The impact of the COVID-19 pandemic on the presentation status of newly diagnosed melanoma: A single institution experience.

To the Editor: The COVID-19 pandemic has had a significant impact on cancer care. Some have projected up to a 10% increase in mortality for specific malignancies due to delays in care caused by the COVID-19 pandemic, but the pandemic's impact on melanoma has yet to be defined. , Delays in diagnosis could result in thicker melanomas at presentation and profound effects on patient outcomes. This study evaluates the presentation status of melanoma lesions before and after a period of pandemic restrictions, which limited dermatologic evaluation to define the pandemic's impact on melanoma care.

Patients referred to the University of Pennsylvania's Dermatopathology Department for pathologic slide review and/or Division of Endocrine and Oncologic Surgery (handling most of the institutional resection volume) for definitive resection of nonmetastatic primary melanomas were identified from a 2-month period after clinical resurgence at our institution (June 15-August 15, 2020; the COVID-19 era cohort) and a corresponding period in the pre–COVID-19 era (June 15-August 15, 2019). Patient and tumor characteristics were analyzed by univariate analyses. All tests were 2-sided, and P values less than .05 were considered statistically significance. Analyses were performed in Stata for Windows version 16.1.

Of all melanomas evaluated at our institution, 358 and 298 patients were evaluated in the pre–COVID-19 era cohort and COVID-19 era cohort, respectively. There were no differences in patient characteristics and tumor type (invasive melanoma versus melanoma in situ) between the 2 cohorts. After exclusion of melanoma in situ lesions, 172 and 153 patients with invasive melanoma were evaluated in the pre–COVID-19 and COVID-19 era cohorts, respectively (Table I ). Patients in the COVID-19 era cohort were more likely to have satellitosis (3.9% vs 0%, P = .001) compared with pre–COVID-19 era patients. Among patients evaluated by the oncologic surgery department, specifically, COVID-19 era (N = 56) patients had higher median tumor Breslow depth (1.4 mm vs 0.87 mm; P = .013) and a higher proportion of patients with mitotic count greater than 1/mm2 (58.9% vs 35.3%; P = .018), satellitosis (8.9% vs 0%; P = .029), and pT3/pT4 tumors (35.7% vs 19.1%; P = .037) compared with pre–COVID-19 era patients (N = 68) (Table II ).

Table I

Patient and tumor characteristics of all melanomas examined by dermatopathology and oncologic surgery from June 15 to August 15, 2019 and June 15 to August 15, 2020

	Pre–COVID-19	COVID-19	P value
	N = 172 (52.9%)	N = 153 (47.1%)
Age (median, IQR)	68 (16.5)	68 (18)	.518
<50 y	24 (14.0)	23 (15.0)
50-59 y	22 (12.8)	30 (19.6)
60-69 y	48 (27.9)	38 (24.8)
70-79 y	56 (32.6)	43 (28.1)
≥80 y	22 (12.8)	19 (12.4)
Sex			.757
Male	96 (55.8)	88 (57.5)
Female	76 (44.2)	65 (42.5)
Race			.257
White	138 (80.2)	116 (75.8)
Black	2 (1.2)	0 (0.0)
Asian	0 (0.0)	1 (0.7)
Unknown	32 (18.6)	36 (23.5)
Immune compromise	1 (0.6)	4 (2.6)	.137
Tumor depth (median, IQR)	0.5 (0.7)	0.6 (0.9)	.171
pT staging group			.900
1/2	147 (85.5)	130 (85.0)
3/4	25 (14.5)	23 (15.0)
Clark level			.880
Level II	57 (33.1)	47 (30.7)
Level III	52 (30.2)	41 (26.8)
Level IV	56 (32.6)	58 (37.9)
Level V	4 (2.3)	4 (2.6)
Unknown	3 (1.7)	3 (2.0)
Lymphovascular invasion	6 (3.5)	4 (2.6)	.092
Unknown	5 (2.9)	0 (0.0)
Ulceration	14 (8.1)	22 (14.4)	.165
Unknown	3 (1.7)	4 (2.6)
Tumor-infiltrating lymphocytes			.537
Brisk	18 (10.5)	14 (9.2)
Nonbrisk	92 (53.5)	88 (57.5)
Unknown	28 (16.3)	17 (11.1)
Vertical growth	114 (66.3)	98 (64.1)	.673
Unknown	8 (4.7)	5 (3.3)
Regression	48 (27.9)	32 (20.9)	.162
Unknown	4 (2.3)	8 (5.2)
Satellitosis	0 (0.0)	6 (3.9)	.001∗
Unknown	5 (2.9)	14 (9.2)
Perineural invasion	3 (1.7)	4 (2.6)	.080
Unknown	8 (4.7)	1 (0.7)
Mitotic count			.240
None	97 (56.4)	79 (51.6)
≤1	33 (19.2)	24 (15.7)
>1	42 (24.4)	50 (32.7)
Residual tumor	50 (29.1)	47 (30.7)	.691
Unknown	58 (33.7)	56 (36.6)
Source			.587
Dermatopathology only	104 (60.5)	97 (63.4)
Surgery	68 (39.5)	56 (36.6)

IQR, Interquartile range.

Indicates significance.

Table II

Patient and tumor characteristics of all melanomas examined by oncologic surgery from June 15 to August 15, 2019 and June 15 to August 15, 2020

	Pre–COVID-19	COVID-19	P value
	N = 68 (54.8%)	N = 56 (45.2%)
Age (median, IQR)	65 (19)	66.5 (14.5)	.699
<50 y	11 (16.2)	7 (12.5)
50-59 y	10 (14.7)	12 (21.4)
60-69 y	22 (32.4)	17 (30.4)
70-79 y	18 (26.5)	17 (30.4)
≥80 y	7 (10.3)	3 (5.4)
Sex			.504
Male	36 (52.9)	33 (58.9)
Female	32 (47.1)	23 (41.1)
Race			.085
White	59 (86.8)	54 (96.4)
Black	2 (2.9)	0 (0.0)
Asian	0 (0.0)	1 (1.8)
Unknown	7 (10.3)	1 (1.8)
Immune compromise	0 (0.0)	3 (5.4)	.053
Tumor depth (median, IQR)	0.8 (1.0)	1.4 (3.0)	.013∗
pT staging group			.037∗
1/2	55 (80.9)	36 (64.3)
3/4	13 (19.1)	20 (35.7)
Clark level			.006∗
Level II	14 (20.6)	3 (5.4)
Level III	23 (33.8)	11 (19.6)
Level IV	28 (41.2)	36 (64.3)
Level V	3 (4.4)	3 (5.4)
Unknown	0 (0.0)	3 (5.4)
Lymphovascular invasion	0 (0.0)	2 (3.6)	.130
Unknown	2 (2.9)	0 (0.0)
Ulceration	12 (17.7)	15 (26.8)	.327
Unknown	1 (1.5)	0 (0.0)
Tumor-infiltrating lymphocytes			.764
Brisk	11 (16.2)	6 (10.7)
Nonbrisk	38 (55.9)	36 (64.3)
Unknown	10 (14.7)	7 (12.5)
Vertical growth	50 (73.5)	46 (82.1)	.492
Unknown	6 (8.8)	4 (7.1)
Regression	24 (35.3)	12 (21.4)	.239
Unknown	1 (1.5)	1 (1.8)
Satellitosis	0 (0.0)	5 (8.9)	.029∗
Unknown	1 (1.5)	0 (0.0)
Perineural invasion	2 (2.9)	3 (5.4)	.734
Unknown	2 (2.9)	1 (1.8)
Mitotic count			.018∗
None	29 (42.7)	12 (21.4)
≤1/mm2	15 (22.1)	11 (19.6)
>1/mm2	24 (35.3)	33 (58.9)
Residual tumor	24 (35.3)	24 (42.9)	.390
Pathologic stage			.183
I	51 (75.0)	34 (60.7)
II	13 (19.1)	13 (23.2)
III	4 (5.9)	9 (16.1)
SLNB performed	44 (64.7)	45 (80.4)	.054
Positive SLN	3 (4.4)	5 (8.9)	.308

IQR, Interquartile range; SLN, sentinel lymph node; SLNB, sentinel lymph node biopsy.

Indicates significance.

During the COVID-19 pandemic, to reallocate clinical resources and control viral transmission, outpatient health care services were limited for patients from March to mid-June. We investigated whether absence of routine dermatologic evaluation during this time resulted in advanced tumor presentation status after clinical resurgence. There was no difference noted in median thickness or pT staging group in melanomas evaluated overall. Among surgical patients specifically, there was an increase in median tumor depth, the proportion of pT3/pT4 lesions, and lesions with satellitosis. This finding may reflect a goal among clinicians to remove thin melanomas at clinics locally, minimizing the need for patient travel. The increase in median thickness of melanomas and absolute number of pT3/pT4 lesions (>50% increase) referred for surgical evaluation raises concerns for delay in diagnosis. Although this study is limited as a single-institution study over a short period, further study is warranted to better define the impact of the pandemic on melanoma care nationally.

Conflicts of interest

None disclosed.