Precision Oncology in Melanoma and Skin Cancer Surgery

There has been perhaps no greater advance in the prognosis of solid tumors in the last decade than for patients with metastatic melanoma. This is due to significant improvements in treatment based on two key components of melanoma tumor biology (1) the identification of driver mutations with therapeutic potential and (2) the mechanistic understanding of a tumor-specific immune response. With breakthrough findings in such a relatively short period of time, the treatment of patients with metastatic melanoma has become intensely personalized.

Neoadjuvant Simultaneous Integrated Boost Radiation Therapy Improves Clinical Outcomes for Retroperitoneal Sarcoma

PURPOSE

Neoadjuvant radiation therapy (RT) with standard techniques (ST) offers a modest benefit in retroperitoneal sarcoma (RPS). As the high-risk region (HRR) at risk for a positive surgical margin and recurrence is posterior and away from radiosensitive organs at risk, using a simultaneous integrated boost (SIB) allows targeted dose escalation to the HRR while sparing these organs. We hypothesized that neoadjuvant SIB RT can improve disease control compared with ST, without increasing toxicity.

METHODS AND MATERIALS

We retrospectively identified patients with resectable nonmetastatic RPS from 2000 to 2021 who received neoadjuvant RT of 180 to 200 cGy/fraction to standard volumes. SIB patients received 205 to 230 cGy/fraction to the appropriate HRR. Clinical endpoints included abdominopelvic control (APC), recurrence-free survival (RFS), overall survival (OS), and acute toxicity.

RESULTS

With a median follow-up of 57 months (95% confidence interval [CI], 50-64), there were 103 patients with RPS who received either ST (n = 69) or SIB (n = 34) RT. Median standard volume dose was 5000 cGy (ST) and 4500 cGy (SIB), with a median HRR SIB dose of 5750 cGy. Liposarcomas (79% vs 53%; P = .004) and cT4 tumors (59% vs 19%; P < .001) were more common in the SIB cohort, without a significant difference in the rate of resection (82% vs 81%; P = .88) or R1 margin (53.5% vs 50%; P = .36); there were no R2 resections. SIB was associated with a significant improvement in 5-year APC (96% vs 70%; P = .046) and RFS (60.2% vs 36.3%; P = .036), with a nonsignificant OS difference (90.1% vs 67.5%; P = .164). On multivariable analysis, SIB remained a predictor for APC (hazard ratio, 0.07; 95% CI, 0.01-0.74; P = .027) and RFS (hazard ratio, 0.036; 95% CI, 0.13-0.98; P = .045). SIB showed no significant detriment in toxicity, albeit with a lower rate of overall grade 3 acute toxicity (3% vs 22%; P = .023) compared with ST.

CONCLUSIONS

In RPS, dose escalation with neoadjuvant SIB RT may be independently associated with improved APC and RFS, without a detriment in toxicity, compared with ST. With the addition of standard RT having only a modest benefit compared with surgery alone, our study suggests that future prospective studies evaluating for the benefit of SIB RT should be considered.

179 CD8+CD69+ Expanded Tumor Infiltrating Lymphocytes from Soft Tissue Sarcoma Have Increased Tumor-Specific Functional Capacity

Background

Adoptive cell therapy (ACT) utilizing tumor infiltrating lymphocytes (TIL) has demonstrated durable responses in patients with metastatic melanoma and offers potential for other solid tumors. Preclinical experience with expanded TIL from soft tissue sarcoma (STS) demonstrates less frequent tumor-specific reactivity compared to melanoma samples, limiting the potential for efficacy.1 We hypothesized that CD69+ TIL have increased tumor-specific reactivity, which can be manipulated in culture, thereby offering an opportunity to enhance the antitumor effect of this cellular immunotherapy product.

Methods

Patients were enrolled on an IRB-approved protocol and TIL were expanded from fresh surgical specimens. After enzymatic digestion, tumor single cell suspensions were cultured in media containing 10% human serum and IL-2 (6000IU/mL). Expanded TIL were then enriched for CD8+ using magnetic bead isolation and CD69+ by flow cytometry cell sorting (FACS). After co-culture with autologous tumor digest, functional capacity was compared between bulk TIL and enriched TIL by evaluation of IFN-gamma (IFNg) and Granzyme B (GzB) secretion. Capacity for direct tumor cytotoxicity was assessed by Cr51 assay after co-culture of autologous immortalized cell lines with expanded TIL subpopulations after enrichment.

Results

Following co-culture with autologous tumor digest, CD69+ TIL demonstrated increased IFNg secretion compared to CD69- TIL in 6 samples (1.4–4.2x, p<0.05). CD8+ enriched TIL (75% compared to bulk) had higher relative IFNg secretion in both CD69+ and CD69- subsets (4.2 and 5.8x, respectively, p<0.001). Maximal IFNg secretion was seen from TIL that were both CD69+ sorted and CD8+ enriched, demonstrating an synergistic effect (16.3x vs Bulk CD69-, 4.2x vs Bulk CD69+, 2.8x vs CD8 enriched CD69- ; p<0.001). Functional capacity was also assessed by GzB secretion with similar results. CD69+ TIL had increased relative secretion (1.8–2.2x) compared to CD69- TIL (p< 0.01). CD8+ enriched TIL had increased relative GzB secretion in both CD69- and CD69+ sorted fractions (1.4x, 1.2x, respectively, p<.05). CD69+ sorted and CD8+ enriched TIL demonstrated an additive effect (2.6x vs Bulk CD69-, p<0.01; 1.2x vs Bulk CD69+, p<0.05; 1.8x vs CD8 enriched CD69-, p<0.01). CD8+ enriched CD69+ sorted TIL had greater relative cytotoxicity (3x, p<0.05) at 40:1 E:T ratio against autologous tumor cell lines compared to bulk expanded TIL (figure 1).

Abstract 179 Figure 1

Functional capacity of CD69+ TIL is demonstrated by increased secretion of GzB (A) and IFNg (B) after co-culture with autologous tumor digest. CD69+ TIL have greater cytotoxicity against autologous immortalized cell lines compared to bulk TIL at 40:1 E:T ratio (C).

Conclusions

TIL expanded from STS demonstrate greater tumor-specific functional capacity and cytotoxicity after CD8 enrichment and CD69+ FACS compared to bulk expanded TIL. These data validate the strategy to enhance CD8+CD69+ TIL during culture to yield a more efficacious cellular immunotherapy product.

Acknowledgements

This work was funded by NIH K08CA252642

Trial Registration n/a

Reference

1. Mullinax JE, Hall M, Beatty M, Weber AM, Sannasardo Z, Svrdlin T, Hensel J, Bui M, Richards A, Gonzalez RJ, Cox CA, Kelley L, Mulé JJ, Sarnaik AA, Pilon-Thomas S. Expanded Tumor-infiltrating Lymphocytes From Soft Tissue Sarcoma Have Tumor-specific Function. J Immunother 2021 Feb-Mar 01;44(2):63–70.

Ethics Approval

Abstract cites IRB-approved protocol in methods section.

Consent n/a

384 An investigation into the role of CD4+ tumor-infiltrating lymphocytes (TIL) in metastatic melanoma patients with a complete response to adoptive cell therapy

Background

Immunotherapy for cancer has long been focused on the generation of CD8+ cytotoxic T lymphocyte responses, independent of their dynamic CD4+ T cell counterpart. One promising approach, adoptive cell transfer (ACT) of tumor-infiltrating lymphocytes (TIL), has yielded response rates ranging from 28–55%.1–2 Investigation into the role of CD4+ TIL in this setting remains critically underexplored as an opportunity to improve upon these successes.

Methods

Two metastatic melanoma patients (PT1 and PT2) were treated with TIL on a completed clinical trial at Moffitt Cancer Center (NCT01005745). Tumor recognition by TIL was assessed via co-culture with tumor. Whole exome (WES) and RNA Sequencing were performed on cryopreserved tumor sections and mutant peptide-MHC binding was predicted. TIL were stimulated with antigen presenting cells (APCs) loaded with neoantigen-derived 25mer peptides and sorted based on 41BB/OX40 upregulation, followed by functional immunologic assays. TCR sequencing was conducted on patient peripheral blood as well as isolated neoantigen-specific TIL clones to determine persistence in vivo and cognate peptide-MHC targets were determined empirically.

Results

PT1, infused with predominantly CD4+ TIL (88%), achieved a complete response (CR) despite lack of IFNγ detection with conventional in vitro tumor co-culture methods. Infusion product TIL were sorted by upregulation of OX40 and 41BB upon stimulation with APCs loaded with the mutant peptide pool. Neoantigen reactivity arose from a single peptide sequence, which conferred recognition by a CD4+ TIL clone, which comprised 17% of the infusion product and enriched to greater than 80% after sorting via FACS. These CD4+ TIL produced IFNγ, TNFα, and granzyme B in response to peptide-loaded APCs in an HLA-DR dependent manner. TCRβ overlap revealed this CD4+ clone peaked at two weeks post-infusion (40%) and persisted after infusion for at least six weeks. PT2 was infused with highly reactive, primarily CD8+ (88%) TIL and also achieved a CR. Isolated CD4+ TIL were also responsive to tumor antigens in the context of MHC Class II in vitro. Tumor-reactive CD4+ TIL were enriched by IFNγ capture and delayed xenograft growth in vivo (p<0.01). Neoantigen peptides stimulated predominantly CD4+ TIL to upregulate OX40/41BB and produce IFNγ, TNFα, and granzyme B.

Conclusions

Investigation of these case studies demonstrated evidence of CD4+ TIL involvement in complete clinical responses after ACT. Ongoing studies will define the precise role of tumor-reactive CD4+ T cells in the anti-tumor immune response and provide the framework for future investigation into their function and therapeutic efficacy.

Trial Registration

NCT01005745

References

Bailey SR, et al. Human CD26high T cells elicit tumor immunity against multiple malignancies via enhanced migration and persistence. Nat Commun 2017;8(1):1961.

Tran E, et al. Cancer immunotherapy based on mutation-specific CD4+ T cells in a patient with epithelial cancer. Science 2014;344(6184):641–5.

Ethics Approval

Approved by USF IRB approval number Ame5_107905. All participants gave informed consent before taking part.

178 Expansion of Tumor-Infiltrating Lymphocytes and Marrow-Infiltrating Lymphocytes from Pediatric Malignant Solid Tumors

Background

High-risk non-CNS pediatric malignant solid tumors (pMST) have unsatisfactory outcomes, and novel therapies are warranted. Adoptive cellular therapy (ACT) using tumor-infiltrating lymphocytes (TIL) has produced durable responses in melanoma, and improvements in TIL expansion have made ACT-TIL feasible for other solid tumors.1–3 Preclinical mouse models suggest that T-cells from bone marrow (marrow-infiltrating lymphocytes, MIL) have antitumor reactivity offering another source for ACT.4 5 To demonstrate feasibility of ACT in pMST we hypothesized that TIL/MIL can be expanded from these patients.

Methods

Patients ≤21 years old undergoing standard-of-care pMST resection were enrolled on an IRB approved protocol. Fresh tumor (≥1 cm3) was collected and bone marrow (10 mL) was obtained when accessible from standard of care procedures. TIL/MIL were cultured in media containing IL-2 (6000 IU/mL). TIL were expanded from tumor fragment cultures (TFC, >1 mm3) or tumor digest. Select TIL samples were further expanded using a rapid expansion protocol (REP). Phenotype of expanded TIL (CD3, CD4, CD8 and CD56) was evaluated using flow cytometry. IFN- γ secretion, measured by ELISA assay, measured tumor-specific reactivity after co-culture with autologous tumor and TIL.

Results

Twenty samples were obtained between March 2019-May 2021. Two samples were ineligible (final pathology not pMST), leaving 18 samples for analysis. Five marrow samples were collected. TIL were expanded from 14/18 samples (78%) through TFC with median 5.17 x 10^6 cells (range 1.86 x 10^6–3.21 x 10^8). Average phenotype (%) of TFC-TIL were CD3 (63.17), CD4 (21.46), CD8 (46.19) and CD56 (32.68). 9/10 (90%) of samples successfully underwent REP with median 9.35 x 10^7 cells(range 2.49 x 10^7–5.86 x 10^8) final viable TIL and average fold-change 718.6 (median 458.6). Average phenotype (%) of post-REP TIL were CD3 (96.04), CD4 (75.04), CD8 (19.17) and CD56 (0.43). TIL were expanded from TFC of therapy-naïve (8/10, 80%) and pretreated (chemotherapy and checkpoint immunotherapy) samples (5/8, 63%). Seven samples had sufficient tissue to test tumor-specific reactivity; all were non-reactive. MIL pre-REP was expanded from four samples with median 9.55 x 10^6 cells (range 8.00 x 10^5–1.00 x 10^7). Average phenotype of expanded MIL (%) were CD3 (45.17), CD4 (24.46), CD8 (36.15) and CD56 (28.21) (table 1).

Results of TIL and MIL expansion from 18 pMST samples. Abbreviations: Dx: diagnosis, pre-REP: pre-rapid expansion protocol, post-REP: post-rapid expansion protocol, PBMC: peripheral blood mononuclear cells, GNB: ganglioneuroblastoma, WT: Wilms tumor, OS: osteosarcoma, NB: neuroblastoma, IMT: inflammatory myofibroblastic tumor; ASPS: alveolar soft part sarcoma, SS: synovial sarcoma, ERMS: embryonal rhabdomyosarcoma, N: no systemic therapy, C: chemotherapy, I: immunotherapy, DNG: did not grow, N/A: not applicable, NR: non-reactive

Abstract 178 Table 1

Expansion of TIL from pMST

Conclusions

This study demonstrates feasibility of pMST TIL expansion ex vivo. Due to tissue volume constraints inherent in pMST sampling, anti-tumor reactivity testing was not feasible for most patients. Determining optimal strategy for TIL-ACT in pMST will require further investigation regarding techniques for expanding tumor-specific TIL.

Acknowledgements

The authors would like to thank Swim Across America (www.swimacrossamerica.org) and the Ocala Royal Dames (www.ocalaroyaldames.org) for their generous support of this work.

References

Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immunotherapy for human cancer. Science 2015;348(6230):62–68.

Hall M, Mullinax JE, Royster E, et al. Expansion and characterization of tumor-infiltrating lymphocytes from human sarcoma. Journal of Immunotherapy of Cancer 2015;3(Suppl. 2):19.

Mullinax JE, Hall M, Beatty M, et al. Expanded tumor-infiltrating lymphocytes from soft tissue sarcoma have tumor-specific function. J Immunother 2021;44(2):63–70.

Feuerer M, Beckhove P, Bai L, et al. Therapy of human tumors in NOD/SCID mice with patient-derived reactivated memory T cells from bone marrow. Nat Med 2001;7(4):452–458.

Feuerer M, Rocha M, Bai L, et al. Enrichment of memory T cells and other profound immunological changes in the bone marrow from untreated breast cancer patients. Int J Cancer 2001;92(1):96–105.

Ethics Approval

This study was approved by the Johns Hopkins All Children’s Hospital IRB (#IRB00193453). Consent was obtained from the patient or parent, as appropriate for age, prior to participating in this study.

Adrenal biopsy, as a diagnostic method, is associated with decreased overall survival in patients with T1/T2 adrenocortical carcinoma: A propensity score-matched analysis

INTRODUCTION

The standard diagnosis for adrenocortical carcinoma (ACC) is clinical diagnosis (CD) based on radiographic and biochemical studies. Biopsy diagnosis (BD) is seldom required for the suspicion of secondary malignancy. We aim to study the impact of BD in the context of underlying T1/T2 ACC on overall survival (OS) compared with CD.

METHODS

National Cancer Database (NCDB) for endocrine malignancies was utilized. Only patients with non-metastatic ACC, whose method of diagnosis and local disease extension were reported, and received a surgical adrenalectomy with curative intent were included. Patients were divided by disease stage into T1/T2, T3, and T4 groups. A propensity score match was applied to those with T1/T2 disease who received CD versus BD and the Kaplan-Meier method was used to compare OS.

RESULTS

In total, 4000 patients with ACC were reported in the database, 1410 met selection criteria. Eight hundred and thirty patients had T1/T2, 365 had T3, and 162 had T4 ACC. Of patients with T1/T2 ACC, 742 (89.4%) received CD versus 88 (11.6%) with BD. A propensity score was calculated per a multivariable regression model with 79 patients matched from each group. Exact matching was applied for margin status and adjuvant therapies. Kaplan-Meier analysis showed a significant difference in median OS between CD versus BD patients in the matched data set (103.89 ± 15.65 vs. 54.93 ± 8.22 months; p = 0.001). In all comers, patients with T1/T2 ACC and BD had comparable median OS to that of patients with T3 ACC (52.21 ± 9.69 vs. 36.01 ± 3.33 months; p = 0.446).

CONCLUSION

BD in T1/T2 ACC could be associated with disease upstaging and worse OS outcomes.

Abstract 1513: A case study investigation into the role of CD4+ tumor-infiltrating lymphocytes in a metastatic melanoma patient with a complete response to adoptive cell therapy

Immunotherapy for cancer has long been focused on the generation of CD8+ cytotoxic T lymphocyte responses, independent of their dynamic CD4+ T cell counterpart. One promising approach, adoptive cell transfer (ACT) of tumor-infiltrating lymphocytes (TIL), has yielded response rates ranging from 28-55%. Although lasting and complete responses have been achieved, there is substantial opportunity for improvement. Investigation into the role of CD4+ TIL in this setting remains critically underexplored. CD4+ T cells recognize tumor antigen presented on MHC Class II either directly on tumor cells or indirectly through antigen presenting cells (APCs) and are able to elicit potent anti-tumor responses under the appropriate conditions. Here, we present a case study of a metastatic melanoma patient who received adoptive transfer of a predominantly (88%) CD4+ TIL product. This patient demonstrated a complete response (CR) to therapy despite a lack of detection of IFNg in the infusion product in vitro when these TIL were co-cultured with autologous tumor prior to ACT. Tumor recognition was also absent when CD8+ TIL were isolated and stimulated directly with HLA-matched tumor lines, indicating a lack of recognition of shared melanoma antigens presented on MHC Class I. Longitudinal analysis of the peripheral blood of this patient confirmed that the infused CD4+ TIL persisted after therapy for at least six weeks. Whole exome sequencing (WES) performed on the TIL surgical specimen discovered 88 non-synonymous single nucleotide variants (SNVs) as candidate neoantigens. Predicted binding of the resulting mutant peptides to autologous HLA molecules generated a predominantly MHC Class II restricted profile, with 81.8% of variants capable of MHC Class II presentation and greater than half exclusive to MHC Class II only. CD4+ TIL were screened for tumor antigen recognition by upregulation of OX40 and 41BB after stimulation with autologous APCs loaded with mutant peptides. Nearly half (49.2%) of CD4+ TIL responded to tumor-derived peptides. These CD4+ TIL were then sorted into tumor-reactive and non-reactive subsets for further clonal analysis of phenotype and transcriptional profile (scRNASeq) of these T cells in order to characterize the nature of the CD4+ TIL response to tumor antigen. Overall, thorough interrogation of this patient's case study demonstrated evidence of CD4+ TIL involvement in a complete clinical response after ACT. Ongoing studies will define the precise role of tumor-reactive CD4+ T cells in the anti-tumor immune response and provide the framework for future investigation into their function and therapeutic efficacy.

Citation Format: MacLean Hall, Pat Innamarato, Amy Mackay, Holly Branthoover, Alex Alfaro, Carolyn Rich, Allison Richards, Jonathan Hensel, Matthew Beatty, John Mullinax, Amod A. Sarnaik, Shari Pilon-Thomas. A case study investigation into the role of CD4+ tumor-infiltrating lymphocytes in a metastatic melanoma patient with a complete response to adoptive cell therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1513.

Abstract CT056: Durable complete responses to adoptive cell transfer using tumor infiltrating lymphocytes (TIL) in non-small cell lung cancer (NSCLC): A phase I trial

Study Objectives: A hallmark of tumor infiltrating lymphocytes (TIL) in melanoma is the potential for a complete response (CR) which can last for decades. This long-lived effect is attributed to persistence of memory T cells. However, clinical efficacy in metastatic non-small cell lung cancer (mNSCLC) has not previously been reported. We launched a phase I trial with the objective to evaluate the safety and efficacy in mNSCLC after evidence of progression on nivolumab.

Methodology: Full eligibility criteria is described in trial registry (NCT03215810). Patients were required to have mNSCLC with at least 1 RECIST measurable lesion and also 1 lesion safely accessible for excisional biopsy, usually a supraclavicular lymph node or pleural nodule. This histologically confirmed metastasis was resected and TIL cultured. Autologously reactive cultures were pooled and cryopreserved. TIL was successfully expanded in 95% of patients, to a median dose of 103 billion CD3+ cells. Patients received nivolumab for 4 cycles with two serial CT scans. Patients with tumor enlargement or new lesions proceeded to lymphodepletion cyclophosphamide/fludarabine (Cy/Flu), TIL, and attenuated IL-2. After first CT scan, patients resumed nivolumab for up to 11 doses to augment TIL persistence.

Results: Of 32 patients screened, 20 were eligible and enrolled. Median age 54 yrs (range 38 - 75), median PD-L1 proportion score 6%, median estimated TMB 5 mut/MB, 20% EGFR-mutant. The majority had bulky disease with mean sum of target lesion diameters of 7.0 cm. Of 20 enrolled, 13 had evidence of progression on nivolumab and received subsequent Cy/Flu/TIL/IL-2. Two more are progressing on nivolumab and set to receive TIL within next 2 months. Common non-hematologic adverse events (AEs) of the Cy/Flu/TIL/IL2 regimen included hypoalbuminemia (92%), hypophosphatemia (85%), nausea (77%), hyponatremia (69%), diarrhea (62%), the majority of these AEs resolving by Day +10 post-TIL. Initial tumor regression occurred in most patients at their first post-TIL CT scan, with the median best overall change in sum of target lesion diameters of -38% (range +20 to -100). Median time-on-trial post-TIL is 1.4 years. Two patients have achieved durable CRs which are still ongoing almost 1 year post-TIL. One is a never-smoker with EGFREx19 mutation and PD-L1 1%. Three additional patients were able to maintain a clinical remission by local ablative therapy of an isolated new lesion performed 6 to 17 months post-TIL. Persistence of the infused TCR-Vβ clonotypes at post-infusion timepoints was associated with clinical benefit. Neoantigen-specific T cells were detected in the TIL and post-TIL peripheral blood.

Conclusions: TIL has manageable toxicity and capacity to achieve durable remission in mNSCLC after nivolumab treatment. TIL may be a promising option for fit mNSCLC patients.

Citation Format: Ben Creelan, Chao Wang, Jamie Teer, Eric Toloza, John Mullinax, Jiqiang Yao, John Koomen, Sungjune Kim, Alberto Chiappori, James Saller, Leighann Montoya, Ana Marie Landin, Tawee Tanvetyanon, Bin Fang, Zachary Thompson, Xiaoping Yu, Andreas Saltos, Dung-Tsa Chen, Jose Conejo-Garcia, Eric Haura, Scott Antonia. Durable complete responses to adoptive cell transfer using tumor infiltrating lymphocytes (TIL) in non-small cell lung cancer (NSCLC): A phase I trial [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT056.

Tumor Microenvironment Composition and Severe Cytokine Release Syndrome (CRS) Influence Toxicity in Patients with Large B-Cell Lymphoma Treated with Axicabtagene Ciloleucel

PURPOSE

One of the challenges of adoptive T-cell therapy is the development of immune-mediated toxicities including cytokine release syndrome (CRS) and neurotoxicity (NT). We aimed to identify factors that place patients at high risk of severe toxicity or treatment-related death in a cohort of 75 patients with large B-cell lymphoma treated with a standard of care CD19 targeted CAR T-cell product (axicabtagene ciloleucel).

EXPERIMENTAL DESIGN

Serum cytokine and catecholamine levels were measured prior to lymphodepleting chemotherapy, on the day of CAR T infusion and daily thereafter while patients remained hospitalized. Tumor biopsies were taken within 1 month prior to CAR T infusion for evaluation of gene expression.

RESULTS

We identified an association between pretreatment levels of IL6 and life-threatening CRS and NT. Because the risk of toxicity was related to pretreatment factors, we hypothesized that the tumor microenvironment (TME) may influence CAR T-cell toxicity. In pretreatment patient tumor biopsies, gene expression of myeloid markers was associated with higher toxicity.

CONCLUSIONS

These results suggest that a proinflammatory state and an unfavorable TME preemptively put patients at risk for toxicity after CAR T-cell therapy. Tailoring toxicity management strategies to patient risk may reduce morbidity and mortality.

Intercontinental collaborative experience with abdominal, retroperitoneal and pelvic schwannomas

Background

Schwannomas are rare tumours that pose a significant management challenge in the abdomen, retroperitoneum and pelvis. No data are available to inform management strategy.

Methods

A collaborative international cohort study, across specialist sarcoma units, was conducted to include adults presenting between 2000 and 2017 with histopathologically confirmed schwannomas within the abdomen, retroperitoneum or pelvis.

Results

Of 485 patients across 12 centres, 38 (7·8 per cent) were discharged without follow-up, 199 (41·0 per cent) underwent early resection and 248 (51·1 per cent) had radiological monitoring. Of these 248 patients, 96 (38·7 per cent) eventually had surgery, giving an overall resection rate of 60·8 per cent (295 of 485). At baseline, median tumour volume was 90·1 (i.q.r. 26·5–262·0) cm3. The estimated growth rate was 10·5 (95 per cent c.i. 9·4 to 11·6) per cent per year, and was consistent in the short term (within 2 years of diagnosis) and long term (beyond 2 years) (ρ = 0·405, P = 0·021). A decision to operate was more common in symptomatic patients (P &lt; 0·001) and for rapidly growing tumours (growth rate more than 20 per cent per year) (P = 0·025). R0/R1 resection was achieved in 91·6 per cent of patients (263 of 287). Kaplan–Meier long-term recurrence rates after R0/R1 resection were 2·3 and 6·7 per cent at 3 and 5 years respectively.

Conclusion

Specific recommendations include: indications for early surgery, prediction of growth from radiological monitoring, promotion of selective submacroscopic resection and cessation of postoperative imaging surveillance.

Expansion of tumor infiltrating lymphocytes (TIL) from bladder cancer

Advanced bladder cancer patients have limited therapeutic options resulting in a median overall survival (OS) between 12 and 15 months. Adoptive cell therapy (ACT) using tumor infiltrating lymphocytes (TIL) has been used successfully in treating patients with metastatic melanoma, resulting in a median OS of 52 months. In this study, we investigated the feasibility of expanding TIL from the tumors of bladder cancer patients. Primary bladder tumors and lymph node (LN) metastases were collected. Tumor specimens were minced into fragments, placed in individual wells of a 24-well plate, and propagated in high dose IL-2 for four weeks. Expanded TIL were phenotyped by flow cytometry and anti-tumor reactivity was assessed after co-culture with autologous tumor digest and IFN-gamma ELISA. Of the 28 transitional cell bladder or LN tumors collected, 14/20 (70%) primary tumors and all of the LN metastases demonstrated TIL expansion. Expanded TIL were predominantly CD3⁺ (median 63%, range 10-87%) with a median of 30% CD8 + T cells (range 5-70%). TIL secreted IFN-gamma in response to autologous tumor. Addition of agonisitic 4-1BB antibody improved TIL expansion from primary bladder tumors regardless of pre-treatment with chemotherapy. This study establishes the practical first step towards an autologous TIL therapy process for therapeutic testing in patients with bladder cancer.

T cell mediated immunity after combination therapy with intralesional PV-10 and blockade of the PD-1/PD-L1 pathway in a murine melanoma model

Intralesional (IL) injection of Rose Bengal (PV-10) induces regression of injected and uninjected lesions in several murine tumor models. In this study, we investigated the anti-tumor response of combining IL PV-10 with blockade of the PD-1 / PD-L1 pathway and the role of immune cell populations in eliciting this response. To investigate the role of T cell subsets in mediating an immune response, B16 or M05 melanoma-bearing mice received combination therapy as well as CD8+, CD4+, or CD25+ depleting antibodies. Tumor growth was measured. T cells were collected from spleens or tumors, and phenotype, activation markers, and reactivity were measured. Splenocytes from mice treated with combination therapy had increased OVA antigen-specific CD8+ T cells in M05-tumor-bearing mice. Depletion of CD4+ T cells or regulatory T cells (Tregs) in combination with IL PV-10 and anti-PD-1 antibody treatment resulted in an enhanced anti-tumor effect. Treatment with CD8+ depleting antibody abrogated anti-tumor immunity. These results support a clinical study for the safety and anti-tumor immune responses with combination therapy of IL PV-10 and PD-1/PD-L1 blockade.

Expansion of tumor infiltrating lymphocytes (TIL) from bladder cancer

142

Background: Patients with advanced bladder cancer have limited therapeutic options resulting in a median overall survival (OS) between 12 and 15 months. At our institution, adoptive cell therapy (ACT) using tumor infiltrating lymphocytes (TIL) has resulted in a durable median OS of 52 months in patients with metastatic melanoma. Immune-mediated anti-tumor responses have been previously shown in bladder cancer, therefore we investigated the phenotype and function of TIL expanded from bladder tumors to establish feasibility of ACT for the treatment of bladder cancer. Methods: Tumor specimens, including primary bladder tumors and lymph node metastases, were collected from 29 bladder cancer patients having standard of care tumor resection, who also had consented to an IRB-approved protocol for TIL generation. The tissue was minced into fragments, placed in individual wells of a 24-well plate, and propagated in high dose IL-2 for four weeks. TIL were considered expanded if they propagated to fill ≥2 wells. The remaining tumor material was digested into a single cell suspension and frozen. TIL were phenotyped by flow cytometry and assessed for autologous tumor reactivity through co-culture with tumor digest and IFN-gamma ELISA. Results: Transitional cell bladder tumors were cultured from 23 patients, of whom 19 (83%) demonstrated TIL expansion. Microbial contamination precluded TIL growth in six specimens. TIL were cultured from 9/12 (75%) patients with preceding chemotherapy and 10/11 (91%) who were chemotherapy naive. Expanded TIL were predominantly CD3+(median 63%, range 10-87%) with a median of 30% CD8+ T cells (range 5-70%). Eight of 15 tested samples (53%) contained TIL that secreted IFN-gamma in response to autologous tumor. Conclusions: The study establishes the practical first step towards an autologous TIL therapy process for therapeutic testing in patients with bladder cancer. Human bladder cancer tissue can be used to expand tumor-specific TIL in vitro. TIL were also expanded from patients that received chemotherapy prior to tumor resection. Future efforts will explore the ability to further expand bladder TIL cultures to clinically meaningful numbers to develop novel ACT strategies for patients with this diagnosis.

Final report of a pilot trial combining ipilimumab and adoptive cell therapy

147

Background: We previously confirmed the feasibility of adoptive cell therapy (ACT) with tumor infiltrating lymphocytes (TIL) for patients with advanced melanoma. Patients successfully treated had a 38% response rate, but patient attrition due to progression before ACT (21%) resulted in a response rate of only 26% based on intent to treat (IIT) analysis. We hypothesized that combination immunotherapy would decrease attrition due to progression. Here we present the final report of a pilot trial combining ipilimumab (IPI) and ACT. Methods: Thirteen patients with metastatic melanoma were accrued to an IRB-approved trial. All had >1 cm3 of soft tissue/nodal metastases amenable to resection leaving residual disease. Patients received 3 mg/kg of IPI two weeks prior to tumor resection for TIL generation. One week following resection, a second dose of IPI was administered. ACT included pre-conditioning chemotherapy and TIL infusion followed by IL-2. Two additional doses of IPI were given 2 and 5 weeks after ACT. Feasibility was a primary endpoint and considered achieved if ≥ 2 doses of IPI and TIL were infused to at least 60% of all patients. The clinical responses were assessed by RECIST 1.1 criteria at 12 weeks following TIL transfer. Results: All patients received at least 2 doses of IPI, and 12 of the 13 patients (92%) received TIL. One patient (7%) dropped out due to progression before TIL infusion. TIL were expanded from 47.2% of tumor fragments and 58.7% of these fragments generated TIL reactive to autologous tumor. Median number of infused TIL was 6.5e10 (2.29e10-1.04e11) and median 85% were CD8+ (27-99). At 12 weeks following infusion, of the 13 patients enrolled, there were 6 responders (46%). Median progression-free survival was 7.4 months (1.4-42.2). Grade ≥3 immune-related adverse events included colitis (1), uveitis (1), and hypothyroidism (1). Conclusions: IPI combined with ACT with TILs for patients with advanced melanoma is feasible with decreased attrition due to progression and is associated with promising clinical results based upon ITT analysis. This combination approach of TIL cell therapy and co-inhibitory blockade serves as a model for future efforts to improve the efficacy of immunotherapy for patients with cancer. Clinical trial information: NCT01701674.

Abstract B12: Immune checkpoint blockade to improve T cell infiltration and function in colorectal cancer

Adoptive transfer of tumor infiltrating lymphocytes (TIL) has emerged as a promising approach to induce effective anti-tumor immunity and tumor regression in various types of cancers. Current studies suggest that the presence of TIL play a role in patient outcomes and can predict tumor stage, node positivity, and in some reports, disease free and overall survival in patients with colorectal cancer. TIL may be inactivated in patients due to factors in the local tumor microenvironment, such as regulatory T cells, myeloid derived suppressor cells and the expression of a number of co-inhibitory immune checkpoints including PD-L1, BTLA, PD-1, CTLA-4, Lag-3 and Tim-3. We first examined the feasibility of expanding TIL from the tumors of three colorectal patients. Tumors were minced into fragments and cultured in media supplemented with 6000 IU IL2 / mL. TIL growth was observed in tumor fragments from each of the three patients with a total of 12 out of 41 fragments demonstrating TIL expansion. TIL grown out of these fragments were predominantly CD3+ with 32% CD4+T cells and 46% CD8+ T cells and 18.5% NK cells. Seven of the 12 fragments tested were tumor specific as demonstrated by IFN-gamma production in response to autologous tumor. CD8+ T cells within the patient TIL expressed PD-1, BTLA, Lag-3 and Tim-3 and a co-stimulatory molecule, 41BB. In addition, we found high levels of PD-L1 expression on tumors of colorectal patients and human colorectal cell lines. Since we observed high levels of immune checkpoints expression in colorectal patients, we wanted to develop a preclinical colorectal model to investigate the effect of immune checkpoint blockade on tumor growth and TIL function. A previous study from our lab has shown that TIL from our preclinical MC38 colorectal murine model express high levels of co-inhibitory checkpoints and that blockade of PD-L1 leads to enhanced infiltration and anti-tumor specificity of TIL. However, given the dynamic nature of immune responses to tumors and expression of multiple immune checkpoints, it may be necessary to block additional checkpoints to improve clinical efficacy. In this study, we examined whether blockade of the immune checkpoint receptor BTLA had an effect on TIL infiltration and tumor growth. Mice bearing MC-38 tumors treated with anti-BTLA antibody demonstrated a delay in tumor growth (150mm2, p&lt;0.01) on day 28 post tumor injection compared to isotype control (300mm2). We observed an increase in CD8+ T cell infiltration (30%, p&lt;0.001) in the tumors of anti-BTLA antibody treated mice compared to mice treated with isotype control (13%). TIL isolated from the anti-BTLA antibody treated mice demonstrated increased IFN-gamma and granzyme B secretion compared to isotype controls (p&lt;0.001). Dual immune checkpoint blockade with anti-PD-L1 and anti-BTLA antibodies had a synergistic effect in delaying tumor growth. We also measured a significant increase in CD8+CD107a+IFN-gamma+ secreting T cells (37%) compared to TIL from mice treated with anti-PD-L1 (27%) or anti-BTLA antibody alone (25%) (p&lt;0.002). These findings suggest that targeting immune checkpoint molecules can improve T cell infiltration and anti-tumor immunity in a preclinical model of colorectal cancer. In addition, we demonstrate that tumor-specific TIL can be expanded from the tumors of colorectal patients. This study presents a rationale for the potential synergy between blockade of multiple immune checkpoints and adoptive transfer of TIL for the treatment of colorectal cancer.

Citation Format: Krithika N. Kodumudi, James Charles, Carrie Luu, John Mullinax, Julie Li, MacLean Hall, Amy Weber, Amod A. Sarnaik, Shari Pilon-Thomas. Immune checkpoint blockade to improve T cell infiltration and function in colorectal cancer. [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer: From Initiation to Outcomes; 2016 Sep 17-20; Tampa, FL. Philadelphia (PA): AACR; Cancer Res 2017;77(3 Suppl):Abstract nr B12.

Distant recurrence risk with prospective use of the 21-gene assay at a single institution

131

Background: Distant recurrence (DR) is the cause of most breast cancer deaths. The 21-gene assay (ODX) Recurrence Score (RS) result predicts both 5 and 10-yr DR risk and can guide adjuvant chemotherapy (CT) recommendations to mitigate this risk. This study analyzed the use of the RS result to guide adjuvant treatment decisions in a large single-institution, prospective cohort of patients (pts). Methods: This is an IRB-approved review of a prospective database of pts receiving ODX on an initial primary breast cancer. Data collected included demographics, primary operation, margin status, receptor status, RS, adjuvant treatment, recurrence, and survival. Pts were stratified as low risk (RS < 18), intermediate risk, or high risk (RS > 30). The primary analysis computed Kaplan-Meier estimates for rate of DR at 5 yrs when pts were stratified by RS. Results: From 2003 to 2009, a RS result was obtained on 606 pts. Median follow up was 2.9 yrs (0.1-9.7) and median age was 58 yrs (27-84). Median RS result was 16 (0-63); 344(57%) pts were low, 212(35%) intermediate, 50(8%) high. Endocrine therapy was given to 92.4%, 94.3%, and 87.5% low, intermediate, and high risk pts, respectively. Adjuvant CT was given to 8.6%, 47.6%, and 70.8% low, intermediate, and high risk pts, respectively. There were 8 DR events with 1.8% 5-yr estimated risk of DR. The 5-yr estimated risk of DR was 0.7% for low risk (344) pts, 3.4% for intermediate risk (211) pts, and 2.6% for high risk (50) pts. Among node negative [N(-)] pts (502), the 5-yr estimated risk of DR was 0.8% for low risk (287) pts, 3.7% for intermediate risk (174) pts, and 3.3% for high risk (41) pts. Among node positive [(N+)] pts (54) there was only 1 DR, which was in a high risk pt. Of pts with unknown nodal status (50), there were no DRs. Conclusions: The RS result is predictive of DR at 5 yrs as shown in historical datasets (Table). The use of ODX to guide adjuvant treatment recommendations in our contemporary, prospective cohort resulted in a much lower 5-year DR rate and thus supports its use to guide adjuvant treatment decisions. [Table: see text]

Jejunal diverticulosis found in a patient with long-standing pneumoperitoneum and pseudo-obstruction on imaging: a case report

Small bowel diverticulosis is a rare finding within the general population and jejunal diverticulosis, specifically, is even rarer. Clinical manifestations can range from post-prandial pain, constipation and malabsorption to serious complications, such as gastro-intestinal hemorrhage, perforation and acute intestinal obstruction. Here we describe the case of an 81-year-old gentleman who presented with a three-year history of abdominal pain and weight loss. Despite unremarkable physical examination and laboratory tests, persistent pneumoperitoneum and dilated loops of small bowel were found on imaging. Having been given a diagnosis of small bowel bacterial overgrowth, the patient underwent capsule endoscopy study for further evaluation of his small bowel. The capsule did not reach the colon and the patient never noted passing the capsule in his stool so, six months post-procedure, a computed tomography (CT) scan seemed to reveal the retained capsule. Subsequent exploratory laparotomy revealed 200 cm of atonic, dilated jejunum with impressive diverticula along the anti-mesenteric border. This case report is an example of an unusual set of presenting signs and symptoms of jejunal diverticulosis, including persistent pneumoperitoneum, pseudo-obstruction and small bowel bacterial overgrowth. A literature review has revealed that these signs have been present in other cases of jejunal diverticulosis, although the etiology and pathophysiology is not clearly understood.

Surgical shunting versus transjugular intrahepatic portasystemic shunting for bleeding varices resulting from portal hypertension and cirrhosis: a meta-analysis

Surgical shunting was the mainstay in treating portal hypertension for years. Recently, transjugular intrahepatic portasystemic shunting (TIPS) has replaced surgical shunting, first as a "bridge" to transplantation and ultimately as first-line therapy for bleeding varices. This study was undertaken to examine evidence from trials comparing TIPS with surgical shunting to reassess the role of surgery in treating portal hypertension. The National Library of Medicine and the National Institutes of Health were searched for clinical trials comparing surgical shunting with TIPS. Meta-analysis using the fixed effects model was undertaken with end points of 30-day and 1- and 2-year survival and shunt failure (inability to complete shunt, irreversible shunt occlusion, major rehemorrhage, unanticipated liver transplantation, death). Three prospective randomized trials and one retrospective case-controlled study were identified. Analysis was limited to patients of Child Classes A or B. Significantly better 2-year survival (OR 2.5 [1.2-5.2]) and significantly less frequent shunt failure (OR 0.3 [0.1-0.9]) were seen in patients undergoing surgical shunting compared with TIPS. Meta-analysis promotes surgical shunting relative to TIPS because of improved survival and less frequent shunt failure. Surgical shunting should be accepted as first-line therapy for bleeding varices resulting from portal hypertension.

Surgical Shunting versus Transjugular Intrahepatic Portasystemic Shunting for Bleeding Varices Resulting from Portal Hypertension and Cirrhosis: A Meta-Analysis

Surgical shunting was the mainstay in treating portal hypertension for years. Recently, trans-jugular intrahepatic portasystemic shunting (TIPS) has replaced surgical shunting, first as a “bridge” to transplantation and ultimately as first-line therapy for bleeding varices. This study was undertaken to examine evidence from trials comparing TIPS with surgical shunting to reassess the role of surgery in treating portal hypertension. The National Library of Medicine and the National Institutes of Health were searched for clinical trials comparing surgical shunting with TIPS. Meta-analysis using the fixed effects model was undertaken with end points of 30-day and 1- and 2-year survival and shunt failure (inability to complete shunt, irreversible shunt occlusion, major rehemorrhage, unanticipated liver transplantation, death). Three prospective randomized trials and one retrospective case-controlled study were identified. Analysis was limited to patients of Child Classes A or B. Significantly better 2-year survival (OR 2.5 [1.2-5.2]) and significantly less frequent shunt failure (OR 0.3 [0.1-0.9]) were seen in patients undergoing surgical shunting compared with TIPS. Meta-analysis promotes surgical shunting relative to TIPS because of improved survival and less frequent shunt failure. Surgical shunting should be accepted as first-line therapy for bleeding varices resulting from portal hypertension.

Surgery residency training programmes have greater impact on outcomes after pancreaticoduodenectomy than hospital volume or surgeon frequency

BACKGROUND

Hospital volume of pancreaticoduodenectomy (PD) and surgeon frequency of PD have been shown to impact outcomes. The impact of surgery residency training programmes after PD is unknown. This study was undertaken to determine the impact of surgery training programmes on outcomes after PD, as well as their importance relative to hospital volume and surgeon frequency of PD.

METHODS

The State of Florida Agency for Healthcare Administration Database was queried for patients undergoing PD during 2002-2007. Measures of outcome were compared for patients undergoing PD at centres with vs. without surgery residency training programmes.

RESULTS

A total of 2345 PDs were identified, of which 1478 (63%) were undertaken at training centres and 867 (37%) were performed at non-training centres. Patients undergoing PD at training centres had shorter lengths of stay, lower hospital charges and lower in-hospital mortality. Relative to surgeon frequency of PD, training centres had a greater favourable impact on hospital length of stay, hospital charges and in-hospital mortality (P < 0.001 for each, ancova). Relative to hospital volume of PDs undertaken, training centres had a greater impact on hospital charges (P < 0.001, ancova).

CONCLUSIONS

Surgery residency training programmes have a favourable effect on outcomes following PD and their impact on outcome is greater than the impact of hospital volume or surgeon frequency of PD.

Pancreaticoduodenectomy for suspected periampullary cancers: the mimes of malignancy

BACKGROUND

Pancreaticoduodenectomies are often undertaken with suspicion of malignancy. We undertook this study to determine if and how unnecessary pancreaticoduodenectomies can be avoided.

METHODS

Data from patients undergoing pancreaticoduodenectomy were prospectively collected. Operative indications, including presenting symptoms and results with imaging, with or without biopsy, were reviewed.

RESULTS

From 1996 through to 2007, 551 patients underwent pancreaticoduodenectomy at our institution. Chronic pancreatitis was the operative indication in 3% of patients; premalignant/malignant lesions were present in 86% of patients. Eleven per cent of patients underwent 'unnecessary' pancreaticoduodenectomies with presumptive diagnoses of cancer but were without premalignant/malignant disease on final report by Pathology [pancreatitis in 63%, serous cystadenomas (<4 cm) in 14%]. Of the unnecessary resections, 20% had histories and imaging sufficient to diagnose pancreatitis, 18% had inaccurate preoperative brushings/biopsies 'documenting' cancer, 11% had clear misinterpretations of their imaging studies and 7% had inadequate preoperative evaluations. However, 45% had signs/symptoms of cancer with a pancreatic head mass/biliary stricture.

CONCLUSION

Only a small minority of patients undergoing pancreaticoduodenectomy for suspicion of periampullary cancer do so unnecessarily. Preoperative review of biopsies, better considerations of pancreatitis and careful evaluation of imaging, particularly for cystic masses, will decrease unnecessary pancreaticoduodenectomies.