Radiotherapy and Immune Checkpoint Blockade for Melanoma: A Promising Combinatorial Strategy In Need of Further Investigation

Abstract

Radiotherapy and immune checkpoint blockade are effective treatments for melanoma. Recent preclinical studies have suggested an interaction of radiotherapy and immune checkpoint blockade. Retrospective clinical studies, as well as a small number of prospective studies have been undertaken to explore this interaction in patients with melanoma. In this review, we present the results of clinical studies combining radiotherapy and immune checkpoint blockade and conclude that this is a promising strategy worth of further investigation.

Introduction

Radiotherapy has long played a role in the treatment of melanoma, as a definitive treatment of in situ melanoma, as an adjuvant therapy after resection of high risk primary tumors or lymph node metastases, and as a palliative intervention for patients with distant metastases.¹

While the primary mechanism of action of radiotherapy is thought to occur through DNA damage and subsequent cell death, there is emerging evidence suggesting that radiation can modulate the host immune system. Case reports of the abscopal effect, which describes treatment response in a non-irradiated metastasis following radiotherapy, support a paradigm shift from radiation functioning as a local therapy to one that might have systemic effects.² It should be noted that attributing an abscopal effect to a local therapy such as radiation, particularly when combined with systemic therapy, is problematic to study. Nevertheless, attempts to identify the underlying mechanisms of radiation and immune system interactions which may lead to anticancer immune effects, as well as how to maximally enhance them, are underway in preclinical and clinical studies.

Preclinical studies of radiotherapy and immune checkpoint blockade in melanoma have demonstrated several important observations. Using mouse models of melanoma (and other cancers), investigators have observed improvements in irradiated tumor control with the combination of immune checkpoint blockade and radiotherapy, over either treatment alone. In preclinical models that incorporated two tumors (one tumor irradiated and the other observed for response), an improvement in control of the non-irradiated tumor was observed, over either treatment alone. In many situations, this translated into longer survival, cure, and immunologic memory. Various radiotherapy approaches have been used, with no clear or consistent approach demonstrating superiority across studies regarding the optimal dose, fractionation, target for treatment, or timing of radiotherapy in relation to immune checkpoint blockade. The reader is referred to recent reviews on preclinical studies of immune checkpoint blockade and radiotherapy for further discussion.^{3, 4}

Clinical studies of radiotherapy and immunotherapy for melanoma have been previously reviewed.⁵ With the recent success and FDA approval of immune checkpoint blocking antibody therapy for melanoma, these agents have become standard first line therapy for metastatic and unresectable melanoma. Therefore, the present review will focus on the published prospective and retrospective studies (Tables 1-3), currently ongoing prospective trials (Table 4) and associated reports with a focus on patients with melanoma.

Table 1. Reports of retrospective studies evaluating of extracranial radiotherapy combined with immune checkpoint blockade in patients with melanoma.

Study	N	Patient description	Treatment groups	End point(s)	Outcomes		Safety
					Clinical	Correlative
Postow et al.14‡	1	Patient with metastatic melanoma receiving SBRT to paraspinal metastasis	Ipi+28.5Gy in 3 fractions Ipi dosing: 3mg/kg or 10mg/kg every three weeks	RR, abscopal response	Marked decrease in SBRT treated site as well as non-irradiated thoracic and splenic metastases 4 months post treatment	Combined ipi and RT resulted in anti-tumor immune activation increased titers of anti-NY-ESO antibodies increase in CD4+ ICOShigh cells decrease in myeloid derived suppressor cells	Asymptomatic hypothyroidism requiring thyroid hormone supplementation
Hiniker et al.33‡	1	Patient with melanoma receiving SBRT to liver metastasis	Ipi+ 54Gy in 3 fractions	RR, abscopal response	Complete radiographic resolution of all non-irradiated liver metastases with combined ipi and SBRT after having experienced POD on ipi alone	Not reported	Autoimmune hypophysitis treated with prednisone
Stamell et al.15‡	1	Patient with metastatic melanoma receiving electron beam radiotherapy to scalp primary	Ipi+24Gy in 3 fractions Ipi dosing: 3mg/kg or 10mg/kg every three weeks	RR, abscopal response	Resolution of all non-irradiated in-transit metastases 8 months following initial therapy	When patient recurred and was treated with SRS, anti-melanoma antigen A3 (MAGEA3) antibodies were observed, implying radiation-induced immune activation	Not reported
Barker et al.17	29	Metastatic melanoma irradiated for extracranial metastases	Ipi+RT (median dose 30 Gy in 5 fractions) Ipi dosing ranged from 3mg/kg to 10mg/kg every three weeks for four doses	OS, safety	Median survival: RT during induction/maintenance Ipi: 9mo/39mo Differences likely due to selection bias	Decrease in absolute lymphocyte count noted in most patients after radiotherapy	Ir-AEs in10mg/kg versus 3mg/kg ipi dose: 25% versus 7% (p=0.005) RT related adverse events higher with EQD2 of >100Gy (α/β = 0.6)
Schiavone et al.16‡	4	Mucosal melanoma of vagina (n=3) and cervix (n=1)	Ipi+RT (6Gy × 5) (n=3) Ipi+RT (2.15Gy × 28) (n=1) Ipi dosing: 3mg/kg or 10mg/kg every three weeks	RR	3 patients taken to post-RT surgery, 1 exhibited pCR, all exhibited complete radiographic response	Not reported	CTCAE grade 3 colitis and rash in 2 patients
Qin et al.18	88	Unresectable stage III or IV melanoma irradiated for extracranial metastases	Ipi +/- RT (variable dose/fractionation) Ipi dose: not reported	OS, PFS, RR	Median Survival: Ipi : 24.8 mo Ipi + RT: 17.9 mo Irradiated tumor response improved if Ipi administered prior to RT (74.7%) versus (44.8%) at 12 mo (P=0.01). No differences in ablative or conventional RT	Not reported	No differences in toxicities across treatment groups
Theurich et al.19	127	Stage IIIC and IV melanoma with cranial and extracranial metastases	Ipi +/- local peripheral therapy (radiation or electrochemotherapy) Ipi dosing: 3mg/kg or 10mg/kg every three weeks	OS, PFS, safety, immune response	Median survival: 13.8 mo Ipi : 10.5 mo Ipi + local peripheral therapy: 23.3 mo (P = 0.0028)	On multivariable analysis, local peripheral therapy associated with statistically significant survival benefit (P=0.05)	Ir-AEs not increased with combination treatment

^‡Denotes case report/series; CR: complete response; PR: partial response; SD: stable disease; POD: progression of disease; LC: local control; LF: local failure; OS: overall survival; RR: response rate; pCR: pathological complete response; ipi: ipilimumab; nivo: nivolumab; SRS: stereotactic radiosurgery; NS: not significant; AE: adverse event ir-AE: immune related adverse events (ir-AE); RT: radiotherapy

Table 3. Reports of prospective trials evaluating of radiotherapy combined with immune checkpoint blockade in patients with melanoma.

Study & Sponsor	Phase	N	Population	Trial groups	End point(s)	Outcomes		Safety
						Clinical	Correlative
Twyman-Saint Victor et al. 11 University of Pennsylvania	I	22	Multiply metastatic melanoma	RT (6Gy × 2 8Gy × 2 6Gy × 3 8Gy × 3) + Ipi (dose not reported) every 3 weeks for four doses	1° endpoint: safety 2° endpoint: tumor response	Median OS: 10.7 mo PFS: 3.8 mo POD: 64%	Pre-clinical analysis suggests that CTLA-4 blockade predominantly inhibits T-regs, PD-L1 blockade reverses dysfunction in T-cell proliferation and effector function (e.g. T-cell exhaustion), and radiation promotes T-cell receptor diversity	No dose limiting toxicities as defined by > Gr 4 immune related or >Gr 3 non-immune-related
Hiniker et al.12 NCT01449279 Stanford University	Pilot	22	Stage IV melanoma	Ipi (3mg/kg) every 3 weeks for 4 doses. Radiation started within 5 days of Ipi start (dose at MD discretion)	1° endpoint: safety, efficacy 2° endpoint: tumor response, biomarker exploration	Median OS: 55 weeks (13.8 mo) PFS: 26 weeks (6.5mo) 3/22 patients CR (32, 55, 65 weeks duration)	Trend for elevated pro-inflammatory markers in responders Possible abscopal tumor size decrease in 6 patients Higher baseline central memory T-cells between patients with CR/PR versus PD	No additional toxicity with combined therapy 14% grade 3-4 toxicity
Tang et al.13 NCT02239900 MD Anderson Cancer Center	I/II	1*	Stage IV uveal melanoma	Ipi (3 mg/kg) every 3 weeks for 4 doses; 50Gy in 4 fractions at weeks 5-6	1° endpoint: safety, efficacy 2° endpoint: tumor response, biomarker exploration	OS: 21 months (AWD as of 8/2016) SD for 6mo, then POD

^*Trial involves many solid malignancies with lung or liver metastases, one patient with uveal melanoma was included in trial; OS: overall survival; AWD: alive with disease; CR: complete response; SD stable disease; POD: progression of disease; PFS: progression free survival

Table 4. Ongoing clinical trials combining radiotherapy with checkpoint blockade for patients with melanoma.

Study Phase	Title	Eligibility (planned patient enrollment)	Treatment Regimen	Sponsor	Registry number
Pilot	Early Biomarkers of Tumor Response in High Dose Hypofractionated Radiotherapy Word Package 3: Immune Response	Patient requiring hypofractionated radiotherapy (≥3 fractions, ≥9Gy per fraction) with hepatocellular carcinoma or hepatic colorectal cancer lesion; metastatic melanoma or renal cell carcinoma (30)	Collection of blood samples before, during and after radiotherapy	Centre Oscar Lambret	NCT02439008
Pilot	Pilot Ipilimumab in Stage IV Melanoma Receiving Palliative Radiotherapy	Unresectable metastatic melanoma with failed 1 systemic therapy (planned 20; actual = 22)	Ipi (3mg/kg) every 3 weeks for 4 doses; radiation starting 2 days after first dose	Stanford University	NCT01449279
Pilot/I	A Pilot Study of Ipilimumab and Radiation in Poor Prognosis Melanoma	resected or unresected high risk, or recurrent cutaneous or mucosal melanoma	RT+Ipi (no dose specified)	Duke University	NCT01996202
I	A Pilot (Phase 1) Study to Evaluate the Safety and Efficacy of Combination Checkpoint Blockade (Ipi and Nivo) Plus External Beam Radiotherapy in Subjects With Stage IV Melanoma	Unresectable stage IV melanoma with at least 1 radiotherapy amenable lesion; ≥28 days from prior treatment (18)	Concurrent Ipi (3 mg/kg) and Nivo (1 mg/kg) every 3 weeks for 4 doses, followed by Nivo monotherapy (3 mg/kg) every 2 weeks. Radiotherapy (Cohort A: 3Gy × 10, Cohort B: 9Gy × 3) will be initiated between the 1st and 2nd dose of immunotherapy.	Ludwig Institute for Cancer Research	NCT02659540
I	Phase I Study of Ipilimumab Combined With Whole Brain Radiotherapy or Radiosurgery for Melanoma	Metastatic melanoma with brain metastasis (24)	Ipi (dose escalation) every 3 weeks for 4 doses; Arm A: whole-brain radiation first two weeks of Ipi; Arm B: Stereotactic radiosurgery on same day as first dose of ipilimumab	Thomas Jefferson University	NCT01703507
I	Phase I Trial of Stereotactic Body Radiotherapy With Concurrent Fixed Dose Immune Checkpoint Inhibitors in Metastatic Melanoma: Dose Limiting Toxicity and Abscopal Effect	Metastatic melanoma with at least 3 extracranial metastases; ≥28 days from prior treatment (21)	Ipi (3 mg/kg) every 3 weeks for 4 doses (days 1, 22, 43 and 64); Stereotactic radiosurgery (dose escalation: 8Gy × 3, 10Gy × 3, 12Gy × 3) on days 39, 41 and 43	University Hospital, Ghent	NCT02406183
I	An Exploratory Study to Investigate the Immunomodulatory Activity of Radiotherapy (RT) in Combination With MK-3475 in Patients With Recurrent/Metastatic Head and Neck, Renal Cell Cancer, Melanoma and Lung Cancer	Recurrent head and neck, lung cancer, renal cell carcinoma, skin cancer; stage III or IV renal cell carcinoma; Stage IV lung or cutaneous melanoma (40)	Four arms, two radiation doses (8Gy ×1 v. 5Gy × 4) either preceded by or followed by dose of pembro on the first day of radiation and repeated doses of pembro every 21 days if no progression	Thomas Jefferson University	NCT02318771
IB	A Phase 1b, Open-label, Multicenter, Multidose, Dose-escalation Study of BMS-936558 (MDX-1106) in Combination With Ipilimumab in Subjects With Unresectable Stage III or Stage IV Malignant Melanoma	Unresectable stage III or IV melanoma (136)	8 arms, varying doses and sequence of nivo and ipi in combination or alone	Bristol-Myers Squibb	NCT01024231
I/II	RADVAX: A Stratified Phase I/II Dose Escalation Trial Of Hypofractionated Radiotherapy Followed By Ipilimumab In Metastatic Melanoma	Metastatic melanoma with index lesion between 1 and 5 cm (40; 23 enrolled prior to recruitment cessation)	Stereotactic radiotherapy (dose escalation) followed by ipi (no dose listed)	University of Pennsylvania	NCT01497808
I/II	Phase I/II Trial of Ipilimumab (Immunotherapy) and Hypofractionated Stereotactic Radiotherapy in Patients With Advanced Solid Malignancies	Many solid malignancies with lung or liver metastases; one uveal melanoma (120)	Ipi (3mg/kg) every 21 days for 4 doses with concurrent (days 1-4, cycle 1) or sequential (days 29-33) stereotactic body radiotherapy (12.5Gy × 4, 6Gy × 10)	M.D. Anderson Cancer Center	NCT02239900
II	Concurrent Ipilimumab and Stereotactic Ablative Radiotherapy (SART) for Oligometastatic But Unresectable Melanoma (SART)	Stage III and IV melanoma with 5 or fewer metastatic sites (50)	Ipi (10mg/kg) every 3 weeks for 4 doses, then every 12 weeks; Stereotactic ablative radiotherapy (dose not defined) between 1st and 3rd doses of ipi	Comprehensive Cancer Centers of Nevada	NCT01565837
II	A Phase 2 Study Using Stereotactic Ablative Radiotherapy and Ipilimumab in Patients With Oligometastatic Melanoma	Stage IV melanoma metastatic to <4 sites (30)	Ipi (3 mg/kg) every 3 weeks for 4 doses; Stereotactic radiotherapy at week 5-6; Patients with SD, PR or CR at week 12 may receive additional ipi	Ohio State University Comprehensive Cancer Center	NCT02107755

Ipi: ipilimumab; nivo: nivolumab; pembro: pembrolizumab

Cytotoxic T-lymphocyte antigen 4 (CTLA-4) blockade combined with radiotherapy for melanoma

Cytotoxic T-lymphocyte antigen 4 is a key negative regulator of T-cell activation in the context of antigen presentation. Inhibition of this signal results in a net activation of cytotoxic T-cell antitumor response. Ipilimumab was the first CTLA-4 inhibitor approved by the FDA for the treatment of unresectable and metastatic cutaneous melanoma based on a significant overall survival benefit observed in two randomized clinical trials.^{6, 7} Ipilimumab has also been approved for use in the adjuvant treatment of patients with high-risk stage III melanoma.⁸ Some patients receiving immunotherapy will have an initial apparent increase in tumor volume which in some cases has been attributed to tumor cell infiltrates. Based upon these occasional clinical findings, immune-related response criteria have been created to avoid denoting these tumor volume increases as progression of disease.⁹

Several prospective and retrospective studies evaluating the safety and efficacy of ipilimumab combined with radiation in unresectable and metastatic melanoma have been published. Most of the studies have focused specifically either on extracranial radiotherapy or brain radiotherapy, and have been separated for discussion and in Tables 1 and 2, respectively.

Table 2. Reports of retrospective trials evaluating of brain radiotherapy combined with immune checkpoint blockade in patients with melanoma.

Study	N	Treatment groups	End point(s)	Outcome		Safety
				Clinical	Correlative
Knisely et al.20	77	Ipi +/- SRS (some also received WBRT or repeat SRS; doses not reported) Ipi dose: not reported	OS	Median survival SRS 4.9 mo SRS+ ipi 21.3 mo No difference in survival noted whether ipi given pre- or post-SRS	Not reported	Not reported
Bot et al.34‡	1	Ipi+WBRT (4Gy × 5) for leptomeningeal disease Ipi dose: 3mg/kg every 3 weeks for a planned four doses	OS, RR	1.5 year survival Complete radiographic response to CNS metastases	Decrease in size of abscopal lung metastases following WBRT	Not reported
Du Four et al.35‡	3	Ipi + RT (3 patients had Stereotactic RT and 2 had WBRT; all had >1 RT course) Ipi dose: 3mg/kg every 3 weeks for a planned four doses	RR, AE	First report of radionecrosis in patients treated with Ipi and radiotherapy	Not reported	Focal radiation necrosis noted
Silk et al.21	70	WBRT (30–37.5 Gy in 10–13) or SRS (14–24 Gy in 1–5 fractions)+/- Ipi Ipi dose: 3 mg/kg every 3 weeks for a planned four doses	OS, RR, AE	Median survival WBRT: 3 mo SRS: 18.3 mo SRS+ipi: 19.9 mo Subset analysis showed SRS+ipi associated with improved overall survival (p=0.009)	More patients with PR or SD in patients receiving ipi before RT	Intratumoral hemorrhage with RT
Mathew et al.22	58	SRS (mean 20Gy) +/- Ipi Ipi dose: 3 mg/kg every 3 weeks for a planned four doses	OS, LC, freedom from new metastasis, AE	Median survival: SRS: ∼5mo SRS+ipi: ∼7mo (NS) Local control and freedom from new metastases not different between SRS and SRS+ipi	No differences noted in outcomes based on when ipi was administered relative to RT	No differences in intracranial hemorrhage
Du Four et al.36‡	4	Ipi + RT (SRS or WBRT+SRS) (3-20Gy per 1-10 fractions) Ipi dose: 3 mg/kg every 3 weeks for a planned four doses (n=3; dose blinded for 1 patient)	RR, AE	Time to histopathologically confirmed radionecrosis (six metastases total) following ipi and RT was 15 mo and 11 mo, respectively	Not reported	Symptomatic radiation necrosis in all patients
Tazi et al.23	10	SRS (dose not reported)+ IpiIpi dose: not reported	OS, AE	Median survival: 29.3 mo Disease-specific graded prognostic assessment (DS-GPA) estimated mean survival was 9.1 months	Not reported	One patients with grade 3 GI toxicity One with hypopituitarism
Patel et al. 24	54	SRS (15-21Gy in 1 fractions or hypofractionated in 3-5 fractions if cavity >40mm) +/- Ipi (within 4 months of SRS) Ipi dose: not reported	OS, LC, AE	Median survival†: SRS: 8mo SRS+ipi: 8mo Local control rates similar in both groups	No differences noted in outcomes based on when ipi was administered relative to RT	Radiation necrosis and intracranial hemorrhage not different in both groups
Schoenfeld et al.37	16	WBRT (median 36Gy) or SRS (median 22 Gy) + Ipi Ipi dose: 3mg/kg (n=14) or 10mg/kg Ipi (n=2)	OS, abscopal response, ir-AE	Median survival: 14.4mo SRS before ipi: 26mo SRS after ipi: 6mo (p < 0.001)	63% of patients receiving cranial RT and ipi within 3mo demonstrated a size decrease in non-irradiated extracranial index metastasis	No significant ir-AEs
Gerber et al.25	13	WBRT (median 30Gy in 10 fractions)+ Ipi Ipi dose: 3 mg/kg (n=12), 10 mg/kg (n=1)	OS, RR, AE	Median survival: 4mo 4/9 evaluable patients demonstrated PR or SD by Modified WHO criteria	5/9 evaluable patients demonstrated PR and SD by immune-related criteria	Grade 3 cognitive change (n=1); All evaluated patients demonstrated new or increased intratumoral hemorrhage (n=10)
Kiess et al.38	46	SRS (median 21 Gy in 1 fraction) + Ipi Ipi dose: 3 mg/kg (n=25) or 10 mg/kg (n= 21) every three weeks for 4 doses (induction), then maintenance every 3 mo (n=13)	OS, LC, AE	Median survival: 12.4 mo OS: significantly worse in the SRS after Ipi cohort (P=0.008) LC: no differences based on timing of Ipi before, during or after SRS	Treated tumors increased to >150% pre-SRS size in 50% in SRS before or during Ipi versus 13% in patients treated with SRS after Ipi 1 patient demonstrated abscopal responses in pelvic and lung metastases	Grade 3/4 toxicities occurred in 20% of patients Intracranial hemorrhage in 40% of patients treated with SRS during Ipi
Shen et al39	193 (36 melanoma primary)	SRS (15-24Gy in 1 fraction, 21-24 in 3 fractions, 25Gy in 5 fractions) + systemic therapy (including 20 with immunotherapy)	OS, RR, AE	Median survival for patients with melanoma brain metastasis: SRS : 11.3 mo SRS+ any systemic therapy: 25.2 mo (P<0.05) Patients with metastatic melanoma treated with SRS were not specifically characterized	Not reported	Higher CNS toxicity with combined SRS and immune therapy
Ahmed et al.26	26 (73 metastases)	+/- resection + nivo + SRS (most common 21 or 24 Gy in 1 fraction) within 6 mo of nivo Analysis of subset of patients within larger trials NCT01176461 and NCT01176474	OS, LC, LF, AE	Median Survival: approximately 12 months from treatment start	Not reported	Hemorrhage and edema noted in failures

^‡Denotes report/series;

^†obtained from Kaplan-Meier curve; CR: complete response; PR: partial response; SD: stable disease; POD: progression of disease; LC: local control; LF: local failure; OS: overall survival; RR: response rate; pCR: pathological complete response; ipi: ipilimumab; nivo: nivolumab; SRS: stereotactic radiosurgery; NS: not significant; AE: adverse event ir-AE: immune related adverse events (ir-AE); RT: radiotherapy

CTLA-4 blockade combined with radiotherapy for extracranial melanoma metastases

Investigators at the University of Pennsylvania were the first to report a prospective phase I trial investigating the feasibility and efficacy of radiotherapy (fractional dose and number of fractions escalated from 6 to 8 Gy and 2 to 3 fractions, respectively) targeting bone, liver, lung and subcutaneous metastases followed by immune checkpoint blockade with ipilimumab (3 mg/kg every 3 weeks for up to 4 doses) in 22 patients with metastatic cutaneous melanoma. The authors did not report a greater than expected rate of radiotherapy or ipilimumab related adverse events at the radiation dose levels assessed. Response rates, overall and progression-free survival in patients treated with the combination of radiotherapy and ipilimumab were relatively similar to patients that received ipilimumab monotherapy in other prior studies.^{6, 10} Correlative immunologic analyses in a small number of patients on the study suggested that outcomes were superior in patients with low levels of PD-L1 expression on melanoma cells, which corresponded to patients that had an increase in the percent of Ki67⁺ GzmB⁺ cells among PD-1⁺ Eomes⁺ CD8 T cells. Because the majority of patients progressed, the investigators attempted to characterize the mechanism of cooperativity between checkpoint blockade and radiation in a companion preclinical study. They found that CTLA-4 blockade predominantly inhibits T-regs, PD-L1 blockade reverses dysfunction in T-cell proliferation and effector function (e.g. T-cell exhaustion), and radiation promotes T-cell receptor diversification.¹¹ The preclinical study suggested that combined blockade of CTLA-4 and PD-1/PD-L1 along with radiotherapy might be an effective regimen to evaluate in future clinical trials.

In another prospective pilot clinical trial of patients with melanoma, investigators at Stanford University reported on 22 patients treated with ipilimumab (3 mg/kg every 3 weeks for up to 4 doses) with radiotherapy initiated after the first dose of ipilimumab. Radiation doses and targets varied widely, with a median dose of 30 Gy in 5 fractions delivered. Eleven patients demonstrated stable disease (n=5), partial response (n=3) or complete response (n=3). Overall, the efficacy profile was favorable compared to historical experience with ipilimumab alone and the prior study from the University of Pennsylvania. When assessing patients' sera for immune markers, investigators noted increased levels of IL-2 producing CD8+ T cells and central memory CD8+ T cells in patients who had partial or complete response when compared with those who progressed.¹²

Most recently, Tang et al. published results of a prospective trial of patients with solid tumor malignancies metastatic to lung or liver treated with ipilimumab (3 mg/kg every 3 weeks for up to 4 doses) and radiotherapy (50 Gy in 4 fractions) after the first or second dose of ipilimumab. While no cutaneous melanoma patients were permitted in the trial, one uveal melanoma patient received 50Gy in 4 fractions to a lung metastasis and demonstrated no appreciable tumor response. This patient ultimately progressed after 6 months, but remains alive with disease 21 months from diagnosis (C. Tang, personal communication, 9/26/16, NCT02239900).¹³

Our group described an abscopal response in a patient following palliative radiotherapy to 28.5 Gy in 3 fractions targeting a paravertebral metastasis while receiving ipilimumab as part of a clinical trial. This case was considered a compelling example of the abscopal effect, as progression of metastatic melanoma was noted many months after the patient initiated ipilimumab, and four months after radiotherapy, follow up imaging showed a decrease in size of both the irradiated metastasis and non-irradiated metastases elsewhere in the thorax and spleen. Correlative immunologic analyses demonstrated changes in several immune parameters including an increase in antibody titers to specific epitopes of the cancer testis antigen NY-ESO-1, an increase in CD4+ ICOS^high cells, and a decrease in myeloid derived suppressor cells.¹⁴ In another case, a patient receiving palliative radiotherapy to an unresectable primary tumor treated to a dose of 24 Gy in 3 fractions demonstrated resolution of all non-irradiated in-transit metastases 8-months following initial radiotherapy. Furthermore, when the patient recurred in the brain and was treated with SRS and ipilimumab, increases in his titer of anti-melanoma antigen A3 (MAGEA3) antibodies were observed. Whether the immunologic changes were a result of SRS, ipilimumab, or both remains unknown, but suggests that immunologic changes can be seen when these treatment modalities have been combined.¹⁵

A case series reported from our center described four patients with mucosal melanoma of the vagina (n=3) and cervix (n=1) treated with initial ipilimumab and radiation, followed by surgery. One patient had a complete pathologic response to ipilimumab and radiation noted at the time of surgery, and others had disease responses.¹⁶ While no unexpected toxicities were reported, it remains unknown which treatment modality is responsible for the observed responses in these patients.

When our group originally investigated adverse events retrospectively in patients receiving radiotherapy during ipilimumab therapy, the likelihood of immune related adverse events did not seem unusually high. The dose of ipilimumab (3mg/kg vs. 10mg/kg) was the primary factor associated with the development of high-grade immune-related adverse events. Likewise, radiotherapy related side effects were primarily related to radiotherapy dose, and did not appear to occur more frequently than would be expected with radiotherapy alone.¹⁷

The timing of radiation treatment relative to the administration of ipilimumab may be an important determinant for response rates, with one study suggesting better irradiated tumor response when ipilimumab was administered prior to radiation (74.7% v 44.8%, p=0.01). While, this and some other studies have not demonstrated a survival benefit of ipilimumab combined with radiation over ipilimumab alone, no randomized trial of ipilimumab with or without radiotherapy has yet been conducted, thus all reported studies to date suffer from selection bias.^{17, 18, 19}

In summary, two small, prospective early phase clinical trials have been conducted to assess the safety of ipilimumab and radiotherapy. The University of Pennsylvania report did not indicate that accrual was complete at the highest radiation dose levels. Moreover, the dose levels that were evaluated were relatively low, and responses of the combination appeared to be no better than ipilimumab alone. In the Stanford trial, radiation doses were slightly higher, and although preliminary, the efficacy appeared favorable. Notably, in the Stanford study, radiotherapy was initiated after immunotherapy, while in the University of Pennsylvania study, radiotherapy was initiated first. Further clinical trials are needed to determine if the radiotherapy dose or sequencing in relation to immunotherapy is associated with the probability of response. Both of the prospective clinical trials reported associations between immunologic responses and clinical responses, but neither reported that a biomarker predicting response is available.

CTLA-4 blockade combined with brain radiotherapy for melanoma brain metastases

In a retrospective analysis, Knisley et al. reported on 77 patients with melanoma brain metastases treated with stereotactic radiosurgery (SRS), with or without ipilimumab. This study demonstrated an overall survival of 4.9 months among patients not receiving immunotherapy, and 21.3 months in patients receiving combination therapy. Notably, the sequencing of ipilimumab in reference to SRS did not impact overall survival.²⁰

Silk et al. retrospectively evaluated 70 patients with melanoma brain metastases treated with either whole brain radiotherapy (WBRT) or SRS in the presence or absence of ipilimumab. Overall survival for patients receiving ipilimumab with or without SRS was 19.9 months versus 18.3 months, respectively. On subset analysis, the authors demonstrated that the combination of SRS with ipilimumab was associated with improved overall survival. ²¹

Mathew et al. reported on 58 patients with melanoma brain metastases treated with either SRS alone (n= 33) or SRS and ipilimumab (n=25) for limited brain metastases. Results demonstrated no treatment timing-dependent difference in overall survival or local control. Additionally, there were no significant differences in intracranial hemorrhage or radiation necrosis noted between treatment groups.²²

Tazi and colleagues published a retrospective analysis of 10 patients diagnosed with brain metastases and treated with combined ipilimumab and SRS when compared to 21 patients with metastatic melanoma without brain metastases who were receiving ipilimumab therapy, reporting similar median survival of 29.3 months versus 33.1 months, respectively (HR = 0.93, p= 0.896). The median survival observed in the cohort of patients treated with ipilimumab and SRS was considerably greater that what would have been expected based on the disease-specific graded prognostic assessment (DS-GPA) estimated mean survival of 9.1 months.²³

Another retrospective study evaluated 54 patients with melanoma brain metastases treated at Emory University, 20 of whom received ipilimumab within 4 months of SRS. Authors found no statistical differences in overall survival, local control, or radiation toxicities (e.g. radiation necrosis or hemorrhage) between the patients who received SRS alone versus SRS and ipilimumab.²⁴

Our group evaluated a small cohort (n=13) of patients treated for metastatic melanoma with whole-brain radiation therapy within 30 days of ipilimumab administration. Although whole brain radiotherapy is often thought to be an ineffective therapy in melanoma, this analysis revealed radiographically stable disease or partial response in five of the nine evaluable patients.²⁵

Finally, our group also reported on a series of 46 patients with melanoma brain metastases who all received ipilimumab and underwent as SRS for brain metastases. Survival was longest amongst the patients selected to receive SRS during ipilimumab immunotherapy. A longer overall survival was associated with DS-GPA score and concurrent administration of SRS and ipilimumab, but not age, KPS, lactate dehydrogenase or dose of ipilimumab.

In summary, several retrospective studies have been conducted to assess the combination of ipilimumab and brain radiotherapy in patients with metastatic melanoma. The majority of the studies suggest a possible improvement in response with this combination, compared to radiotherapy alone. Moreover, most studies suggest that the sequence of therapy may be important, with concurrent administration of radiotherapy and ipilimumab yielding the highest response rates. Nevertheless, a formal prospective assessment of the safety and efficacy of this combination in the brain has not yet been reported. Furthermore, interpretation of these studies has been limited by challenges of retrospectively assessing efficacy due to inherent selection biases, similar to the retrospective experiences assessing extracranial radiotherapy and ipilimumab as previously mentioned.

Programmed death 1 (PD1) blockade combined with radiotherapy for melanoma

The success of ipilimumab in melanoma has resulted in an increased interest in developing T cell costimulatory and coinhibitory drugs. Of these, antibodies blocking the programmed death-1 (PD-1) and programmed death ligand-1 (PD-L1) interaction have demonstrated significant promise and have since been approved for melanoma, as well as other malignancies. Pembrolizumab and nivolumab, both monoclonal antibodies against PD-1 were approved by the FDA in 2014 for unresectable or metastatic melanoma, and both have demonstrated improvements in overall survival.

Investigators at Moffitt Cancer Center reported on a retrospective series of 26 patients with 73 melanoma brain metastases treated with or without resection followed by SRS within 6 months of nivolumab. Authors reported overall survival of approximately 12 months from treatment start in both groups, and that treatment was generally well tolerated. Of note, these patients were part of a prospective phase I trial of nivolumab with or without multipeptide vaccines.²⁶

Our group retrospectively investigated patients participating in a phase I clinical trial of ipilimumab and nivolumab who also received radiotherapy. In a small cohort of patients (n=9), there appeared to be no unexpected adverse events after treatment. Among patients who did not achieve a complete response to immunotherapy alone, those who received radiotherapy after initial response assessment appeared to go on to have higher rates of response as best overall response compared to the initial response assessment.²⁷ Based on these results and those of the preclinical study by Twyman-Saint Victor et al., a prospective clinical trial investigating ipilimumab, nivolumab and radiotherapy is currently underway at our center as well as others (NCT 02659540).

To better determine the synergistic effects of radiation when combined with CTLA-4 and/or PD-1 blockade, several clinical trials have been initiated. Some of these trials simultaneously are examining the effect of radiation fractionation on immune modulation (Table 4).

Future Directions

While several retrospective studies have been published, only two prospective trials for patients with metastatic melanoma receiving combined radiotherapy and immune checkpoint blockade have been reported to date.^{11, 12} Response rates vary widely and may reflect analyses of small heterogeneous cohorts and treatment regimens. Importantly, adverse event rates do not appear dramatically different from those seen in either radiation or immune checkpoint blockade alone, indicating a relatively acceptable safety profile, especially when compared to cytotoxic chemotherapy or even other molecularly targeted therapy for melanoma.

As prospective studies mature, we will hopefully be able to better understand the mechanism of synergy between radiation and immunotherapy with the goal of identifying predictive biomarkers of treatment response. Initial attempts to define reliable companion diagnostic tests to predict patient response to blockade of the PD-1/PD-L1 axis, for example, were centered on immunohistochemical staining of tumor sections for PD-L1. A positive PD-L1 test has been positively associated with benefit from PD-1 approaches, but since patients with PD-L1 negative tumors can also benefit, the clinical utility of this biomarker in melanoma remains unclear.²⁸ Furthermore, whether PD-L1 expression is more relevant on tumor cells or on immune cell infiltrates remains to be clarified.^{29, 30} A promising finding within the realm of functional genomics is the observation that the non-synonymous mutational burden appears to be associated with outcomes following CTLA-4 and PD1 blockade in the context of melanoma and non-small cell lung cancers.^{30, 31, 32}

As we learn more about the interplay between radiation and its role in immune modulation, we will be able to more rationally combine it with immunotherapy. By fully characterizing those individuals who exhibit extraordinary responses to these therapies, we might learn which biological pathways are dominant and possibly manipulate these to shift patients from non- and partial-responders into complete responders.

Partnerships between clinicians (medical and radiation oncologists) and research scientists (cancer biologists, immunologists, radiobiologists) are essential to fully understand how the immune response to cancer may be manipulated through checkpoint modulating antibodies, and how radiotherapy may further augment the anti-cancer immune response. Collaboration with the pharmaceutical industry to leverage biologic insights into clinically effective treatments for patients with metastatic melanoma will be essential. While the outlook for patients with metastatic melanoma has dramatically improved in the past few years with the advent of more effective systemic immunotherapy, there is still significant room for improvement. Combinations of radiotherapy and immune checkpoint blockade are one such area for further investigation.