow cytometry.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.