A Class Act: Antitumor Treatments
MUSC Immunologists are Optimizing Adoptive Cell Transfer for Antitumor Treatments
By Kimberly McGhee
Photography by Brennan Wesley
Viewers of The Emperor of All Maladies, a recent PBS documentary about cancer produced by Ken Burns, may remember two visionary scientists who harnessed the power of the immune system to achieve high rates of objective and sometimes complete response in a subset of patients with metastatic cancer.
Steven A. Rosenberg, M.D., Ph.D., and his colleagues at the National Cancer Institute (NCI), Carl H. June, M.D., and his colleagues at the University of Pennsylvania, along with researchers at other institutions including Sloan Kettering, Loyola, Children’s Hospital of Philadelphia, MD Anderson, and Baylor, are using adoptive cell transfer (ACT) to reinforce the ranks of a patient’s own T cells—the body’s warriors—and train them to better target the patient’s cancer.
ACT involves harvesting T cells from a patient; expanding, conditioning, and sometimes reengineering them in vitro; and then reinfusing them into the same patient where they can direct an antitumor response. Some have wondered whether this personalized approach to cancer care is feasible, but the impressive and durable clinical responses seen in some patients with ACT therapy have drawn the interest of the pharmaceutical industry, which is now partnering with academic centers to translate these promising therapies to the clinic.
The Medical University of South Carolina (MUSC) has been actively engaged in T cell immunology for almost two decades, since former Chair of Surgery and current MUSC President David J. Cole, M.D., who studied with Rosenberg at the NCI, opened an immunology laboratory focused on ACT in the early 1990s and recruited talented immunologists such as Shikhar Mehrotra, Ph.D., and Mark P. Rubinstein, Ph.D. They work closely with researchers in the Department of Microbiology and Immunology, chaired by Zihai Li, M.D., and in particular with Chrystal M. Paulos, Ph.D., who studied with June and the NCI’s Nicholas P. Restifo, M.D., to continue to optimize ACT therapy through preclinical and clinical studies. Translation of innovative laboratory findings into the clinic is fostered by the Center for Cellular Therapy, which features a cGMP Class 6–compliant clean room suite, directed by Mehrotra. Such a facility, which provides a sterile environment in which the patient’s cells can be manipulated ex vivo before reinfusion, is key for developing ACT and makes MUSC an ideal site for clinical trials of immune-based therapies.
The T cells used for ACT include naturally occurring tumor-infiltrating lymphocytes (TILs) and T cells that are genetically engineered with antigen specificity via a chimeric antibody receptor (CAR) or aT cell receptor (TCR).
TILs are naturally occurring T cells that attack cancer early but somewhat ineffectually—they quickly become exhausted in the immunosuppressive microenvironment of the tumor and cancer cells quickly learn to evade them. Rosenberg recognized that TILs harvested from the patient’s excised tumor and expanded ex vivo—outside the immunosuppressive environment of the tumor—had the potential to direct an effective anti-cancer immune response. As early as 1994, Rosenberg and Restifo reported that a third of patients with metastatic melanoma receiving TILs through ACT showed an objective response, with a subset showing a complete response.1
Unfortunately, the infused T cells often did not persist and so many responses were not sustained. The addition of a lymphodepletion regimen, typically chemoablation alone2 or together with total body irradiation (TBI),3 enabled better persistence of the T cells because it removed immune agents that contribute to the immunosuppressive tumor microenvironment. Rosenberg, Paulos, and colleagues showed a direct correlation between increased intensities of TBI and the treatment efficacy of ACT.4
Of the 194 patients with metastatic melanoma treated with TILs grown from fragments of melanoma tumor plus IL-2 at the NCI, 107 (55 percent) have shown objective responses; a significant percentage of those treated with a chemoablative regimen or TBI had a complete response (20 percent and 40 percent, respectively).3 A subset of patients has remained recurrence-free at five to ten years of follow-up.3 As the field continues to develop, the benefits of high-dose TBI for improved treatment efficacy will need to be weighed against its associated risks.4
Lion Biotechnologies, which is partnering with the NCI to take TIL therapy into the clinic, is opening a phase 2 trial (NCT02360579) to study the efficacy and safety of using autologous TILs (LN-144) followed by IL-2 in patients with metastatic melanoma who have undergone a preparatory chemoablative regimen for lymphodepletion prior to reinfusion.
Engineering Tumor Specificity into T Cells
Although TILs can be harvested from any sort of tumor, those harvested from melanoma have shown the strongest antitumor efficacy thus far. To help extend the benefit of ACT to other types of cancer, Rosenberg, June, and others began engineering antigen receptors on the surface of T cells to improve their targeting of a patient’s cancer. In essence, either CARs or TCRs are transduced via a viral vector onto the surface of the harvested T cells, and then the T cells are amplified and conditioned before being reinfused.
Chimeric Antibody Receptors
CARs combine the best antitumor traits of antibodies and TCRs. They are hybrids, composed of the targeting moiety of an antibody and the signaling component of a TCR. Like an antibody, they can efficiently target tumor antigen or other cellular components on the cell surface. Like a TCR, they can activate T cells to direct a robust immune response.
Rosenberg, June, Michel Sadelain, M.D., Ph.D., of Memorial Sloan Kettering Cancer Center, and others developed a CAR targeting the CD19 protein that is overexpressed in many B-cell cancers such as chronic lymphocytic leukemia (CLL). After reinfusion, the CAR-expressing T cells expanded more than a thousand fold, and each of them was estimated to kill 1,000 CLL cells. Two of the three treated patients went into complete remission.5 In a more recent study, 27 of 30 children and adults with acute lymphocytic leukemia (ALL), including 15 who had previously undergone stem cell transplant, achieved complete remission, with responses lasting up to 24 months.6
Unfortunately, all B cells in the body, including those in healthy tissue, express CD19 and so were targeted by the CAR-expressing T cells as well. Indeed, in the ALL study, 73 percent of patients developed B cell aplasia but were successfully treated with immunoglobulin replacement therapy. Approximately 27 percent experienced the cytokine release syndrome (CRS), a consequence of the release of cytokines from the large number of infused T cells and the cells they target. Fever, nausea, chills, low blood pressure, an abnormally rapid heart rate, and shortness of breath are among the symptoms that characterize CRS. The symptoms can be easily managed, but a few patients may experience a more severe and potentially fatal “cytokine storm,” which can be treated effectively with tocilizumab, an antibody targeting the interleukin (IL)-6 receptor.
TCR-transduced T cells
In contrast to CARs, which are man-made, TCRs are naturally occurring. The body contains millions of different types of TCRs, a few of which can effectively target the patient’s tumor but are likely present in too few numbers to effectively destroy it. In TCR-transduced T cell ACT, the receptors that best target the patient’s tumor are transduced into harvested T cells using a viral vector. After transduction, the T cells, which now express the selected receptors in great number on their surface, are amplified and conditioned before being reinfused. Unlike CAR-expressing T cells, which target only antigen or other components on the tumor cell’s surface, TCR-transduced T cells can recognize and destroy antigen that has been processed from internal proteins and presented on the cell surface by a major histocompatibility complex (MHC).
In a recent pilot study, T cells transduced with an NY-ESO-1–reactive TCR were adoptively transferred into patients with advanced metastatic synovial cell sarcoma or melanoma. The NY-ESO-1 antigen is expressed on the tumor cells of 70 percent to 80 percent of patients with synovial cell sarcoma and 25 percent of those with melanoma. Eleven (61 percent) of 18 patients with NY-ESO-1+ synovial cell carcinoma and 11 (55 percent) of 20 patients with NY-ESO-1+ melanoma showed an objective response to this treatment, with three- and five-year survival rates of 38 percent and 15 percent for the former and of 33 percent for the latter.7
A recent preclinical study at MUSC using a unique mouse model developed by Mehrotra showed that the high-affinity TIL 1383I TCR, isolated from MHC class I–restricted CD4+ T cells obtained from the TILs of a patient with metastatic melanoma, could control the growth of melanoma when expressed in CD3+ T cells.8 MUSC immunologists have an ongoing collaboration with researchers at Loyola University, where this high-affinity TCR is being used in an ACT trial (NCT01586403; PI, Michael I. Nishimura, Ph.D.). Efforts are being made to obtain funding for similar clinical trials at MUSC.
A number of academic and industrial partnerships have formed to bring CAR- and TCR-transduced T cell ACT therapy to the clinic, including those between the University of Pennsylvania and Novartis; Baylor, Bluebird Bio, and Celgene; Memorial Sloan Kettering Cancer Center, the Fred Hutchinson Cancer Research Center, and Juno Therapeutics; the National Cancer Institute and Kite Pharma; and the Cellular Biomedicine Group and the Chinese PLA General Hospital.9
Optimizing ACT Therapy
Although the results obtained with ACT therapy have been impressive, not all treated patients respond or have a durable response. As TIL therapy and the first of the CARs and transduced TCRs make their way through the clinical trial process, intensive research is underway to optimize ACT so that more patients can benefit. Three areas of special interest are identifying new populations of T cells, conditioning the harvested T cells ex vivo with cytokines or co-stimulators to improve efficacy and persistence, and finding ways to counteract the immunosuppressive microenvironment of the tumor.
Identifying New Types of T Cells for ACT
Most commonly, a combination of cytotoxic (CD8+) and helper T cells (CD4+) are used for ACT. The CD4+ cells not only help activate CD8+ cells but have been shown to mediate powerful tumor immunity in mice and in humanized models of solid tumors.10 Of growing interest are Th17 cells and Tc17 cells, which are IL-17-secreting CD4+ and CD8+ cells, respectively. Th17 cells can either promote or suppress tumor growth, and research by Paulos, Li, Mehrotra, and others to identify the factors that destine them to one or the other of these fates is ongoing.11 These cells are of great interest because they maintain stemness and have the potential to be much more persistent after ACT than cells that have already differentiated into cytotoxic effector cells. According to Paulos, “We found that these cells last in the body a long time, longer than other subsets of T cells.”
Improving Conditioning Regimens
Harvested T cells are exposed to cytokines and costimulatory molecules to preferentially expand a subset of T cells or to improve efficacy and persistence. IL-2 is the cytokine most commonly used for T cell conditioning, but many others are under investigation. Recently, Rubinstein, Mehrotra, Paulos, and Cole reported that IL-12 improves the transduction efficiency of CD8+ cells12 and enhances their antitumor efficacy.13 Paulos, Rubinstein, Cole, and colleagues also showed that TBI-induced IL-12 enhances Tc17 cell–mediated immunity, suggesting that IL-12 may also be crucial for expanding antitumor Tc17 cells for ACT.14 Paulos, June, and colleagues found that the inducible costimulatory molecule (ICOS) preferentially expands Th17 cells,15 and Paulos, Rubinstein, Mehrotra, Cole, and colleagues reported that it also expands Tc17 cells.16
Counteracting the Tumor Microenvironment
When reinfused T cells encounter the highly immunosuppressive tumor microenvironment, many do not survive to direct a robust immune response. For instance, the oxidative stresses of the tumor microenvironment prove too much for some of the heterogeneous T cells that result from expansion ex vivo. Mehrotra and colleagues have shown a direct correlation between a T cell’s antioxidant capacity and its antitumor efficacy17 and have demonstrated better antitumor responses with ACT using tumor-reactive cells with higher surface expression of antioxidant-reduced thiols.18
The Way Forward
Immune checkpoint inhibitors, which take the brakes off the immune system, are rapidly becoming standard of care for some patients with advanced cancer (for more information on these inhibitors, see Part I of this article). Because these inhibitors help mitigate the immunosuppressive tumor microenvironment, it is thought they could be combined with ACT to improve the persistence of the infused T cells.
“We think that, in many cases, adoptively transferred T cells become dysfunctional as a result of the suppressive environment of the tumor,” says Rubinstein. “Immune checkpoint inhibitors may help prevent that.”
Studies in the Rubinstein laboratory have shown that the combination of immune checkpoint blockade and ACT achieves better results than either alone. Rubinstein is partnering with MUSC Health hematologist/oncologist John M. Wrangle, M.D., to treat lung cancer patients with a combination of PD1 blockade therapy and IL-15/IL-15Rαcytokine complexes in an effort to bolster memory T cells long term.
Combining immune checkpoint blockade with ACT therapy likely holds the greatest promise for achieving objective and durable responses in a larger proportion of patients and a wider variety of cancers. “Combination therapy is the future, and immunologists and clinicians at MUSC are working hard to bring it forward,” says Paulos.
1 Rosenberg SA, et al. J Natl Cancer Inst. 1994 Aug 3;86(15):1159-1166.
2 Dudley ME, et al. Science. 2002 Oct 25;298(5594):850-854.
3 Rosenberg SA, Restifo NP. Science. 3 April 2015;348(6230):62-68.
4 Wrzesinski C, Paulos CM, Kaiser A, et al. J Immunother. 2010 Jan;33(1):1-7.
5 Kalos M, et al. Sci Transl Med. 2011 August 10; 3(95): 95ra73.
6 Maude SL, et al. N Engl J Med 2014;371:1507-1517.
7 Robbins PF, et al. Clin Cancer Res. 2015 Mar 1;21(5):1019-27.
8 Mehrotra S, et al. J Immunol. 2012 Aug 15;189(4):1627-1638.
11 Bailey SR, et al. Front Immunol. 2014 Jun 17;5:276.
12 Andrijauskaite K, et al. Cancer Gene Ther. 2015 Jul;22(7):360-367.
13 Rubinstein MP, et al. Cancer Immunol Immunother. 2015 May;64(5):539-49.
14 Bowers JS, et al. Clin Cancer Res. 2015 Jun 1;21(11):2546-57.
15 Paulos CM, et al. Sci Transl Med. 2010 Oct 27;2(55):55ra78.
16 Nelson MH, et al. J Immunol. 2015 Feb 15;194(4):1737-47.
17 Kesarwani P, et al. Oncoimmunology 2015 Feb 3;4(1):e985942.
18 Kesarwani P, et al. Cancer Res. 2014 Nov 1;74(21):6036-6047.