TIL Therapy: Facts and Hopes

In the late 1980s, almost a decade before the discovery of immune checkpoints, Rosenberg and colleagues showed promising results of adoptive cell therapy (ACT) with tumor-infiltrating lymphocytes (TIL) for patients with metastatic melanoma (1). These early promising results were confirmed with a clinical trial published in 1994, but most clinical responses were only transient (2). In three consecutive studies with optimized cell preparation and extensive lymphodepletion, response rates reached 49% to 72% (3–5), including several durable complete responses. Over the last 30 years, several hundred patients with melanoma were treated in TIL-based ACT (TIL therapy) clinical trials (6), including a recent phase III (7). With long-term follow-up, there is clear evidence of long-lasting complete responses (3). Efforts for protocol modifications toward improved access and commercialization have been ongoing for almost a decade. In addition, TIL therapy is now expanding to other solid tumor types, however, with mixed outcomes.

Despite the apparent antitumor potential and high efficacy rates in melanoma, TIL therapy has not yet reached regulatory approval. Moreover, with the introduction of checkpoint inhibitors (CPI) and BRAF/MEK inhibitors as standard treatment options, TIL therapy trials have moved to later treatment lines. Still, there is an unmet need for effective therapies for most patients with metastatic melanoma (8, 9). Recently, TIL therapy has demonstrated activity also in the post–PD-1 setting in phase II and III trials (7, 10, 11). Hence, this treatment is expected to enter the mainstream of cancer care shortly (12).

This review presents an overview of the current status of TIL therapy for melanomas and other solid cancers. In addition, we discuss the toxicity, regulatory and practical barriers, and possible approaches for the larger-scale clinical application of TIL therapy.

Since the first preliminary studies performed by the group of Steven Rosenberg (1, 2), several phase I and II TIL therapy studies for patients with melanoma have been conducted. Table 1 summarizes the most epoch-making studies within the last 30 years. These studies illustrated that clinical responses could be obtained despite applying different lymphodepleting regimens, IL2 doses, or number of infused cells. However, the melanoma treatment landscape has changed remarkably in the past decade. The most significant achievement has been the approval and large-scale introduction of CTLA-4 and PD-1 inhibitors as a standard, first-line treatment for patients with metastatic melanoma.

Consequently, ACT trials with TILs have primarily been conducted in second- or later-line treatment during the last decade. Therefore, it was an obvious concern that tumors with resistance to PD-1 inhibition might also be resistant to TIL therapy. However, studies conducted in the post-CPI era still showed similar (7, 11) or lower but still significant (10, 13) response rates compared with those obtained before the widespread use of CTLA-4- and PD-1 inhibitors. Notably, Sarnaik and colleagues (11) showed convincing efficacy of centrally manufactured TIL therapy in patients with unresectable stage III to IV melanoma. In total, 66 heavily pretreated patients received TIL therapy followed by high-dose IL2 (600,000 IE/Kg) for up to 6 doses. An objective response rate of 36% was reported, with 2 (3%) complete responses and 22 (33%) partial responses (11). In addition, in a recently conducted multicenter, randomized phase III trial (7), we compared TIL therapy followed by high-dose IL2 (600,000 IE/Kg) to standard treatment with ipilimumab in a post–PD-1 setting (86% previously treated with anti–PD-1). In total, 168 patients were randomized to TIL therapy or four cycles of ipilimumab at a standard dose of 3 mg/kg. TIL therapy was superior to ipilimumab, with a significantly improved progression-free survival of 7.2 months compared with 3.1 months, respectively. In addition, complete responses were observed in 20% of TIL-treated patients compared with 7.1% of ipilimumab-treated patients (7).

Metastatic non-skin melanomas represent a significant therapeutic challenge. New therapies provided modest improvements, and thus there is a high unmet need for improved treatment options in mucosal (14) and uveal melanomas (15). With only little effect of CPI treatment, it has been speculated that metastases from uveal melanoma have poor immunogenic potential (16). Nevertheless, TILs can be successfully expanded from uveal melanoma metastases, and encouraging results of TIL therapy have been reported by Chandran and colleagues. In a cohort of 20 patients, an objective response rate of 35% was reported (17). Even though responses appeared less frequent and durable compared with cutaneous melanoma, TIL therapy represents a possible future treatment option for this hard-to-treat disease.

A high intratumoral T-cell infiltrate is a good prognostic factor predicting longer disease-free survival (after surgical resection of the primary tumor) and/or better overall survival in most tumor types studied to date (18). These findings suggest a potential for TIL therapy in several solid tumors. However, robust results in nonmelanoma cancer types are still awaited (19). To date, the most promising results beyond melanoma have been observed in patients with HPV-associated malignancies (20) and non–small cell lung cancer (21) and, more recently, in breast cancer (22). Selected published studies within the last two decades are listed in Table 2. With an overall response rate ranging from 8% to 24% in these studies, objective response rates appear lower in nonmelanoma cancers. It is still unclear why response rates are higher in melanoma. However, melanoma has a very high average mutational burden (23) and extensive T-cell infiltration (24). Notably, most CD8⁺ TILs in human tumor infiltrates are not cancer antigen-specific (25, 26). However, the proportion of bystander TILs in melanoma may be lower than in other cancer types, regardless of resistance to anti–PD-1 (26). Overall, these findings explain the lower success of TIL therapy in nonmelanoma cancer types, suggesting that future research should focus on strategies to rescue the few tumor-reactive TILs.

New approaches to improve responses in nonmelanoma solid tumors are under development. The identification and selective expansion of TILs targeting specific mutated tumor antigens (or neoantigens) is an area of growing interest. Tumor neoantigens can be identified by whole-genome sequencing, and neoantigen-specific T cells can be isolated and expanded for infusion (27). Early promising results have been reported by Tran and colleagues in gastrointestinal cancers through the successful expansion of TILs targeting KRAS and ERBB2IP (28, 29). Similarly, neoantigen-specific TILs can be expanded from breast cancer, leading to clinical responses when infused with anti–PD-1 antibodies (22). A potential strategy to target a shared but highly tumor-specific antigen is based on the selective expansion of TILs naturally reactive against p53 mutations. Although feasible and safe, convincing clinical results with this strategy are still awaited (30). As listed in Supplementary Table S1, multiple clinical trials investigating TIL therapy in nonmelanoma solid tumors, focusing on various approaches to increase the efficacy of TIL therapy, are ongoing. In particular, the combination of TIL therapy and CPI is highly promising and is currently under investigation in several clinical studies (NCT03645928, NCT01174121, NCT05681780, NCT03449108, NCT04611126, and NCT03108495). Recent preliminary results from the trials NCT03108495 and NCT03645928 (TIL therapy combined with anti–PD-1) indicated an overall response rate of 43% in head and neck squamous cell carcinomas and 57% in cervical cancer (31). Another interesting approach is the generation of T cells genetically engineered to express neoantigen-specific TCRs. In a newly published case report, TCR therapy targeting mutant KRASG12D induced a partial response in a patient with pancreatic adenocarcinoma (32). Similarly, in a patient with breast cancer, TILs genetically engineered to express a p53 neoantigen-specific TCR resulted in a partial response lasting 6 months (30). Therefore, further research in TCR therapy is needed to elucidate the potential of this treatment as a possible strategy to overcome the challenges of T-cell therapy in nonmelanoma solid tumors. In conclusion, even though the efficacy and durability of TIL therapy in most nonmelanoma tumors are still not comparable with the results obtained in melanoma, promising and innovative strategies warrant further development.

TIL production is a good manufacturing practice (GMP) compliant, ex vivo expansion process based on the surgical resection of tumor tissue from the patient. Metastases available with minimally invasive surgery, such as lymph node metastases or subcutaneous tumors, are preferred, but the generation of TILs has been successful regardless of the site of the resected metastasis (33). The tumor tissue is enzymatically digested or mechanically fragmented. Individual tumor fragments are typically cultured in IL2-containing media. Over 2 to 6 weeks, the lymphocytes will migrate into the culture media and outgrow any competing cells. Minimally expanded, or "Young" TILs are massively expanded, usually within a standard, 14-day rapid expansion protocol (REP), using anti-CD3, irradiated feeder cells (of autologous or allogeneic origin), and high doses of IL2. To retain a continuous availability of oxygen and a sufficient removal of waste products, bioreactors or gas-permeable vessels can be used during the REP (34, 35). Lastly, the expanded TILs are harvested, washed, and tested before release and administration to lymphodepleted patients. In classic TIL therapy protocols, the administration of TILs is followed by the infusion of high-dose bolus IL2 (HD IL2) every 8th hour, but the administration of low-dose or decrescendo IL2 regimens has also been tested with moderate success (36, 37). Although variations across TIL manufacturing and administration centers can be observed, a common pathway is outlined in Fig. 1.

Side effects from TIL therapy have mainly been attributed to chemotherapy or IL2, leading to Common Terminology Criteria for Adverse Events (CTCAE) grade 3–4 treatment-related adverse events (TRAE) in virtually all patients treated. Conditioning chemotherapy with cyclophosphamide (60 mg/kg) and fludarabine phosphate (25 mg/m²) results in reversible pancytopenia with the need for repeated transfusions with red blood cells (RBC) and thrombocytes in most patients, whereas neutropenia puts the patients in high risk of infection (38). The reconstitution of the immune system after lymphodepleting chemotherapy has been evaluated by Kverneland and colleagues (39) in a cohort of melanoma and nonmelanoma patients receiving TIL therapy and low-dose IL2. In this study, the median duration of neutropenia was 6 days when granulocyte colony-stimulating factor (G-CSF) was administered. However, a considerable proportion of the patients (26%) suffered from recurring neutropenia (grades 3–4) for 2 to 3 months after discharge. In another study, the duration of neutropenia tended to be slightly higher in patients receiving HD IL2 (40). The reported median duration of grade 3–4 lymphopenia has been reported to be 18 days, but it is speculated to be even longer for most patients (39). Thus, the lymphodepleting regimen induces significant immunologic subtype alterations for several months following treatment. However, TIL therapy has not been associated with a clinically significant prolonged infectious risk post-TIL infusion.

As illustrated in Fig. 2A, acute toxicities from HD IL2 can affect almost every organ. The most frequently reported toxicities are electrolyte derangement, hypotension, fluid retention, fever, and dyspnea, but these TRAEs are transient and respond to standard interventions. Bolus infusion of IL2 is continued for up to 15 cycles or until dose-limiting toxicity, which, in the majority of cases, limits HD IL2 treatment to a few days. So far, no clear correlation between the number of IL2 doses and objective tumor response has been reported. Given the deleterious role of IL2 via induction of regulatory T cells (41), these data suggest the potential for a more conservative IL2 dosing strategy to reduce the toxicity of the treatment. The central management strategies and stopping rules for treatment with HD IL2 are outlined in Table 3. At the time of discharge, all toxicities from IL2 treatment have resolved or are reduced to grade 1 or 2 (42).

As illustrated in Fig. 2B, there is a noteworthy difference in the onset and resolution of TRAE related to TIL therapy and CPI. TRAEs can potentially affect a patient's self-reported quality of life (QoL), but understanding this interaction is complex. Although it is evident that severe and acute TRAEs are burdensome, former studies have suggested that CTCAE scoring might fail to capture mild but chronic toxicities, which can substantially affect QoL, especially when these toxicities are long-lasting or interfere with daily activities (43). Nevertheless, recent data showed that patients treated with TILs scored significantly higher on global health-related QoL 6 months posttreatment than ipilimumab-treated patients (7). Therefore, these differences in patient-reported QoL must be considered when comparing the treatment modalities.

The evidence around TIL therapy generated to date fuels the debate of whether it is prime time to implement TIL therapy as a standard of care for patients with metastatic melanoma. However, TIL manufacturing and administration is logistically complicated, practically demanding, and regulatory challenging. Experiences from the approval of CAR-T therapy confirm that these barriers can complicate and procrastinate the successful implementation of advanced therapies (44).

We discuss the most significant barriers to implementation.

No commercial TIL products have yet been licensed by the United States Food and Drug Administration (FDA) or the European Medicines Agency (EMA). The regulatory pathway for nonprofit academic developers of TIL manufacturing is presently unexplored. In the European Union (EU), hematologic stem cell products for transplantation have been manufactured and managed by hospitals in the EU under EU Tissue and Cells Directive (45). Nevertheless, academic TIL therapy is facing a regulatory mountain to climb. Political awareness of this barrier for academic sponsors developing advanced therapy medicinal products (ATMPs) led EMA to launch a pilot initiative to support the translation of ATMPs (46). Importantly, for this initiative to succeed, regulatory flexibility is required.

In the United States, the increasing awareness of the regulatory and practical challenges associated with the implementation and approval of advanced therapies fueled the interest of the FDA. A recently published draft guidance contains CAR-T cell–specific recommendations on chemistry, manufacturing and control, pharmacology and toxicology, and clinical study design (47). This guidance will hopefully facilitate further implementation of CAR-T therapies and serve as a template for implementing TIL therapy. With an increasing interest in personalized therapies, such as TIL therapy, the application of the pharmaceutical concept of potency assays can represent a significant regulatory challenge due to the individualized nature of the TIL products, including but not limited to variability in targeted antigens and other functional characteristics as reviewed by Hulen and colleagues (48).

TIL manufacturing requires highly specialized GMP-controlled processes necessitating certified clean rooms, validated robust procedures, and qualified technical personnel. In addition, current treatment protocols are complex and require experience with in-patient administration of high-dose IL2. Therefore, local TIL production and infusion of TILs are currently restricted to a few centers across advanced countries, and a strategy for a more widespread application of TIL therapy is needed. This opens the discussion of centralized manufacturing of TILs as compared with on-site (point-of-care) and whether administration of TILs should be restricted to a few specialized cancer centers or be implemented on a large scale in mid-size cancer centers.

Centralized manufacturing facilities fit commercial strategies very well. Centralized manufacturing ensures a uniform, streamlined, consistent process, easier quality control, and transparency in evaluating regulatory requirements. This would further ensure that all patients are treated according to the same standards as with current models of commercial CAR-T therapy. However, centralized manufacturing requires complex logistics with a high level of coordination between surgical (tumor removal), manufacturing (TIL isolation and expansion), and treatment (patient admission and TIL administration) centers. In addition, all products are delivered—at best—within several hours from release or cryopreserved. Cryopreservation adds manufacturing complexity; however, it was demonstrated feasible in a phase II study (11).

On-site manufacturing fits nicely in an academic setting, but while large comprehensive institutions have the theoretical and physical framework to implement a GMP-approved manufacturing facility, few centers have the required economic and logistical resources. Thus, establishing academic manufacturing facilities demands local fundraising for start-up funding and a laborious organizational effort, including personnel training and establishment of local (small) teams with advanced expertise. On the other hand, an obvious advantage of point-of-care manufacturing is the direct pathway from the laboratory to the administration of the final cell product without the need for cryopreservation and long-distance transportation. Further, point-of-care manufacturing would allow for greater flexibility within the production, with an opportunity to accommodate patient-specific needs. Significantly, point-of-care manufacturing would facilitate continuous research across academic institutions in a highly collaborative fashion.

The current setup of TIL therapy is remarkably specialized, with high-dose chemotherapy and HD IL2 requiring trained medical personnel and quick access to intensive care assistance. Further, the treatment administration is logistically complex, requiring coordination between the manufacturing laboratory (centralized or on-site) and the clinic. Consequently, even though numerous patients could benefit from the treatment, access to TIL therapy has so far been reserved for very few patients within clinical trials worldwide. Centralizing TIL administration on highly specialized, reference cancer centers would maintain this unequal access to therapy since time, and traveling costs would possibly prevent many patients from receiving the therapy. Thus, discrimination based on socioeconomic and geographical factors will be maintained. CAR-T therapy faced similar challenges when approved by the FDA in the past decade, but still, today, it is believed that the geographical accessibility of the treatment limits its widespread use (44).

An appealing compromise is a semi-large-scale implementation, making TIL therapy available for a broader population at specialized mid-sized oncology centers. Once established, this strategy would ensure access to TIL therapy for more patients regardless of socioeconomic factors and, to some degree, place of residence. Nevertheless, the administration of TIL therapy at mid-size centers would demand that patient flow is continuous, or involved teams and facilities capable of adjusting to fluctuating demands. Most likely, such an approach will prove cost-effective via collaboration with centralized manufacturing facilities, which might facilitate reliable TIL production.

Table 4 summarizes the characteristics of centralized versus point-of-care manufacturing and centralized versus semi-large-scale administration of TIL therapy.

When implementing TIL therapy, economic aspects must be considered. The one-off expense of establishing GMP facilities is significant, but once implemented, the cost per product will be relatively modest compared with the per-patient acquisition of commercial products. Trained personnel, including sufficient numbers of technicians, quality assurance (QA), and certified qualified persons (QP), will also be a prerequisite. At the National Center for Cancer Immune Therapy of Denmark (CCIT-DK), we have estimated direct expenses on "academic" TIL therapy (including production and hospitalization) to be roughly $110,000 compared with $160,000 for drug expenses only for four cycles of ipilimumab (standard list price for an average 80 kg North American) as outpatient therapy (49). Because expenses for diagnosis and treatment of autoimmune TRAEs related to CPI are challenging to estimate, it is expected that TIL therapy could be, at least, comparable with ipilimumab costs. However, it is still unclear what the final costs of TIL therapy will be. Although an academic nonprofit setting may help reduce the financial burden, commercial actors may be necessary for widespread implementation. Final costs have been a barrier to CAR-T therapy's implementation and widespread application, and TIL therapy could face similar challenges. Nevertheless, initiatives to lower the cost of CAR-T therapy, and especially strategies to move the treatment to the outpatient setting, can help reduce final costs (50, 51).

Heterogeneous national and regional reimbursement systems may contain additional in-build obstacles for the implementation of TIL therapy and advanced cell therapies in general. Restrictions on reimbursement represent a significant barrier to CAR-T therapy access in the United States (52). These challenges should be considered while preparing for the implementation of TIL therapy. Despite repeated revisions of the reimbursement strategy covering CAR-T therapy in the Medicare program, the current model still results in substantial losses for providers, leaving hospitals with an economic deficit (51). No standard reimbursement strategy exists for the nonmedicare population covered by commercial insurance companies (51). Alternative reimbursement strategies are being explored, but no ideal method has yet been found (44). Also, in Europe, the reimbursement of CAR-T costs has been a challenge, supporting the need for streamlined reimbursement schemes across highly heterogeneous national/regional-specific systems (53, 54).

Still, studies evaluating the cost–benefit of TILs versus standard care and strategies facilitating manufacturing, administration, and reimbursement are needed to make TIL therapy attractive from a health economic perspective.

So far, treatment with autologous TILs has been reserved for patients included in clinical trials with narrow eligibility criteria. Thus, TIL therapy has been offered to younger patients (age ≤ 75 years) with no significant comorbidities and a good Eastern Cooperative Oncology Group Performance Status of 0–1. Also, clinically significant or unstable brain metastases have excluded patients from participating in TIL trials.

A potential setting of application will be for metastatic disease in patients with metastatic melanoma who progressed on or after anti–PD-1-based immunotherapy (second- or later-line treatment).

However, in recent years anti–PD-1 antibodies were introduced in the adjuvant setting, and still a considerable fraction of the patients relapse, the majority within the first 12 months of treatment (55). So far, there is no clear guidance on managing these patients, and possible strategies include combination therapy with CTLA-4 and PD1 inhibitors, monotherapy with CLTA-4 or PD1 inhibitors, or BRAF/MEK inhibitors for patients with BRAF mutations. With no consensus in the field and demonstrated efficacy of TILs in the post–PD-1 setting (7, 10, 11, 13), TIL therapy can be a reasonable first-line treatment for patients with metastatic melanoma who received and progressed on or after adjuvant therapy with anti–PD-1.

Emerging evidence supports the efficacy of neoadjuvant immunotherapy in resectable but clinically evident (e.g., palpable) disease. The OpACIN trial was the first to show a potential benefit of neoadjuvant checkpoint inhibition, with high rates of complete pathologic responses (56, 57). In addition, in a newly conducted randomized phase II trial testing adjuvant versus neoadjuvant treatment with pembrolizumab in patients with stage IIIB–IV melanoma, a significantly higher event-free survival was found in the neoadjuvant cohort (58). With these findings, it is evident that the timing of the treatment significantly affects the outcome, and it is disputable what will be the role of adjuvant anti–PD-1 in resectable, clinically evident stage IIIB–IV melanoma. This area requires further research, but it is a nearby thought that TIL therapy could become a first-line treatment option for patients relapsing after neoadjuvant therapy.

TIL therapy proved effective in patients with CPI-resistant metastatic melanoma. In addition, early promising results support the application of TIL therapy in nonmelanoma cancer types. In the near future, studies to support large-scale implementation in melanoma, the application of TIL therapy in other cancer types, and strategies to optimize and improve TIL manufacturing and therapy are highly warranted.

The widespread implementation of TIL therapy in melanoma is challenging, requiring substantial logistic, regulatory, and economic efforts. To ensure fast access to this potentially life-saving treatment, we recommend implementing tailored approaches to the distinct national and regional healthcare systems, with large-scale administration of TIL therapy in mid-size centers through centralized manufacturing in some countries or regions. In other locations where GMP capability is already established at reference oncological centers, point-of-care manufacturing plus centralized administration may be adopted. Regulatory requirements can be addressed via fast-track, ad-hoc authorization based on hospital exemption schemes or, in the long run, centralized regulatory approval by agencies such as EMA or FDA. Health economics aspects should be addressed in the context of local insurance or national healthcare reimbursement policies.

Development of platforms to further improve TIL therapy is under way. Although approaches to reduce toxicities, streamline the treatment process, and improve the efficacy of classic TIL therapy with combinations are highly valuable, most current efforts are directed at improving the quality of TIL products. Although there is no clear consensus on how to identify the most potent TIL subpopulations, most approaches rely on TIL selection followed by selective expansion (reviewed in Granhøj and colleagues, ref. 59). Alternative approaches for potency improvement are the use of gene modification for arming TILs with genes coding for cytokines or receptors or knocking out checkpoints (reviewed in Hulen and colleagues; ref. 48). Obviously, the use of these strategies would add further complexity to the manufacturing and regulatory processes.

Overall, although "classic" TIL therapy has demonstrated efficacy in melanoma, improved next-generation approaches may secure the extension of efficacy to other cancer types where classic approaches did not prove highly effective (19).

In summary, as classic TIL therapy is about to enter the mainstream of clinical care for melanoma, there is a need for further research and political prioritizing to make the treatment available to a broader population of patients. Strategies to improve treatment efficacy and its extension to other indications are ongoing.

T.H. Borch reports personal fees from Bristol Myers Squibb outside the submitted work. I.M. Svane reports grants from BMS, Adaptimmune, IO Biotech, Lytix Biopharma, Enara Bio, and TILT Biotherapeutics and personal fees from Novartis, Pierre Fabre, and TILT Biotherapeutics outside the submitted work. M. Donia reports other support from Bristol Myers Squibb and Genetech and personal fees from Achilles Therapeutics during the conduct of the study; in addition, M. Donia has a patent for WO2013167136A1 issued. No disclosures were reported by the other author.

Anders Kverneland and Christopher Aled Chamberlain, National Center for Cancer Immune Therapy (CCIT-DK), are acknowledged for useful scientific discussions.