SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 788 

RESEARCH LETTER 
 

 

Epidemiology and Mortality Risk Factors of Sebaceous 
Carcinoma: A SEER – Based Population Study 
 

Joseph Haquang1, Rocky Li1, Kevin Mai1, Kuang Cheng Chen1, Fardad Sisan1, Kuo YuLing, 
MD2 

 
1 College of Osteopathic Medicine of the Pacific, Western University of Health Sciences - Pomona, CA 
2 National Taiwan University Hospital– Department of Dermatology, Taipei City, Taiwan 
 

 

 
 

 
 
Sebaceous carcinoma (SC) is a rare and 
potentially aggressive cutaneous malignancy 
most commonly derived from the sebaceous 

glands, the glands of Zeis, or meibomian 
glands. The current literature search does not 
report which demographic and 
socioeconomic factors are associated with 
increased mortality risk.1 Our article aims to 
build upon the limited prognostic research by 

ABSTRACT 

Background: Sebaceous carcinoma is a rare and potentially aggressive cutaneous 
malignancy. It is derived from the adnexal epithelium of sebaceous glands and tends to hold 
a diverse clinical presentation. Although it is often reported in the periocular region, it can 
manifest from any sebaceous unit in the skin. Due to the rarity of this condition, prognostic 
and demographic factors are largely indeterminate. Hence, we sought to assess the 
prognostic impact of demographic and socioeconomic factors on the outcome of patients with 
sebaceous carcinoma.  
Methods: A retrospective analysis was performed utilizing data from the Surveillance, 
Epidemiology, and End Results (SEER) database. From the registries, cases of sebaceous 
carcinoma from the years 2000 to 2018 were analyzed. Univariate and multivariate cox 
regression analyses were used to analyze the significance of socioeconomic and 
demographic factors on the survival of sebaceous carcinoma. Tumor grade and extent were 
included in the multivariate cox regression to minimize confounding. 
Results: A total of 4154 cases of sebaceous carcinoma were analyzed within this study. 
Socioeconomic and demographic factors analyzed includes age, sex, race, and income. On 
multivariate analysis including tumor grade, tumor extent, age, sex, race and income, African 
American race was a significant risk indicator for survival (hazard ratio [HR], 1.9; P=.007). 
Increased age of 70+ were also identified as a significant risk indicator for survival (HR, 5.86; 
P<0.001). Female sex was identified as a protective indicator for survival (HR, 0.82; P=0.03). 
Income status did not significantly influence the survival outcome of sebaceous carcinoma.  
Conclusion: Although income status did not show any significant influence on the survival 
outcome of sebaceous carcinoma, sex, race, and age characteristics did. The etiology behind 
these prognostic factors is unclear but may be related to access to medical care or lack of 
social support. 

INTRODUCTION 



SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 789 

evaluating current data from 2000-2018 with 
a more extensive case number. Our analysis 
reports the associated SC mortality risk of 
race, age, sex, and income. 
 

 
 
Patients diagnosed with SC from 2000 to 
2018 were identified in SEER 18 registry 
using SEER Rare Cancer classification 
variable “40.2 Sebaceous adenocarcinoma”. 
The following data were extracted: age, year 
of diagnosis, race, vital status, site, combined 
summary stage, grade (degree of 
differentiation), extension, metastasis at 
diagnosis, tumor size, lymph nodes, and 
median household income inflation adjusted 
to 2019. Survival was measured in months 
from time of diagnosis to loss of follow up 
which included all-cause mortality. 
Multivariate Cox regression model was 
conducted using covariates deemed clinically 
significant or with P value of less than 0.25 in 
univariate analysis to include all potential 
factors. Age categories, sex, race, grade, 
tumor extent, and income were included in 
the final model.2-4  
 

 
 
A total of 4154 patients with primary SC were 
identified. Patient demographics and tumor 
characteristics are summarized in Table 1. 
The majority of patients diagnosed were over 
70 years of age, with a male predominance 
(61.7%). 85.8% of the patients were white. 
Median and mean months of follow-up were 
50 and 64. 
 
The results from the multivariate Cox 
Regression model are shown in Table 2. 
Older age at diagnosis is associated with 
lower survival, with highest mortality among 
those 70 and older (HR = 5.86 [95% 

confidence interval (CI) 3.62 - 9.51]). Female 
sex is a protective prognostic predictor, with 
18% reduction in mortality compared to their 
male counterparts (HR = 0.82 [95% CI 0.69 - 
0.98]). Black individuals with primary SC 
have lower survival than white individuals 
(HR = 1.9 [95% CI 1.19 - 3.03]). Other 
variables such as income, location, and 
tumor extent do not have statistically 
significant association with survival in this 
model. 
 

 
 
We corroborate previous literature describing 
a higher incidence of SC in white individuals. 
Our results demonstrated a statistically 
significant HR of 1.9 in individuals identifying 
as black, implying a higher mortality risk 
despite having a lower overall incidence of 
SC. Various factors discussed by Shao et al. 
regarding other skin cancers, such as Acral 
Lentiginous Melanoma, could similarly 
explain the higher mortality in SCC. For 
instance, reduced access to care leads to 
decreased early skin examination/detection 
rates, less awareness of melanoma, and 
misdiagnosis from varied clinical 
presentation.[5] Geographical differences 
may also play a role leading to increased UV 
radiation exposure.[6] Furthermore, age >70 
had a significant HR of 5.86, and those 50-69 
had a HR of 2.09. Advanced age is a 
recognized risk factor for cutaneous 
malignancies and could be related to 
impaired immunologic and DNA repair 
mechanisms that decline with age. 
Interestingly, we could not conclude an 
association between socioeconomic status 
and mortality. Possibly, the rarity of the 
condition requires a larger sample size to 
fully assess the association. Finally, our 
study described a predominance in male sex. 
However, females had a hazard ratio of 0.82 
which suggests a lower risk of mortality. This  

METHODS 

RESULTS 

DISCUSSION 



SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 790 

Table 1. Patient demographics and tumor characteristics. 
 

Characteristic 
Number 

(Percent) 

Sex  
Female 1592 (38.3%) 

Male 2562 (61.7%) 

Age  

0-49 300 (7.2%) 

50-69 1459 (35.1%) 

>70 2395 (57.7%) 

Race  

White 3566 (85.8%) 

African American 115 (2.8%) 

Asian/Pacific Islander 213 (5.1%) 

American Indian/Alaska Native 29 (0.7%) 

Unknown 231 (5.6%) 

Tumor extent  

Localized 2651 (63.8%) 

Distant 141 (3.4%) 

Regional 128 (3.1%) 

Unknown 1234 (29.7%) 

Site  

Lip 38 (0.9%) 

Eyelid 924 (22.2%) 

External ear 154 (3.7%) 

Face 1414 (34.0%) 

Scalp and neck 506 (12.2%) 

Trunk 728 (17.5%) 

Upper limb and shoulder 262 (6.3%) 

Lower limb and hip 87 (2.1%) 

Overlapping lesion 6 (0.1%) 

Skin, NOS 25 (0.6%) 

Scrotum 10 (0.2%) 

Grade  

I 515 (12.4%) 

II 240 (5.8%) 

III 340 (8.2%) 

IV 39 (0.9%) 

Unknown 3020 (72.7%) 

Metastasis   

Metastasized 129 (3.1%) 

Not Metastasized 2978 (71.7%) 



SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 791 

Unknown 1047 (25.2%) 

Income  

< $35,000 48 (1.2%) 

$35,000 - 39, 999 70 (1.7%) 

$40,000 - 44, 999 148 (3.6%) 

$45,000 - 49, 999 236 (5.7%) 

$50,000 - 54, 999 267 (6.4%) 

$55,000 - 59, 999 275 (6.6%) 

$60,000 - 64, 999 770 (18.5%) 

$65,000 - 69, 999 549 (13.2%) 

$70,000 - 74, 999 354 (8.5%) 

$75,000+ 1434 (34.5%) 

  



SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 792 

Table 2. Cox regression multivariate analysis of overall survival of sebaceous carcinoma. 
 

Characteristics HR (95% CI) 

Age  
0-49 1 

50-69 2.09 (1.26-3.45) 

>70 5.86 (3.62-9.51) 

Sex  
Male 1 

Female 0.82 (0.69-0.98) 

Race  
White 1 

African American 1.9 (1.19-3.03) 

Asian/ Pacific Islander 0.66 (0.43-1.02) 

American Indian/ Alaska Native 1.21 (0.49-2.97) 

Grade  
Well differentiated 1 

Moderately differentiated 0.8 (0.63-1.02) 

Poorly differentiated 1.21 (0.99-1.48) 

Undifferentiated 1.26 (0.83-1.92) 

Tumor Extent  
Localized 1 

Regional 1.21 (0.80-1.83) 

Distant 1.35 (0.93-1.96) 

Income  
< $35,000 1 

$35,000 - 39, 999 2.15 (0.74-6.23) 

$40,000 - 44, 999 1.01 (0.37-2.77) 

$45,000 - 49, 999 0.88 (0.34-2.31) 

$50,000 - 54, 999 1.58 (0.62-4.0) 

$55,000 - 59, 999 1.12 (0.44-2.87) 

$60,000 - 64, 999 1.24 (0.50-3.05) 

$65,000 - 69, 999 1.16 (0.47-2.88) 

$70,000 - 74, 999 0.91 (0.36-2.32) 

$75,000+ 1.19 (0.49-2.91) 

HR: Hazard Ratio; CI: Confident Interval; * indicates statistically significant P value 
(<0.05) 

 
  



SKIN 
 

May 2023     Volume 7 Issue 3 
 

(c) 2023 THE AUTHORS. Published by the National Society for Cutaneous Medicine. 793 

finding could be explained by the influence of 
estrogen, behavioral differences, or other 
confounders.[7] Of note, limitations in our 
interpretation arise from constraints in the 
SEER database as recorded mortality 
includes deaths from any cause and losses 
to follow-up. Thus, these factors could cause 
selection bias and skewing of our analysis. 
 
Overall, increased data collection can lead to 
better clinical understanding and policy 
changes for SC patients. A systematic 
analysis incorporating several multi-regional 
studies could paint a more accurate picture of 
the influence of demographic and 
socioeconomic factors on SC. 
 
Conflict of Interest Disclosures: None 
 
Funding: None 
 
Corresponding Author: 
Joseph Haquang 
College of Osteopathic Medicine of the Pacific, 
Western University of Health Sciences - Pomona, CA 
Email: joseph.haquang@westernu.edu 

 
 
References: 
1. Dasgupta, T., L.D. Wilson, and J.B. Yu, A 

retrospective review of 1349 cases of 
sebaceous carcinoma. Cancer, 2009. 115(1): 
p. 158-65. 

2. Team, R.C., R: A language and environment 
for statistical computing. R Foundation for 
Statistical Computing, Vienna, Austria. 2012. 
2018. 

3. Team, R., RStudio: integrated development 
environment for R. RStudio, PBC, Boston. 
2020. 

4. Lin, H. and D. Zelterman, Modeling survival 
data: extending the Cox model. 2002, Taylor 
& Francis. 

5. Shao, K. and H. Feng, Racial and Ethnic 
Healthcare Disparities in Skin Cancer in the 
United States: A Review of Existing 
Inequities, Contributing Factors, and 
Potential Solutions. J Clin Aesthet Dermatol, 
2022. 15(7): p. 16-22. 

6. Becker, G. and E. Newsom, Socioeconomic 
status and dissatisfaction with health care 
among chronically ill African Americans. 

American journal of public health, 2003. 
93(5): p. 742-748. 

7. Morgese, F., et al., Gender Differences and 
Outcomes in Melanoma Patients. Oncology 
and Therapy, 2020. 8(1): p. 103-114.