Rosacea is a chronic inflammatory skin condition that predominantly manifests as persistent facial flushing, irritation, and acneiform papules or pustules. Although the etiology of rosacea remains unclear, a variety of factors both genetic and environmental seem to play a role in its pathogenesis. It has been hypothesized that elements of immune system dysregulation, deregulation of neurovascular signaling, and overgrowth of cutaneous pathogens are involved [1]. These abnormal inflammatory processes impair the skin’s ability to act as a protective barrier for the body [2]. A genetic component has been suggested as well given that rosacea tends to appear more frequently in patients of Northern European and Celtic ancestry [1]. As the pathogenesis of rosacea is known to involve an abnormal inflammatory response, studies have been conducted to investigate its co-occurrence with other systemic diseases. Recent studies have suggested that rosacea is associated with a heightened risk of various chronic systemic diseases, including hypertension, autoimmune disease, cardiovascular disease, gastrointestinal disorders, and dyslipidemia [3], as well as multiple psychiatric comorbidities (major depressive disorder, persistent mood disorders, adjustment disorder, and generalized anxiety disorder) [4]. The complex relationship between inflammation and mental health is yet to be clearly understood, but there is evidence that chronic low-grade inflammation can contribute to mental illnesses such as depression [5]. Therefore, the inflammatory qualities of a chronic disease such as rosacea can potentially have far-reaching effects on patients’ mental and physical health.

In addition to the aforementioned illnesses, there have also been many studies that have examined the association between rosacea and various cancers. The current understanding of the pathophysiology of this relationship is limited. However, it is hypothesized that patients with rosacea may have an increased risk of skin cancer because of inflammatory changes in the skin barrier, including reduced epidermal levels of photoreceptive transurocanic acid [6].

Studies that have investigated the relationship between rosacea and cancer have reported inconsistent results. A recent study in Denmark [3,6] found that patients with rosacea had an increased risk of developing glioma, nonmelanoma skin cancer (NMSC), breast cancer, and hepatic cancer, whereas another study in the United States noted that females who had a history of rosacea had a subsequently increased risk of thyroid and basal cell carcinoma (BCC) [7]. Dupont [8], in contrast, found no substantial association between rosacea and skin cancers. Given the ambiguity of these findings, we conducted a systematic review of the current published work to evaluate the relationship between rosacea and various commonly studied cancers in the literature, including BCC; squamous cell carcinoma (SCC); Merkel cell carcinoma; melanoma; glioma; and hepatic, breast, and thyroid carcinomas.

In accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we conducted a review of observational studies on the association between rosacea and various cancers. Case reports, case series, case-control studies, cross-sectional studies, and cohort studies were included.

The Cochrane, PubMed, Embase, and Ovid databases were searched for relevant studies from inception to March 3, 2021. The search was then updated on June 9, 2023. The search terms are provided in detail in Multimedia Appendix 1.

Studies were included or excluded from a meta-analysis based on the criteria outlined in Textbox 1.

All studies were screened by 2 independent reviewers (LT and JX), and eligibility for inclusion was determined by screening the titles and reviewing the full texts. Any conflicts of eligibility were resolved by a third-party reviewer (WG). Using information obtained from the articles, rosacea and each cancer were categorized as likely or unlikely associated, whereas cancers with conflicting results were categorized as having an unclear association.

Data extracted from the included studies comprised author, publication year, title, study location, study type, total number of patients, number of patients with rosacea vs control patients without rosacea, general mean age, mean age of patients presenting with rosacea, general percentage of female patients, percentage of female patients presenting with rosacea, number of patients who presented with a history of smoking, drinking, confounding diseases, type of rosacea, inclusion and exclusion criteria used for the study, method of diagnosis of rosacea, mean duration of rosacea, socioeconomic status association with prevalent rosacea, number of patients with cancer in general, number of patients with cancer and rosacea, P value, odds ratio (OR; 95% CI), hazard ratio (HR), incidence rate ratio, risk ratio (RR), and any adjustments made to the statistical values. To assess the risk of bias of the included studies, the Newcastle-Ottawa Quality Assessment Scale was used.

Among the 3004 articles originally identified, 2786 (92.74%) remained after deduplication and were screened. After initial screening, of the 2786 remaining articles, 104 (3.73%) full-text articles were assessed for eligibility, of which 39 (37.5%) were eligible for our systematic review (Table 1). A total of 62.5% (65/104) of the full texts were excluded for reasons such as being in a non-English language, being the wrong publication type (including conference abstracts), having the wrong study design (including research papers), or looking at the wrong outcomes (Figure 1). Owing to a lack of available studies, we were not able to conduct a meta-analysis.

The risk of bias of the studies included for the statistical analysis is summarized in Figures 2 [3,6,7,9-11,13] and 3 [12]. The Newcastle-Ottawa Quality Assessment Scale was used to assess the bias found in a total of 21% (8/39) of the studies included for analysis. Of these 8 studies, 7 (88%) were rated as having a low risk of bias based on a Newcastle-Ottawa Quality Assessment Scale for cohort studies score of ≥5 (Figure 2). There was 1 study for which we had to use the Newcastle-Ottawa Quality Assessment Scale for case-control studies and that was also determined to have a low risk of bias with a score of ≥5 (Figure 3).

Our search returned 41% (16/39) of papers examining the relationship between rosacea and BCC. Of these 16 papers, 2 (12%) were cohort studies, the first investigating the impact of a personal history of rosacea on the risk of developing BCC in females in the United States and the second investigating the impact of a personal history of rosacea on the risk of developing facial BCC in females in the Providence, Rhode Island, region [7]. In the first study, 424 patients with rosacea and 4552 age- and sex-matched controls without rosacea were followed up for reports of the development of cancer. The mean age of the study sample was 37.6 (SD 4.1) years for patients with rosacea and 36.2 (4.7) years for the control group. The prevalence of BCC was found to be significantly higher among patients with rosacea compared with controls [7].

In the second study, of the 4537 patients diagnosed with BCC from October 2016 to November 2020 in the Rhode Island region, 2453 had BCC on the face, and of them, 267 had a history of rosacea. A multivariate model with adjustments for age, sex, smoking history, skin color, and other cancer risk factors was developed. Results for chi-square and R statistic analysis indicated that facial BCC in patients with a history of rosacea was significantly lower than in patients without rosacea (3.80 vs 5.07 per 100 patients; P<.001). When comparing BCC of the body, no significant difference was found between patients with a history of rosacea and patients without rosacea (3.95 vs 4.22 per 100 patients; P<.58) [9].

The other 88% (14/16) of papers on rosacea and BCC comprised case series or case reports. Notably, all reported cases detailed male patients with rhinophymatous rosacea who were later diagnosed with BCC in the area of the nodular lesion (Table 2). Lazzeri et al [15] presented a case series and literature review in which they found a total of 46 patients (including 3 new cases of their own) with rhinophyma that went on to develop cutaneous cancer. Of these patients, 28 were diagnosed with BCC, 11 were diagnosed with SCC, 4 were diagnosed with SCC and BCC, and 1 was found to have angiosarcoma. Common presenting symptoms in these patients included sudden progressive enlargement of long-standing rhinophyma in both BCC and SCC cases and additional symptoms of ulceration, malodorous drainage, and serous discharge in SCC cases [15]. Interestingly, BCC was less likely than SCC to present with sudden ulceration, bleeding, or serous discharge, although such symptoms have certainly been observed [16-20].

Our search returned 33% (13/39) of studies examining the relationship between rosacea and SCC. Of these, one was a cohort study investigating the impact of a personal history of rosacea on the risk of developing SCC in females in the United States. In this study, 577 cases of SCC were isolated from a population of 90,238 females. A Cox proportional hazard model and multivariate model with adjustments for age and other cancer risk factors were developed. The results indicated that rosacea was significantly associated with overall cutaneous SCC (RR 1.40, 95% CI 1.02-1.93). In this study, the location of SCC was also divided into 2 groups—head and neck and non–head and neck—which were also compared with the history of rosacea. The results indicated that rosacea was significantly associated with head and neck SCC (RR 1.71, 95% CI 1.09-2.69). Findings indicated no significant statistical associations between rosacea and non–head and neck SCC (RR 1.21, 95% CI 0.78-1.90) [10].

The second study investigated various cancer risks among females in the United States with a history of rosacea. Among a total of 75,088 females, 6015 patients were diagnosed with rosacea from 1991 to 2011, with 452 SCC diagnoses documented. A Cox proportional hazard model and multivariate model with adjustments for age, BMI, alcohol consumption, smoking history, and other cancer risk factors were developed. The results indicated that there were no statistically significant associations between rosacea and SCC (HR 1.30, 95% CI 0.90-1.88) [7].

The other 85% (11/13) of the studies were case series and reports that examined the association between rosacea and SCC. Again, all studies identified SCC development in patients with long-standing rhinophyma (Table 2). Notably, Kesty and Baldwin [31] reported the case of a patient aged 67 years with no previous history of cancer who presented with an enlarging left nasal mass of 4 months. An initial punch biopsy diagnosed rhinophyma, and the patient was sent home. The patient presented back to the hospital several weeks later with new reports of ulceration, drainage, and rapid growth of the mass and had another biopsy indicating invasive SCC. Although a total rhinectomy could be performed, the patient was ultimately found to have metastases to the lungs and submandibular area [31]. Although such cases are rare and no overall association can be made, these reports may serve as a warning to physicians to remain vigilant for the possibility of carcinoma in patients with rhinophyma, particularly those with rapidly growing and ulcerating masses.

Our search returned 5% (2/39) of studies that did not differentiate between NMSCs. In the first case, Egeberg et al [6] conducted a cohort study analyzing data from the nationwide Danish registry. The study included 49,475 patients with rosacea with an average age of 53.7 (SD 16.5) years and a reference population of 4,312,213 patients without rosacea with an average age of 48.6 (SD 18.0) years. Their findings noted that patients diagnosed with rosacea were statistically more likely to develop NMSC compared with patients in the reference population (HR 1.36, 95% CI 1.26-1.47) [6]. In the second case, Cho et al [11] conducted a nationwide population-based retrospective cohort study in South Korea. The study included 11,420 patients compared using a multivariable stratified Cox proportional hazard model. Their findings noted that patients with rosacea had a significant association with the development of NMSC compared with the reference population (HR 2.66, 95% CI 1.53-4.61) [11].

A total of 8% (3/39) of the studies analyzed the association between rosacea and breast cancer. Of these 3 studies, 2 (67%) were cohort studies. Egeberg et al [6] conducted a cohort study that analyzed data from the nationwide Danish registry and found that patients with rosacea were more likely to develop breast cancer compared with patients without a history of rosacea (HR 1.25, 95% CI 1.15-1.36). However, Li et al [7] found no significant increase in the occurrence of breast cancer in patients with rosacea (RR 1.03, 95% CI 0.89-1.20).

Li et al [12] also conducted a case-control study to determine the present epidemiological status of rosacea in China. From a total of 10,095 patients enrolled in the study, 351 were observed to have rosacea, 290 of whom were female and 61 of whom were male. The overall average age of patients with rosacea was 38.4 (SD 11.3) years, whereas the average age of the control population was 35.5 (SD 19.1) years. A previous diagnosis of melasma, hypertension, coronary heart disease, hyperthyroidism, diabetes, chronic gastritis, peptic ulcer, gastrointestinal cancer, breast cancer, and gynecological cancer was then identified in patients in both the control and rosacea populations. It was found that female patients with a diagnosis of rosacea were more likely to have had breast cancer (OR 8.453, 95% CI 1.638-43.606), among other diseases such as melasma, hypertension, and hyperthyroidism. Regarding male patients with rosacea, there was no significant increase in breast cancer incidence. A meta-analysis of observational studies on the association between rosacea and breast cancer was ultimately not conducted because of heterogeneity between the studies; however, it would seem that most of the current data available point to a positive correlation [12].

Owing to the directionality of the studied relationship, we were unable to include in the meta-analysis a case-control study by Long et al [38] that looked at rosacea incidence in various cancers in China. However, they found that patients with breast cancer had a significantly higher incidence of rosacea compared with individuals without rosacea (OR 5, 95% CI 4.02-6.2) [38]. Although this is the only study that was found of the inverse relationship, the potential for a bidirectional association lends further credence to the hypothesis that these 2 conditions are related.

A total of 8% (3/39) of the studies analyzed the association between rosacea and malignant melanoma. Egeberg et al [6] found no increased occurrence of melanoma in patients with rosacea (HR 1.10, 95% CI 0.95-1.27). Similarly, Li et al [7] also found no significant increase in the development of malignant melanoma in patients with rosacea (RR 0.96, 95% CI 0.57-1.62). Finally, the findings of Cho et al [11] noted that patients with rosacea had no association with the development of melanoma compared with the reference population (HR 1.69, 95% CI 0.25-11.37).

Although excluded from the meta-analysis because of not being in full text, an abstract by Erickson et al [13] investigated sex-specific differences in patients with rosacea and their likelihood of developing certain cancers. The study included 11,466 patients with rosacea, 8676 of whom were female, and 175,363 patients who did not have rosacea, 929 of whom were female. Interestingly, this abstract found that rosacea was inversely associated with the development of melanoma in female patients (RR 0.63, 95% CI 0.47-0.85). No significant relationship was found in male patients with rosacea [13]. Overall, despite the small number of studies available, there was no evidence of an association between the 2 conditions.

A total of 5% (2/39) of studies were found that examined the association between rosacea and hepatic cancer. Egeberg et al [6] noted that patients with rosacea have an increased risk of developing hepatic cancer (HR 1.42, 95% CI 1.06-1.90). The findings of Cho et al [11] noted that patients with rosacea had no association with the development of hepatic cancer compared with the reference population (HR 1.32, 95% CI 0.89-1.95).

In total, 3% (1/39) of studies were found that examined the association between rosacea and glioma. Egeberg et al [3] conducted a cohort study on Danish citizens aged >18 years on January 1, 1997, and followed them through to December 31, 2011. Patients with rosacea or glioma at baseline were excluded to allow for the study of the temporal relationship between the conditions. The reference population for the study was 5,416,138 individuals with a mean age of 40.8 (SD 19.7) years, and the rosacea diagnosis group contained 68,372 patients with a mean age of 42.2 (SD 16.5) years. The reference population consisted of 2,684,509 male patients and 2,732,029 female patients. Although the correlation was weak, the study found a significantly increased risk of glioma in patients with rosacea (incidence rate ratio 1.36, 95% CI 1.18-1.58). Interestingly, it was found that the rosacea-associated increased risk of glioma was greater in males than in females [3].

In addition, the excluded case-control study by Long et al [38] found that patients with glioma were more likely to have had rosacea (OR 2.16, 95% CI 1.12-4.17), pointing to a potential bidirectional relationship between the 2 conditions.

A total of 5% (2/39) of the studies analyzed the association between rosacea and thyroid cancer, reporting conflicting results. Egeberg et al [6] found no significant increase in the development of thyroid cancer in patients with rosacea (HR 1.06, 95% CI 0.68-1.65), whereas Li et al [7] found an increase in the development of thyroid cancer in patients with rosacea (RR 1.59, 95% CI 1.07-2.36).

Our review found that, based on the available literature, there is a positive association between rosacea and glioma, NMSC, and breast cancer. Rosacea was not found in any study to be significantly associated with melanoma. Regarding BCC, SCC, and thyroid and hepatic cancers, a clear conclusion could not be drawn because of conflicting results across 2 studies.

Our review found conflicting evidence regarding the association between BCC and rosacea. Although Li et al [7] found a positive association between BCC and rosacea in a cohort study of female patients, Lin et al [10] found no association between BCC of the body and rosacea in female patients and an inverse association between BCC of the face and rosacea in female patients. More studies will be needed to determine whether an association truly exists. Interestingly, across our reviewed cohort studies, no positive associations were found between rosacea and facial BCC or facial SCC. This may be due to patients with rosacea adopting better sun-protective measures such as sunscreen and broad hats to avoid flaring their rosacea, simultaneously reducing their risk of facial NMSC. It should be noted that there are currently no comparative studies analyzing the association between BCC and rosacea in male patients. Our review also found 14 case reports and case series discussing the presence of BCC in male patients with rhinophyma. It is uncertain whether the increased prevalence of BCC in this population is due to the male sex, the increased severity of rosacea in patients with rhinophyma, or both. Although rosacea is more commonly seen in females, rhinophyma, a subtype of rosacea, is found to be more common in White males aged >50 years. Although the reason for this is unknown, it is important to note that, according to the current literature, males with rhinophyma are 3% to 10% more likely to develop some form of skin cancer at the site of the nodular lesion [39]. Characteristics to be watchful for in patients with long-standing rhinophyma include sudden changes in ulceration of the lesion and rapid growth associated with malodorous drainage [15].

Our review found a positive association between rosacea and NMSC. It should be noted that actinic keratosis was also found to be associated with rosacea in one cohort study [11]. UV exposure is thought to be a common pathogenic factor in the development of skin cancer, actinic keratosis, and rosacea. It is thought that patients with rosacea have an altered skin barrier and are more likely to have had higher UV exposure at early ages, which may predispose them to skin cancers such as SCC and BCC [3,7]. UV radiation is a known risk factor for skin cancers and also plays an important chronic role in rosacea development through the generation of reactive oxygen species and cathelicidin expression, in addition to being a known acute trigger for rosacea outbreaks [40]. It should be noted that, although NMSC had a positive association with rosacea, SCC showed conflicting results. Although Lin et al [10] found an overall positive association between SCC and rosacea, Li et al [7] found no such association. The conflicting findings between studies on NMSC versus BCC and SCC when it comes to rosacea may be because all current comparative studies regarding BCC, SCC, and rosacea have exclusively used female patients. In comparison, Cho et al [11] found an increased risk of NMSC in patients with rosacea in a cohort that included male patients with SCC and BCC. It is possible that there are gender-specific differences when it comes to rosacea and NMSC, perhaps because of different sun-avoidant practices when having rosacea. In addition, although many studies on rosacea, NMSC, BCC, and SCC controlled for age, smoking status, BMI, and alcohol intake among other potential confounders, race and socioeconomic status were brought up as additional possible confounders that were not adjusted for. More studies are needed to analyze the impact of sex, racial, and socioeconomic differences when it comes to rosacea and NMSCs.

It should be noted that, regarding melanoma, all comparative studies (3/3, 100%) analyzed in our review found no association between melanoma and rosacea. Wu et al [41] proposed that the risk of SCC and BCC is positively associated with cumulative UV exposure, whereas melanoma tends to be more associated with intermittent UV exposure. Therefore, it is possible that cumulative UV radiation may be an important confounding factor in the development of both skin cancers and rosacea, accounting for the seemingly positive association that we observed between rosacea and NMSC or SCC but not melanoma.

Other than one study indicating no association, the relationship between rosacea and breast cancer seemed to be positive. Given the prevalence of both rosacea and breast cancer in females, there may be a component of hormonal changes and estrogen mediating some common inflammatory and immune-related causes [38]. However, there was conflicting evidence regarding the association between hepatic cancer and rosacea. Egeberg et al [6] found a positive association between hepatic cancer and rosacea; however, this result was potentially confounded by the fact that the rosacea group also reported greater alcohol consumption than the controls. Regarding glioma, underlying the increased odds of glioma development in patients with rosacea may be common inflammatory pathways dependent on matrix metalloproteinases and the activation of interleukin-17. In particular, matrix metalloproteinase-9 plays an important role in both rosacea pathogenesis and regulation of cell invasion in malignant glioma [3]. One study found increased expression of matrix metalloproteinase-9 in tumor tissue specimens from 76% of patients with glioblastoma, the most common and aggressive malignant form of glioma [42]. Interleukin-17 upregulation is also recognized as a hallmark of rosacea and may play a role in immune suppression in glioma [43,44].

Studies examining the association between rosacea and thyroid cancer reported conflicting results. Accumulating evidence demonstrates that chronic inflammation plays a pivotal role in the pathogenesis of thyroid cancer [45,46]. Therefore, inflammation may be a potential link between rosacea and thyroid cancer. However, more studies need to be conducted to clarify this relationship.

Our study was not able to conduct a meta-analysis because of the statistical heterogeneity between the studies. A meta-analysis would have further clarified the association between rosacea and the various cancers. In addition, there were a small number of studies on most of the cancers included in our review. Other limitations include a lack of studies examining the association between rosacea and cancer in skins of color. Most of the research described in this review was conducted on White, middle-aged, and female populations; further studies are needed regarding rosacea and cancer incidence in skins of color and male populations. In addition, the field of literature regarding BCC and rhinophyma or ocular rosacea is shallow and only comprises case reports/case series. These case series or case reports were included; however, they have an inherent bias because of the absence of a comparative group.

Our review of the current literature found that rosacea is significantly associated with NMSC, glioma, and breast cancer. An association between rosacea and thyroid cancer as well as between rosacea and hepatic cancer was also reported, but more studies are needed because of the limited amount of data. Rosacea does not appear to be associated with melanoma. Further studies should be conducted to determine whether there is an association between thyroid cancer and rosacea.