Tumor infiltrating lymphocyte therapy for ovarian cancer and renal cell carcinoma

Abstract

Personalized cancer immunotherapy based on infusion of T cells holds the promise to specifically target a patient’s individual tumor. Accumulating evidence indicates that the T cells mediating these tumor regressions after cancer immunotherapies may primarily target patient-specific mutations expressed by the patients’ tumors and that the presence of these “neo-antigen” specific T-cells may be related to a high number of mutations in the tumor. In melanoma, treatment with autologous tumor-infiltrating lymphocytes (TILs) can mediate durable complete responses. Previous trials investigating TIL therapy in solid tumors other than melanoma have shown limited success, however none of these early trials used current preparative chemotherapy regimens, and the methods for in vitro lymphocyte expansion have changed considerably. New advances and understandings in T cell based immunotherapies have stimulated the interest in developing this approach for other indications. Here, we summarize the early clinical data in the field of adoptive cell transfer therapy (ACT) using tumor-infiltrating lymphocytes for patients with renal cell carcinoma (RCC) and ovarian cancer (OC). In addition we describe the major advances in the characterization and application of TIL therapy for patients with RCC and OC.

Keywords:

• Adoptive cell transfer
• adoptive T cell therapy
• ovarian cancer
• renal cell carcinoma
• tumor infiltrating lymphocytes

Abbreviations

ACT	=	Adoptive cell therapy
TIL	=	tumor-infiltrating lymphocytes
RCC	=	renal cell carcinoma
OC	=	ovarian cancer
IL-2	=	interleukin-2
PD-1	=	programmed cell death-1
IFN-γ	=	interferon-gamma
Tregs	=	regulatory T cells
LAK cells	=	lymphokine-activated killer cells
CTLA-4	=	cytotoxic T lymphocyte antigen-4

Introduction

Adoptive cell therapy (ACT) with autologous tumor infiltrating lymphocytes (TILs) has a long history with extensive clinical experience in patients with metastatic melanoma. TIL therapy takes advantage of naturally existing tumor-reactive T cells already present within the tumor and represents the ultimate personalized therapy since TILs are isolated from the patient’s own tumor tissue and re-infused after in vitro activation and expansion. TIL infusion are preceded by preconditioning lymphodepleting chemotherapy and followed by infusions of interleukin-2 (IL-2). Objective response rates of up to 50% including complete tumor regression in 10–20% of patients have been reported from several independent centers.^1-4

Previous trials investigating TIL therapy in other solid tumors have shown limited success, but recent advances in immunotherapy strategies have stimulated the interest in developing this approach for other indications. TIL therapy has now shown clinical activity in other selected solid tumors,^5,6 and is being tested in other histologies including ovarian cancer (OC), renal cell carcinoma (RCC), colorectal cancer, pancreatic cancer, hepatocellular carcinoma, cholangiocarcinoma and gastric cancer (ClinicalTrials.gov identifier NCT01174121). Accumulating evidences support OC and RCC as obvious candidates for implementation of TIL therapy. The immune system has the ability to generate endogenous immune responses against OC and RCC.^7,8 An intermediate mutational load⁷ qualify them as good targets for strategies enhancing the pre-existent endogenous immune response to neo-antigens.⁹ Boosting and strengthening these responses with immunotherapy could potentially lead to complete tumor regression with potential cure. Here, we describe the major advances in the characterization and application of TIL therapy for patients with RCC and OC.

Ovarian Cancer (OC)

Current treatment strategies for advanced OC

In 2014, approximately 22,000 new cases of OC were diagnosed in the US alone. Most patients are diagnosed with advanced disease, thus OC is currently the most deadly gynecological cancer resulting in more than 14.000 deaths every year in the US.¹⁰ The current standard first-line therapy for OC consists of cytoreductive surgery followed by adjuvant chemotherapy with taxanes and platinum-based compounds.¹¹ This treatment is very efficient at reducing tumor burden, but virtually all patients with advanced disease relapse and current second-line therapies are generally not curative. Five-year survival rate for stage III and IV is well below 50%. During the past decades no substantial improvement in survival rates has been achieved and new treatments with curative potential are highly warranted.

TILs in OC

The prognostic significance of TILs in OC has been known for many years. Zhang and colleagues was the first to show that the presence of TILs in OC is a strong prognostic factor for a favorable clinical outcome.¹² It was later reported that in particular intraepithelial CD8⁺ TILs (T cells found within the malignant tumor epithelium) were associated with improved survival as opposed to stromal TILs.¹³ More recently, several studies including a meta-analysis have confirmed that the presence of TILs is a predictor of good clinical outcome in OC.¹⁴

The preferential intraepithelial location of CD8⁺ TILs suggests an anti-tumor effect through direct contact with tumor cells. These intraepithelial CD8⁺ TILs express the integrin CD103 that binds E-cadherin found on epithelial cells.²¹ CD8⁺CD103⁺ TILs are characterized by co-expression of programmed cell death-1 (PD-1) and produce abundant interferon-gamma (IFN-γ) when un-specifically activated in vitro.²² Interestingly, CD20⁺ mature memory B cells were suggested to enhance the prognostic significance of CD8⁺ TILs in OC¹⁷ and it was suggested that these B cells might serve as antigen-presenting cells, organizing tertiary lymphoid structures and secreting polarizing cytokines.¹⁷

As observed in many other solid tumors CD4⁺FoxP3⁺CD25⁺ regulatory T cells (Tregs) have been identified as a negative prognostic factor in OC.¹³ However, some studies have correlated Treg cells with favorable prognosis.¹⁸ The reason for this discrepancy is not clear. However, it is generally believed that Tregs in the tumor microenvironment hamper the ability of the immune system to eliminate cancer cells.

Two recent studies have demonstrated the presence of tumor-antigen specific T cells in the tumor microenvironment of OC patients. Wick et al. characterized the tumor mutanome from 3 OC patients and found a highly specific CD8⁺ T cell response against a nonsynonymous mutation in one patient.¹⁹ Ye Q. et al showed that up-regulation of CD137 after T cell recognition of autologous tumors can be used as a biomarker for naturally occurring tumor reactive T cells in OC. Consequently, isolation of CD137⁺ TILs during TIL expansion could potentially be applied in TIL transfer therapy to enrich for tumor-specific T cells.²⁰

Previous experiences with TIL therapy in OC

Early human trials using TIL therapy have showed mixed results. In 1991 Aoki et al. published the first TIL therapy trial using TILs expanded in IL-2 for patients with advanced or recurrent OC.²¹ Clinical responses were observed in 5/7 patients receiving TILs without chemotherapy and in 9/10 patients receiving chemotherapy before TIL infusion. In 1994, Freedman et al. treated 11 OC patients with intraperitoneal injections of TILs followed by IL-2 but no clinical responses were observed.²² A group from Japan treated OC patients after previous debulking surgery with combination chemotherapy and TIL infusion and found that patients receiving TILs had an improved overall survival compared with the group of patients whose TILs did not expand to sufficient numbers and thus only received combination chemotherapy after debulking surgery.^23,24 However, the use of patients lacking sufficient numbers of TILs as a kind of control group may have favored the experimental group, since having TILs is generally a good prognostic factor in OC. No recent clinical TIL therapy trials have been performed in OC. Clinical trials of TIL therapy for OC are shown in Table 1.

Table 1. Clinical TIL therapy trials in ovarian cancer (OC)

Author	Phase	Pre-conditioning chemotherapy	TIL	IL-2 dose and regimen	OC patients receiving TIL	ORR (%)	CR (%)	Comments
Aoki et al. 1991	III	Cisplatin or FCAP	i.v.	None	17	82	0	Stage III-IV or recurrent. 2–3 TIL infusions per pt
Freedman et al. 1994	I	None	Bolus Intraperitonal	Low-dose IL-2 intraperitoneal	11	0	0	Ascites regression (2 pts), tumor and CA-125 reduction (1 pt)
Ikarashi et al. 1994	I/II	FCAP or CAP	i.v.	None	12	–	–	Stage II, III or IV, after debulking surgery, 100% survival after 26.5 months
Fujita et al. 1995	I/II	FCAP or CAP	i.v.	None	13	–	–	Stage II, III or IV, after debulking surgery, 100% disease-free after 33,6 months
Freedman et al. 2000	I	None	Bolus Intraperitonal	IL-2 or IFN-γ	2	0	0	Only 2 of 22 enrolled pts received TIL. One pt with >50% reduction in serum CA-125

Abbreviations: TIL, tumor infiltrating lymphocytes; FCAP, 5-fluorouracil, cyclophosphamide, adriamycin, cisplatin; CAP, cyclophosphamide, Adriamycin, cisplatin; IL-2, interleukin-2; IFN-γ, interferon-gamma; ORR, overall response rate; CR, complete response; pt(s), patient(s).

Renal cell carcinoma (RCC)

Current treatment strategies for metastatic RCC

Like in melanoma, cytokine-based immunotherapy can induce complete responses in RCC. This has been used in the past as a proof of its immunogenicity and the presence of natural T cell responses to RCC-antigens. Several independent clinical trials with high dose IL-2 have demonstrated response rates of around 20%. Due to durable complete responses observed in 5–8% of treated patients, high-dose bolus IL-2 was approved by the US Food and Drug Administration (FDA) for metastatic RCC in 1992.²⁵ IL-2 has played a major part of the treatment of RCC for many years but within the last decade 7 molcularly targeted agents mainly targeting the vascular endothelial growth factor (VEGF) pathway have been approved for use in metastatic RCC (mRCC). Even though these agents represent a major improvement in the treatment of mRCC and despite impressive response rates, tumor regression is rarely durable and most patients eventually progress and die.²⁶ Recently, monoclonal antibodies against immune check-point blockade molecules, such as PD-1, have shown encouraging clinical activity in RCC.²⁷

TILs in RCC

The role of TILs in RCC is controversial. Whereas studies in the majority of solid cancers suggest a positive correlation between the amount of TIL and favorable prognosis,²⁸ some studies in RCC have shown that a high density of CD8⁺ TILs is correlated with poor clinical outcome.^29,30 Nakano et al. showed that only proliferating CD8⁺ TILs were correlated with increased survival. Furthermore, lymphocyte infiltration was found to correlate with higher tumor grade, which could possibly explain the negative correlation between T cell infiltration and survival.³⁰ As for most other cancer types, the frequency of Tregs in peripheral blood and in TILs has been shown to correlate with poor prognosis in RCC.³¹

Early trials demonstrated that RCC tumors contain tumor-reactive T cells. More recently, Markel et al. showed that TILs from primary RCC lesions could be expanded in numbers sufficient for clinical application and that TILs from some patients showed anti-tumor reactivity.³² In some cases, the TILs secreted IFN-γ upon recognition of autologous tumors, while only killing activity was observed in other cases. Surprisingly, none of the TIL cultures demonstrated simultaneous killing and IFN-γ secretion. A study from Wang et al.³³ suggested that tumor-reactive TILs from RCC showed proliferative exhaustion, driven by tumor-expression of CD70.

A recent study from Baldan et al. demonstrated generation of TILs from RCC using anti-CD3/anti-CD28 coated paramagnetic beads. With this method they showed IFN-γ secretion from expanded TILs co-cultured with uncultured autologous tumor cells in about 50% of patients.³⁴ They also reported that an improved tumor-disaggregation method produced targets cells that were more consistently recognized by expanded TILs and thus may lead to a more reliable potency assay.

CD4⁺ T cell responses in RCC patients against shared tumor antigens have been reported³⁵ and CD4⁺ T cells may have a functional role within the TIL population. Induction or upregulation of MHC class II expression on RCC tumor cells after exposure to IFN-γ or IFN-α has also been reported, supporting that CD4⁺ tumor-reactive T cells may potentially directly target RCC tumor cells.³⁶

Previous experiences with cell-based therapy in RCC

The history of ACT for mRCC goes back more than 3 decades. Two effector cell types, T lymphocytes and natural killer (NK) cells, have been evaluated for ACT in RCC. The first cell based therapy for mRCC consisted of infusion of IL-2 and lymphokine-activated killer (LAK) cells isolated form peripheral blood.³⁷ Several clinical trials including more than 500 RCC patients were performed and pooled data revealed an objective response rate of 22%. However, neither clinical response rate nor overall survival were significantly greater than in patients treated with IL-2 alone³⁸ and consequently, LAK cell therapy was abandoned.

Previous clinical trials with TIL-based ACT for RCC have only showed modest success. Several TIL therapy trials have been performed with highly varying response rates (Table 2). A single-center study involving 36 patients treated with TIL and low-dose IL-2 achieved an overall response rate of 34.6%.³⁹ A phase III trial comparing TIL plus IL-2 with IL-2 alone provided no evidence of greater benefit of TIL therapy. However, in this multi-center trial there was a high TIL production failure (41%), which resulted in premature discontinuation of the trial. This left open the question of whether TIL therapy might have clinical benefit, as indicated by earlier phase I/II trials. No recent clinical TIL-based ACT trials for RCC patients have been performed. None of the early ACT-TIL trials in RCC used the preparative chemotherapy regimen used today and the methods used for ex vivo lymphocyte expansion have changed considerably.

Table 2. Clinical TIL therapy trials in renal cell carcinoma (RCC)

Author	Phase	Pre-conditioning chemotherapy	TIL	IL-2 dose and regimen	RCC Patients receiving TIL	ORR (%)	CR (%)	Comments
Topalian et al. 1988	I	Cyclophosphamide	i.v.	High-dose bolus i.v. every 8h	4	25	0	Varying doses and combinations of TIL
Kradin et al. 1989	I	None	i.v.	i.v. continuous	7	29	0	TILs were infused every second day
Bukowski et al. 1991	I	None	i.v.	i.v. continuous	18	0	0	TILs were infused days 1 and 8, with or without IL-2
Thiounn et al. 1994	III	None	i.v.	None	6	30	30	Patients were pretreated with IL-2
Goedegebuure et al. 1995	I	None	i.v.	Moderate bolus i.v. every 8h	4	0	0	Two patients with minor response
Figlin et al. 1997	I/II	None	i.v.	i.v. continuous + IFNα-2a	55	34.6	9	CD8^+ TIL or cytokine-primed TIL
Figlin et al. 1999	III	None	i.v.	i.v. continuous	39	8	NR	Randomized study of IL-2 plus CD8^+ TIL vs IL-2 alone
Dillman et al. 2004	Retro-spective	None	i.v.	Hybrid bolus and i.v. continuous	9	11	0	Increased survival of patients receiving intermediate or high dose TIL

Abbreviations: TIL, tumor infiltrating lymphocytes; IL-2, interleukin-2; IFN-α, interferon-α-2a; ORR, overall response rate; CR, complete response.

TIL recognition of neo-antigens

The major advantage of using TIL therapy for metastatic cancer is the wide range of undefined and defined tumor antigens recognized by the T cells. Despite the known association between TILs and clinical outcome, there is still limited knowledge about the underlying antigens recognized by TILs. In melanoma, T cell reactivity against non-mutated, shared antigens have been observed in the majority of patients, and broadening of this T cell response has been documented following both TIL therapy and anti-cytotoxic T lymphocyte antigen-4 (CTLA-4) treatment.⁴⁰ However, it has become clear that “known” antigens only account for a small percentage of TIL responses in melanoma.

The development of deep-sequencing technologies has made it possible to identify the mutations present within the exome (the protein-encoding part of the genome) of an individual tumor. It is becoming evident that melanoma represents a unique target for TIL therapy as it is a highly mutated form of cancer.⁷ Robbins and colleagues used whole exome sequencing to identify mutations in tumor samples from 3 patients that had obtained an objective response to autologous TIL therapy.⁴¹ They identified 3 mutated antigens that were recognized by TILs and the presence of these antigens correlated with objective response. Additional studies have shown that somatic mutations can facilitate a wide array of neo-antigen specific T-cell responses at the tumor site.⁴² This demonstrates that whole exome sequencing can be used as a new platform to identify potential T-cell reactive mutated antigens that could be potential predictive biomarkers for TIL therapy.

Melanomas contain around 300 exomic mutations in average,⁷ which appear to be sufficient for the formation of immunogenic neo-antigens that can be frequently recognized by T cells. Accordingly, formation of neo-antigens that can potentially be recognized by autologous T cells is expected also to be common for other tumors with a high mutational load, e.g. lung cancer and colorectal cancer,⁷ whereas tumors with low mutational burden may be less likely to express neo-antigens that can be recognized by T-cells.⁷ Recent studies in melanoma and non-small cell lung cancer showed that the mutational burden within the tumor exome was predictive for response to ipilimumab⁴³ and pembrolizumab,⁴⁴ respectively. However, a high mutational load may not necessarily result in an equivalent amount of neo-antigens and the presence of neo-antigens does not equal the induction of T cell reactivity. In fact, the vast majority of mutations within the expressed genes do not lead to the formation of tumor antigens that are recognized by autologous T cells.⁴⁵ Most importantly, immune surveillance can itself exert selective pressure and lead to immune-based depletion of neo-antigens. Rooney et al. have shown that this phenomenon may occur at high frequency in RCC.⁴⁶

Young-TIL expansion and tumor-reactive TILs

At Herlev Hospital, so far over 40 patients with metastatic melanoma have been treated with TIL therapy in the context of 4 clinical trials, of which 3 are still ongoing (ClinicalTrials.gov identifier NCT00937625 NCT02278887, NCT02379195, NCT02354690).

The implementation in most TIL transfer therapy trials of the “Young-TIL” expansion methods,⁴⁷ using non-selected, short-term cultured T cells, has resulted in a younger and therefore improved phenotype of the expanded TILs and a significantly shortened preparation time leading to a reduced dropout rate.⁴⁷ In addition, clinical grade TIL cultures could be prepared from over 80% of tumor specimens.⁴⁸

Using the “Young-TIL” expansion method we concurrently expanded Young-TIL cultures from tumor samples from RCC, OC and melanoma. Preliminary data confirmed that TILs from RCC and OC can recognize autologous tumors and be expanded in sufficient numbers for clinical application, as established from previous experiences with melanoma (data unpublished).

Combination therapies and future perspectives

The development of deep-sequencing technologies provides the possibility to better understand what TILs recognize and to increase the efficacy of immune based therapy. Based on the data obtained over the past few years, it is very likely that neo-antigen specific T cell reactivity is responsible for the majority of clinical responses after TIL therapy in melanoma patients. In line, the major advantages of TIL therapy compared to other T-cell therapy strategies using engineered T cell receptors (TCRs) or chimeric antigen receptors (CARs) is that TIL therapy targets multiple neo-antigens simultaneously which can probably reduce the risk of tumor escape. Indeed, the tumor-microenvironment in RCC and OC is known to be highly immune-suppressive with frequent expression of PD-1 on TILs. Check-point antibodies blocking PD-1 or its ligand PD-L1, designed to remove the brakes from T cells subjected to tumor suppression, have indeed shown encouraging results in RCC²⁷ and OC.⁴⁹ The combination of TIL therapy and these checkpoint inhibitors may result in the optimal anti-tumor scenario where the tumor is no longer able to suppress through these pathways. Thus, T cell checkpoint blockade after infusion of TILs is a potentially promising approach to increase response rates after TIL transfer therapy and induce long-term durable survival.

An additional exciting development in the field of TIL transfer therapy is the observation that CD8⁺ TILs expressing 4–1BB (CD137)²⁰ and or PD-1⁵⁰ are highly enriched for autologous tumor specificity. The use of IL-7 and IL-15 have shown to be more optimal cytokines than IL-2 for the activation, survival and expansion of tumor-reactive TILs for clinical cell production²⁰ and this has led to efforts to isolate these cells in a clinical setting, followed by a their selective expansion. Such efforts may highly enrich tumor-specific cells and is consequently a promising method in cancers with lower frequency of naturally occurring tumor-reactive T-cells such as OC and RCC.

Conclusion

TIL therapy is emerging as one of the most powerful strategies for treatment of cancer. Early clinical trials of TIL therapy for OC and RCC have been largely disappointing but the TIL technology and knowledge of factors important for effective TIL therapy has grown drastically since. Cell culture methods have changes considerably now considering the need for shortening cell culture time and the pre-conditioning chemotherapy regimen used today has shown to be an important factor for T cell persistence after infusion.

Preliminary data from our lab show that TILs from RCC and OC can indeed be expanded to clinical relevant numbers using the Young-TIL expansion methods and tumor-recognition was observed in the majority of patients were autologous tumor cell lines were available.

By use of technological advances that have supported the development of TIL therapy in melanoma we intend to revisit TIL therapy for OC and RCC. Consequently, we are initiating 2 small clinical pilot trials testing TIL-based adoptive cell therapy for OC patients and RCC patients. Young-TILs expanded in the REP will be infused after standard preconditioning chemotherapy and followed by an attenuated continuous decrescendo IL-2 regimen (ClinicalTrials.gov identifier NCT02482090).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.