Anorectal Mucosal Melanoma in the Era of Immune Checkpoint Inhibition: Should We Change Our Surgical Management Paradigm?

Abstract

BACKGROUND:

The advent of immune checkpoint inhibition therapy has dramatically improved survival in patients with skin melanoma. Survival outcomes after resection of anorectal melanoma treated with immune checkpoint inhibition have not been reported.

OBJECTIVE:

To compare survival outcomes following surgical resection of anorectal melanoma between patients who received immune checkpoint inhibition and patients who did not.

DESIGN:

Retrospective analysis of data from a prospectively maintained database.

SETTING:

Comprehensive cancer center.

PATIENTS:

Patients who underwent surgery for anorectal melanoma between 2006 and 2017. Patients were stratified according to the use of immune checkpoint inhibition.

MAIN OUTCOME MEASURES:

Overall and disease-specific survival.

RESULTS:

Of the 47 patients included in the analysis, 29 (62%) received immune checkpoint inhibition therapy. Twenty-two (76%) of the 29 patients received immune checkpoint inhibition after detection of metastasis or disease progression rather than in the neoadjuvant or adjuvant setting. Overall survival did not differ significantly between patients who received immune checkpoint inhibition therapy and patients who did not (median, 52 and 20 months, respectively; 5-year rate, 41% vs. 35%, respectively; p = 0.25). Disease-specific survival also did not differ significantly. Our analysis did not identify any clinical or pathological features associated with response to immune checkpoint inhibition therapy or with survival.

LIMITATIONS:

Relatively small sample and retrospective design. Heterogenous treatment regimen in the immune checkpoint inhibition group.

CONCLUSIONS:

Immune checkpoint inhibition therapy by itself does not appear to improve survival in patients who undergo resection or excision of anorectal melanoma. Combinations of immune checkpoint inhibition with other therapeutic modalities warrant further investigation. See Video Abstract at http://links. lww.com/DCR/B499.

INTRODUCTION

Mucosal melanoma is a rare group of diseases, accounting for approximately 1% of all melanomas among Caucasians.¹ Most commonly, primary mucosal melanoma tumors arise in the head and neck, anorectal, and vulvovaginal regions. Anorectal melanomas constitute approximately 24% of all mucosal melanomas and 4% of all anal canal tumors.² Diagnosis can be challenging, as lesions are often amelanotic and symptoms are usually nonspecific.³ Like other mucosal melanomas, anorectal melanoma is classified as stage I if the disease is localized, stage II in the case of lymph node metastasis, and stage III in the case of distant metastasis.^4,5

Anorectal melanoma is poorly understood and highly lethal. Most patients experience distant recurrence, with 5-year recurrence-free survival of 16–17%.⁶ Overall survival at 5 years is estimated to be 20–22%. Based on retrospective data, median survival for stages I, II, and III is estimated to be 24, 17, and 8 months.⁷

Despite the use of multimodal therapy including surgery, radiation, and chemotherapy, improvement in survival has been marginal. Aggressive surgery with abdominoperineal resection may be associated with better local control, but survival outcomes do not differ significantly from those of local excision.^6,8-10 Currently, abdominoperineal resection is selectively used for large, advanced, resectable, nonmetastatic tumors that are not amenable to local excision.^7,11

Melanoma is a classic example of a tumor that responds well to ICI.^12-15 With the recent advancements in ICI therapy, including CTLA-4- and PD-1-targeted therapies, there has been a dramatic improvement in survival of some patients with advanced cutaneous melanoma.^14-19 There have been reports of favorable response of mucosal melanoma to ICI.^20-25 However, the rarity of anorectal mucosal melanoma has precluded adequate study of ICI therapy for this disease type, and the survival benefit of ICI for patients with anorectal mucosal melanoma requires additional investigation. Although cases of anorectal mucosal melanoma were included in ICI trials, the outcome data were not reported separately by disease site, and the treatment effect on anorectal mucosal melanoma is still unclear.

The purpose of this study was to compare survival outcomes following surgical resection of anorectal melanoma between patients who received ICI therapy and patients who did not and to examine whether the advent of ICI has changed the surgical approach to and outcomes associated with locoregional anorectal melanoma.

METHODS

After receiving institutional review board approval (IRB), a prospectively maintained database at Memorial Sloan Kettering Cancer Center was retrospectively queried to identify patients with locoregional anorectal melanoma who underwent surgery between January 2006 and December 2017. Forty-seven patients underwent resection with curative intent during that period (eight other patients underwent nonoperative management and were not included). The following data were examined: clinical characteristics, extent of resection, tumor thickness and maximum diameter, nodal involvement, margin involvement, lymphovascular invasion, perineural invasion, and ulceration. Tumors were classified as perianal (distal to dentate line), anorectal (involving the dentate line), or rectal (proximal to the dentate line). We staged tumors according to the Ballantyne system, with clinically localized disease and regional nodal involvement corresponding to stages I and II, respectively.⁴

Clinicopathological factors were compared by the χ² test. Overall survival and disease-specific survival from the time of diagnosis were analyzed using the Kaplan-Meier method, and significance was assessed by the log-rank test. All analyses were performed with SPSS software (version 25). p < 0.05 was considered significant.

RESULTS

Patient Demographics and Primary Tumor Characteristics

Of the 47 patients, 29 received ICI therapy and 18 did not. Of the 29 patients who received ICI, 5 received a CTLA-4 inhibitor, 8 received a PD-L1 inhibitor, and 16 received a combination of CTLA-4 and PD-L1 inhibitors. Twenty-two (76%) of the 29 patients received ICI therapy after development of metastasis or local disease progression, and seven patients (24%) received it in the adjuvant (nonmetastatic) setting. The median patient age in the ICI group was 58 years (min-max, 32–77), and 47% were women. Molecular testing for BRAF, NRAS and KIT was performed in 35 patients. However, in the non-immunotherapy (NI) group there was missing information for 10/18 patients, precluding a meaningful comparison between the groups. For the 35 patients with available information on all three genes, the molecular subtypes were, in decreasing frequency, triple negative (51.4%), KIT mutation (34.3%), NRAS mutation (14.3%) and BRAF mutation (2.9%). One patient had a mutation in both KIT and BRAF. Patient and tumor characteristics were comparable between the two groups of patients (Table 1).

TABLE 1.

Demographic and tumor characteristics of patients who received immunotherapy and patients who did not

Characteristic	Immunotherapy (n = 29)	No immunotherapy (n = 18)	p
Age, median (range)	58 (32–77)	65 (47–84)	NS
Sex, n (%)			NS
Male	16 (55)	9 (50)
Female	13 (45)	9 (50)
Breslow thickness (mm), median (95% CI)	10 (6–12)	9.5 (3.5–18)	NS
Diameter (cm), median (95% CI)	2.3 (2–3)	1.9 (1.3–3.8)	NS
Site or origin, n (%)			NS
Perianal	7 (24)	5 (28)
Anorectal	15 (51)	11 (61)
Rectal	7 (24)	2 (11)
Ulceration, n (%)	21 (70)	12 (67)	NS
LVI, n (%)	11 (37)	10 (56)	NS
PNI, n (%)	3 (8)	2 (11)	NS

NS, not significant; CI, confidence interval; LVI, lymphovascular invasion; PNI, perineural invasion.

Staging and Management

Baseline staging was performed by PET/CT in 30 of 47 (64%) of the patients (ICI – 18 of 29 (62%), NI- 12 of 18 (67%), p=0.75). The rest of the patients underwent baseline staging by CT. Thirty (64%) of the 47 patients presented with stage I disease, and 18 (60%) of the 30 received ICI therapy. The majority of the patients in the study underwent local excision of the primary tumor (21 [73%] of 29 patients in the ICI group and 12 [67%] of 18 patients in the NI group. R0 resection was achieved in 12 (41%) of 29 patients in the ICI group and 9 (50%) of 18 patients in the NI group. The ICI and NI groups did not differ significantly in clinical staging at presentation or types of surgery (Table 2). No patient had a complete clinical response to ICI, and among the patients for whom abdominoperineal resection was planned, none underwent a local excision instead. Systemic non-ICI chemotherapy was given to 13 of 29 patients (45%) in the ICI group and 6 of 18 patients (33%) in the NI group (p=0.44; table 2). In both groups the most common type of chemotherapy was temozolomide either alone or in conjunction with platinum-based chemotherapy (ICI- 6 of 29 (21%), NI- 4 of 18 patients (22%), p=0.90). Tyrosine kinase inhibitors (TKI) were given to 5 patients in the ICI group (17%) but were not given to any patients in the NI group. A flow chart of the management of the patients in the study can be seen in figure 1.

TABLE 2.

Disease stage and treatment characteristics for patients who received immunotherapy and patients who did not

Characteristic	No. of patients (%)		p
	Immunotherapy (n = 29)	No immunotherapy (n = 18)
Ballantyne stage			NS
I	18 (62)	12 (67)
II	11 (38)	6 (33)
Type of surgery			NS
APR	8 (28)	6 (33)
LE	21 (72)	12 (67)
Systemic chemotherapy	13 (45)	6 (33)	NS
Locoregional radiation	8 (27)	2 (11)	NS
R0 resection	12 (42)	9 (50)	NS

NS, not significant; APR, abdominoperineal resection; LE, local excision.

FIGURE 1.

Flow chart of patients in the study. ARMM, anorectal mucosal melanoma; LE, local excision; APR, abdominoperineal resection. ICI, immune checkpoint inhibition therapy; NI, no immunotherapy received.

Survival

Median follow-up was 51 months for survivors. For the ICI and NI groups, median overall survival was 52 and 20 months, respectively, and 5-year overall survival was 41% and 35%, respectively (p = 0.25; Fig. 2A). Disease-specific survival did not differ significantly between the two groups (Fig. 2B)

FIGURE 2.

Survival in patients with mucosal melanoma who received immune checkpoint inhibition therapy and patients who did not. Censoring is indicated by tick marks.

We analyzed multiple clinical and pathological variables for potential association with survival in the immunotherapy group. We found no statistically significant associations. Characteristics that approached statistical significance were, local excision, and absence of lymphovascular and perineural invasion (Table 3). Median overall survival and 5-year overall survival for stage I patients were 52 months and 44%, compared with 21 months and 26% for stage II patients (p = 0.13; Fig. 3A). For patients who underwent local excision, median overall survival and 5-year overall survival were 52 months and 46%, compared with 28 months and 20% for patients who underwent abdominoperineal resection (p = 0.17; Fig. 3B).

TABLE 3.

Univariable analysis of factors potentially associated with disease-specific survival in patients who received immunotherapy (n = 29)

Variable (n)	Median survival (mo.)	5-year survival (%)	p
Age
≤60 y (16)	54.8	42	0.52
>60 y (13)	NR	55
Sex
Female (13)	37.5	37.5	0.94
Male (16)	88.5	51
Presentation
Anemia (14)	54.8	41	0.45
Mass (11)	88.5	57
Pain (1)	NR	NR
Colonoscopy (3)	18.9	NR
Site or origin
Anorectal (15)	88.5	70	0.42
Perianal (7)	54.8	NR
Rectum (7)	30.6	43
Ballantyne stage
I (18)	88.5	52	0.58
II (11)	NR	51
Type of Surgery
APR (8)	27.6	19	0.19
LE (21)	88.5	65
Margins
R0 (15)	37.5	48	0.49
R1/2 (13)	NR	58
LVI
Yes (11)	23.4	50	0.17
No (7)	NR	71
PNI			0.16
Yes (3)	NR	33
No (9)	30.6	77
Depth
≥10 mm (13)	NR	55	0.58
<10 mm (12)	54.8	20
Systemic chemotherapy
Yes (13)	30.6	39	0.43
No (16)	88.5	61
Locoregional radiation
Yes (8)	88.5	52	0.96
No (21)	NR	52

NR, not reached; APR, abdominoperineal resection; LE, local excision; LVI, lymphovascular invasion; PNI, perineural invasion.

FIGURE 3.

Survival in relation to disease stage (A) and type of surgery (B). Censoring is indicated by tick marks. LE, local excision; APR, abdominoperineal resection.

DISCUSSION

Our study found that overall survival and disease-specific survival did not differ significantly between patients with anorectal melanoma who underwent surgical resection and received ICI therapy and patients who underwent surgical resection and did not receive ICI.

ICI is one of the most important new advances in the treatment of cancer. Recently, a pooled analysis of 889 patients evaluated the efficacy and safety of nivolumab, alone or in combination with ipilimumab, and showed median progression-free survival of approximately 3 months for monotherapy and 6 months for combination therapy for patients with unresectable or metastatic mucosal melanoma.²⁰ In addition, multiple other studies reported efficacy of ICI therapy for mucosal melanoma, with comparable response rates for CTLA-4 and PD-1 inhibitors in patients with metastatic mucosal melanoma.^21-25

The relatively small number of patients in the current study and perhaps the limited effect of ICI may have precluded the detection of differences in survival or recurrence. The difference in median survival between the groups which at first glance seems considerable (ICI −52 months, NI −20 months), is based on very few events and may only suggest a trend which could be substantiated in a larger study. The ICI group included 5 patients treated with ipilimumab monotherapy, which is no longer used as a frontline systemic treatment. TKI were prescribed to 5 patients in the ICI group, whereas no patient in the NI group was treated with TKI. Although this may introduce a bias, it is a bias that favors the ICI group and therefore would not change the outcome of the study. The retrospective nature of the study which was performed at a single center limits the generalizability of our findings. However, our findings are consistent with those of a recent large population database study of 1333 patients with anorectal melanoma,²⁶ which found no difference in survival between patients who received ICI therapy and patients who had surgery alone.

The findings of previous studies have made it clear that the choice of the type of surgery for mucosal melanoma does not confer a survival advantage.^27-29 Because it can help preserve the patient’s quality of life, local excision is the treatment of choice, and more radical surgery is selectively used when local excision is not feasible.^8,30-32 Local excision was performed in a majority (68%) of patients in our cohort, similarly to previous cohorts from our institution.²⁷ Although our study was underpowered to detect small differences, our findings provide no indication that surgical management of mucosal melanoma should be changed.

We found no association between survival in ICI patients and the clinicopathological factors we analyzed. Somewhat longer disease-specific survival on average was noted for patients with anorectal tumors than patients with rectal or perianal tumors, similarly to a previous study from our institution.³³ The presence or the depth of lymphovascular or perineural invasion was not associated with survival.

The findings of our study suggest that the availability of ICI therapy should not alter our standard practice of reserving radical resection for tumors not amenable to local excision. ICI therapy should nevertheless be considered for inclusion in a multimodal treatment plan, in view of the dramatic and even complete responses in some patients. A better understanding of the disease initiation and evolution of melanoma in organs with low sun exposure may help explain in part the limited effect of ICI therapy. Further research is needed to characterize the mechanisms of resistance and identify both clinical and molecular predictors of response, which can be used to improve patient selection. Promising results have been obtained for cutaneous melanoma by combining ICI therapy with other targeted therapies, including MEK and BRAF inhibitors. These approaches, as well as locoregional immune-triggering therapies, such as radiation and oncolytic virus injections, should be explored for mucosal melanoma.^34-38