Comorbid immunosuppression in Merkel cell carcinoma: A retrospective database study.

To the Editor: Merkel cell carcinoma (MCC) is an aggressive neuroendocrine carcinoma associated with immunosuppression. A better understanding of the outcomes of patients with comorbid immunosuppression is needed in light of excellent responses to immunotherapy, which generally requires immune competence.

We performed an institutional review board-exempt cohort study using the National Cancer Database (NCDB) from 2004 to 2015. All P values were 2-sided, and we used survival analysis censored at 50 months of follow-up. We identified 3691 MCC patients with complete staging, immunologic, survival, and demographic data, of whom 423 (11%) were known to be immunosuppressed. Of these, 14 patients had a diagnosis of HIV/AIDS, 120 had a diagnosis of chronic lymphocytic leukemia, 114 were solid organ transplant recipients (SOTRs), 77 had non-Hodgkin's lymphoma, ≤ 10 had a combination of immunosuppressive diagnoses, and 94 had another or unspecified cause for immunosuppression. Immunosuppressed patients were more likely to present with node-positive disease (44% vs 34%; P < .001) and more advanced T stage (T2-4 in 41% vs 34%; P = .018) but not metastatic disease (6% vs 5%; P = .83) (Table I).

Table I

Patient characteristics

Clinical/demographic characteristic		Not immune suppressed∗	Immunosuppressed∗	P value†
Number of patients		3268 (88.5)	423 (11.5)
Age, y	Median [IQR]	76.00 [67.00, 83.00]	72.00 [66.00, 79.00]	<.001
Female sex		1180 (36.1)	117 (27.7)	.001
Race	Asian	21 (0.6)	2 (0.5)	.119
	Black	29 (0.9)	9 (2.1)
	Other	21 (0.6)	3 (0.7)
	White	3197 (97.8)	409 (96.7)
Regional income quartile	1	340 (10.4)	57 (13.5)	.1
	2	656 (20.1)	71 (16.8)
	3	844 (25.8)	118 (27.9)
	4	1428 (43.7)	177 (41.8)
Regional high school Graduation quartile	1	374 (11.4)	60 (14.2)	.311
	2	733 (22.4)	95 (22.5)
	3	995 (30.4)	131 (31.0)
	4	1166 (35.7)	137 (32.4)
Rural		62 (1.9)	11 (2.6)	.428
Charlson Comorbidity Index	0	2323 (71.1)	279 (66.0)	.033
	1	684 (20.9)	99 (23.4)
	2	182 (5.6)	26 (6.1)
	3	79 (2.4)	19 (4.5)
T stage	0	187 (5.7)	18 (4.3)	.018
	1	1961 (60.0)	232 (54.8)
	2	827 (25.3)	121 (28.6)
	3	168 (5.1)	24 (5.7)
	4	125 (3.8)	28 (6.6)
N Stage	0	2171 (66.4)	238 (56.3)	<.001
	1	1001 (30.6)	160 (37.8)
	2	96 (2.9)	25 (5.9)
M Stage	1	173 (5.3)	24 (5.7)	.832

IQR, Interquartile range.

Data are presented as number (%).

Bold P values are statistically significant.

The vast majority of patients diagnosed with MCC underwent local therapy with some form of surgical intervention (91% of nonimmunosuppressed patients and 92% of immunosuppressed patients; P = .75). Patients with immunosuppression were more likely to receive external beam radiation therapy than immune competent patients (60% vs 55%, respectively; P = .012), but there was no significant difference in the use of systemic therapy or immunotherapy (P = .214, P = .878, respectively).

Overall, immunosuppression was associated with significantly reduced overall survival (univariable hazard ratio [HR], 2.34 [2.06-2.65]; P < .001; multivariable HR, 2.37 [2.17-2.81]; P < .001 with adjustments for stage, age, sex, comorbidity, and socioeconomic factors). Compared to chronic lymphocytic leukemia, immunosuppression related to being a SOTR was associated with significantly worse outcomes (HR, 1.57 [1.17-2.13]; P = .003), and HIV/AIDS trended toward better outcomes (HR, 0.42 [0.17-1.04]; P = .062) (Fig 1).

Fig 1

Survival by types of immune suppression.

Our findings provide additional detail on prior work, suggesting inferior oncologic outcomes for patients with immunosuppression, with the largest differences seen with SOTRs.,, SOTRs, generally require long-term immunosuppressive therapy and are commonly excluded from therapeutic advances as the field naturally moves toward immunotherapy, such as in the ongoing phase III STAMP study of adjuvant pembrolizumab (NCT03712605). This problem is expected to grow, as the number of patients with solid organ transplants is rising.

There are several limitations to using a hospital-based cancer registry, including errors in coding, bias in missing data, lack of complete clinical data (including details on other or combined causes of immunosuppression), selection bias, and residual confounding. Additionally, the lack of laboratory values limits our ability to quantify the degree of immunosuppression in conditions like HIV (eg, viral load, CD4 count). Finally, the Charlson Comorbidity Index includes several conditions related to immunosuppression, which significantly limits correction for medical comorbidity between immunosuppressed and competent immune patients.

Overall, our findings highlight the gap in outcomes among patients with MCC and comorbid immunosuppression. Therapeutic advances are needed for these patients as the treatment landscape shifts toward immunotherapy.

Conflicts of interest

Dr Riviere discloses consulting fees from Peptide Logic, LLC. Dr Gao was a consultant for EMD Serono (2015) and received research support from (2015-2017). Drs Brazel, Goshtasbi, Kuan, and Harris have no conflicts of interest to declare.