Safety and potential increased risk of toxicity of radiotherapy combined immunotherapy strategy

1 INTRODUCTION

Immune checkpoints are essential to maintain balance between activation and quiescence of the immune system and to suppress self-reactive T cells.¹ The concept of T-cell exhaustion mediated through immune checkpoints is widely accepted to play a key role in immune quiescence. With the realization that overexpression of immune checkpoint molecules in the tumor microenvironment (TME) perform a significant function in antitumor immunity evasion,² immune checkpoint inhibitors (ICI) began to expand the scope of tumor treatment. Blocking these checkpoints via monoclonal antibodies (mAbs) designed to bind these regulatory receptors or their intended ligands “removes the brakes” on the immune system and has been proven to be effective with varying degrees of success in several tumor types.³ After decades of effort, attempts to enlist the aid of the immune system in malignant tumor treatment have begun to bear fruit. So far seven drugs in ICI class have been approved for use against various malignancies to date.³ Breakthroughs with checkpoint blockade therapy including mono-antibodies against programmed cell death protein 1 (anti-PD1 antibodies: nivolumab, pembrolizumab, and cemiplimab⁴), programmed cell death 1 ligand 1 (anti-PDL1 antibodies: atezolizumab, durvalumab, and avelumab), and cytotoxic T lymphocyte-associated protein 4 (anti-CTLA4 antibody: ipilimumab) have demonstrated the clinical potential of cancer immunotherapy and brought about a major paradigm shift in cancer treatment. However, despite the improved outcome of the cancer treatment, the efficacy of ICI alone remains still limited and tumor regression have not been obtained in many cancer patients. In many types of solid tumors, radiotherapy (RT) is widely accepted as a mainstay of first-line treatment. More than one half of patients with malignant tumors would receive RT as part of their treatment.⁵ Recently the RT experienced historical changes as with the widely acceptance of the abscopal effect where antitumor effects could also be noticed to occur in tumor lesions outside the irradiated site, allowing the emerging field of radio-oncoimmunology to be an area of active investigation. Emerging notion was widely accepted that radiation can do much more in addition to being a powerful cytotoxic agent.⁶ Substantial researches reveal that in addition to induce DNA double-strand breaks, which mediate direct or indirect cell death, RT can modulate the immunogenicity of tumors by stimulating the release of antigenic peptides from tumors,⁷ promoting maturation, antigen presentation, and homing ability of antigen presenting cells including dendritic cells. Besides, RT can also remodel the immune contexture of the TME and enhance T-cell priming, antitumor T-cell recognition and activity,^8-12 which have been observed in renal clear cell carcinoma, melanoma, hepatic carcinoma, and many other tumor types.^13-18 Given that convincing evidence has accumulated showing that local RT could stimulate a profound systematically immunoreaction, RT is increasingly viewed as a potential and promising combination partner with immune checkpoint inhibitors and some other immunotherapy drugs. Hence, the combined strategy of immunotherapy and RT, especially single high-dose RT, is a research hotspot, which is considered to have the potential to bring about huge breakthrough in cancer treatment. There has been heightened enthusiasm in exploring the synergy of RT and immunotherapy. Presently, emerging preclinical data and prospective clinical studies demonstrated the combined use of RT with ICIs, either concomitant or sequential, induces synergistic antitumor responses and show favorable results.¹⁹ Besides, cancer treatment usually require multimodality therapy where a combined concept, either concurrent or sequential is inevitable for patients in the real world. The excitement surrounding the possible synergistic antitumor effect of RT and immune checkpoint inhibitors may work as a double-edged sword, which need to be seriously considered. The cost of the increased efficacy provided by RT+ ICI (inhibition of both CTLA-4 and PD1 or PDL1) is a concomitant increase in AE.¹⁹ The mechanisms used by radiotherapy to enhance the immune response are different to those derived from ICI. Therefore, the toxicities of RT and ICI are not totally overlapping. Attention should be paid about the side effects of ICI+RT combination therapy. The RT-induced AE is derived from the DNA damage of irradiated nonmalignant tissues. The incidence and severity of AE depend on the anatomical site that is irradiated, the presence of comorbid conditions, the volume of tissue irradiated, and the dose/fractionation strategy. In theory, to reduce RT-induced AE is supposed to optimize all these parameters. While in clinical settings, the primary strategies to reduce the RT-induced AE is to minify the irradiated volume of normal tissue. The development of technologies such as intensity modulated RT (IMRT), image-guided RT (IGRT), adaptive RT (ART),^{20, 21} respiratory gating,²² and proton therapy²³ enabling the delivery of more conformal doses of RT has reduced the risk of RT-induced AE by decreasing the dose delivered to nonmalignant tissues while still enabling the delivery of tumoricidal RT doses to the tumor.^{24, 25} Dose de-escalation is another strategy to minify the dose delivered to nonmalignant structures, especially for radiosensitive tumor histologies such as human papillomavirus (HPV) associated carcinoma of the head and neck.²⁶

ICIs can trigger autoimmune-like manifestations in multiple organ systems, called immune-related adverse events (irAEs). ICI have a wide spectrum of AE including dermatologic, endocrine, neurologic, gastrointestinal, respiratory, and musculoskeletal toxicities.²⁷ IrAE usually occurred 3–6 months after the beginning of treatment. Sometimes, it happens 1 year after the beginning of anti-PD1 antibody treatment.²⁸ Most irAEs have mild symptoms, but myocarditis, pneumonia, colitis, and neurotoxicity may be fatal. Since adverse reactions of ICI drugs tend to involve multiorgan, last for a long time, would not disappear immediately after drug withdrawal and severe irAE would be life-threatening,^{29, 30} early detection, and appropriate treatment based on the principle of multidisciplinary team (MDT) are essential. According to the National Comprehensive Cancer Network (NCCN) Guideline on the management of irAEs published in 2018, there are some general principles regardless of the tumor type. ICI+RT combination therapy could often continue in the presence of Grade 1 irAEs with close monitoring. Grade 2–3 irAEs need suspension of ICI and initiation glucocorticoid. Grade 4 irAEs require permanent cessation of ICI, high-dose steroids administration, and additional immuno-suppression with biologic agents such as infliximab for patients with refractory syndromes.³¹

Given that a certain degree of overlap exists between the AE profile of RT and ICI, and the phenomenon that RT can induce immunogenic damage or death of nonmalignant tissue cells, concerns exist that RT+ICI combination may increase both the incidence and severity of irAEs. To address this concern, we systematically go over the available evidence about side effects of the RT and ICI combination strategy and particularly discussed the challenges of RT combined with ICI.

2 RADIATION COMBINED WITH CHECKPOINT BLOCKADE IN CENTRAL MALIGNANCE

RT is one of the most important treatment strategy of central malignant tumors. Neurotoxicity is a major concern after RT for brain metastases as it delivers higher doses of radiation (to smaller volumes) than standard. The acute toxicities of RT for intracranial tumors include brain edema, somnolence, and headache, which can usually be alleviated by symptomatic support during treatment. Chronic toxicities include cognitive decline, treatment-associated brain necrosis (TABN), and so on. Among them, TABN usually occurs 11 months after RT, which is one of the most common central delayed toxicities. The incidence rates of TABN were 5.9–17.5% after RT and the risk of TABN correlated with tumor size. Table 1 summarize the studies involving the toxicity profile of RT combined with ICIs available for analysis. Among them, there are series of single-institution retrospective studies revealing whether or not the risk of central toxicity in patients receiving RT+ICIs is higher than those who receive monotherapy resulting in similar conclusions. Martine³² analyzed 480 patients with cerebral metastatic lesions secondary to non–small cellular lung carcinoma (NSCLC), melanoma, and renal cell carcinoma exposed to stereotactic RT (SRT). A total of 115 patients received SRT+ICIs (ipilimumab, pembrolizumab, or nivolumab) and 365 received SRT monotherapy. Target lesions that were 0–2 cm, 2–3 cm, and >3 cm were exposed to 1,800 to 2,000 cGy in one fraction, 1,800 cGy in one fraction, and 2,500–3,000 cGy in five fractions, respectively. In the SRT group and the SRT + ICI group, the median follow-up duration was 25 and 23 months, respectively. Symptomatic TABN occurred in 7% of patients receiving SRT monotherapy and 20% in patients receiving SRT + ICI (p < 0.05). Similarly, Kaidar³³ reported the results of brain SRT for patients with melanoma cerebral metastatic lesions who were treated with ICIs compared with patients who did not receive ICIs. ICI strategy was mostly ipilimumab, which was received 3 mg/kg intravenously every 3 weeks for four doses. RT doses and fractionations were decided based on the size and location of tumor lesions. Treatment was delivered using CyberKnife as: postoperative bed 700 cGy×3 fractions or 500 cGy×5 fractions. Intact lesions (>3 cm) were treated by 600 cGy×5 fractions. The median follow-up time was 21 (7–51) months. Eight patients had TABN, which totally occurred in patients receiving SRT + ICI. Nine patients had hemorrhage, of which seven occurred in the ICI+SRT group. None of the patients receiving SRT monotherapy had TABN. TABN was seen in 28% patients receiving SRT + ICI (p < 0.05). Hemorrhage occurred in SRT alone and SRT + ICI patients was 6.89% and 25%, respectively (p < 0.05). These retrospective researches concluded that treatment of ICI + SRS might be a risk factor for developing TABN for brain malignance. Some researches support this opinion from other perspectives. Colaco et al.³⁴ compared the central toxicity between intracranial SRT + chemotherapy, SRT + target therapy, and SRT + ICIs by conducting a retrospective study enrolling 180 patients with cerebral metastases from different primary tumors. All patients from a single-institution were irradiated with Gamma Knife. The median SRT dose for each lesion was 2,000 cGy (range 1,500–2,400 cGy) to the 50% isodose surface. The median follow-up time was 11.7 months. 16.9% in patients receiving RT + chemotherapy, 25.0% in patients receiving RT + target therapy, and 37.5% in patients receiving RT + ICI present TABN (p = 0.03). In contrast to these studies suggesting higher risks of TABN with combination therapy than with monotherapy, there are some studies showing no such trend. For instance, in the retrospective study conducted by Patel,³⁵ 54 consecutive melanoma brain metastases patients were analyzed. Patients were treated with ipilimumab 3 mg/kg. Brain metastases up to 20 mm in diameter were typically treated to 2,100 cGy, 21 to 30 mm in diameter to 1,800 cGy, and 31 to 40 mm in diameter to 1,500 cGy. Patients with large metastatic cavities (typically >40 mm in diameter) were treated with SRT over three to five fraction. The median follow-up duration is 7.3 months. TABN presented in 21% patients receiving SRT and 30% patients receiving SRT + ICIs (ipilimumab; p > 0.05); symptomatic TABN occurred in 15% patients receiving SRT and 15% patients receiving SRT + ipilimumab (p > 0.05). Of note, Diao et al.³⁶ found even though SRT + ipilimumab lead to higher rate of hemorrhage and necrosis from image examination, the incidence rates of radiation necrosis proved by pathology were less than 2%. Fang³⁷ revealed that for patients receiving SRT + ICI strategies, immunotherapy type (anti-CTLA-4 and/or anti-PD1 Ab) and timing proximity to SRT were not associated with TABN by reviewing 137 melanoma patients treated with SRS + ICI in a single-institution retrospective study. SRS was delivered using Gamma Knife with median prescription dose of 2,000 cGy (range 1,200–3,000 cGy) to the 50% isodose line (range 40–90%). Median tumor/target volume was 122 mm³ (range 3.3–25,290 mm³). The patients were followed up for 9.8 months (range 0–104.6 months) from all SRT treatment sessions. Multivariate analysis revealed risk of TABN was not associated with ICI type, ICI doses, or timing of ICI relative to SRT. It was associated with the incorporation of chemotherapy within 6 months of ICI + SRT (p < 0.05) and with an increased number of lesions treated (p < 0.05). A prospective phase I study³⁸ confirmed ipilimumab 10 mg/kg with SRT was safe and none of the patients experienced >grade 3 adverse effects.

A meta-analysis was conducted to evaluate the safety and efficacy, as well as the optimal timing of the SRS and ICI combination therapy.³⁹ A total of 534 patients were included. The most commonly used ICI was ipilimumab. The most commonly used SRS dose was 20 Gy (18–24 Gy). The 1-year OS was 64.6% and 51.6% for concurrent and non-concurrent therapy (p < 0.001). Combination of SRS and ICI does not appear to be associated with untoward rates of necrosis.³⁹ This research recommends that clinicians cautiously continue to treat these patients with combination of SRS and ICI, and should strongly consider administering these therapies within 1 month of one another if the otherwise presumed risk of toxicity is low (i.e., low tumor volume irradiated, etc.).

In summary, although some retrospective researches indicated ICI + RT strategy increase the risk of TABN, the risk of TABN proved by pathology are fairly low. We think the combined regimen of SRT and ICI for central malignant tumors is clinically appropriate because the relative safety of ICI + SRT strategy has been confirmed by the prospective phase I research, meta-analysis, and some other retrospective researches. The SRS + ICI strategy concurrently induces a better prognosis than non-concurrent therapy. Concurrent combination therapy is more recommended if the presumed risk of toxicity is low.

3 RADIOTHERAPY COMBINED WITH IMMUNE CHECKPOINT INHIBITORS IN PULMONARY MALIGNANCE

Anti-PD1/PDL1 mAbs have the potential to efficiently interdict the immune inhibitory signal pathway by blocking the interaction between PD1 and PDL1, which reboots the priming of tumor-antigen-specific T cells and activating the immune response against tumor.⁴⁰ The combined regimen of anti-PD1/PDL1 mAbs with thoracic RT has already been regarded as the standard of care in the treatment of locally advanced NSCLC. Even though most researches combining anti-PD1/PDL1 mAbs and RT have confirmed an acceptable safety profile, the high-quality clinical evidence about the pulmonary safety of this strategy still remains deficient. Table 2 summarized several representative retrospective studies and prospective clinical trials of safety data, which were mainly based on the analysis of radiation pneumonitis of combined therapy.

Two phase II clinical studies showed that combination therapy increased lung-related toxicity, but did not increase other irAE. In the LUNG-14-179 study⁴¹, which included 93 patients with locally NSCLC treated with Pembrolizumab maintenance therapy after concurrent chemoradiotherapy (cCRT), 19.6% patients stopped taking drugs due to toxicities. The most common toxicities were fatigue (46.2%, ≥grade 3: 4.3%), dyspnea (21.5%, ≥grade 3: 5.4%) and treatment-related pneumonia (17.2%, ≥grade 3: 6.5%). Other grade 3 or more adverse events did not exceed 5%. Besides, a phase II trial evaluate the safety and efficacy of nivolumab combined with CRT in stage III NSCLC (NICOLAS study).⁴² Patients received three cycles of platinum-based chemotherapy and concurrent RT (66 Gy/33 fractions). Nivolumab started concurrently with RT. Eighty-two patients were recruited with median follow-up of 13.4 months. The most frequent adverse events (AEs) were anemia, fatigue, and pneumonitis. Early safety analysis provides evidence that the addition of nivolumab to concurrent CRT is safe and tolerable regarding the 6-month rate of pneumonitis grade ≥3 at the one-sided significance level of 5%. No unexpected adverse events or increased risk for severe pneumonitis observed.

To confirm whether the combined therapy act as the enhancing factor for the adverse effect compared with monotherapy, several retrospective studies compared head-to-head on this issue and conflicting results are reported. Numerous studies suggest that the combination use of ICIs and RT would not lead to higher incidence of toxicity.^43-45 Although in contrast, some other retrospective studies claim that NSCLC patients treated with immune checkpoint inhibitors and preceding thoracic RT might lead to higher risk of pulmonary toxicities.^{46, 47} Overall, these retrospective evidence lacks robust randomized data. Potential confounding including insufficient information, various median time-lag between RT and ICI treatments, and combined analysis of various doses and treatment intents attribute to the controversial results. Compared with retrospective studies, prospective clinical trials would offer more objective and reliable evidence. A prospective secondary analysis)post hoc analysis(of the phase I trial (KEYNOTE-001)⁴⁸ determine the effect of preceding thoracic RT on pulmonary toxicity with pembrolizumab. Ninety-seven eligible patients presented with metastatic NSCLC were included in this analysis. Patients were intravenously administered pembrolizumab at a dose of 2 or 10 mg/kg of bodyweight every 2–3 weeks. Pulmonary adverse effects included dyspnea, coughing, wheezing, pneumonitis, and respiratory failure. Pembrolizumab + preceding thoracic RT had a higher incidence of all grade pulmonary toxicities (13% vs. 1%, p < 0.05). No statistically significant difference in risk of > grade 3 severe pneumonitis was found between the thoracic RT group and the non-RT group (4% vs. 1%, p = 0.44). However, detailed information regarding RT dose, fractionated scheme, and plan dosimetry parameters were not available for many patients in this study, which hampered the further analysis. A prospective, randomized, double blind, placebo controlled phase III study^49-51 (the PACIFIC study, NCT02125461), which enrolled 713 patients with locally advanced, unresectable NSCLC compare the efficiency and safety of definitive CRT ± consolidation durvalumab. Two or more cycles of platinum-based chemotherapy (containing etoposide, vinblastine, vinorelbine, a taxane [paclitaxel or docetaxel], or pemetrexed) concurrently with definitive RT (5,400–6,600 cGy). Durvalumab was administered 1 to 42 days after completion of chemoradiotherapy at a dose of 10 mg/kg of body weight. The most common toxicities resulting in discontinuation of durvalumab/placebo were ir-pneumonitis or radiation pneumonitis (in 6.3% and 4.3%) and pneumonia (in 1.1% and 1.3%). Patients in the durvalumab arm and placebo arm experienced similar pneumonitis of any grade (33.9% vs. 24.8%, p > 0.05) and grade 3 and 4 pneumonitis (3.4% vs. 2.6%, p > 0.05). The most common toxicities of any grade that were associated specially with durvalumab/placebo were diarrhea (18.3% and 18.8%), pneumonitis (12.6% and 7.7%), rash (12.2% and 7.3%), and pruritus (12.2% and 4.7%). The results of the PACIFIC trial provide inestimable prospective validation, which suggest serial administration of definitive thoracic RT and immune checkpoint inhibitors would not obviously augment the incidence of ≥3 grade pneumonitis. Miura conducted the first real-world study⁵² from a single institution of the administration of chemoradiotherapy followed by durvalumab in patients with unresectable advanced NSCLC. Radiation dose of 60 Gy/30F as concomitant chemoradiotherapy. Any adverse events during chemoradiotherapy and durvalumab were observed in 32 patients (78.0%), whereas three patients (7.3%) experienced grade 3 toxicities. Twenty-five (61.0%) patients experienced pneumonitis, four (9.8%) thyroid dysfunction, three (7.3%) myopathy, two (4.9%) rash or eruption, one (2.4%) bowel disease, and one (2.4%) malaise. Among patients experienced pneumonitis, grade 1, 2, and 3 pneumonitis were 52.0% (13/25), 26.8% (11/25), and 4.0%, (1/25). All patients who received corticosteroids of 0.5 mg/kg experienced an improvement in their pneumonitis. If grade 2 pneumonitis occurs, the cessation of durvalumab is considered. Retreatment with durvalumab was reconsidered according to individual conditions. Two of six patients who received retreatment with durvalumab experienced re-exacerbation of pneumonitis. The data in this real-world study and PACIFIC trial suggested that durvalumab exhibited manageable toxicities after chemoradiotherapy. Other  PD-1 inhibitor such as nivolumab is undergoing clinical trials as well. For instance, RTOG 3505 is a randomized trial to evaluate the efficiency and toxicity of cisplatin and etoposide plus thoracic RT followed by nivolumab or placebo for locally advanced NSCLC.⁵³ Besides combining chemoradiotherapy with immunotherapy as consolidation therapy, there are numerous ongoing trials investigating the efficacy and safety of combining RT with ICI as neoadjuvant therapy in resectable NSCLC before surgery, even though no results have yet been reported.

In addition, the PEMBRO-RT study⁵⁴ is the first randomized trial to assess whether stereotactic body RT on a single tumor site preceding pembrolizumab treatment enhances tumor response in patients with metastatic NSCLC. In the experimental arm, SBRT was delivered as 8 Gy×3F on alternate days. Pembrolizumab was administered intravenously at 200 mg every 3 weeks. The first course was given within 7 days after completion of SBRT. In the control arm, no SBRT was delivered. The median follow-up time was 23.6 months (0.1-34.4 months).The ORR at 12 weeks was 18% in the control arm vs 36% in the experimental arm (p  =  0.07). The most common adverse events were fatigue (28/72,39%), flulike symptoms (23/72,32), and cough (20/72,28%). Fatigue (10/37, 27% vs. 18/35, 51%; p  = 0.05) and pneumonia (3/37, 8% vs. 9/35, 26%; p = 0.06) occurred more often in the experimental arm than in the control arm. Pembrolizumab-related toxic effects were primarily fatigue (18%), flulike symptoms (15%), and pruritus (14%). Grade 3 to 5 pembrolizumab-related toxic effects were reported in 12 patients (17%), with no significant differences between arms. Additionally, there are two meta-analysis compared the safety profile of PD1 versus PDL1 blockage and suggest anti-PD1 mAbs lead to a higher incidence of pneumonitis as compared to anti-PDL1 mAbs, especially regarding to the grade 3–4 pneumonitis.^{55, 56} The potential mechanisms accounting for the higher risk of anti-PD1 mAbs induced pneumonitis lie in the total breaking of interaction between PDL2 and PD1 by PD1 blockade.^57-59

These data support the perspective that combined strategies increase pulmonary toxicities to some extent. Although discontinuation of treatment due to AEs was higher with ICI, rates of grade 3/4 pneumonitis or radiation pneumonitis were low and comparable between arms. These prospective clinical trials mentioned above indicated the combined use of RT with ICIs in pulmonary malignance had a clinically manageable AEs. For patients receiving ICI+chemoradiotherapy strategy, the most general AE observed in chemoradiotherapy including neutropenia, esophagitis, fatigue should be addressed in time. Fatal consequence of chemoradiotherapy such as neutropenic sepsis should be predicted in advance, early identified and properly managed. For patients who are prospected to be high risk of pulmonary toxicity, PDL1 antibody may be a favorable choice compared with PD1 antibody.

4 RADIATION COMBINED WITH CHECKPOINT BLOCKADE IN PELVIC MALIGNANCE

RT is a first-line treatment for pelvic malignant tumors such as cervical cancer, prostate cancer, and locally advanced rectal cancer. There are several studies on immunotherapy combined with pelvic RT. Prospective researches provide a relatively homogeneous sample with prospectively gathered data, contributing to robustness of the clinical evidence. In a multi-institutional phase I trial (GOG 9929⁶⁰), 21 patients with node-positive cervical cancer received chemoradiotherapy followed by ipilimumab therapy at a maximum tolerated dose of 10 mg/kg. Patients with cervical cancer were administered with cisplatin, 40 mg/m² intravenously weekly for 6 doses, concurrent with extended-field RT to treat the pelvic lymph nodes and para-aortic lymph nodes (PALNs) to 4,500 cGy with an RT boost and brachytherapy (LDR 4000 cGy or HDR 3000 cGy). If there was no progression of disease, withing 2 weeks after RT completion, sequential ipilimumab was given at the following dose levels every 3 weeks: dose level 1: 3 mg/kg, level 2: 10 mg/kg, and an expansion cohort of 10 mg/kg. All patients completed chemoradiotherapy, and of the 21 patients who received at least two cycles of ipilimumab, 18 (86%) completed four cycles of ipilimumab, and three (14%) completed two cycles. The maximum tolerated dose was 10 mg/kg. Most of the acute toxic effects were grade 1 or 2 diarrhea, dermatitis, and endocrinopathies, with 1 grade 3 gastrointestinal tract toxic effect of abdominal pain. Two of the 21 patients (9.5%) who received ipilimumab had self-limiting grade 3 toxic effects (lipase increase; dermatitis). With a median follow-up of 14.8 months, there were no grade 4 or 5 adverse events. GOG9929 is the first study to show tolerability of ipilimumab as a part of the definitive chemoradiotherapy of locally advanced cervical cancer. Another phase I/II, nonrandomized, open-label, multicenter study ( NCT00323882)⁶¹ explored the efficacy and safety of ipilimumab monotherapy versus combination with RT in patients with metastatic castration-resistant prostate cancer. RT was given focally at a single dose of 8 Gy per target bone lesion for up to three bone lesions per patient at 24–48 h before the first ipilimumab dose. In dose escalation, 33 patients received escalating doses of ipilimumab at 3–10 mg/kg + RT. There were no dose limited toxicities during the 5-week assessment period. Then the 10-mg/kg cohorts were expanded to 50 patients (ipilimumab monotherapy, 16; ipilimumab + RT, 34). Common (≥15%) immune-related adverse effects of any grade in the 10 mg/kg ± RT group were diarrhea, colitis, rash, and pruritus. Other common treatment-treated AEs were fatigue, nausea, vomiting, and decreased appetite. Sixteen patients (32%) reported immune related AEs of ≥grade 3, most commonly colitis (16%, all grade 3), diarrhea (8%, all grade 3), and hepatitis (4%, grade 3; 6%, grade 4). Any grade treatment-related AEs in 10 mg/kg group versus 10 mg/kg+ RT were 100% (16/16) and 85% (29/34). Any grade immune-related AE in 10 mg/kg group versus 10 mg/kg+ RT were 100% (16/16) and 71% (24/34). ≥3 grade treatment-related AEs in 10 mg/kg group versus 10 mg/kg+ RT were 63% (10/16) and 38% (13/34). ≥3 grade immune-related AEs in 10 mg/kg group versus 10 mg/kg+ RT were 63% (10/16) and 18% (6/34). One treatment-related mortality (5 mg/kg group) occurred on the patient with grade 3 colitis who died from aspergillosis. Among patients receiving 10 mg/kg ± RT, eight had PSA declines of ≥50% (duration: 3–13+ months), one had complete response (duration: 11.3+ months), and six had stable disease (duration: 2.8–6.1 months). This study revealed that in metastatic castration-resistant prostate cancer patients, ipilimumab 10 mg/kg combined with RT concurrently could result in clinical antitumor activity with disease control and manageable AEs. Besides, a phase III prospective, multicenter, randomized controlled study (NCT00861614) to evaluate the sequential use of ipilimumab after RT in patients with metastatic castration-resistant prostate cancer that progressed after docetaxel chemotherapy.⁶² A total of 799 patients with metastatic prostate cancer were included. All of them were advanced in the RT of bone metastatic focus 800 cGy (no more than five target lesions). Then they were randomly assigned to ipilimumab group or placebo group at 1:1, once every 3 weeks, until the disease progressed, serious adverse events occurred, other causes died or the patients voluntarily left the group. Grade 3–4 AEs occurred in 26% patients in the ipilimumab group and 3% patients in the placebo group. The most frequent grade 3–4 adverse events included diarrhea (16% in the ipilimumab group vs. 2% in the placebo group), fatigue (11% vs. 9%), anemia (10% vs. 11%), and colitis (5% vs. 0.1%). Mortality occurred owing to the adverse effects of the agent, all in the ipilimumab group. In the analysis of pelvic RT subgroup, it was observed that among ipilimumab + pelvic radiotherapy group, placebo + pelvic radiotherapy group and ipilimumab + nonpelvic radiotherapy group, grade 3–4 gastrointestinal toxicity was 25.5% (36/141), 10.7% (13/122), and 21.8% (55/252), respectively. The main adverse event was diarrhea, and the data of the three groups were 19.9% (28 /141), 2.5% (3/122), and 14.3% (36/252), respectively.

In summary, the commonly used ICI in pelvic malignance is ipilimumab. The current evidence suggests that receiving ipilimumab after the exposure of RT would result in AE no more severe than administrating ipilimumab alone. The occurrence of severe AE was more likely attributed to ipilimumab instead of the combined strategy. Pelvic radiotherapy would not augment the incidence of grade 3–4 gastrointestinal toxicities together with the use of ipilimumab.

5 RADIATION COMBINED WITH IMMUNE CHECKPOINT INHIBITORS IN HEAD AND NECK MALIGNANCE

Head and neck malignancies are the sixth largest malignancies in the world. RT is an important treatment method for head and neck tumors.⁶³ The common adverse effects caused by RT include acute reactions (dermatitis, mucositis, and xerostomia) as well as chronic reactions (trismus, dysphagia, tissue atrophy, and osteoradionecrosis).⁶⁴ Currently some clinical experiences are available to evaluate the tolerance of RT combined with ICI treatment. In this part, we maintain our focus on clinical trials which aimed at addressing this question exactly in head and neck cancer. Single arm studies such as RTOG 3504⁶⁵ trial demonstrated the safety of the combined treatment of nivolumab concurrently with four chemoradiotherapy regimens for patients with newly diagnosed head and neck squamous carcinoma(HNSCC). Ten enrolled patients for cohort 1 (weekly cisplatin). According to the safety data for cohort 1, seven of eight patients received >70% of prescribed cisplatin, which was discontinued early in three patients due to AEs unrelated to nivolumab. No DLT was observed. Severe AEs included anaphylaxis to cisplatin (n = 1), cholecystitis (1), but none attributable to nivolumab. Grade ≥3 toxicities attributable to nivolumab included fatigue (n = 1), anorexia (1), leukopenia (2), neutropenia (1), mucositis (1), and lipase elevation (1). Seven continued on to maintenance nivolumab. This trial indicated nivolumab is safe and feasible to administer concurrently with a weekly cisplatin-RT regimen for patients with newly diagnosed HNSCC. Apart from this, the 2018 ASCO conference reported preliminary safety results of GORTEC 2015-01 (“PembroRad”) study.⁶⁶ This is a phase II multicenter clinical randomized controlled study. A total of 133 patients with advanced stage III, IVa, and IVb inoperable head and neck squamous cell carcinoma who were not suitable for cisplatin chemotherapy were randomly divided into two groups (1:1 ratio). Group A was cetuximab combined with RT (400 mg/m² loading dose and 250 mg/m² weekly), and Group B was RT combined with Pembrolizumab (200 mg three times a week during RT). In both groups, patients received IMRT (6996 cGy in 33 fractions). Compared with patients administered with cetuximab, the incidence of grade 3 dermatitis, rash, and mucositis were significantly reduced and rates of dysthyroidism were obviously increased in the pembrolizumab arm. Besides, Wise-Draper et al. carried out a study on pembrolizumab in the treatment of locally advanced head and neck squamous cell carcinoma (LA-HNSCC), which included 80 patients.⁶⁷ Two hundred milligrams of pembrolizumab was administered 1 week before surgery, and then radical surgery was performed. Adjuvant concurrent pembrolizumab (three times a week for six doses) and RT (6,000–6,600cGy) were conducted, along with weekly cisplatin (40 mg/m²) for patients with high-risk feature (positive resection margin or extracapsular invasion of tumor lymph nodes). The 2018 ASCO conference reported interim analysis. None of the patients in the pembrolizumab arm present grade 4 adverse effects or discontinuation of treatment due to dose-limiting toxicity. In addition, the pembrolizumab-cisplatin arm had no >grade 3 toxicities recorded in the 19 patients, leading the authors to get a conclusion that the combination strategy is tolerable. Based on the current researches, immune checkpoint inhibitors could be safely administered concurrently or sequentially with RT without exacerbation of expected adverse effects.

6 CHALLENGES OF COMBINATION THERAPY FROM THE PERSPECTIVE A RADIO-ONCOLOGIST

From the perspective of a radio-oncologist, the real challenges of combination therapy include the timing of RT, optimal dose, and fractionations, RT target and target volume, how to select patients and whether all these factors influence radio-immunotherapy toxicity and if there are any potential biomarkers to predict toxicity.

7 CONCLUSION

Caution is warranted and the toxicity profile should be paid attention to when optimizing the ICI and RT combination therapy. Given that accumulating preclinical and clinical data have documented that immunotherapy can augment radiation-mediated local tumor response and radiation can magnify the systemic effects of immuneotherapy,⁹⁸ concerns exist that the combined regimen of RT with immunotherapy would potentially augment the incidence and severity of adverse effects. Even though there have been numerous prospective clinical trials started in recent years to evaluate the safety and efficiency of RT + ICI combining regimens, most of them are still ongoing and the safety results are pending (Table 3). Till now only a small amount of data about this combining strategy are available for us to analyze. By and large, results of the researches available right now suggest that the combing use of RT and ICIs seem to show manageable safety without a notable life-threatening increase in the risk of toxicities. It is important to note that little is known about the potential late effects of RT combined ICI therapy because very few studies have 5 year follow-up right now. Late toxicity, particularly those occurring after 5–10 years is poorly documented and certainly highly underestimated. The acute toxicity with ICI is much less predictable and often life threatening and in some can give rise to permanent effects. Besides, the optimal dose/ fractionated regimen, the sequence of RT combined with ICIs and the RT target/target volume are reviewed in this article. Based on the limited evidence, preliminary conclusion are formed which need to be validated by more researches. Till now no universal biomarker has been identified for the ICI and RT combined therapy, which merit further investigation.

AVAILABILITY OF DATA AND MATERIAL

The data that support the findings of this study are available from the corresponding author upon reasonable request.