How can we increase the value of clinical trials with immunotherapy?

Abstract

In the 40 years since the initial studies with interleukin-2 led to the Society for Immunotherapy of Cancer’s creation, immunotherapy has become a crucial pillar of cancer therapy that is extending and improving countless lives worldwide. Checkpoint inhibitor immunotherapy has become a standard of care treatment for over 20 distinct types of cancer, and other immunotherapy approaches such as novel engineered cytokines, T-cell engagers, oncolytic viruses, personalized cancer vaccines and various cellular therapies are in active development and have already or may eventually receive regulatory approval. This commentary examines what we have learned during this remarkable four-decade drug development journey and how we can best leverage that knowledge to efficiently develop, license and apply future immunotherapy approaches.

Checkpoint inhibitor (CPI) immunotherapy has become a standard of care treatment for over 20 distinct cancers. In many cases, the regulatory approvals for CPIs have been based on improved efficacy in comparison to prior standard-of-care (SOC) tumor-targeted therapies or their efficacy in settings where no SOC therapy exists. Other immunotherapy approaches such as novel engineered cytokines, T-cell engagers, oncolytic viruses, personalized cancer vaccines and various cellular therapies are in active development and have already or may eventually receive regulatory approval. Therefore, immunotherapy has clearly become a crucial pillar of cancer therapy that is extending and improving countless lives worldwide. Given this quite amazing accomplishment, it is perhaps enlightening to take a look at this 40-year journey from the start of the Society for Immunotherapy of Cancer (SITC, nee the Society for Biologic Therapy), and reflect on what we have learned and how we can best apply that knowledge to efficiently develop, license and apply more effective immunotherapy approaches.

The interleukin-2 story

The first effective immunotherapy for solid tumors, recombinant high-dose (HD) interleukin (IL)-2, was developed and studied by the National Cancer Institute (NCI) Surgery Branch and the Cytokine Working Group (nee Extramural IL-2/LAK Working Group) in the mid-1980s.¹ HD IL-2 received US Food and Drug Administration approval in 1992 and 1998 for patients with metastatic renal cell cancer and melanoma, respectively, because of its ability, shown in phase II trials, to produce durable responses in a subset (10–20%) of patients with these diseases for whom no other effective treatment existed.^{2 3} Unfortunately, because of its limited efficacy and the generalized activation of the immune system related to systemic administration leading to many toxicities referred to as “cytokine storm” or “capillary leak syndrome” HD IL-2 therapy was only applicable to highly selected patients, treated at a few specialized centers.⁴ Efforts to improve the therapeutic index of IL-2 by giving it at lower doses, combining it with other agents (eg, interferon (IFN) alpha or chemotherapy), or attempting to dissociate toxicity from efficacy with steroids or selective toxicity blockers were ineffective.5,9 Efforts to limit treatment to those most likely to benefit, identified immune infiltration and programmed death-ligand 1 (PD-L1) expression as associated with treatment response.¹⁰ Therefore, although HD IL-2 was ultimately of limited clinical utility, it served as “proof of principle” that the immune system, if sufficiently activated, could eliminate solid tumors causing durable remissions, thereby keeping the field of immunotherapy alive through the 1990s and early 2000s.

Lessons from IL-2 and subsequent CPI immunotherapy development

This work also established several principles about immunotherapy that aided the discovery and investigation of novel immunotherapies. Many of these principles are still applicable to current immunotherapy development.

These principles include:

1. Immunotherapy treats or replaces the immune system so that it can treat the cancer.

2. Because the immune system can attack many tumor targets simultaneously and broaden and deepen over time, it can eliminate the last cancer cell producing long-term remissions or even cures.

3. The hallmark of an effective immunotherapy is therefore plateaus on the Kaplan-Meier curves for duration of response (DOR), progression-free survival (PFS), overall survival (OS) that can be maintained off treatment yielding treatment-free survival (TFS).¹¹

4. Elimination of the last tumor cells requires potent immune activation within the tumor microenvironment (TME) so that any potentially resistant tumor cells will die in the perforin/granzyme-rich crossfire.

5. Immune-related adverse events are a consequence of activating the immune system and are frequently associated with response,^{12 13} and therefore, the ability of systemic immunotherapy to cure patients depends on the tumor cells being the most immunogenic cells within the body.

6. The fully activated immune system works very quickly in destroying tumor cells (within 3–12 weeks) and its effects are not fully captured on routine imaging;¹⁴ therefore, treatments can frequently be stopped in responding patients without compromising efficacy.¹⁵ De-escalating or stopping treatment not only reduces costs, but can also turn “survivors” into “thrivers” returning such patients to their precancer lives, although wiser and often grateful for having survived what previously was considered a lethal disease.

7. Treatment approaches that compromise or limit immune activation, for example, prior or concurrent T-cell damaging systemic therapies,⁵ immunosuppressive treatments,⁶ and lower doses⁷ can reduce the curative potential of immunotherapy.

8. If there are insufficient numbers of tumor-specific T cells in the TME, then novel approaches need to be employed to expand the intratumoral antitumor T-cell repertoire, for example, neoadjuvant versus adjuvant therapy,^{14 16} personalized cancer vaccines,¹⁷ Tumor-infiltrating lymphocytes (TIL), oncolytic viruses, T-cell engagers, Stimulator of Interferon Genes (STING) agonists, in order to create and/or expand the antitumor immune response.

Applying these lessons to current immunotherapy development

How can these principles be used to further expand the application of current immunotherapies, efficiently develop the next generation of immunotherapies and maximize disease cures?

In our opinion, several factors/properties should be kept in mind in the clinical development of novel immunotherapy agents and combination regimens, some of which may also reduce drug development and ultimately treatment costs. Please note, these approaches are meant to guide, rather than be universally applied, as there are likely exceptions to the approaches outlined.

1. Wherever possible, move pure immune therapies to the frontline or even the neoadjuvant setting, where they can be most effective in creating remissions and curing patients.

2. Use immunotherapy endpoints such as DOR, landmark PFS, landmark OS, and TFS in trials, to measure the impact of treatments, rather than simply focusing on median PFS or early looks at median OS. These traditional endpoints are suboptimal measures of immunotherapy efficacy relative to tumor-targeted therapies.

3. Attempt to dissociate toxicity from efficacy through approaches that selectively activate antitumor immunity within the TME (eg, conditionally-active cytokines or CPIs) or that can block unique toxicity mechanisms (eg, IL-6).¹⁸

4. Avoid approaches that combine immunotherapies with treatments that could compromise the full activation of the immune response, for example, cytotoxic chemotherapy, radiation treatment, antiangiogenic therapy or molecularly targeted therapies. While such combination regimens have frequently shown improvements in ORR, median PFS and median OS relative to the non-immunotherapy SOC, they have also often increased treatment expense and toxicity while failing to improve or even compromising the immunotherapy endpoints mentioned above relative to pure immunotherapy regimens.¹⁹

5. If such tumor-targeted approaches are necessary to gain control of the disease, they should be offered for a short period of time before initiating the best pure immunotherapy for the patient, so as to not compromise the subsequent intratumoral immune activation and thus, the curative potential of the regimen.^{20 21}

6. Trials should include early treatment stopping rules based on novel surrogate biomarkers of efficacy such as: RECIST Complete Response (CR) or depth of response >80%; pathologic CR or near CR, ctDNA clearance or evidence of minimal residual disease; PET-CT scan normalization (see SITC Early and Alternative Endpoints Workshop) that can lead to early efficacy readouts, reduce costs and maximize TFS.

7. Trials should use predictive markers of response to stratify and segregate patients. Those patients whose tumors lack tumor-specific CD8 T cells (or are PD-L1 and IFN-γ signature negative) should receive therapies aimed at expanding such cells (often in combination with CPI) and the results of the combination should be compared with CPI therapy alone. Effective combinations should improve the durable remission rates in those patients destined not to respond to CPI alone.

8. When studying immunotherapy regimens in patients who have failed first-line CPI therapy, one should add another immunotherapy, particularly one that might be effective at overcoming the mechanism of resistance to the anti-PD-(L)1-based first-line therapy. If a therapy is effective in patients with anti-PD-(L)1 resistant disease, it should be quickly moved to the frontline setting and particularly studied in a likely anti-PD-(L)-1 resistant population.

Maximizing the efficiency and value of pivotal immunotherapy trials

Because of the durable response and prolonged OS properties of immunotherapy, initial trials with anti-PD-(L)1-based treatments frequently showed significant benefits when compared with SOC tumor-targeted therapies or when applied in situations where no SOC therapy existed. However, more recently, several novel combination immunotherapy regimens have shown no benefit when compared with the SOC immunotherapy treatments.22,24 Such negative trials are “dry wells” that waste financial, personnel and precious patient resources as well as time, and in general have hindered investment in and progress of subsequent novel immunotherapy development.

In order to minimize such “dry wells”, investigators should endeavor to adhere to the SITC Checklist for moving immunotherapy regimens into randomized phase III trials²⁵ (figure 1). As noted in the SITC MVP3 Taskforce Checklist, the highest priority for choosing regimens to further develop should be given to those that have shown benefit over SOC in randomized phase II trials, especially for pure immunotherapy regimens that have shown reproducible durable responses that are sustained off-treatment, particularly in patients with either anti-PD-(L)1 resistant tumors or tumor types that do not typically respond to PD-1 pathway blockade. Other regimen properties to take into consideration, particularly where these Checklist high priority results do not exist, include preclinical evidence of single agent activity, clinical activity in SOC PD-(L)-1 resistant tumors and agents with non-overlapping toxicities. Ideally, such trials should also incorporate translational correlative studies and an integrated biomarker program that can help define which patients are benefiting from the novel regimen and choose trial endpoints, as mentioned above, that best represent that curative potential of pure immunotherapies.

Figure 1. Framework for evaluating planned phase III trials in immuno-oncology. This framework assumes that fundamental practical and institutional considerations for a study are satisfied and is intended to assess the immuno-oncology aspects of a planned trial. The highest priority is placed on clinical data, particularly results of randomized phase II trials and evidence of durable responses in PD-(L)-1 resistant tumors and in tumor types that do not typically respond to checkpoint inhibitor therapy. (Reference Atkins et al²⁵ with permission). PD-1, programmed cell death protein-1; PD-L1, programmed death-ligand 1.

Concluding thoughts

Most importantly, as we as an immuno-oncology community move forward in developing new regimens for patients with high-risk and metastatic cancers, our focus should not merely be on turning cancer into a chronic disease, but instead we should strive to make cancer a “curable” disease. We can best do so by choosing the regimen that is able to maximize the activation of antitumor immunity for a particular patient as early as possible within the disease process.

In doing so we will be efficiently exploiting the unique properties of immunotherapy to create durable remissions and thereby optimally turn cancer survivors into thrivers. Focusing on the above principles will not only advance clinical outcomes, but also reduce both the overall cost of drug development and the cost of using agents once they are approved.