Immunotherapy meets William Shakespeare

There is no question that the use of checkpoint inhibitors to boost the immune response of patients with certain cancers has dramatically changed the landscape in the treatment of solid tumors. The number of drugs and the list of responding tumor types continue to expand, and by now many, if not most, of us have seen patients who have had an extraordinarily positive response. But hospital-based physicians and subspecialists across internal medicine have also cared for some of these same patients who have had strikingly severe treatment-triggered autoimmune disease, true examples of what Nobel Laureate Paul Ehrlich termed “horror autotoxicus.”

The current checkpoint inhibitors, which are designed to ease off the natural brakes of the immune cascade, do not represent the first attempts at immunotherapy for cancer. Attempts to stimulate immune-mediated destruction of cancers predate the era of immune checkpoint inhibitors and included cytokine therapy (eg, recombinant interleukin-2, interferon-alfa), cancer vaccines, and adoptive cell transfer. The latter 2 approaches were optimistically intended to target putative tumor-specific antigens. These earlier approaches aimed to enhance the activation and proliferation of tissue-based macrophages to eliminate tumor cells. All were limited by modest efficacy, and some had significant toxicity due to nonspecific immune activation. But the antitumor effects of bacillus Calmette–Guérin therapy and adjuvants, mediated through activation of both innate and adaptive immune responses, demonstrated that nonspecific immune activation could result in clinically meaningful antitumor effects in select malignancies. A twist of this work was that, in the search for immune cell–derived tumor-killing cytokines, tumor necrosis factor was discovered. Years later, its role as a tumoricidal agent became relatively irrelevant, but it came to be a major target of anti-inflammatory biologic therapies.

Through extensive research in tumor biology, it became clear that many cancers develop ways to mitigate the host antitumor response by blunting proinflammatory intercellular communication. By exploiting and stimulating the normal “checkpoint” ligand-receptor pairs that have evolved to prevent uncontrolled inflammation in the normal host every time there is an infection and to maintain normal immune tolerance, tumors have developed a mechanism to promote their own protection. Tumors usurping the normal immune checks and balances to enhance their survival and growth has been termed adaptive immune resistance. Once recognized, a great deal of research went into finding ways to reverse this tumor effect. What has been developed thus far is a series of monoclonal antibodies that target the normal checkpoint pathways that have been commandeered by tumors (but not the commandeering action of the tumor per se).

The development of immune checkpoint inhibitors marked a paradigm shift by targeting these normal but, in patients with cancer, augmented cell functions (specifically cytotoxic T-lymphocyte–associated protein 4 [CTLA-4], programmed cell death protein 1 [PD-1], and programmed cell death ligand 1 [PD-L1] pathways) that can blunt tumoricidal activity. Normally these checkpoint restraints maintain self-tolerance and limit the development and expression of autoimmunity. By blocking these checkpoints, therapies such as ipilimumab (anti-CTLA-4) and nivolumab (anti-PD-1) reverse tumor-induced immune resistance and permit the inflammatory tumor-infiltrating T cells and macrophages to aggressively pursue their tumoricidal functions.

But, inhibition of checkpoint pathways also disrupts the normal homeostasis of the immune system in areas away from the tumor. Normal beneficial checkpoint controls are diminished and regulatory T-cell functions are reduced, permitting increased activity of autoreactive T and B cells to emerge. This reduces peripheral tolerance, resulting in the emergence of autoimmunity: immune-related adverse events (irAEs) that mimic classic autoimmune diseases and can affect any organ system with a potentially fatal outcome. Release of preexisting auto-reactive T cells from their normal checkpoint restraints contributes to the risk of irAEs, which may also develop through non-checkpoint mechanisms, including as a direct result of some monoclonal antibodies binding to their target antigen expressed on nonimmune cells.

Cutaneous irAEs (rash, pruritus) are the most frequent, occurring in up to 46% to 62% of patients.¹ Gastrointestinal irAEs, particularly colitis, as discussed by Kundu et al² in this issue of the Journal, are more common with anti-CTLA-4 therapy, while thyroid dysfunction (thyroiditis) is more common with anti-PD-1 and anti-PD-L1 agents. Hepatitis and hypophysitis are also notable, with hypophysitis being particularly associated with anti-CTLA-4 therapy, as the monoclonal antibody can bind directly to CTLA-4 expressed on the surface of pituitary cells.³ Musculoskeletal toxicities such as inflammatory arthritis and myositis can occur, as can myasthenia gravis and autoimmune cytopenias.

Thus, immune checkpoint inhibitors are potentially life-saving therapies that can trigger a wide range of autoimmune pathology. The relationship between irAEs and antitumor immune responses is nuanced and may depend on the affected organ, the tumor, the gut microbiome, and the patient. Severe irAEs are typically treated by interrupting checkpoint inhibitor treatment and administering empirically selected systemic immunosuppressive agents, generally including corticosteroids and, ironically, in the case of colitis, tumor necrosis factor inhibitors.

So where is Shakespeare in all of this? In Hamlet he wrote the phrase that, according to the Cambridge English Dictionary, has become synonymous with “to suffer harm from a plan by which you intended to harm someone else”: …tis the sport to have the engineer/Hoist with his own petard (Shakespeare W. Hamlet. Act 3, scene 4, line 229). An apt description of checkpoint inhibition.

• Copyright © 2025 The Cleveland Clinic Foundation. All Rights Reserved.