Novel T-Cell Therapies Make Inroads Into Solid Tumors

Steven A. Rosenberg, MD, PhD

Today, the term immunotherapy is ubiquitous in discussions about cancer, but when Steven A. Rosenberg, MD, PhD, and colleagues at the National Cancer Institute (NCI) cured 33-year old Linda Taylor’s metastatic melanoma with infusions of interleukin 2 (IL-2) in 1984, immune-oncology was a nascent field.

The NCI’s early clinical trials of IL-2 helped transform immunotherapy from the theoretical to the practical, and insights gleaned about T-cell responses laid the foundation for an immunotherapy approach known as adoptive cell therapy (ACT). T cell—based ACT uses autologous tumor-infiltrating lymphocytes (TILs) selected for their antitumor reactivity or autologous T cells genetically engineered with T-cell receptors (TCRs) or chimeric antigen receptors (CARs), which are infused back into the patient to induce antitumor effects. An emerging approach to cellular therapy uses natural killer cells engineered with CARs. Overall, these strategies form a major area of research in the evolving immune-oncology field (Figure¹).

In clinical trials, ACT has shown strong efficacy in blood cancers, and the FDA recently approved 2 anti-CD19 CAR T—cell therapies for certain relapsed B-cell malignancies in adults and children, tisagenlecleucel (Kymriah) and axicabtagene ciloleucel (Yescarta). The success of CAR T cell–therapies in hard-to-treat leukemias and lymphomas led the American Society of Clinical Oncology to designate ACT the Advance of the Year in 2018.

However, achieving consistent success with ACT in solid tumors has proved more elusive. Investigators are exploring a variety of novel strategies to expand the utility of these groundbreaking technologies. These efforts were highlighted during the Society for Immunotherapy of Cancer Annual Meeting (SITC 2018) that took place in Washington, DC, in November. Rosenberg and other experts discussed the challenges of applying ACT to solid tumors and the progress made toward overcoming them in interviews with OncologyLive^® in advance of the conference.

Figure. Categories of T-Cell—Based Immuno-Oncology Therapies¹

Identifying the Right Targets

Expectations that such efforts will eventually succeed are high. “I think that not only the American Society of Clinical Oncology, with its 40,000 members, but [also] the oncology field as a whole is looking to ACT for some major improvements for people who have cancers that cannot be successfully treated today,” said Rosenberg, a 2013 Giants of Cancer Care^® award winner.Solid tumors present several barriers to ACT efficacy that are largely absent in B-cell malignancies: heterogeneous antigen expression, a hostile immunosuppressive microenvironment, and sites that are difficult for the infused T cells to track to and infiltrate. The range of antigens expressed in solid tumors poses problems for both TIL and CAR T-cell therapies.

Chantale Bernatchez, PhD, a member of the TIL laboratory at The University of Texas MD Anderson Cancer Center in Houston, said the goal of TIL therapy, which uses antitumor reactive TILs isolated from tumor tissue and expanded in the laboratory for reinfusion, is to boost a patient’s preexisting tumor response. The infused TIL populations target multiple tumor epitopes, but “we don’t know what antigens are recognized by the majority of the cells infused,” she said.

Rosenberg and colleagues were the first group to use TIL therapy in metastatic melanoma, one of the most mutagenic cancers.² After years of refinement, TIL therapy today can achieve complete durable regressions in about 30% of patients with metastatic melanoma. DNA studies of TILs reactive against melanoma have shown that they recognize differentiation antigens, expressed by healthy melanocytes and melanoma; cancer germline antigens, primarily limited to germ and cancer cells; and neoantigens, which are protein fragments found just on the surface of cancer cells.

“Every DNA mutation we target gives rise to a protein that does not exist in normal tissues. Those are called neoantigens because they arise ‘newly’ from mutations in the DNA in that patient’s tumor,” Rosenberg said. Advances in whole exome and RNA sequencing facilitated the identification of neoantigens and opened the door to applying ACT to the common epithelial cancers, which he said caused about 550,000 deaths in the United States in 2018. “Most people think the common cancers are not immunogenic, but… virtually all the epithelial cancers contain mutations that are recognized by the patient’s autologous immune system,” he said.

In June 2018, Rosenberg and fellow NCI investigators published a letter in Nature Medicine describing the successful treatment of metastatic breast cancer with autologous TILs.³ Before enrolling in the ACT trial, the patient had received 5 or 6 prior treatments, including multiple lines of chemotherapy and hormonal therapies. “She had tumor growing out of her chest wall, multiple large lesions in her liver, and lymph node metastases,” Rosenberg said. Six weeks after receiving an infusion of TILs selected for reactivity against 4 unique neoantigens expressed by her cancer, her primary tumor burden was reduced by half. Subsequent imaging showed complete disease regression, and she remains cancer-free almost 2.5 years later.

“This is a treatment that can work. We’ve shown that it can work in patients with bile duct tumors, colon cancers, cervical cancers, and now breast cancer,” he said, acknowledging that response rates are low. “But we now have a blueprint for attacking the unique mutations in a cancer to elevate it to a treatment that can be used in a more widespread way for people with these otherwise untreatable cancers.”

ACT with CARs or TCRs is also antigendependent. Bernatchez explained that T cells collected from the peripheral blood are engineered to express synthetic receptors capable of recognizing 1 antigen expressed on the surface of a cancer cell (CAR) or within it (TCR). “[After infusion,] the engagement of the CAR construct with tumor antigen triggers potent killing by the CAR T cell,” she said. An advantage of using CARs over TCRs is that, unlike TCRs, CAR T cells are not dependent on human leukocyte antigen for antigen recognition.

Since joining the CAR T-cell research team at City of Hope in Duarte, California, 16 years ago, Christine Brown, PhD, has focused on developing CAR T cells for glioblastoma and other brain cancers. She considers identification of target antigens to be the biggest challenge when applying CAR T-cell therapy to solid tumors. The heterogeneous antigen expression in solid tumors provides numerous potential targets, but few antigens are exclusive to tumor cells. “We need to understand more about how the possible antigen targets in the tumor are expressed on normal essential tissues,” she said.

Challenges With Trafficking and Infiltration

Some investigators who used CARs or TCRs modified to target antigens overexpressed in tumors but also expressed in normal tissues reported severe on-target/off-tumor toxicities.⁴ Autoimmune toxicity is rare with TIL therapy, however, probably because neoantigens are the main drivers of the immune response. Using TCRs and CARs that target neoantigens would be safer, but CARs bind just to cell-surface proteins, which are rarely mutated, Rosenberg said. Cancers change as they grow or in response to selective pressure, and antigen loss or escape in patients treated with CAR T-cell therapy illustrates the need for CARs that target multiple antigens.For ACT to succeed, T cells need access to the tumor tissue at quantities sufficient to induce a response. The mechanisms regulating T-cell trafficking, infiltration, and activity are tumor specific, according to Brown. “How these T cells get to the brain and infiltrate the brain may be very different than how they traffic to systemic tumors,” she said.

Although most conventional therapies cannot cross the blood-brain barrier, Brown said that the findings of early trials of CD19 CARs in patients with hematologic malignancies showed that T cells could reach the central nervous system. More evidence that systemically administered CAR T cells can reach the brain came from a recently published study by Tawbi et al in which checkpoint inhibitors were administered intravenously to patients with melanoma metastatic to the brain.⁵ “If a patient responded systemically, that patient also generally responded in the brain,” Brown said.

Brown’s team at City of Hope has been evaluating locoregional delivery of CAR T cells to brain tumors, injecting them directly into the patient’s tumor site or lateral ventricles, where cerebral spinal fluid is made. “Our research suggests that if you deliver the cells locally, they’re more efficacious on a per-cell-dose basis,” she said. In 2016, Brown and colleagues published the case of a man with highly aggressive recurrent glioblastoma treated with several intracranial infusions of CAR T cells engineered to target the interleukin- 13 receptor α 2 antigen.⁶

His local tumors regressed completely, but after 7.5 months, he developed new lesions. He did not experience any serious adverse events, and Brown’s team is in the early stages of testing the hypothesis that local delivery reduces the risk of on-target/off-tumor toxicities compared with systemic delivery. They are also conducting preclinical studies to determine whether CAR T cells can be genetically modified with a homing molecule that targets the brain, as Samaha and colleagues recently reported.7

Role of the Microenvironment

In another study, investigators at City of Hope are testing a HER2-specific CAR T-cell therapy in patients with HER2-positive breast cancer that has spread to the brain or leptomeninges. The CAR T cells will be administered via intraventricular delivery directly to the brains of participating patients in an outpatient setting. The phase I trial, which seeks to enroll 30 participants, is the first use of a CAR therapy in HER2-positive patients with brain metastases and the first with the novel route of administration, City of Hope said in announcing the trial in October.⁸Once CAR T cells traffic to the tumor site, several features of the microenvironment inhibit T-cell infiltration, proliferation, and efficacy. Stromal elements may present a physical barrier to infiltration.

The microenvironment is hypoxic and nutritionally depleted and is characterized by an acidic pH and oxidative stress. It contains an abundance of suppressive immune cells (eg, T regulatory lymphocytes, tumor-associated macrophages, and myeloid-derived suppressor cells) that overexpress chemokines, cytokines, and other suppressive factors. “The lymphodepleting regimen administered immediately prior to TIL infusion may remodel the tumor microenvironment favorably by eliminating suppressive cells and making the tumor more permissive to T-cell infiltration,” Bernatchez said.

The microenvironment’s complexity makes it a challenging problem, according to Brown. “Thinking of a tree, you can’t touch every branch—you want to identify the truncal pathways that have the biggest impact,” she said. Several groups, including hers, have been striving to develop better mouse models “to elucidate the pathways most likely to improve therapy and to ensure that the microenvironment does not suppress the CAR T cells and thwart their efficacy,” she said.

Most ACT research has used immune-compromised models with human T cells, but immune-competent mouse models, featuring a full immune system and mouse T cells, are now available, creating some excitement. “We’re also doing studies to ‘humanize’ the mice, giving them human immune systems for when we test our CAR-T therapies,” Brown said.

Combinations that may enhance the immune response also are being explored. City of Hope has planned a trial that will treat patients with a combination of CAR T-cell therapy and nivolumab (Opdivo), an immune checkpoint inhibitor that blocks PD-1. “The idea is to stimulate not only the effectiveness of the CAR T therapy but also an endogenous immune response,” Brown said.

Improving the Manufacturing Process

The NCI and MD Anderson are also conducting trials that pair ACT with an anti—PD-1 agent. “Our institution is evaluating the response to TIL therapy in metastatic melanoma patients in combination with pembrolizumab [Keytruda], with randomization to low-dose or high-dose IL-2 post-TIL infusion,” Bernatchez said. She described additional steps that might be needed to enrich TILs from tumor types that are “T-cell poor and may not have a sufficient amount of effector cells in place to benefit from the brakes,” which she said may have more aggressive immunosuppression mechanisms.In ACT, before CAR T cells or TILs can be reinfused into the patient, they must be expanded in the laboratory until the population of antitumor cells is sufficient for treatment. Bernatchez said that TIL therapy requires infusion of far more cells than does CAR T-cell therapy to produce an objective clinical response (10-150 x 109/kg vs 1-2 x 106/kg). The manufacturing process is time-consuming and complex, and she called it the biggest barrier to TIL therapy. The T-cell therapies typically produced at the investigating academic institutions use a lengthy manual process that drastically limits the number of patients who can be treated and the timeliness of treatment. Restricting selection to neoantigen- targeting TILs would create a more potent population, allowing maximal response with fewer cells. “We need to develop T cell—intrinsic markers of response that can be used to select ‘better TIL’ at the onset of the culture to focus the expansion to the active subset,” she said.

“We can’t always control what cells grow in vitro,” even when a tumor-reactive T cell is selected for expansion or genetically modified to express certain receptors, Rosenberg said. Nonreactive T cells may proliferate during expansion, reducing the likelihood of response to TIL therapy. This past September, he and his colleagues received approval from the FDA and the Institutional Review Board at the National Institutes of Health to try a novel approach to T-cell enrichment. “We take TCRs that recognize neoantigens, encode them in a retrovirus, and then put them into the patient’s circulating peripheral blood cells for treatment. This allows us to get enrichment of tumor-reactive cells up to 70% or 80%,” Rosenberg said.

Bernatchez said that MD Anderson is investigating an approach that will allow the team to culture desired CD8-positive T cells more effectively from tumors with a highly immunosuppressive microenvironment, such as pancreatic ductal adenocarcinoma (PDAC). It is easier to grow CD8-positive cells from melanoma tumors, which have a relatively permissive environment for T-cell infiltration, she said: “Since our data in melanoma showed a better outcome in patients treated with products containing a high CD8 proportion, we have adapted our methodology to facilitate expansion of CD8-positive TIL through the inclusion of an agnostic anti—4-1BB antibody.” Her team is also using IL-2 and anti- CD3 for the initial T-cell expansion from tumor tissue.9 “We have seen success rates of over 90% in more than 8 different tumor types tested, including PDAC and ovarian cancer, uniformly producing CD8-rich TIL products,” she said.

Key Takeaways

Although the manufacturing process for CAR T cells is simpler than the process for TILs, Brown said that when she started working with CAR T-cell therapy, it took the team 3 to 5 months per patient to manufacture the treatment. The process now takes about 2 weeks, she said: “Newer trials are trying to move that closer to a week or even shorter, because the less these cells are expanded ex vivo, the younger and more potent they are upon adoptive transfer and the better able they are to retain their function.” She said researchers are also looking at ways to manipulate the conditions used to expand the cells so that they retain their young phenotype, function, and potency longer.Each expert summarized the key takeaway points from their emerging research. “We’re at an early stage in applying CAR T-cell therapy to solid tumors,” Brown said, adding that she hoped that sharing what her institution has learned in early-stage clinical trials of CAR T-cell therapy in patients with brain cancer will help others who are trying to make a difference for this population. “I also think it is important to understand why we think CAR T-cell therapy has potential for a disease as challenging as brain tumors,” she said.

“Overall, we need to be able to go beyond treating a handful of patients per month,” Bernatchez said. “Process improvements made through the years have greatly streamlined TIL manufacturing, but the process is still manual, and the final product is largely uncharacterized.” She suggested automating the TIL generation process: “Ideally, the final process would include up-front selection and expansion of TIL with optimal phenotype and specificity.”

Despite the many advances in cancer treatments that have occurred in the 44 years since he joined the NCI as chief of surgery, Rosenberg said, “it is still virtually impossible to cure any solid tumor that has metastasized using a systemic treatment.” He knows that ACT has curative potential for solid tumors because he has patients with melanoma, renal cell carcinoma, and various malignancies who remain cancer-free years or decades after TIL therapy. The ability to identify and target neoantigens represents an important step forward in bringing ACT to the bedside of more patients. “It’s somewhat ironic that the very mutations that cause cancer will end up becoming its Achilles’ heel,” he said.

References

1. Houot R, Schultz LM, Marabelle A, Kohrt H. T-cell—based Immunotherapy: adoptive cell transfer and checkpoint inhibition. Cancer Immunol Res. 2015;3(10): 1115-1122 doi: 10.1158/2326-6066.CIR-15-0190.
2. Rosenberg SA, Packard BS, Aebersold PM, et al. Use of tumor-infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. N Engl J Med. 1988;319(25):1676-1680. doi: 10.1056/NEJM198812223192527.
3. Zacharakis N, Chinnasamy H, Black M, et al. Immune recognition of somatic mutations leading to complete durable regression in metastatic breast cancer. Nat Med. 2018;24(6):724-730. doi: 10.1038/s41591- 018-0040-8.
4. Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immunotherapy for human cancer. Science. 2015;348(6230):62-68. doi: 10.1126/science.aaa4967.
5. Tawbi HA, Forsyth PA, Algazi A, et al. Combined nivolumab and ipilimumab in melanoma metastatic to the brain. N Engl J Med. 2018;379(8):722- 730. doi: 10.1056/NEJMoa1805453.
6. Brown CE, Alizadeh D, Starr R, et al. Regression of glioblastoma after chimeric antigen receptor T-cell therapy. N Engl J Med. 2016;375(26):2561-2569. doi: 10.1056/NEJMoa1610497.
7. Samaha H, Pignata A, Fousek K, et al. A homing system targets therapeutic T cells to brain cancer. Nature. 2018;561(7723):331-337. doi: 10.1038/s41586-018-0499-y.
8. City of Hope opens first-of-its-kind CAR T clinical trial for patients with HER2-positive breast cancer that has spread to the brain [press release]. Duarte, Calif: City of Hope; October 30, 2018. cityofhope. org/news/car-t-trial-for-her2-positive-breast-cancer-to-brain. Accessed November 8, 2018.
9. Sakellariou-Thompson D, Forget MA, Creasy C, et al. 4-1BB agonist focuses CD8+ tumor-infiltrating T-cell growth into a distinct repertoire capable of tumor recognition in pancreatic cancer. Clin Cancer Res. 2017;23(23):7263-7275. doi: 10.1158/1078-0432.CCR-17-0831.