- Email: [email protected]
Ipilimumab in patients with melanoma and brain metastases: an open-label, phase 2 trial
Articles
Ipilimumab in patients with melanoma and brain metastases: an open-label, phase 2 trial Kim Margolin, Marc S Ernstoff, Omid Hamid, Donald Lawrence, David McDermott, Igor Puzanov, Jedd D Wolchok, Joseph I Clark, Mario Sznol, Theodore F Logan, Jon Richards, Tracy Michener, Agnes Balogh, Kevin N Heller, F Stephen Hodi
Summary Background Brain metastases commonly develop in patients with melanoma and are a frequent cause of death of patients with this disease. Ipilimumab improves survival in patients with advanced melanoma. We aimed to investigate the safety and activity of this drug specifically in patients with brain metastases. Methods Between July 31, 2008, and June 3, 2009, we enrolled patients with melanoma and brain metastases from ten US centres who were older than 16 years into two parallel cohorts. Patients in cohort A were neurologically asymptomatic and were not receiving corticosteroid treatment at study entry; those in cohort B were symptomatic and on a stable dose of corticosteroids. Patients were to receive four doses of 10 mg/kg intravenous ipilimumab, one every 3 weeks. Individuals who were clinically stable at week 24 were eligible to receive 10 mg/kg intravenous ipilimumab every 12 weeks. The primary endpoint was the proportion of patients with disease control, defined as complete response, partial response, or stable disease after 12 weeks, assessed with modified WHO criteria. Analyses of safety and efficacy included all treated patients. This trial is registered with ClinicalTrials.gov, number NCT00623766. Findings We enrolled 72 patients: 51 into cohort A and 21 into cohort B. After 12 weeks, nine patients in cohort A exhibited disease control (18%, 95% CI 8–31), as did one patient in cohort B (5%, 0·1–24). When the brain alone was assessed, 12 patients in cohort A (24%, 13–38) and two in cohort B (10%, 1–30) achieved disease control. We noted disease control outside of the brain in 14 patients (27%, 16–42) in cohort A and in one individual (5%, 0·1–24) in cohort B. The most common grade 3 adverse events in cohort A were diarrhoea (six patients [12%]) and fatigue (six [12%]); in cohort B, they were dehydration (two individuals [10%]), hyperglycaemia (two [10%]), and increased concentrations of serum aspartate aminotransferase (two [10%]). One patient in each cohort had grade 4 confusion. The most common grade 3 immune-related adverse events were diarrhoea (six patients [12%]) and rash (one [2%]) in cohort A, and rash (one individual [5%]) and increased concentrations of serum aspartate aminotransferase (two [10%]) in cohort B. One patient in cohort A died of drug-related complications of immune-related colitis. Interpretation Ipilimumab has activity in some patients with advanced melanoma and brain metastases, particularly when metastases are small and asymptomatic. The drug has no unexpected toxic effects in this population. Funding Bristol-Myers Squibb.
Introduction Estimates of frequency of brain metastases in patients with advanced melanoma range from 10% to 50%; autopsy series suggest the value could be as high as 66–75%.1 Median overall survival is only 4 months after diagnosis of brain metastases.2,3 The resistance of melanoma to radiotherapy and cytotoxic chemotherapies, and its frequent metastasis to the brain, render its management particularly challenging. Treatment options for patients with brain metastases are surgery, stereotactic radiosurgery, and whole-brain radiation. In an analysis of 686 patients with melanoma and brain metastases, Fife and colleagues3 reported median survival of 9 months after surgery with or without radiotherapy, but whole-brain irradiation did not extend life. Retrospective data from several other reports4–7 of treatment for melanoma metastatic to the brain suggest that stereotactic radiotherapy is more advantageous than is whole-brain irradiation and that surgery and stereotactic radiotherapy have similar www.thelancet.com/oncology Vol 13 May 2012
effects. However, the patient cohorts in these reports were heterogeneous in size, number of brain lesions, extracranial disease, and treatment variables. The effect of concomitant corticosteroids might be to improve symptoms, but they could also be immunosuppressive, and data from published trials do not distinguish the outcomes by corticosteroid use. However, it is generally assumed that in older series the dose and duration of corticosteroid treatment might have been related to the size, number, and symptoms of brain metastases and peritumoral oedema.7,8 In the USA, temozolomide is the most widely used systemic treatment for patients with melanoma and brain metastases. Although this drug can cross the blood–brain barrier, clinical responses are noted in roughly 10% as a single agent or combined with wholebrain irradiation or other agents.9,10 The development of immune-modulating agents creates new possibilities. Previously, immune treatments were thought to be ineffective in patients with brain metastases because they
Lancet Oncol 2012; 13: 459–65 Published Online March 27, 2012 DOI:10.1016/S14702045(12)70090-6 See Comment page 434 University of Washington, Seattle, WA, USA (Prof K Margolin MD); Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA (M S Ernstoff MD); The Angeles Clinic and Research Institute, Santa Monica, CA, USA (O Hamid MD); Massachusetts General Hospital Cancer Center, Boston, MA, USA (D Lawrence MD); Beth Israel Deaconess Medical Center, Boston, MA, USA (D McDermott MD); Vanderbilt University Medical Center, Nashville, TN, USA (I Puzanov MD); Memorial Sloan-Kettering Cancer Center, New York, NY, USA (J D Wolchok MD); Loyola University, Chicago, IL, USA (Prof J I Clark MD); Yale Cancer Center, New Haven, CT, USA (M Sznol MD); Indiana University Simon Cancer Center, Indianapolis, IN, USA (T F Logan MD); Oncology Specialists SC, Park Ridge, IL, USA (J Richards MD); Bristol-Myers Squibb, Plainsboro, NJ, USA (T Michener PharmD); Bristol-Myers Squibb, Braine-l’Alleud, Belgium (A Balogh MS); Bristol-Myers Squibb, Lawrenceville, NJ, USA (K N Heller MD); and Dana-Farber Cancer Institute, Boston, MA, USA (F S Hodi MD) Correspondence to: Prof Kim Margolin, Seattle Cancer Care Alliance, Seattle, WA 98109, USA [email protected]
459
Articles
would not cross an intact blood–brain barrier, but research has now shown that activated T cells can pass through this barrier.11,12 Therefore, treatments that stimulate T-cell responses might be effective against brain metastases of melanoma.13,14 Phase 2 studies15–17 have shown that ipilimumab is active in advanced melanoma. Phase 3 trials have been designed with peptide vaccine18 or dacarbazine19 comparators and established that groups given ipilimumab have improved survival, supporting approval of the drug as monotherapy at 3 mg/kg in the USA and Europe. In one of these studies,18 61 (11%) of 540 patients in two treatment groups given ipilimumab had brain metastases. Hazard ratios for death in patients with brain metastases were 0·70 (95% CI 0·41–1·20) for 46 individuals given ipilimumab plus gp100 peptide vaccine and 0·76 (0·38–1·54) for 15 patients given ipilimumab alone. These values are similar to those of patients without brain metastases.20 The safety profile and pattern of immune-related adverse events (irAEs) were also similar in patients with or without brain metastases. Case studies of two patients with melanoma— one from a phase 2 study21 and one given ipilimumab on a compassionate-use basis22—also showed benefits of ipilimumab in treatment of brain metastases. Studies of ipilimumab combined with either temozolomide or fotemustine in this setting are in progress.23 We aimed to assess efficacy and safety of ipilimumab specifically in patients with melanoma and brain metastases in a prospective clinical trial. We investigated ipilimumab’s effects in a cohort of patients who did not need treatment with corticosteroids, before testing the drug in a less favourable group of individuals who needed corticosteroids for clinical or radiological control of their brain metastases.
Methods Study design and patients Between July 31, 2008, and June 3, 2009, we undertook an open-label, phase 2 trial with two parallel cohorts. We used a two-stage, modified Gehan design.24 In stage one, we enrolled asymptomatic patients from ten centres in the USA into cohort A to assess the effect of ipilimumab monotherapy on brain and extracranial metastases. This stage met efficacy parameters (three partial responses on the basis of modified WHO [mWHO] criteria25) with acceptable tolerability. The study proceeded to stage two, in which accrual to cohort A continued and enrolment into cohort B, consisting of patients with symptomatic metastases controlled with corticosteroids, began at the same US centres. Patients older than 16 years with histologically confirmed metastatic melanoma were eligible. Patients had to have at least one measurable index brain metastasis of 0·5–3 cm in diameter, or two measurable lesions larger than 0·3 cm visible on contrast MRI, or both. Patients in cohort A were neurologically asymptomatic 460
and had received no systemic corticosteroid treatment in the 10 days before start of ipilimumab treatment. Patients given concurrent systemic corticosteroids for control of brain-metastasis-related symptoms (judged by the treating physician) or oedema were enrolled into cohort B. Participants’ Eastern Cooperative Oncology Group (ECOG) scores had to be 0 or 1. At least 28 days must have elapsed since the last systemic treatment. Previous focused radiotherapy and whole-brain irradiation had to have been completed at least 14 days before start of ipilimumab. The stereotactic radiotherapy field should not have included the brain index lesion; alternatively, the lesion had to be progressive and measurable in two dimensions after any irradiation. Patients with active autoimmune disease were excluded. Participants could have had any number of previous treatments, but patients were ineligible when they had received other immunomodulatory antibodies (eg, other anti-CTLA-4, anti PD-1 or PD-L1, agonistic anti-CD40 or anti-OX40, or anti-CD137). The protocol was approved by institutional review boards or independent ethics committees of participating centres. The study was done in accordance with the Declaration of Helsinki and with good clinical practice as defined by the International Conference on Harmonization. All participating patients gave voluntary, written informed consent.
Procedures Patients were to receive one dose of 10 mg/kg intravenous ipilimumab every 3 weeks to a total of four doses (designated weeks 1, 4, 7, and 10; induction). Patients who were clinically stable at 24 weeks were eligible to continue with treatment with ipilimumab 10 mg/kg every 12 weeks (maintenance). Treatment continued until progression of the global tumour burden (brain and nonCNS lesions) measured with immune-related response criteria (irRC),18,19,26 a related adverse event necessitating discontinuation of ipilimumab treatment, clinical deterioration (as per protocol), or withdrawal of consent. Imaging of brain lesions and non-CNS tumour lesions was done every 6 weeks during induction and every 12 weeks thereafter (or more frequently when clinically indicated) to assess tumour response. The primary endpoint was the proportion of patients with disease control (complete response, partial response, or stable disease) after 12 weeks, assessed with mWHO criteria. Secondary endpoints were the proportion of patients with an objective response (a complete or partial response), median progression-free survival, overall survival, safety, and tolerability. At every tumour assessment, response was assessed in brain, non-CNS, and global (brain and non-CNS) anatomical compartments. We used both mWHO criteria and irRC to assess endpoints. Both systems use the sum of the products of two-dimensional tumour measurements; however, with irRC the measurements of new lesions are www.thelancet.com/oncology Vol 13 May 2012
Articles
Cohort A (n=51)
Cohort B (n=21)
Age (years) Median (range)
Cohort A (n=51)
Cohort B (n=21)
mWHO
mWHO
irRC
irRC
59 (33–79)
57 (30–74)
CR
0
0
0
Male
33 (65%)
11 (52%)
PR
5 (10%)
5 (10%)
1 (5%)
1 (5%)
Female
18 (35%)
10 (48%)
SD
4 (8%)
8 (16%)
0
1 (5%)
PD†
40 (78%)
36 (71%)
0
25 (49%)
14 (67%)
2* (4%)
2* (4%)
1
26 (51%)
7 (33%)
CR
0
0
1 (5%)
1 (5%)
CNS
48 (94%)
19 (90%)
PR
8 (16%)
8 (16%)
0
0
Lung
22 (43%)
12 (57%)
SD
4 (8%)
5 (10%)
1 (5%)
1 (5%)
Liver
13 (25%)
8 (38%)
PD†
39 (76%)
38 (75%)
19 (90%)
19 (90%)
Other (visceral)
15 (29%)
12 (57%)
Unknown
Lymph node
13 (25%)
11 (52%)
Non-CNS
Soft tissue
Sex
ECOG performance status
Site of index lesion
Global
Unknown
20 (95%)
0
19 (90%)
0
0
CNS
0
0
0
0 0
12 (24%)
11 (52%)
CR
0
0
0
Skin
3 (6%)
2 (10%)
PR
7 (14%)
7 (14%)
1 (5%)
1 (5%)
Other
7 (14%)
2 (10%)
SD
7 (14%)
10 (20%)
0
1 (5%)
PD†
35 (69%)
32 (63%)
Any
40 (78%)
15 (71%)
2 (4%)*
2 (4%)*
Immunotherapy
24 (47%)
5 (24%)
Interferon
19 (37%)
4 (19%)
Interleukin 2
16 (31%)
3 (14%)
Previous systemic treatment
Biochemotherapy
2 (4%)
0
Targeted agent
8 (16%)
3 (14%)
Experimental protocol
10 (20%)
1 (5%)
Chemotherapy
22 (43%)
14 (67%)
16 (31%)
13 (62%)
Carboplatin
2 (4%)
4 (19%)
Paclitaxel
3 (6%)
4 (19%)
Vinblastine
2 (4%)
4 (19%)
Temozolomide
Cisplatin
3 (6%)
Unknown
20 (95%)
19 (90%)
0
0
mWHO=modified WHO criteria. irRC=immune-related response criteria. CR=complete response. PR=partial response. SD=stable disease. PD=progressive disease. *No non-CNS lesions were noted for these two patients; therefore, both non-CNS and global response were classified as “unknown”. †PD includes patients lost to follow-up or for whom no tumour assessment was done (eg, primary lesion resected or only brain MRI completed).
Table 2: Disease response by mWHO and irRC criteria after 12 weeks
Cohort A (n=51) mWHO
Cohort B (n=21) irRC
mWHO
irRC
9 (18%, 8–31)
13 (25%, 14–40)
1 (5%, 0·1–24)
2 (10%, 1–30)
4 (19%)
Global disease control
12 (24%, 13–38)
13 (25%, 14–40)
2 (10%, 1–30)
2 (10%, 1–30)
14 (27%, 16–42)
17 (33%, 21–48)
1 (5%, 0·1–24)
2 (10%, 1–30)
Abraxane
2 (4%)
3 (14%)
CNS disease control
Dacarbazine
3 (6%)
0
Non-CNS disease control
Previous radiotherapy
Global objective response
5 (10%, 3–21)
5 (10%, 3–21)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
Any
CNS objective response
8 (16%, 7–29)
8 (16%, 7–29)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
Non-CNS objective response
7 (14%, 6–26)
7 (14%, 6–26)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
To the brain Whole brain Gamma knife or targeted
31 (61%) 21 (41%) 17 (33%) 4 (8%)
9 (43%) 10 (48%) 5 (24%)
Data are n (%, 95% CI). mWHO=modified WHO criteria. irRC=immune-related response criteria.
0 Table 3: Disease control and objective response after 12 weeks
Data are n (%) unless otherwise stated. ECOG=Eastern Cooperative Oncology Group.
Table 1: Demographic and baseline characteristics
included in assessment of tumour burden, and new lesions alone do not define progressive disease. Adverse events were graded according to National Cancer Institute’s Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0. irAEs—defined as any adverse events causally related to drug exposure and consistent with an immune-mediated mechanism—and brain-related events were assessed. irAEs have been attributed to immune or inflammatory reactions that are a result of ipilimumab’s mechanism of action.18,19 Serious adverse events were reported from time of consent. The safety analysis was based on the frequency www.thelancet.com/oncology Vol 13 May 2012
and severity of adverse events for all patients who received at least one dose of ipilimumab. Worst toxicity grades per patient were tabulated by treatment cohort for adverse events and laboratory measurements with CTCAE. Safety data were reported between the first dose and 70 days (five half-lives) after last dose. All toxicity noted was summarised by grade, attribution, and time of onset.
Statistical analysis In stage one, we planned for enrolment of 21 patients into cohort A. If fewer than two objective responses assessed with mWHO criteria were recorded or fewer than eight patients were progression-free at 12 weeks in 461
Articles
Cohort A Cohort B
100
Overall survival (%)
80
60
40
20
0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
11 3
11 3
11 3
11 2
10 1
7 0
4 0
2 0
2 0
2 0
2 0
1 0
0 0
Time after start of treatment (months) Number at risk Cohort A 51 Cohort B 21
49 19
43 13
38 11
33 10
28 10
27 8
23 7
21 6
18 5
18 5
15 5
14 4
13 4
13 4
13 4
12 4
12 4
11 4
11 4
Figure: Overall survival Crosses indicate censored patients.
these 21 patients, the study would be terminated early because of insufficient activity. Otherwise, we would accrue a further 20 patients, and cohort B would open to enrol 21 patients. All treated patients were included in efficacy and safety analyses. We computed the proportion of patients with disease control and the proportion with an objective response with both mWHO and irRC criteria. We calculated exact two-sided 95% CIs with the method of Clopper and Pearson.27 Time-to-event endpoints such as duration of response, progression-free survival, and overall survival were estimated with the Kaplan-Meier product limit method, together with a two-sided 95% CI for the median, calculated with the method of Brookmeyer and Crowley.28 Statistical analyses were done in SAS (version 9.2). This study is registered with ClinicalTrials.gov, number NCT00623766.
Role of the funding source The trial was designed jointly by senior academic researchers and the sponsor, Bristol-Myers Squibb. Data were collected by the sponsor and analysed in collaboration with senior academic authors, who vouch for the completeness and accuracy of the data and analyses. All interpretation of the data was by KM and academic coworkers. KNH, TM, and AB had access to raw data; KNH served as medical monitor; TM was protocol manager; and 462
AB was study statistician. An initial draft of the report was prepared jointly by academic researchers and a professional medical writer employed by the sponsor. All authors contributed to subsequent drafts and had final responsibility for the decision to submit for publication.
Results We enrolled 72 patients: 51 into cohort A and 21 into cohort B. Ten patients remained in the study on April 25, 2011: eight in cohort A and two in cohort B. Table 1 shows demographic and baseline characteristics of the enrolled participants. 15 patients in cohort A completed induction. 28 individuals had discontinued treatment because of early progression of disease or death, five requested discontinuation without documented progression, and three had adverse events. 11 patients in cohort A started maintenance treatment; of the other four, three had progressed and one was taken off treatment because of adverse events before this stage began. While on maintenance treatment, one patient underwent resection of a brain metastasis and discontinued ipilimumab, and one had progression of disease. Five patients in cohort B completed induction. 11 patients had discontinued treatment because of disease progression or death, three requested discontinuation, and two had adverse events. Two patients received maintenance treatment; of the other three, one did not www.thelancet.com/oncology Vol 13 May 2012
Articles
start because of adverse events, one because of progressive disease, and one had died. The median number of doses of induction treatment in cohort A was three (range one to four) and in cohort B two (one to four). Patients in cohort A received a mean of six doses of maintenance treatment (one to ten); both patients in cohort B who entered the maintenance phase received seven doses. The proportion of patients who achieved disease control varied when mWHO criteria or irRC were used. Four patients in cohort A had progressive disease with mWHO, but stable disease with irRC, as did one in cohort B (table 2). After week 12, nine patients in cohort A exhibited disease control (18%, 95% CI 8–31), as did one patient in cohort B (5%, 0·1–24) by mWHO criteria; by irRC, 13 patients achieved disease control (25%, 14–40) in cohort A, as had two (10%, 1–30) in cohort B (tables 2, 3). Control of nonCNS lesions and brain lesions is shown in table 3. The proportion of patients achieving an objective response after 12 weeks was the same with mWHO or irRC (table 3). No patients had a discordant (brain vs nonCNS) response status. No patient had a global complete response, but partial responses were recorded in both cohorts (table 2). 36 participants in cohort A and 19 in cohort B died. Median overall survival was 7·0 months (95% CI 4·1–10·8) in cohort A and 3·7 months (1·6–7·3) in cohort B (figure). For patients in cohort A, overall survival was 55% (95% CI 41–68) at 6 months, 31% (18–44) at 12 months, and 26% (14–39) at both 18 months and 24 months (figure). Those in cohort B had an overall survival of 38% (17–59) at 6 months, 19% (2–36) at both 12 months and 18 months, and 10% (0–22) at 24 months (figure). Table 4 shows duration of progression-free survival. The most common adverse events in both cohorts were fatigue, diarrhoea, nausea, headache, rash, and pruritus (table 5). The most common grade 3 events overall were diarrhoea, fatigue, dehydration, hyperglycaemia, and increased concentrations of serum aspartate aminotransferase (table 5). One patient in each cohort experienced grade 4 confusion. Three patients in cohort A had brain oedema (one grade 1 and two grade 2). A 66-year-old man in cohort A died for reasons related to the study drug 24 days after his second dose of ipilimumab because of gangrenous colitis, despite treatment with corticosteroids. The incidence of irAEs (all grades) was similar in both cohorts (table 5), with diarrhoea, rash, pruritus, and increased serum aspartate aminotransferase the most common. There were no grade 4 irAEs and the most common grade 3 irAEs were diarrhoea in cohort A and increased serum aspartate aminotransferase in cohort B (table 5). The most common CNS-related events were grade 1 and 2 headache and dizziness, which were possibly related to ipilimumab. One patient had grade 4 brain haemorrhage attributed to metastatic melanoma, but ipilimumab could have been implicated. www.thelancet.com/oncology Vol 13 May 2012
Cohort A
Cohort B
mWHO
irRC
mWHO
Overall
1·4 (1·2–2·6)
2·7 (1·6–3·7)
1·2 (1·2–1·3)
irRC 1·3 (1·2–2·5)
Brain
1·5 (1·2–2·5)
1·9 (1·2–2·9)
1·2(1·2–1·3)
1·2 (1·2–1·3)
Non-CNS
2·6 (1·3–4·1)
3·3 (2·6–4·7)
1·3 (1·2–2·5)
1·3 (1·2–2·5)
Data are months (95% CI). mWHO=modified WHO criteria. irRC=immune-related response criteria.
Table 4: Median progression-free survival
Cohort A (n=51) Any grade* Grade 3
Cohort B (n=21) Grade 4
Any grade* Grade 3 Grade 4
Any event Diarrhoea
25 (49%)
6 (12%) 0
9 (43%)
0
0
Nausea
22 (43%)
3 (6%)
0
4 (19%)
0
0
Vomiting
13 (25%)
3 (6%)
0
1 (5%)
0
0
0
0
4 (19%)
0
0
6 (12%) 0
12 (57%)
1 (5%)
0
Constipation Fatigue Oedema (peripheral)
8 (16%) 28 (55%) 4 (8%)
1 (2%)
0
5 (24%)
0
0
Headache
18 (35%)
2 (4%)
0
6 (29%)
0
0
Dizziness
11 (22%)
0
0
2 (10%)
0
0
Rash
19 (37%)
1 (2%)
0
7 (33%)
1 (5%)
0
Pruritus
16 (31%)
0
0
6 (29%)
0
0
Decreased appetite
14 (27%)
2 (4%)
0
4 (19%)
0
0
Dehydration
5 (10%)
2 (4%)
0
4 (19%)
2 (10%) 0
Hyperglycaemia
4 (8%)
2 (4%)
0
4 (19%)
2 (10%) 0
8 (16%)
0
0
4 (19%)
1 (5%)
0
11 (22%)
0
0
2 (10%)
0
0
1 (2%)
0
4 (19%)
2 (10%) 0
Back pain Cough Aspartate aminotransferase increased
4 (8%)
Confused state
9 (18%)
1 (2%)
1 (2%)
3 (14%)
1 (5%)
1 (5%)
Insomnia
8 (16%)
0
0
4 (19%)
0
0
Immune-related events Diarrhoea
22 (43%)
6 (12%) 0
8 (38%)
0
0
Rash
17 (33%)
1 (2%)
0
6 (29%)
1 (5%)
0
Pruritus
16 (31%)
0
0
5 (24%)
0
0
3 (6%)
0
0
4 (19%)
2 (10%) 0
Aspartate aminotransferase increased CNS-related events Headache
18 (35%)
2 (4%)
0
6 (29%)
0
0
Dizziness
11 (22%)
0
0
2 (10%)
0
0
Data are n (%). Events occurring in at least 15% of patients in either cohort. *Grades 1–5 with National Cancer Institute’s Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0.
Table 5: Adverse events
Discussion We have established that ipilimumab shows activity in patients with melanoma and brain metastases, particularly when they have stable, asymptomatic metastases that do not need corticosteroid treatment (panel). Although the primary objective of this study was to estimate the proportion of patients for whom ipilimumab controlled disease, we also investigated safety, because initial tumour growth and peritumoral inflammatory changes can cause neurological complications. However, the results of our 463
Articles
Panel: Research in context Systematic review We searched PubMed with the terms “melanoma” and “brain metastases” for reports published in English. We did not use any specific date restrictions, but only reports published in roughly the past decade were included to ensure data were up to date. We established that few effective treatment options are available for patients with melanoma metastatic to the brain.1–3,7–12 Retrospective analyses of patients with advanced melanoma and brain metastases from a phase 3 trial of ipilimumab,18 and case studies of two individuals with brain metastases20,21 suggested that ipilimumab was active in brain metastases, serving as the rationale for our study. Interpretation As far as we are aware, our prospective study was the first to investigate ipilimumab specifically in patients with advanced melanoma and brain metastases. Our data show that the drug is active in these individuals. Antitumour activity, 2-year survival, and safety in this study are similar to what has been reported in patients without brain metastases. We believe that these data are sufficient to support use of ipilimumab in selected patients with brain metastases, particularly metastases that are small and asymptomatic. Additionally, we have provided safety and initial efficacy data that could serve as the foundation for future trials of combinations of ipilimumab and radiotherapy, other immunotherapies, or possibly molecularly targeted agents.
trial strongly suggest that treatment of patients similar to those in our study will not increase morbidity. The number of adverse events noted did not seem to be higher than has been reported previously in patients with melanoma free of brain metastasis, and we did not record unique events. Although the small patient numbers prevented formal analysis, we did not note an association between irAEs and response or long survival times. Additionally, the numbers of patients with tumour regression in our trial and in the others are alike. Differences in characteristics of participants and protocol procedures prohibit direct comparison between trials, but we believe that our data confirm the safety of this approach in patients with brain metastases. Furthermore, a retrospective analysis29 of a phase 2 trial showed that 12 patients with stable brain metastases given ipilimumab had overall survival of 14 months. Two patients with partial responses and one with stable disease survived for longer than 4 years.29 In our trial, different clinical outcomes were recorded in the two cohorts, which could be a result of prognostic variation but could also be related to a negative effect of steroids on ipilimumab activity. Whereas other researchers have reported that patients with both an antitumour response and an irAE do not lose the antitumour effect when treated with corticosteroids,30 the 464
possibility remains that steroid treatment at initiation of ipilimumab could abrogate or downmodulate an immune response to checkpoint blockade. However, the objective responses and survival plateau in cohort B indicate that steroids do not entirely eliminate the possibility of response to immune checkpoint blockade. Nevertheless, the apparent low rate of benefit in this cohort suggests that further study is needed into patients with specific unfavourable prognostic factors and into avoidance of steroid use at the start of ipilimumab treatment. Overall, our trial shows that the activity of ipilimumab is similar in patients with advanced melanoma with and without brain metastases. Further investigations could assess combinations of ipilimumab and chemotherapy (eg, fotemustine in the NIBIT-M1 trial23) and others, such as molecularly targeted small molecules and other immunomodulatory strategies (eg, new checkpointblocking antibodies alone or in combination).31,32 Contributors KM, MSE, DM, IP, JIC, MS, JR, TM, KNH, and FSH designed the study. KM supervised the study. KM, MSE, OH, DL, DM, IP, JDW, MS, TFL, JR, TM, KNH, and FSH collected data. AB and KNH analysed data. KM, MSE, OH, DL, DM, IP, JIC, MS, JR, KNH, and FSH interpreted data. All authors reviewed and provided comment on initial versions, and reviewed and approved the final draft of the report. Conflicts of interest OH and DM have received research grants from Bristol-Myers Squibb. OH, JIC, TFL, and JR have received honoraria as speakers for Bristol-Myers Squibb. DM, JIC, and TFL have participated in advisory boards for and received honoraria from Bristol-Myers Squibb. JDW, MS, and JR have served as consultants for Bristol-Myers Squibb. TM, AB, and KNH are employees of Bristol-Myers Squibb; TM owns stock in the company. FSH has been an unpaid consultant for and has received clinical trial support from Bristol-Myers Squibb. The other authors declare that they have no conflicts of interest. Acknowledgments This trial was funded by Bristol-Myers Squibb. Professional medical writing and editorial assistance were provided by Jamie Zhang (Bristol-Myers Squibb). We thank the patients and investigators who participated in this study. References 1 Bafaloukos D, Gogas H. The treatment of brain metastases in melanoma patients. Cancer Treat Rev 2004; 30: 515–20. 2 Davies MA, Liu P, McIntyre S, et al. Prognostic factors for survival in melanoma patients with brain metastases. Cancer 2011; 117: 1687–96. 3 Fife KM, Colman MH, Stevens GN, et al. Determinants of outcome in melanoma patients with cerebral metastases. J Clin Oncol 2004; 22: 1293–300. 4 Gupta T. Stereotactic radiosurgery for brain oligometastases: good for some, better for all? Ann Oncol 2005; 16: 1749–54. 5 Liew DN, Kano H, Kondiziolka D, et al. Outcome predictors of gamma knife surgery for melanoma brain metastases. J Neurosurg 2011; 114: 767–79. 6 Samlowski WE, Watson GA, Wang M, et al. Multimodality treatment of melanoma brain metastases incorporating stereotactic radiosurgery. Cancer 2007; 109: 1855–62. 7 DiLuna ML, King JT Jr, Knisely JPS, Chiang VL. Prognostic factors for survival after stereotactic radiosurgery vary with the number of cerebral metastases. Cancer 2007; 109: 135–45. 8 Liew DN, Kano H, Kondziolka D, et al. Outcome predictors of gamma knife surgery for melanoma brain metastases. J Neurosurg 2011; 114: 769–79. 9 Agarwala SS, Kirkwood JM, Gore M, et al. Temozolomide for the treatment of brain metastases associated with metastatic melanoma: a phase II study. J Clin Oncol 2004; 22: 2101–07.
www.thelancet.com/oncology Vol 13 May 2012
Articles
10
11
12 13
14
15
16
17
18
19
20
Margolin K, Atkins MB, Thompson JA, et al. Temozolomide and whole brain irradiation in melanoma metastatic to the brain: a phase II trial of the Cytokine Working Group. J Cancer Res Clin Oncol 2002; 128: 214–18. Prins RM, Vo DD, Khan-Farooqui H, et al. NK and CD4 cells collaborate to protect against melanoma tumor formation in the brain. J Immunol 2006; 177: 8448–55. Wilson EH, Weninger W, Hunter CA. Trafficking of immune cells in the central nervous system. J Clin Invest 2010; 120: 1368–79. Schmittel A, Proebstle T, Engenhart-Cabillic R, et al. Brain metastases following interleukin-2 plus interferon-alpha-2a therapy: a follow-up study in 94 stage IV melanoma patients. Eur J Cancer 2003; 39: 476–80. Dudley ME, Yang JC, Sherry R, et al. Adoptive cell therapy for patients with metastatic melanoma evaluation of intensive myeloblative chemoradiation preparative regimens. J Clin Oncol 2008; 26: 5233–39. O’Day SJ, Maio M, Chiarion-Sileni V, et al. Efficacy and safety of ipilimumab monotherapy in patients with pretreated advanced melanoma: a multicenter single-arm phase II study. Ann Oncol 2010; 21: 1712–17. Hersh EM, O’Day SJ, Powderly J, et al. A phase II multicenter study of ipilimumab with or without dacarbazine in chemotherapy-naive patients with advanced melanoma. Invest New Drugs 2011; 29: 489–98. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in patients with pretreated advanced melanoma: a randomised, double-blind, multicentre, phase 2, dose-ranging study. Lancet Oncol 2010; 11: 155–64. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010; 363: 711–23. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med 2011; 364: 2517–26. Lebbe C, McDermott DF, Robert C, et al. Ipilimumab improves survival in previously treated, advanced melanoma patients with poor prognostic factors: subgroup analyses from a phase III trial. Ann Oncol 2010; 21 (suppl 8): 401–07.
www.thelancet.com/oncology Vol 13 May 2012
21
22
23
24
25
26
27 28 29
30
31
32
Schartz NEC, Farges C, Madelaine I, et al. Complete regression of a previously untreated melanoma brain metastasis with ipilimumab. Melanoma Res 2010; 20: 247–50. Hodi FS, Oble DA, Drappatz J, et al. CTLA-4 blockade with ipilimumab induces significant clinical benefit in a female with melanoma metastases to the CNS. Nat Clin Pract Oncol 2008; 5: 557–61. Di Giacomo AM, Ascierto PA, Pittiglio E, et al. A phase II study combining ipilimumab and fotemustine in patients with metastatic melanoma - the NIBIT M1 trial. Eur J Cancer 2011; 47 (suppl 1): 9305 (abstr). Gehan EA. The determination of the number of patients required in a preliminary and a follow-up trial of a new chemotherapeutic agent. J Chronic Dis 1961; 13: 346–53. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 2000; 92: 205–16. Wolchok JD, Hoos A, O’Day S, et al. Guidelines for the evaluation of immune therapy activity in solid tumors: immune-related response criteria. Clin Cancer Res 2009; 15: 7412–20. Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934; 26: 404–13. Brookmeyer R, Crowley J. A confidence interval for the median survival time. Biometrics 1982; 38: 29–41. Weber JS, Amin A, Minor D, Siegel J, Berman D, O’Day SJ. Safety and clinical activity of ipilimumab in melanoma patients with brain metastases: retrospective analysis of data from a phase II trial. Melanoma Res 2011; 21: 530–34. Beck KE, Blansfield JA, Tran KQ, et al. Enterocolitis in patients with cancer after antibody blockade of cytotoxic T-lymphocyte-associated antigen 4. J Clin Oncol 2006; 24: 2283–89. Blank C, Mackensen A. Contribution of the PD-L1/PD-1 pathway to T-cell exhaustion: an update on implications for chronic infections and tumor evasion. Cancer Immunol Immunother 2007; 56: 739–45. Hino R, Kabashima K, Kato Y, et al. Tumor cell expression of programmed cell death-1 ligand 1 is a prognostic factor for malignant melanoma. Cancer 2010; 116: 1757–66.
465
Ipilimumab in patients with melanoma and brain metastases: an open-label, phase 2 trial Kim Margolin, Marc S Ernstoff, Omid Hamid, Donald Lawrence, David McDermott, Igor Puzanov, Jedd D Wolchok, Joseph I Clark, Mario Sznol, Theodore F Logan, Jon Richards, Tracy Michener, Agnes Balogh, Kevin N Heller, F Stephen Hodi
Summary Background Brain metastases commonly develop in patients with melanoma and are a frequent cause of death of patients with this disease. Ipilimumab improves survival in patients with advanced melanoma. We aimed to investigate the safety and activity of this drug specifically in patients with brain metastases. Methods Between July 31, 2008, and June 3, 2009, we enrolled patients with melanoma and brain metastases from ten US centres who were older than 16 years into two parallel cohorts. Patients in cohort A were neurologically asymptomatic and were not receiving corticosteroid treatment at study entry; those in cohort B were symptomatic and on a stable dose of corticosteroids. Patients were to receive four doses of 10 mg/kg intravenous ipilimumab, one every 3 weeks. Individuals who were clinically stable at week 24 were eligible to receive 10 mg/kg intravenous ipilimumab every 12 weeks. The primary endpoint was the proportion of patients with disease control, defined as complete response, partial response, or stable disease after 12 weeks, assessed with modified WHO criteria. Analyses of safety and efficacy included all treated patients. This trial is registered with ClinicalTrials.gov, number NCT00623766. Findings We enrolled 72 patients: 51 into cohort A and 21 into cohort B. After 12 weeks, nine patients in cohort A exhibited disease control (18%, 95% CI 8–31), as did one patient in cohort B (5%, 0·1–24). When the brain alone was assessed, 12 patients in cohort A (24%, 13–38) and two in cohort B (10%, 1–30) achieved disease control. We noted disease control outside of the brain in 14 patients (27%, 16–42) in cohort A and in one individual (5%, 0·1–24) in cohort B. The most common grade 3 adverse events in cohort A were diarrhoea (six patients [12%]) and fatigue (six [12%]); in cohort B, they were dehydration (two individuals [10%]), hyperglycaemia (two [10%]), and increased concentrations of serum aspartate aminotransferase (two [10%]). One patient in each cohort had grade 4 confusion. The most common grade 3 immune-related adverse events were diarrhoea (six patients [12%]) and rash (one [2%]) in cohort A, and rash (one individual [5%]) and increased concentrations of serum aspartate aminotransferase (two [10%]) in cohort B. One patient in cohort A died of drug-related complications of immune-related colitis. Interpretation Ipilimumab has activity in some patients with advanced melanoma and brain metastases, particularly when metastases are small and asymptomatic. The drug has no unexpected toxic effects in this population. Funding Bristol-Myers Squibb.
Introduction Estimates of frequency of brain metastases in patients with advanced melanoma range from 10% to 50%; autopsy series suggest the value could be as high as 66–75%.1 Median overall survival is only 4 months after diagnosis of brain metastases.2,3 The resistance of melanoma to radiotherapy and cytotoxic chemotherapies, and its frequent metastasis to the brain, render its management particularly challenging. Treatment options for patients with brain metastases are surgery, stereotactic radiosurgery, and whole-brain radiation. In an analysis of 686 patients with melanoma and brain metastases, Fife and colleagues3 reported median survival of 9 months after surgery with or without radiotherapy, but whole-brain irradiation did not extend life. Retrospective data from several other reports4–7 of treatment for melanoma metastatic to the brain suggest that stereotactic radiotherapy is more advantageous than is whole-brain irradiation and that surgery and stereotactic radiotherapy have similar www.thelancet.com/oncology Vol 13 May 2012
effects. However, the patient cohorts in these reports were heterogeneous in size, number of brain lesions, extracranial disease, and treatment variables. The effect of concomitant corticosteroids might be to improve symptoms, but they could also be immunosuppressive, and data from published trials do not distinguish the outcomes by corticosteroid use. However, it is generally assumed that in older series the dose and duration of corticosteroid treatment might have been related to the size, number, and symptoms of brain metastases and peritumoral oedema.7,8 In the USA, temozolomide is the most widely used systemic treatment for patients with melanoma and brain metastases. Although this drug can cross the blood–brain barrier, clinical responses are noted in roughly 10% as a single agent or combined with wholebrain irradiation or other agents.9,10 The development of immune-modulating agents creates new possibilities. Previously, immune treatments were thought to be ineffective in patients with brain metastases because they
Lancet Oncol 2012; 13: 459–65 Published Online March 27, 2012 DOI:10.1016/S14702045(12)70090-6 See Comment page 434 University of Washington, Seattle, WA, USA (Prof K Margolin MD); Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA (M S Ernstoff MD); The Angeles Clinic and Research Institute, Santa Monica, CA, USA (O Hamid MD); Massachusetts General Hospital Cancer Center, Boston, MA, USA (D Lawrence MD); Beth Israel Deaconess Medical Center, Boston, MA, USA (D McDermott MD); Vanderbilt University Medical Center, Nashville, TN, USA (I Puzanov MD); Memorial Sloan-Kettering Cancer Center, New York, NY, USA (J D Wolchok MD); Loyola University, Chicago, IL, USA (Prof J I Clark MD); Yale Cancer Center, New Haven, CT, USA (M Sznol MD); Indiana University Simon Cancer Center, Indianapolis, IN, USA (T F Logan MD); Oncology Specialists SC, Park Ridge, IL, USA (J Richards MD); Bristol-Myers Squibb, Plainsboro, NJ, USA (T Michener PharmD); Bristol-Myers Squibb, Braine-l’Alleud, Belgium (A Balogh MS); Bristol-Myers Squibb, Lawrenceville, NJ, USA (K N Heller MD); and Dana-Farber Cancer Institute, Boston, MA, USA (F S Hodi MD) Correspondence to: Prof Kim Margolin, Seattle Cancer Care Alliance, Seattle, WA 98109, USA [email protected]
459
Articles
would not cross an intact blood–brain barrier, but research has now shown that activated T cells can pass through this barrier.11,12 Therefore, treatments that stimulate T-cell responses might be effective against brain metastases of melanoma.13,14 Phase 2 studies15–17 have shown that ipilimumab is active in advanced melanoma. Phase 3 trials have been designed with peptide vaccine18 or dacarbazine19 comparators and established that groups given ipilimumab have improved survival, supporting approval of the drug as monotherapy at 3 mg/kg in the USA and Europe. In one of these studies,18 61 (11%) of 540 patients in two treatment groups given ipilimumab had brain metastases. Hazard ratios for death in patients with brain metastases were 0·70 (95% CI 0·41–1·20) for 46 individuals given ipilimumab plus gp100 peptide vaccine and 0·76 (0·38–1·54) for 15 patients given ipilimumab alone. These values are similar to those of patients without brain metastases.20 The safety profile and pattern of immune-related adverse events (irAEs) were also similar in patients with or without brain metastases. Case studies of two patients with melanoma— one from a phase 2 study21 and one given ipilimumab on a compassionate-use basis22—also showed benefits of ipilimumab in treatment of brain metastases. Studies of ipilimumab combined with either temozolomide or fotemustine in this setting are in progress.23 We aimed to assess efficacy and safety of ipilimumab specifically in patients with melanoma and brain metastases in a prospective clinical trial. We investigated ipilimumab’s effects in a cohort of patients who did not need treatment with corticosteroids, before testing the drug in a less favourable group of individuals who needed corticosteroids for clinical or radiological control of their brain metastases.
Methods Study design and patients Between July 31, 2008, and June 3, 2009, we undertook an open-label, phase 2 trial with two parallel cohorts. We used a two-stage, modified Gehan design.24 In stage one, we enrolled asymptomatic patients from ten centres in the USA into cohort A to assess the effect of ipilimumab monotherapy on brain and extracranial metastases. This stage met efficacy parameters (three partial responses on the basis of modified WHO [mWHO] criteria25) with acceptable tolerability. The study proceeded to stage two, in which accrual to cohort A continued and enrolment into cohort B, consisting of patients with symptomatic metastases controlled with corticosteroids, began at the same US centres. Patients older than 16 years with histologically confirmed metastatic melanoma were eligible. Patients had to have at least one measurable index brain metastasis of 0·5–3 cm in diameter, or two measurable lesions larger than 0·3 cm visible on contrast MRI, or both. Patients in cohort A were neurologically asymptomatic 460
and had received no systemic corticosteroid treatment in the 10 days before start of ipilimumab treatment. Patients given concurrent systemic corticosteroids for control of brain-metastasis-related symptoms (judged by the treating physician) or oedema were enrolled into cohort B. Participants’ Eastern Cooperative Oncology Group (ECOG) scores had to be 0 or 1. At least 28 days must have elapsed since the last systemic treatment. Previous focused radiotherapy and whole-brain irradiation had to have been completed at least 14 days before start of ipilimumab. The stereotactic radiotherapy field should not have included the brain index lesion; alternatively, the lesion had to be progressive and measurable in two dimensions after any irradiation. Patients with active autoimmune disease were excluded. Participants could have had any number of previous treatments, but patients were ineligible when they had received other immunomodulatory antibodies (eg, other anti-CTLA-4, anti PD-1 or PD-L1, agonistic anti-CD40 or anti-OX40, or anti-CD137). The protocol was approved by institutional review boards or independent ethics committees of participating centres. The study was done in accordance with the Declaration of Helsinki and with good clinical practice as defined by the International Conference on Harmonization. All participating patients gave voluntary, written informed consent.
Procedures Patients were to receive one dose of 10 mg/kg intravenous ipilimumab every 3 weeks to a total of four doses (designated weeks 1, 4, 7, and 10; induction). Patients who were clinically stable at 24 weeks were eligible to continue with treatment with ipilimumab 10 mg/kg every 12 weeks (maintenance). Treatment continued until progression of the global tumour burden (brain and nonCNS lesions) measured with immune-related response criteria (irRC),18,19,26 a related adverse event necessitating discontinuation of ipilimumab treatment, clinical deterioration (as per protocol), or withdrawal of consent. Imaging of brain lesions and non-CNS tumour lesions was done every 6 weeks during induction and every 12 weeks thereafter (or more frequently when clinically indicated) to assess tumour response. The primary endpoint was the proportion of patients with disease control (complete response, partial response, or stable disease) after 12 weeks, assessed with mWHO criteria. Secondary endpoints were the proportion of patients with an objective response (a complete or partial response), median progression-free survival, overall survival, safety, and tolerability. At every tumour assessment, response was assessed in brain, non-CNS, and global (brain and non-CNS) anatomical compartments. We used both mWHO criteria and irRC to assess endpoints. Both systems use the sum of the products of two-dimensional tumour measurements; however, with irRC the measurements of new lesions are www.thelancet.com/oncology Vol 13 May 2012
Articles
Cohort A (n=51)
Cohort B (n=21)
Age (years) Median (range)
Cohort A (n=51)
Cohort B (n=21)
mWHO
mWHO
irRC
irRC
59 (33–79)
57 (30–74)
CR
0
0
0
Male
33 (65%)
11 (52%)
PR
5 (10%)
5 (10%)
1 (5%)
1 (5%)
Female
18 (35%)
10 (48%)
SD
4 (8%)
8 (16%)
0
1 (5%)
PD†
40 (78%)
36 (71%)
0
25 (49%)
14 (67%)
2* (4%)
2* (4%)
1
26 (51%)
7 (33%)
CR
0
0
1 (5%)
1 (5%)
CNS
48 (94%)
19 (90%)
PR
8 (16%)
8 (16%)
0
0
Lung
22 (43%)
12 (57%)
SD
4 (8%)
5 (10%)
1 (5%)
1 (5%)
Liver
13 (25%)
8 (38%)
PD†
39 (76%)
38 (75%)
19 (90%)
19 (90%)
Other (visceral)
15 (29%)
12 (57%)
Unknown
Lymph node
13 (25%)
11 (52%)
Non-CNS
Soft tissue
Sex
ECOG performance status
Site of index lesion
Global
Unknown
20 (95%)
0
19 (90%)
0
0
CNS
0
0
0
0 0
12 (24%)
11 (52%)
CR
0
0
0
Skin
3 (6%)
2 (10%)
PR
7 (14%)
7 (14%)
1 (5%)
1 (5%)
Other
7 (14%)
2 (10%)
SD
7 (14%)
10 (20%)
0
1 (5%)
PD†
35 (69%)
32 (63%)
Any
40 (78%)
15 (71%)
2 (4%)*
2 (4%)*
Immunotherapy
24 (47%)
5 (24%)
Interferon
19 (37%)
4 (19%)
Interleukin 2
16 (31%)
3 (14%)
Previous systemic treatment
Biochemotherapy
2 (4%)
0
Targeted agent
8 (16%)
3 (14%)
Experimental protocol
10 (20%)
1 (5%)
Chemotherapy
22 (43%)
14 (67%)
16 (31%)
13 (62%)
Carboplatin
2 (4%)
4 (19%)
Paclitaxel
3 (6%)
4 (19%)
Vinblastine
2 (4%)
4 (19%)
Temozolomide
Cisplatin
3 (6%)
Unknown
20 (95%)
19 (90%)
0
0
mWHO=modified WHO criteria. irRC=immune-related response criteria. CR=complete response. PR=partial response. SD=stable disease. PD=progressive disease. *No non-CNS lesions were noted for these two patients; therefore, both non-CNS and global response were classified as “unknown”. †PD includes patients lost to follow-up or for whom no tumour assessment was done (eg, primary lesion resected or only brain MRI completed).
Table 2: Disease response by mWHO and irRC criteria after 12 weeks
Cohort A (n=51) mWHO
Cohort B (n=21) irRC
mWHO
irRC
9 (18%, 8–31)
13 (25%, 14–40)
1 (5%, 0·1–24)
2 (10%, 1–30)
4 (19%)
Global disease control
12 (24%, 13–38)
13 (25%, 14–40)
2 (10%, 1–30)
2 (10%, 1–30)
14 (27%, 16–42)
17 (33%, 21–48)
1 (5%, 0·1–24)
2 (10%, 1–30)
Abraxane
2 (4%)
3 (14%)
CNS disease control
Dacarbazine
3 (6%)
0
Non-CNS disease control
Previous radiotherapy
Global objective response
5 (10%, 3–21)
5 (10%, 3–21)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
Any
CNS objective response
8 (16%, 7–29)
8 (16%, 7–29)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
Non-CNS objective response
7 (14%, 6–26)
7 (14%, 6–26)
1 (5%, 0·1–24)
1 (5%, 0·1–24)
To the brain Whole brain Gamma knife or targeted
31 (61%) 21 (41%) 17 (33%) 4 (8%)
9 (43%) 10 (48%) 5 (24%)
Data are n (%, 95% CI). mWHO=modified WHO criteria. irRC=immune-related response criteria.
0 Table 3: Disease control and objective response after 12 weeks
Data are n (%) unless otherwise stated. ECOG=Eastern Cooperative Oncology Group.
Table 1: Demographic and baseline characteristics
included in assessment of tumour burden, and new lesions alone do not define progressive disease. Adverse events were graded according to National Cancer Institute’s Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0. irAEs—defined as any adverse events causally related to drug exposure and consistent with an immune-mediated mechanism—and brain-related events were assessed. irAEs have been attributed to immune or inflammatory reactions that are a result of ipilimumab’s mechanism of action.18,19 Serious adverse events were reported from time of consent. The safety analysis was based on the frequency www.thelancet.com/oncology Vol 13 May 2012
and severity of adverse events for all patients who received at least one dose of ipilimumab. Worst toxicity grades per patient were tabulated by treatment cohort for adverse events and laboratory measurements with CTCAE. Safety data were reported between the first dose and 70 days (five half-lives) after last dose. All toxicity noted was summarised by grade, attribution, and time of onset.
Statistical analysis In stage one, we planned for enrolment of 21 patients into cohort A. If fewer than two objective responses assessed with mWHO criteria were recorded or fewer than eight patients were progression-free at 12 weeks in 461
Articles
Cohort A Cohort B
100
Overall survival (%)
80
60
40
20
0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
11 3
11 3
11 3
11 2
10 1
7 0
4 0
2 0
2 0
2 0
2 0
1 0
0 0
Time after start of treatment (months) Number at risk Cohort A 51 Cohort B 21
49 19
43 13
38 11
33 10
28 10
27 8
23 7
21 6
18 5
18 5
15 5
14 4
13 4
13 4
13 4
12 4
12 4
11 4
11 4
Figure: Overall survival Crosses indicate censored patients.
these 21 patients, the study would be terminated early because of insufficient activity. Otherwise, we would accrue a further 20 patients, and cohort B would open to enrol 21 patients. All treated patients were included in efficacy and safety analyses. We computed the proportion of patients with disease control and the proportion with an objective response with both mWHO and irRC criteria. We calculated exact two-sided 95% CIs with the method of Clopper and Pearson.27 Time-to-event endpoints such as duration of response, progression-free survival, and overall survival were estimated with the Kaplan-Meier product limit method, together with a two-sided 95% CI for the median, calculated with the method of Brookmeyer and Crowley.28 Statistical analyses were done in SAS (version 9.2). This study is registered with ClinicalTrials.gov, number NCT00623766.
Role of the funding source The trial was designed jointly by senior academic researchers and the sponsor, Bristol-Myers Squibb. Data were collected by the sponsor and analysed in collaboration with senior academic authors, who vouch for the completeness and accuracy of the data and analyses. All interpretation of the data was by KM and academic coworkers. KNH, TM, and AB had access to raw data; KNH served as medical monitor; TM was protocol manager; and 462
AB was study statistician. An initial draft of the report was prepared jointly by academic researchers and a professional medical writer employed by the sponsor. All authors contributed to subsequent drafts and had final responsibility for the decision to submit for publication.
Results We enrolled 72 patients: 51 into cohort A and 21 into cohort B. Ten patients remained in the study on April 25, 2011: eight in cohort A and two in cohort B. Table 1 shows demographic and baseline characteristics of the enrolled participants. 15 patients in cohort A completed induction. 28 individuals had discontinued treatment because of early progression of disease or death, five requested discontinuation without documented progression, and three had adverse events. 11 patients in cohort A started maintenance treatment; of the other four, three had progressed and one was taken off treatment because of adverse events before this stage began. While on maintenance treatment, one patient underwent resection of a brain metastasis and discontinued ipilimumab, and one had progression of disease. Five patients in cohort B completed induction. 11 patients had discontinued treatment because of disease progression or death, three requested discontinuation, and two had adverse events. Two patients received maintenance treatment; of the other three, one did not www.thelancet.com/oncology Vol 13 May 2012
Articles
start because of adverse events, one because of progressive disease, and one had died. The median number of doses of induction treatment in cohort A was three (range one to four) and in cohort B two (one to four). Patients in cohort A received a mean of six doses of maintenance treatment (one to ten); both patients in cohort B who entered the maintenance phase received seven doses. The proportion of patients who achieved disease control varied when mWHO criteria or irRC were used. Four patients in cohort A had progressive disease with mWHO, but stable disease with irRC, as did one in cohort B (table 2). After week 12, nine patients in cohort A exhibited disease control (18%, 95% CI 8–31), as did one patient in cohort B (5%, 0·1–24) by mWHO criteria; by irRC, 13 patients achieved disease control (25%, 14–40) in cohort A, as had two (10%, 1–30) in cohort B (tables 2, 3). Control of nonCNS lesions and brain lesions is shown in table 3. The proportion of patients achieving an objective response after 12 weeks was the same with mWHO or irRC (table 3). No patients had a discordant (brain vs nonCNS) response status. No patient had a global complete response, but partial responses were recorded in both cohorts (table 2). 36 participants in cohort A and 19 in cohort B died. Median overall survival was 7·0 months (95% CI 4·1–10·8) in cohort A and 3·7 months (1·6–7·3) in cohort B (figure). For patients in cohort A, overall survival was 55% (95% CI 41–68) at 6 months, 31% (18–44) at 12 months, and 26% (14–39) at both 18 months and 24 months (figure). Those in cohort B had an overall survival of 38% (17–59) at 6 months, 19% (2–36) at both 12 months and 18 months, and 10% (0–22) at 24 months (figure). Table 4 shows duration of progression-free survival. The most common adverse events in both cohorts were fatigue, diarrhoea, nausea, headache, rash, and pruritus (table 5). The most common grade 3 events overall were diarrhoea, fatigue, dehydration, hyperglycaemia, and increased concentrations of serum aspartate aminotransferase (table 5). One patient in each cohort experienced grade 4 confusion. Three patients in cohort A had brain oedema (one grade 1 and two grade 2). A 66-year-old man in cohort A died for reasons related to the study drug 24 days after his second dose of ipilimumab because of gangrenous colitis, despite treatment with corticosteroids. The incidence of irAEs (all grades) was similar in both cohorts (table 5), with diarrhoea, rash, pruritus, and increased serum aspartate aminotransferase the most common. There were no grade 4 irAEs and the most common grade 3 irAEs were diarrhoea in cohort A and increased serum aspartate aminotransferase in cohort B (table 5). The most common CNS-related events were grade 1 and 2 headache and dizziness, which were possibly related to ipilimumab. One patient had grade 4 brain haemorrhage attributed to metastatic melanoma, but ipilimumab could have been implicated. www.thelancet.com/oncology Vol 13 May 2012
Cohort A
Cohort B
mWHO
irRC
mWHO
Overall
1·4 (1·2–2·6)
2·7 (1·6–3·7)
1·2 (1·2–1·3)
irRC 1·3 (1·2–2·5)
Brain
1·5 (1·2–2·5)
1·9 (1·2–2·9)
1·2(1·2–1·3)
1·2 (1·2–1·3)
Non-CNS
2·6 (1·3–4·1)
3·3 (2·6–4·7)
1·3 (1·2–2·5)
1·3 (1·2–2·5)
Data are months (95% CI). mWHO=modified WHO criteria. irRC=immune-related response criteria.
Table 4: Median progression-free survival
Cohort A (n=51) Any grade* Grade 3
Cohort B (n=21) Grade 4
Any grade* Grade 3 Grade 4
Any event Diarrhoea
25 (49%)
6 (12%) 0
9 (43%)
0
0
Nausea
22 (43%)
3 (6%)
0
4 (19%)
0
0
Vomiting
13 (25%)
3 (6%)
0
1 (5%)
0
0
0
0
4 (19%)
0
0
6 (12%) 0
12 (57%)
1 (5%)
0
Constipation Fatigue Oedema (peripheral)
8 (16%) 28 (55%) 4 (8%)
1 (2%)
0
5 (24%)
0
0
Headache
18 (35%)
2 (4%)
0
6 (29%)
0
0
Dizziness
11 (22%)
0
0
2 (10%)
0
0
Rash
19 (37%)
1 (2%)
0
7 (33%)
1 (5%)
0
Pruritus
16 (31%)
0
0
6 (29%)
0
0
Decreased appetite
14 (27%)
2 (4%)
0
4 (19%)
0
0
Dehydration
5 (10%)
2 (4%)
0
4 (19%)
2 (10%) 0
Hyperglycaemia
4 (8%)
2 (4%)
0
4 (19%)
2 (10%) 0
8 (16%)
0
0
4 (19%)
1 (5%)
0
11 (22%)
0
0
2 (10%)
0
0
1 (2%)
0
4 (19%)
2 (10%) 0
Back pain Cough Aspartate aminotransferase increased
4 (8%)
Confused state
9 (18%)
1 (2%)
1 (2%)
3 (14%)
1 (5%)
1 (5%)
Insomnia
8 (16%)
0
0
4 (19%)
0
0
Immune-related events Diarrhoea
22 (43%)
6 (12%) 0
8 (38%)
0
0
Rash
17 (33%)
1 (2%)
0
6 (29%)
1 (5%)
0
Pruritus
16 (31%)
0
0
5 (24%)
0
0
3 (6%)
0
0
4 (19%)
2 (10%) 0
Aspartate aminotransferase increased CNS-related events Headache
18 (35%)
2 (4%)
0
6 (29%)
0
0
Dizziness
11 (22%)
0
0
2 (10%)
0
0
Data are n (%). Events occurring in at least 15% of patients in either cohort. *Grades 1–5 with National Cancer Institute’s Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0.
Table 5: Adverse events
Discussion We have established that ipilimumab shows activity in patients with melanoma and brain metastases, particularly when they have stable, asymptomatic metastases that do not need corticosteroid treatment (panel). Although the primary objective of this study was to estimate the proportion of patients for whom ipilimumab controlled disease, we also investigated safety, because initial tumour growth and peritumoral inflammatory changes can cause neurological complications. However, the results of our 463
Articles
Panel: Research in context Systematic review We searched PubMed with the terms “melanoma” and “brain metastases” for reports published in English. We did not use any specific date restrictions, but only reports published in roughly the past decade were included to ensure data were up to date. We established that few effective treatment options are available for patients with melanoma metastatic to the brain.1–3,7–12 Retrospective analyses of patients with advanced melanoma and brain metastases from a phase 3 trial of ipilimumab,18 and case studies of two individuals with brain metastases20,21 suggested that ipilimumab was active in brain metastases, serving as the rationale for our study. Interpretation As far as we are aware, our prospective study was the first to investigate ipilimumab specifically in patients with advanced melanoma and brain metastases. Our data show that the drug is active in these individuals. Antitumour activity, 2-year survival, and safety in this study are similar to what has been reported in patients without brain metastases. We believe that these data are sufficient to support use of ipilimumab in selected patients with brain metastases, particularly metastases that are small and asymptomatic. Additionally, we have provided safety and initial efficacy data that could serve as the foundation for future trials of combinations of ipilimumab and radiotherapy, other immunotherapies, or possibly molecularly targeted agents.
trial strongly suggest that treatment of patients similar to those in our study will not increase morbidity. The number of adverse events noted did not seem to be higher than has been reported previously in patients with melanoma free of brain metastasis, and we did not record unique events. Although the small patient numbers prevented formal analysis, we did not note an association between irAEs and response or long survival times. Additionally, the numbers of patients with tumour regression in our trial and in the others are alike. Differences in characteristics of participants and protocol procedures prohibit direct comparison between trials, but we believe that our data confirm the safety of this approach in patients with brain metastases. Furthermore, a retrospective analysis29 of a phase 2 trial showed that 12 patients with stable brain metastases given ipilimumab had overall survival of 14 months. Two patients with partial responses and one with stable disease survived for longer than 4 years.29 In our trial, different clinical outcomes were recorded in the two cohorts, which could be a result of prognostic variation but could also be related to a negative effect of steroids on ipilimumab activity. Whereas other researchers have reported that patients with both an antitumour response and an irAE do not lose the antitumour effect when treated with corticosteroids,30 the 464
possibility remains that steroid treatment at initiation of ipilimumab could abrogate or downmodulate an immune response to checkpoint blockade. However, the objective responses and survival plateau in cohort B indicate that steroids do not entirely eliminate the possibility of response to immune checkpoint blockade. Nevertheless, the apparent low rate of benefit in this cohort suggests that further study is needed into patients with specific unfavourable prognostic factors and into avoidance of steroid use at the start of ipilimumab treatment. Overall, our trial shows that the activity of ipilimumab is similar in patients with advanced melanoma with and without brain metastases. Further investigations could assess combinations of ipilimumab and chemotherapy (eg, fotemustine in the NIBIT-M1 trial23) and others, such as molecularly targeted small molecules and other immunomodulatory strategies (eg, new checkpointblocking antibodies alone or in combination).31,32 Contributors KM, MSE, DM, IP, JIC, MS, JR, TM, KNH, and FSH designed the study. KM supervised the study. KM, MSE, OH, DL, DM, IP, JDW, MS, TFL, JR, TM, KNH, and FSH collected data. AB and KNH analysed data. KM, MSE, OH, DL, DM, IP, JIC, MS, JR, KNH, and FSH interpreted data. All authors reviewed and provided comment on initial versions, and reviewed and approved the final draft of the report. Conflicts of interest OH and DM have received research grants from Bristol-Myers Squibb. OH, JIC, TFL, and JR have received honoraria as speakers for Bristol-Myers Squibb. DM, JIC, and TFL have participated in advisory boards for and received honoraria from Bristol-Myers Squibb. JDW, MS, and JR have served as consultants for Bristol-Myers Squibb. TM, AB, and KNH are employees of Bristol-Myers Squibb; TM owns stock in the company. FSH has been an unpaid consultant for and has received clinical trial support from Bristol-Myers Squibb. The other authors declare that they have no conflicts of interest. Acknowledgments This trial was funded by Bristol-Myers Squibb. Professional medical writing and editorial assistance were provided by Jamie Zhang (Bristol-Myers Squibb). We thank the patients and investigators who participated in this study. References 1 Bafaloukos D, Gogas H. The treatment of brain metastases in melanoma patients. Cancer Treat Rev 2004; 30: 515–20. 2 Davies MA, Liu P, McIntyre S, et al. Prognostic factors for survival in melanoma patients with brain metastases. Cancer 2011; 117: 1687–96. 3 Fife KM, Colman MH, Stevens GN, et al. Determinants of outcome in melanoma patients with cerebral metastases. J Clin Oncol 2004; 22: 1293–300. 4 Gupta T. Stereotactic radiosurgery for brain oligometastases: good for some, better for all? Ann Oncol 2005; 16: 1749–54. 5 Liew DN, Kano H, Kondiziolka D, et al. Outcome predictors of gamma knife surgery for melanoma brain metastases. J Neurosurg 2011; 114: 767–79. 6 Samlowski WE, Watson GA, Wang M, et al. Multimodality treatment of melanoma brain metastases incorporating stereotactic radiosurgery. Cancer 2007; 109: 1855–62. 7 DiLuna ML, King JT Jr, Knisely JPS, Chiang VL. Prognostic factors for survival after stereotactic radiosurgery vary with the number of cerebral metastases. Cancer 2007; 109: 135–45. 8 Liew DN, Kano H, Kondziolka D, et al. Outcome predictors of gamma knife surgery for melanoma brain metastases. J Neurosurg 2011; 114: 769–79. 9 Agarwala SS, Kirkwood JM, Gore M, et al. Temozolomide for the treatment of brain metastases associated with metastatic melanoma: a phase II study. J Clin Oncol 2004; 22: 2101–07.
www.thelancet.com/oncology Vol 13 May 2012
Articles
10
11
12 13
14
15
16
17
18
19
20
Margolin K, Atkins MB, Thompson JA, et al. Temozolomide and whole brain irradiation in melanoma metastatic to the brain: a phase II trial of the Cytokine Working Group. J Cancer Res Clin Oncol 2002; 128: 214–18. Prins RM, Vo DD, Khan-Farooqui H, et al. NK and CD4 cells collaborate to protect against melanoma tumor formation in the brain. J Immunol 2006; 177: 8448–55. Wilson EH, Weninger W, Hunter CA. Trafficking of immune cells in the central nervous system. J Clin Invest 2010; 120: 1368–79. Schmittel A, Proebstle T, Engenhart-Cabillic R, et al. Brain metastases following interleukin-2 plus interferon-alpha-2a therapy: a follow-up study in 94 stage IV melanoma patients. Eur J Cancer 2003; 39: 476–80. Dudley ME, Yang JC, Sherry R, et al. Adoptive cell therapy for patients with metastatic melanoma evaluation of intensive myeloblative chemoradiation preparative regimens. J Clin Oncol 2008; 26: 5233–39. O’Day SJ, Maio M, Chiarion-Sileni V, et al. Efficacy and safety of ipilimumab monotherapy in patients with pretreated advanced melanoma: a multicenter single-arm phase II study. Ann Oncol 2010; 21: 1712–17. Hersh EM, O’Day SJ, Powderly J, et al. A phase II multicenter study of ipilimumab with or without dacarbazine in chemotherapy-naive patients with advanced melanoma. Invest New Drugs 2011; 29: 489–98. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in patients with pretreated advanced melanoma: a randomised, double-blind, multicentre, phase 2, dose-ranging study. Lancet Oncol 2010; 11: 155–64. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010; 363: 711–23. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med 2011; 364: 2517–26. Lebbe C, McDermott DF, Robert C, et al. Ipilimumab improves survival in previously treated, advanced melanoma patients with poor prognostic factors: subgroup analyses from a phase III trial. Ann Oncol 2010; 21 (suppl 8): 401–07.
www.thelancet.com/oncology Vol 13 May 2012
21
22
23
24
25
26
27 28 29
30
31
32
Schartz NEC, Farges C, Madelaine I, et al. Complete regression of a previously untreated melanoma brain metastasis with ipilimumab. Melanoma Res 2010; 20: 247–50. Hodi FS, Oble DA, Drappatz J, et al. CTLA-4 blockade with ipilimumab induces significant clinical benefit in a female with melanoma metastases to the CNS. Nat Clin Pract Oncol 2008; 5: 557–61. Di Giacomo AM, Ascierto PA, Pittiglio E, et al. A phase II study combining ipilimumab and fotemustine in patients with metastatic melanoma - the NIBIT M1 trial. Eur J Cancer 2011; 47 (suppl 1): 9305 (abstr). Gehan EA. The determination of the number of patients required in a preliminary and a follow-up trial of a new chemotherapeutic agent. J Chronic Dis 1961; 13: 346–53. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 2000; 92: 205–16. Wolchok JD, Hoos A, O’Day S, et al. Guidelines for the evaluation of immune therapy activity in solid tumors: immune-related response criteria. Clin Cancer Res 2009; 15: 7412–20. Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934; 26: 404–13. Brookmeyer R, Crowley J. A confidence interval for the median survival time. Biometrics 1982; 38: 29–41. Weber JS, Amin A, Minor D, Siegel J, Berman D, O’Day SJ. Safety and clinical activity of ipilimumab in melanoma patients with brain metastases: retrospective analysis of data from a phase II trial. Melanoma Res 2011; 21: 530–34. Beck KE, Blansfield JA, Tran KQ, et al. Enterocolitis in patients with cancer after antibody blockade of cytotoxic T-lymphocyte-associated antigen 4. J Clin Oncol 2006; 24: 2283–89. Blank C, Mackensen A. Contribution of the PD-L1/PD-1 pathway to T-cell exhaustion: an update on implications for chronic infections and tumor evasion. Cancer Immunol Immunother 2007; 56: 739–45. Hino R, Kabashima K, Kato Y, et al. Tumor cell expression of programmed cell death-1 ligand 1 is a prognostic factor for malignant melanoma. Cancer 2010; 116: 1757–66.
465