1 Introduction

Immunotherapies such as immune checkpoint blockers (ICBs) are an established therapeutic approach to cancer treatment. It is important that physicians and other healthcare stakeholders who influence treatment decisions involving patient care, reimbursement, and drug access understand how immunotherapies differ from traditional chemotherapies and targeted agents, and the importance of proper patient selection. Knowledge of the efficacy of single-agent and combination therapies and their associated safety profiles will help guide informed decisions.

Multiple therapeutic approaches exist for the treatment of cancer, each with a distinct mechanism of action. Traditional cytotoxic chemotherapy agents interfere with cell proliferation and division by inhibiting molecular mechanisms common across normal and malignant cells, thus directly, but nonspecifically, destroying both healthy and cancerous cells. Targeted agents, such as some tyrosine kinase inhibitors (TKIs), are generally designed to destroy cancer cells directly by targeting specific genetic alterations present in those cells. Conversely, immunotherapies act on cancer cells indirectly through the regulation of the immune system [1]. Over time, tumor cells can develop mechanisms to evade immune system recognition [2, 3]. One method for fighting malignancies is to increase activation of the immune system, which is required for successful destruction of cancer cells [2].

For decades, immunotherapies have been used as cancer treatments, including bacillus Calmette-Guérin in non-muscle invasive bladder cancer [4], high-dose interleukin-2 in metastatic renal cell carcinoma (RCC) and metastatic melanoma [5], and interferon α-2b in adjuvant treatment of melanoma [6]. However, their efficacy has been limited by researchers’ lack of understanding regarding the processes underlying immune regulation. Since 2010, additional immunotherapies have received U.S. Food and Drug Administration (FDA) approval, including sipuleucel-T [7], approved for treatment of asymptomatic or minimally symptomatic metastatic castration-resistant prostate cancer; talimogene laherparepvec (T-VEC) [8], approved for the treatment of unresectable melanoma, recurrent after initial surgery; tisagenlecleucel, approved for the treatment of pediatric and young adult patients with B-cell precursor acute lymphoblastic leukemia [9]; axicabtagene ciloleucel, approved for the treatment of adult patients with large B-cell lymphomas [10]; and ICBs including ipilimumab [11], nivolumab [12], pembrolizumab [13], atezolizumab [14], avelumab [15], and durvalumab [16], approved for a wide range of malignancies, including melanoma, non-small cell lung cancer (NSCLC), RCC, urothelial carcinoma (UC), head and neck squamous cell carcinoma (HNSCC), Hodgkin lymphoma, Merkel cell carcinoma, microsatellite instability-high (MSI-H) or mismatch repair-deficient (dMMR) cancer, hepatocellular carcinoma, and gastric or gastroesophageal junction adenocarcinoma (Table 1). Although not yet approved by the FDA, durvalumab was recently added to the National Comprehensive Cancer Network (NCCN) guidelines for NSCLC as consolidation therapy for patients with unresectable stage III NSCLC who have received two or more cycles of definitive concurrent chemoradiation [70, 71].

Table 1 FDA-approved immunotherapies

ICBs act on cancer cells indirectly by removing the “brakes” that serve to regulate T lymphocytes, the main cells responsible for triggering an anticancer immune response [2, 11,12,13,14,15,16]. ICBs are an established class of immunotherapy that target negative regulators of T-cell activation, specifically the immune checkpoints, cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4), programmed cell death-1 (PD-1), and programmed cell death ligand-1 (PD-L1). Inhibition of these immune checkpoint molecules prevents the downregulation of immune cells, leading to enhanced T-cell activity, which ultimately results in increased antitumor immunity [2].

2 Endpoints to Assess Clinical Outcomes Associated with ICBs

Currently, overall survival (OS) is the gold standard clinical endpoint used to demonstrate direct clinical benefit for novel anticancer agents in support of regular FDA approval [72]. Improvements in median OS associated with ICBs versus other therapies have been reported in several cancer types (Table 2), including RCC treated with nivolumab versus the targeted agent everolimus [28], NSCLC treated with either pembrolizumab or atezolizumab versus the chemotherapeutic agent docetaxel [42, 57], and UC treated with pembrolizumab versus chemotherapy [46]. However, as novel agents extend patient survival times, it becomes increasingly difficult to conduct long clinical trials in order to measure OS [75, 76]. Although the use of ICBs has improved survival in melanoma over standard chemotherapy, with some patients experiencing OS of 3 to 5 years [77, 78], when the follow-up is less than 1 year, median OS is usually not reached [22, 23, 39, 43]. Therefore, there is an interest in validating surrogate endpoints that can accurately predict survival benefit in clinical trials of immunotherapy and using these surrogate endpoints for drug approval [75].

Table 2 Comparison of efficacy between checkpoint blockers as monotherapy and standard of care

The correlation between objective response rate (ORR), time to progression, disease-free survival, or progression-free survival (PFS) and OS is poorly understood [76, 79]. Some studies investigating ICBs in NSCLC, RCC, HNSCC, and UC have demonstrated increased OS in the absence of a PFS benefit [27, 28, 31, 42, 47, 57], whereas other trials in melanoma and NSCLC have demonstrated increased OS, as well as ORR and PFS, compared with standard of care (Table 2) [23, 43].

Several ICBs have gained FDA accelerated approval based on ORR, including atezolizumab, nivolumab, durvalumab, and avelumab in previously treated patients with UC [12, 14,15,16]; pembrolizumab in previously treated patients with HNSCC [13]; combination nivolumab plus ipilimumab in melanoma [80]; and pembrolizumab in NSCLC, as monotherapy or in combination with chemotherapy [13, 41, 52]. PFS has been investigated in several meta-analyses as a surrogate endpoint for OS in metastatic melanoma [75, 81], and has served as the basis for FDA approval of first-line pembrolizumab in patients with NSCLC [13].

Generally, ICBs have been shown to significantly improve ORR when compared with standard therapies, for example in patients with melanoma [22, 23, 39], RCC [28], and NSCLC with high PD-L1 expression [43] (Table 2). ICBs have also been shown to prolong duration of response (DOR) when compared with standard therapies (Table 2) [22, 23, 25, 39, 42, 43, 46]. The use of alternative endpoints as a surrogate for OS is an area of ongoing research, and further knowledge on this topic is likely to emerge in the near future.

3 Immunotherapeutics and Patient Selection

As the indications for approved ICBs expand, and new monotherapies and combination therapies come to market, the identification of biomarkers that predict benefit will be essential in selecting patients who will benefit most from immunotherapy. The immunologic profile of the tumor can be taken into consideration when selecting appropriate patients. The level of PD-L1 expression within tumor cells and/or immune cells is associated with higher ORR or longer OS following treatment with PD-1/PD-L1 blockers in NSCLC and UC, pembrolizumab in HNSCC, and nivolumab in melanoma [23, 24, 27, 32, 41, 42, 44, 49, 54, 60, 62]. However, some patients with low or no levels of PD-L1 expression also respond to ICBs [27], indicating that PD-L1 expression is enriched for responders, but the absence of expression is not an absolute indicator of lack of benefit. Finally, some clinical trials in NSCLC have shown no strong correlation between outcome and baseline PD-L1 status [25].

To identify patients who may receive the most benefit from ICBs, a series of FDA-approved diagnostic assays has been developed to measure the level of PD-L1 expression in tumor and/or immune cells. These assays include one mandatory companion diagnostic with pembrolizumab monotherapy for patients with NSCLC or gastric/gastroesophageal junction adenocarcinoma (PD-L1 IHC 22C3 pharmDX, Dako) [82], and three complementary (optional) diagnostics: PD-L1 IHC 28–8 pharmDX (Dako) for nivolumab (non-squamous NSCLC, HNSCC, and UC) or nivolumab/ipilimumab combination (melanoma) [83], VENTANA PD-L1 SP142 assay for atezolizumab (UC and NSCLC) [84], and VENTANA PD-L1 SP-263 for durvalumab (UC) [85]. Therefore, PD-L1 testing should be used for patient selection only when planning to administer pembrolizumab in patients with NSCLC (except when pembrolizumab is used in first line [1 L] in combination with chemotherapy) or gastric/gastroesophageal junction adenocarcinoma [13]. Despite the development of FDA-approved assays for PD-L1 testing, some clinics use laboratory-developed tests, which can be less costly but can also increase the amount of testing variability [86]. Variability in PD-L1 testing can arise because of the type (tumor cells, immune cells, or a combination) and percentage cutoffs used for positivity, archival versus fresh tissue, primary versus metastatic biopsies, diversity of antibodies utilized, and tumor heterogeneity [86, 87]. Several comparative studies across different PD-L1 assays have been conducted, including collaborative studies between industry and academic institutions [88,89,90,91]. The outcomes of these studies have varied, with two studies showing concordance among assays [88, 90], one study showing equivalence for most assays [91], and one study revealing differences across all of the assays that do not support interchangeability [89]. Based on these preliminary findings, the PD-L1 assays that are currently available are not considered interchangeable.

The presence of tumors that harbor mutations in specific genes can influence therapy decisions. For example, the use of epidermal growth factor receptor (EGFR) TKIs is standard of care in patients with EGFR-mutation-positive NSCLC [92,93,94], and studies suggest that this population may not derive benefit from immunotherapy versus EGFR TKIs [95] or chemotherapy [96]. Therefore, the clinical benefit from monotherapy with anti-PD-1/PD-L1 antibodies remains suboptimal in EGFR-mutation-positive NSCLC, and novel combination and therapeutic approaches are needed [96]. The approval of anti-PD-1 therapy for the treatment of adult and pediatric patients with MSI-H or dMMR solid tumors (pembrolizumab) or colorectal cancer (pembrolizumab and nivolumab) that has progressed, underscores the importance of considering other biomarkers that are not specific to the immune checkpoint pathway when making ICB therapy decisions [13]. Patients with MMR deficiency are associated with a higher mutational burden and tumor neoantigen load than MMR-proficient patients, and these features could be driving clinical benefit of ICBs [33, 97, 98]. In fact, tumor mutational burden, known to enhance neoantigen formation, has been shown to be associated with increased response to ICBs, and in some cases improved OS as well, across tumor types such as melanoma [99, 100], NSCLC [101], and UC [54, 56, 102]. Baseline gene expression profiling has also been correlated with response to ICBs; specifically, interferon gamma (IFNγ) signature, which is indicative of an inflammatory tumor microenvironment, is associated with responsiveness to ICBs in several tumor types, including melanoma [103], UC [32, 54, 104, 105], NSCLC [58, 106], HNSCC [103], and gastric cancer [103].

Patients with autoimmune diseases raise concerns about the risk of immune-mediated toxicity associated with immunotherapy and are often excluded from clinical trials. However, as the use of immunotherapy continues to expand into a broader, real-world population, patients with preexisting autoimmune disorders or immune-mediated adverse events (imAEs) from prior immunotherapy are being considered [107, 108]. In one study, the use of the PD-1 blockers pembrolizumab or nivolumab in 119 patients with advanced melanoma and preexisting autoimmune disorders and/or imAEs from prior ipilimumab monotherapy resulted in an ORR of 37%, although approximately 10% of patients discontinued treatment because of imAEs [108].

Other factors that may influence immunotherapy treatment decisions include performance status, comorbidities that are incompatible with imAEs associated with these agents, and the presence of brain metastases. Although the majority of the clinical trials testing ICBs exclude patients with active brain metastases, pembrolizumab was administered to 36 patients with melanoma or NSCLC and untreated or progressive brain metastases in an investigator-initiated phase 2 trial. Relevant reduction in brain metastases was observed in 28% of patients, warranting further investigation of ICBs in this patient population [109]. In the phase 2 CheckMate 204 study, the combination of nivolumab and ipilimumab was administered to 75 patients with advanced melanoma and untreated brain metastases, and provided an intracranial ORR of 55% and an extracranial ORR of 49% [110].

Modern oncologic therapies are increasingly reliant on biomarkers within the tumor microenvironment. Personalized cancer care in the immediate future will have even greater dependence on predictive biomarkers for optimizing therapeutic options for patients. Therefore, the development and validation of novel biomarkers that identify patients who will benefit from anticancer treatments is critical. Biomarker assays are urgently needed, including assays for circulating biomarkers, which optimize test feasibility, convenience, and accuracy, and are non-invasive, preserving patient safety.

4 Pseudoprogression with ICBs

Measuring clinical outcomes associated with immunotherapies comes with a distinct set of challenges not observed with standard therapies. In some cases, the time required to establish an effective immune response may be delayed compared with standard therapies because of atypical responses reported with immunotherapies that are not observed with targeted agents or chemotherapy [111]. Pseudoprogression, also called tumor flare, is a distinct immune-related pattern of response caused by the infiltration of immune cells to the tumor site that can manifest in the form of an apparent increase in tumor size, the development of new lesions, or a mixed response such as progression and regression of different tumors in the same patient [112, 113]. The development of granulomatous changes in the lymph nodes resembling progression have also been described during immunotherapy treatment [114]. In studies investigating immunotherapies in patients with cancer, the prevalence of pseudoprogression can vary based on tumor type; for example, it has been reported to be 7% to 10% in melanoma [23, 113, 115], 5% to 7% in NSCLC [25, 27], 7% in UC [54], and 0% to 2% in HNSCC [44, 116].

Following the standard RECIST (Response Evaluation Criteria In Solid Tumors) v1.1 criteria [117], findings of pseudoprogression can be initially interpreted as disease progression and may lead to discontinuation of treatment before the potential clinical benefit of immunotherapy is fully realized [111, 112]. Studies have demonstrated that after initial apparent disease progression, some patients derive clinical benefit from continued administration of immunotherapy [22, 38, 57, 111, 118,119,120,121]. In a phase 3 study (CheckMate 025), 69% of patients with metastatic RCC treated with nivolumab beyond first progression subsequently demonstrated tumor reduction in target lesions, and almost half (48%) had a 30% reduction in tumor burden from baseline [111]. In another phase 3 study (CheckMate 037) investigating nivolumab in patients with advanced melanoma, 31% received treatment beyond progression, and 27% of these had a greater than 30% reduction in target lesions [22]. Similar findings were observed in 62 patients with recurrent or metastatic HNSCC treated with nivolumab beyond progression in the phase 3 CheckMate 141, with 24% of these patients experiencing tumor reduction [118], and in 137 patients with advanced or metastatic UC treated with atezolizumab beyond progression in the phase 2 IMvigor 210, with 33% experiencing tumor reduction [120]. In patients from IMvigor 210, prolonged survival was observed in subgroups of patients with favorable baseline prognostic characteristics (Eastern Cooperative Oncology Group performance status 0, lymph node-only disease, or no visceral metastases) [120]. Because of the unique responses observed with these agents, immune-related response criteria (irRC) have been developed to serve as a guide for the evaluation of antitumor responses with immunotherapies [113]. Based on survival analysis from patients with melanoma treated with pembrolizumab in the KEYNOTE 001 trial, the benefit of immunotherapy was underestimated in approximately 15% of patients when assessed by conventional RECIST v1.1 versus irRC [115]. Currently, irRC is often used in clinical trials of immunotherapy as a secondary approach for measuring responses, whereas standard RECIST is more prevalent in clinical practice.

According to the authors’ personal experience, when treating long-term survivors who are experiencing a durable response from immunotherapy, it may be possible to incorporate treatment breaks followed by treatment rechallenge in cases of subsequent disease progression, although treatment breaks are not indicated in the label. In the KEYNOTE-006 study, 104 ipilimumab-naïve patients with advanced melanoma completed 2 years of pembrolizumab treatment: of these patients, 23%, 65%, and 12% had complete response (CR), partial response (PR), and stable disease (SD), respectively, at the time of completion of pembrolizumab treatment [122]. After a median follow-up of nearly 3 years, most (91%) of these 104 patients were progression-free, with ongoing CR, PR, and SD experienced by 22%, 62%, and 10% of patients, respectively [122]. Understanding the role of treatment breaks with immunotherapy is an area in need of further investigation.

5 Immunotherapy-Based Combination Approaches

Combination regimens, including two immunotherapies administered together or immunotherapy combined with either chemotherapy or targeted agents, may increase the number of patients with durable response or longer survival (Table 3). The PD-1/PD-L1 and CTLA-4 blockers target different pathways involved in immune regulation, and the combination of these agents enhances tumor response compared with monotherapy [141]. The initial approval of ipilimumab/nivolumab combination therapy for first-line treatment of melanoma was based on the high ORR reported with this combination versus single-agent ipilimumab in the CheckMate 069 study (Table 3) [35], and was further supported by the phase 3 CheckMate 067 study, which showed significant improvements in median PFS [12, 24]. The accelerated approval of pembrolizumab plus chemotherapy (pemetrexed/carboplatin) for first-line treatment of non-squamous NSCLC was based on the high ORR reported with this combination versus pemetrexed/carboplatin alone in the KEYNOTE-021 trial (Table 3) [52]. Additional immunotherapy-based combination therapies are being tested in phase 3 studies (Table 4), and for some of these combination approaches, preliminary data are available (Table 3).

Table 3 Summary of efficacy endpoints for select trials of combination approaches involving immune checkpoint blockers
Table 4 Ongoing pharma-sponsored phase 3 trials of immunotherapy-based combination approaches for advanced malignancies

The concurrent use of immunotherapies in combination regimens, along with the supportive care required to manage increased toxicity, may contribute to the overall healthcare costs associated with these agents. Based on current labeling for the treatment of melanoma patients, ipilimumab and nivolumab are administered together only for the initial four doses; nivolumab is then given as monotherapy [12]. Alternative dosing regimens for ICBs used in combination are currently under investigation, with the goal of improving the safety profile while maximizing clinical benefit [125, 142, 143].

6 Adverse Events Associated with ICBs

By enhancing immune system function, ICBs can lead to adverse events (AEs) distinct from chemotherapy [144, 145], which include a range of dermatologic, gastrointestinal (GI), endocrine, and hepatic toxicities, as well as other less common inflammatory events [146]. Though imAE onset is variable, most occur during the initial months of therapy [11,12,13,14,15,16]. Whereas imAEs of any grade can occur in up to 90% of patients treated with ICBs as monotherapy [17, 20, 24, 36, 42, 43, 54, 56, 59, 62], the incidence of grade ≥ 3 imAEs can range from 1% to 10% with anti-PD-1/PD-L1 monotherapy [24, 43, 54, 56, 59, 62] and from 15% to 42% with anti-CTLA-4 monotherapy [17, 20, 24, 36]. Combination therapy with anti-CTLA-4 and anti-PD-1 antibodies is associated with a 40% to 45% incidence of grade ≥ 3 imAEs [24, 36]. Although infrequent, life-threatening imAEs can occur with ICBs [11,12,13,14,15,16].

Because severe imAEs can lead to treatment discontinuation, careful monitoring and prompt management are important to ensure patients continue to receive beneficial immunotherapy. Unlike chemotherapy, which can only be tolerated for shorter durations (e.g., 6 cycles), immunotherapy agents can be administered for up to 2 or 3 years in some cases [21, 147, 148]. Although recent analyses on cumulative toxicity associated with ICBs after long-term therapy are needed, an analysis conducted in 306 patients with advanced solid tumors treated for up to 22 months with nivolumab monotherapy in a phase 1 study showed no cumulative toxicity after a minimum of 14 months of follow-up [148]. In a pooled safety analysis of 282 patients with advanced melanoma who were treated with nivolumab monotherapy in two phase 3 and two phase 1 studies and who experienced new treatment-related imAEs, 85% did so within the first 16 weeks of treatment [149]. Based on a long-term safety analysis conducted in 95 patients with metastatic UC treated with atezolizumab in a phase 1a trial, most treatment-related AEs occurred within the first year after treatment initiation, with a 50% reduction in the incidence of these AEs during the second year [150]. Therefore, patient monitoring remains important with long-term therapy due to the rare occurrence of late-onset imAEs.

Guidelines for the management of imAEs have been proposed in expert reviews [144, 145, 151, 152] but are also available within the prescribing information for each agent and in brochures that can be downloaded from the manufacturers’ websites [11,12,13,14,15,16, 153,154,155,156,157]. Most moderate and severe immune-mediated toxicities can be managed effectively with corticosteroids and can be resolved within 6 to 12 weeks [146]. For steroid-refractory cases, other immunosuppressive agents (e.g., mycophenolate mofetil or the tumor necrosis factor alpha antibody, infliximab) may be required to obtain control of the immune mediated toxicity [144, 145]. Patients developing moderate to severe imAEs may require integrated multidisciplinary care that should include specialists in gastroenterology, pulmonology, dermatology, neurology, ophthalmology, endocrinology, or rheumatology, depending on the type of toxicity [153, 155]. In addition, imAE awareness should be raised among healthcare providers outside the oncology team, such as emergency room physicians and nurses, who might be involved in managing patients receiving immunotherapy. In a real-world study investigating ipilimumab in 129 patients with metastatic melanoma, 26% of patients required corticosteroids for the management of AEs, and 5.4% were administered infliximab in the refractory setting [158]. In a large expanded-access program of nivolumab in combination with ipilimumab, which included 732 North American patients with advanced melanoma, grade 3/4 treatment-related AEs (TRAEs) occurred in 50% of patients, and 32% of the patients discontinued treatment due to TRAEs [159]. These results point to a safety profile consistent with clinical trial data.

7 Quality of Life Associated with ICBs

Although clinical outcomes for patients with cancer are often measured in terms of survival and response, patient-reported outcomes and health-related quality of life (HRQoL) are also important considerations from a patient perspective. Treatment with nivolumab or pembrolizumab has been shown to improve or maintain HRQoL compared with standard chemotherapy or targeted agents. An analysis of HRQoL from the phase 2 KEYNOTE-002 trial, which examined global health status and functional scales (quality of life and physical, emotional, cognitive, and social functioning) as well as symptom scales (fatigue, nausea, pain, dyspnea, insomnia, appetite loss, constipation, and diarrhea), showed that pembrolizumab improved or maintained HRQoL when compared with chemotherapy in patients with ipilimumab-refractory melanoma [160]. A recent analysis of HRQoL from the phase 3 KEYNOTE-045 study showed that pembrolizumab improved HRQoL when compared with chemotherapy in patients with platinum-refractory advanced UC [161]. Several phase 3 studies comparing nivolumab with chemotherapy reported similar findings in treatment-naïve patients with melanoma (CheckMate 066) [162] and in patients with recurrent HNSCC (CheckMate 141) [31, 163]. Nivolumab was also associated with HRQoL improvement over the targeted agent, everolimus, in previously treated patients with advanced RCC (CheckMate 025) [164]. The phase 3 CheckMate 067 showed that ipilimumab/nivolumab combination therapy maintained HRQoL in treatment-naïve patients with melanoma; in this study, no clinically meaningful deterioration was observed in patients treated with ipilimumab/nivolumab combination therapy compared with those treated with ipilimumab [165]. Taken together, these findings indicating HRQoL improvement or maintenance with immunotherapy may support the preferred use of immunotherapies over some targeted agents, such as everolimus, or chemotherapy, especially from a patient perspective.

8 Conclusions and Future Directions of Immunotherapy

Immunotherapies are an emerging treatment for many cancer types, with distinct properties that distinguish these anticancer agents from traditional chemotherapy or targeted agents. Unlike chemotherapy or targeted agents, which generally act directly on the tumor cells, cancer immunotherapies generally function by modulating the immune system, thereby indirectly affecting tumor survival. Because of this, a unique pattern of responses has been reported with immunotherapies that includes pseudoprogression or mixed tumor responses, which can result in the perception of disease progression. In randomized controlled trials, ICBs have been consistently associated with durable responses and often increased rates of response compared with standards of care. Observations of improved or maintained HRQoL versus standard of care further add to the clinical benefits of ICB therapy. In addition, treatment with ICBs is associated with a distinct set of imAEs, which have the potential to be serious. Further studies are needed to evaluate the efficacy and safety of checkpoint blockade in special, difficult-to-treat populations, such as patients with preexisting immune-related conditions, low performance status, or brain metastases. ICBs are currently being studied in the neoadjuvant and adjuvant settings as well as in combination with novel investigational agents including other classes of immunotherapy and targeted agents. As the indications for ICBs expand and cancer treatment continues to shift towards a more personalized approach, the ability to identify patients who will derive the most benefit from immunotherapy will continue to evolve.