A total of 475 trials were identified using the key search terms described above. 387 trials met the inclusion criteria (Fig. 1). Of the included trials,147 were phase I, 86 were phase I/II, 130 were phase II, 3 were in phase II/III, 19 were phase III, and 2 trials were in phase IV (Table 1). 199 studies (51.4%) were conducted exclusively in the United States (US), 118 (30.5%) were exclusively international, and 70 (18.1%) were global. 204 (52.7%) trials were specific to patients with melanoma while 183 (47.3%) trials were not specific to melanoma. Of the trials that were specific to melanoma, 17.3% included only cutaneous melanoma, 12.4% included cutaneous or mucosal melanoma (12.4%), 3.1% included only mucosal melanoma, 5.7% included only uveal melanoma, 4.7% of trials explicitly included all types of melanomas and 56.8% did not specify which origin site of melanoma was included. Among the trials meeting inclusion criteria, treatment modalities included: targeted therapy (19.4%), immune therapy (IO) (53.5%), radiation (2.6%), chemotherapy (1.8%), combination therapy (20.9%), and other treatment modalities (1.8%). In regards trial sponsors, pharmaceutical companies sponsored 187 (48.3%) of included studies, while non- pharmaceutical company sources sponsored 200 (51.7%) of included studies.
Table 1
Clinical trial characteristics recorded from ClinicalTrials.gov.
|
All
(n = 387)
|
N
|
%
|
Phase
|
7
|
1.8
|
Early Phase 1
|
Phase 1
|
140
|
36.2
|
Phase 1|Phase 2
|
86
|
22.2
|
Phase 2
|
130
|
33.6
|
Phase 2|Phase 3
|
3
|
0.8
|
Phase 3
|
19
|
4.9
|
Phase 4
|
2
|
0.5
|
US or International
|
199
|
51.4
|
US
|
International
|
118
|
30.5
|
Both
|
70
|
18.1
|
Melanoma specific or Multiple Malignancy
|
204
|
52.7
|
Melanoma
|
Multiple Malignancies
|
183
|
47.3
|
Melanoma Subtype
|
67
|
17.3
|
Cutaneous
|
Cutaneous/mucosal
|
48
|
12.4
|
Mucosal
|
12
|
3.1
|
Uveal
|
22
|
5.7
|
Not specified
|
220
|
56.8
|
All included
|
18
|
4.7
|
Therapy category
|
75
|
19.4
|
Targeted
|
IO
|
207
|
53.5
|
Radiation
|
10
|
2.6
|
Other
|
7
|
1.8
|
Chemo
|
7
|
1.8
|
Combination
|
81
|
20.9
|
Sponsor
|
187
|
48.3
|
Pharmaceutical
|
Non-pharmaceutical
|
200
|
51.7
|
Trial inclusion of patients with MBM was reported based on disease activity, symptoms, and treatment status. Overall, 268 (69.3%) trials excluded patients solely based on their CNS disease status (active, symptomatic, symptomatic and active, treated) and 12.9% of trials did not specify whether patients with CNS disease were included. More specifically, any form CNS disease was excluded in 69 trials (17.8%), active disease was excluded in 101 (26.1%) of trials, symptomatic disease was excluded in 39 (10.1%), symptomatic and active disease was excluded in 41 (10.6%) trials (Table 2). In terms of inclusion, all CNS disease was included in 28 trials (7.2%), exclusively untreated CNS disease was included in 5 (1.3%) trials, and exclusively treated disease was included in 54 (14%) trials. Additionally, 50 (12.9%) trials did not specify the CNS disease activity, symptomatic, or treatment status inclusion criteria (see Table 2). Regarding leptomeningeal involvement of metastatic melanoma, 196 (50.6%) trials excluded patients with any LMD. Of the studies that did allow leptomeningeal involvement 17 (4.4%) included any LMD, 159 (41.1%) allowed LMD with local treatment, 5 (1.3%) excluded active LMD, and 10 (2.6%) of trials did not specify inclusion or exclusion of LMD.
Table 2
Status of CNS metastasis or leptomeningeal inclusion in included clinical trials.
Statement reg CNS disease
|
28
|
7.2
|
Included
|
Active lesions excluded
|
101
|
26.1
|
Symptomatic excluded
|
39
|
10.1
|
Symptomatic and active excluded
|
41
|
10.6
|
Untreated only
|
5
|
1.3
|
Treated only
|
54
|
14.0
|
Not specified
|
50
|
12.9
|
Excluded
|
69
|
17.8
|
Statement on Leptomeningeal disease
|
17
|
4.4
|
Included
|
Excluded
|
196
|
50.6
|
Unknown
|
10
|
2.6
|
Allows with local treatment
|
159
|
41.1
|
Active lesions excluded
|
5
|
1.3
|
Includes leptomeningeal disease
|
32
|
8.3
|
Yes
|
No
|
195
|
50.4
|
Unknown
|
160
|
41.3
|
Preliminary analysis was conducted using all included trials. This analysis found there was no statically significant association between the odds of a study including patients with active melanoma CNS metastasis in regard to study phase, location, melanoma specificity, or melanoma classification. However, both sponsor and therapy type did show a significant association. Trials not using immunotherapy (non-IO)_ therapy had 2.2 times the odds of including active CNS disease with odds ratio 2.23 (95% CI 1.2–4.3; p = 0.0174). Trials with non-pharmaceutical sponsors had nearly 2 times the odds of including patients with active CNS disease with an odds ratio of 1.98 (95% CI: 1.0–3.9; p = 0.0461).
To remove possible confounding, a univariate analysis was performed using data only from trials that reported all variables of interest (335 trials). This analysis found there was no statistically significant difference between the odds of a study including patients with active melanoma brain metastasis regarding study phase, location of trial, melanoma only or other malignancy present, sponsor, or melanoma classification (see Table 3). Significant associations were found between studies including patients with active CNS disease and therapy type as well as inclusion of leptomeningeal disease. Trials utilizing non-IO vs IO therapy were more likely to include CNS disease with an odds ratio of 1.18 (95% CI: 1.07–3.07) and p = 0.026. Inclusion of leptomeningeal disease significantly increased the odds of a trial including patients with active CNS disease with odds ratio 9.16 (95% CI: 4.01–20.94; p = < 0.0001) (Tables 3 & 4).
Table 3. Preliminary analysis of all included trials evaluating associations between collected variables and the odds of a study including patients with active melanoma CNS metastasis
Odds Ratio Estimates for modeling trial inclusion of patients with active melanoma brain metastases
|
|
Odds ratio
|
95% Confidence Limits
|
P-value
|
N used for model
|
Phase of study
|
|
|
|
|
221
|
|
Phase 1 included vs not
|
0.87
|
0.38
|
2.01
|
0.7482
|
|
|
Phase 2 included vs not
|
1.67
|
0.67
|
4.17
|
0.2711
|
|
|
Phase 3/4 included vs not
|
0.99
|
0.20
|
4.82
|
0.9904
|
|
Location of trial
|
|
|
|
|
221
|
|
US included vs not
|
1.23
|
0.43
|
3.50
|
0.6975
|
|
|
International included vs not
|
0.69
|
0.28
|
1.74
|
0.4348
|
|
Therapy type: Non-IO vs IO alone
|
2.23
|
1.15
|
4.32
|
0.0174
|
221
|
Melanoma vs Multi site
|
1.07
|
0.56
|
2.06
|
0.835
|
221
|
Non-pharma vs pharmaceutical
|
1.98
|
1.01
|
3.88
|
0.0461
|
221
|
Classification of melanoma
|
|
|
|
|
98
|
|
Cutaneous included vs not
|
1.27
|
0.33
|
4.82
|
0.7283
|
|
|
Mucosal included vs not
|
1.17
|
0.43
|
3.15
|
0.7563
|
|
|
Uveal included vs not
|
1.76
|
0.49
|
6.30
|
0.3833
|
|
Includes leptomeningeal disease vs not
|
9.16
|
4.01
|
20.94
|
<0.0001
|
221
|
Table 4. Univariate analysis using data only from trials that reported all variables of interest
Odds Ratio Estimates for modeling trial inclusion of patients with active melanoma brain metastases (335 trials included in model)
|
|
Odds ratio
|
95% Confidence Limits
|
P-value
|
Therapy type: non-IO vs IO alone
|
1.18
|
1.07
|
3.07
|
0.026
|
Includes leptomeningeal disease vs not
|
8.56
|
3.71
|
19.76
|
<0.0001
|