Identifying the neural basis for rosacea using positron emission tomography‐computed tomography cerebral functional imaging analysis: A cross‐sectional study

Abstract

Background

The neural basis of rosacea is not well understood. This study aimed to determine whether cerebral glucose metabolism (CGM) changes on ¹⁸F‐fluorodeoxyglucose (¹⁸F‐FDG) positron emission tomography (PET)/computed tomography (CT) scans can detect functional network changes in specific brain areas in patients with rosacea.

Materials and methods

Eight adults with rosacea and 10 age/sex‐matched healthy adults (controls) were enrolled in the study. ¹⁸F‐FDG PET/CT brain images for all eight patients and whole‐body images for two of the patients were analyzed qualitatively and semi‐quantitatively. Differences between the study groups were examined using Fischer's exact test and a Student's t‐test. A voxel‐based analysis using statistical parametric mapping was performed to compare the brain metabolism of the patients with that of the controls.

Results

Compared with the controls, the patients with rosacea showed extensive changes in the CGM signals in the cerebral cortex and limbic system, with less CGM shown in the right superior parietal lobule, right postcentral gyrus, right parahippocampal gyrus, left superior frontal gyrus, and lateral posterior thalamic nucleus and more CGM in the right precentral gyrus, left inferior frontal gyrus, and cerebellar tonsil. No dysmetabolic lesions were found in the whole‐body ¹⁸F‐FDG PET/CT images.

Conclusion

Specific neural functional changes occur in patients with rosacea that may explain its pathogenesis.

1. INTRODUCTION

Rosacea is a common chronic inflammatory disease that affects the skin of the central area of the face and typically manifests as flushing, erythema, papules, pustules, and phymatous changes. ¹ The global prevalence of rosacea is estimated to be 5.41% for women and 3.90% for men; however, it varies from 1% to 22% in the general population. ¹ , ²

The pathophysiology of rosacea has not yet been fully elucidated; however, in recent years, studies have focused on the role of neurovascular homeostasis in the pathogenesis of rosacea. ³ Some patients with rosacea reportedly have a cerebral vascular control deficit that manifests as an exaggerated sympathetic response, along with subsequent flushing and a burning sensation. ⁴ Patients with rosacea may have comorbid cardiovascular or gastrointestinal disease or certain neuropathologic or psychiatric disorders, such as migraines, Alzheimer's disease, Parkinson's disease, or depression. ⁴ The treatments for rosacea include antibiotics, retinoids, acaricidal agents, vasoconstrictors, intense pulsed light therapy, and laser therapy. ¹ , ⁵ Systemic neuroleptic agents and tricyclic and pain‐modifying antidepressants have also been reported to be effective for patients who do not respond to conventional therapy. ⁶ , ⁷ Responses to these treatments suggest that neural functional changes occur in patients with rosacea. However, no neuroimaging studies of rosacea have been reported to date.

¹⁸F‐fluorodeoxyglucose (¹⁸F‐FDG) positron emission tomography (PET)/computed tomography (CT) is an emerging neuroimaging technique that has been used widely for the functional evaluation of glucose metabolism in both the brain and the rest of the body. In the present study, ¹⁸F‐FDG PET/CT was used to explore cerebral glucose metabolism (CGM) in patients with rosacea.

2. MATERIALS AND METHODS

This study had a cross‐sectional design that was reviewed and approved by the ethics committee of the First Affiliated Hospital of Nanjing Medical University (2020‐SRFA‐082). Informed consent in a written form was obtained from all the study participants.

2.1. Clinical data

Eight patients with a history of rosacea were diagnosed according to the standard diagnostic criteria ¹ by two senior dermatologists, who worked independently and were recruited from the Department of Dermatology of the First Affiliated Hospital of Nanjing Medical University between January 2019 and January 2020. All the patients were diagnosed at least 1 year prior to the initiation of the study. Patients younger than 18 years or older than 60 years and/or with a history of cardiovascular, digestive, respiratory, endocrine, neoplastic, or neuropsychiatric disease, and lactating and/or pregnant women were excluded.

The demographic and clinical characteristics of the participants, including age, sex, weight, height, blood pressure, and fasting blood glucose were recorded. Facial images of the patients with rosacea were obtained using a VISIA 6.0 Complexion Analysis System (Canfield Scientific Inc., Parsippany, NJ, USA). The severity of each patient's rosacea was assessed by two senior dermatologists who were blinded to the experiment, according to the 5‐point Investigator Global Assessment scoring system. ⁸

Ten healthy participants, who were matched for age and sex with the patients in the rosacea group, were enrolled in the control group. None of the healthy participants had any history of cardiovascular, digestive, respiratory, endocrine, neoplastic, or neuropsychiatric disease, nor did any of them have any structural abnormalities in the brain or history of cerebral trauma.

2.2. ¹⁸F‐FDG PET/CT

For the eight patients in the rosacea group, ¹⁸F‐FDG PET/CT (Biograph 16 True Point, Siemens, Munich, Germany) imaging of the brain was performed at the Department of Nuclear Medicine. Whole‐body scans were obtained for two of these patients. ¹⁸F‐FDG PET/CT brain and whole‐body scans were also obtained for the participants in the control group. Each of the study participants was required to fast for at least 4 h before the scan, during which time their fasting blood glucose level was evaluated (reference range: 3.9–6.1 mmol/L). All the participants were instructed to rest for at least 30 min before the ¹⁸F‐FDG tracer (3.7–5.55 MBq/kg) was injected subcutaneously into a vein in the dorsum of the hand. After an additional 40 min of rest, all the participants underwent the PET/CT. The brain PET scan (120 s/bed position) included a low‐amperage CT scan (120 kV and 380 mA) that was performed for attenuation correction. The images were acquired in three‐dimensional mode over 10 min. The data were subjected to an iterative reconstruction (matrix: 256 × 256; thickness: 5 mm). Tomography was performed in the transverse, sagittal, and coronal planes, and fusion images were obtained after the iterative reconstruction.

2.3. Analysis of the ¹⁸F‐FDG PET/CT images

The ¹⁸F‐FDG PET/CT images were evaluated visually by two senior nuclear medicine physicians. After exclusion of any lesions in the body and the brain, the brain PET data for the two groups were analyzed using statistical parametric mapping (SPM). All the PET data were converted from the Digital Imaging and Communications in Medicine format to the analyze format. Preprocessing included standardization and smoothing using a Gaussian kernel of 10 × 10 × 10 mm (full width at half maximum).

2.4. Statistical analysis

The clinical data were compared between the study groups using a paired t‐test, and the results are shown as the mean and standard deviation. All the statistical analyses were performed using SPSSAU (Statistical Product and Service Software Automatically) software (SPSS online, Version 22.0, https://spssau.com/, Beijing). A p‐value <0.05 was considered statistically significant.

The statistical analysis of the ¹⁸F‐FDG PET/CT data was performed using the SPM8 software running on MATLAB R2013b. Parameter setting and estimate staging were performed to test the null hypothesis. To explain the differences in the brains of the participants, a grand mean scaled value of 50 was selected as a unified standard. Based on the SPM analysis, and in accordance with a p < 0.001 threshold level and the voxel threshold (K = 25), different metabolic regions in the brain were projected onto a three‐dimensional image with Talairach coordinates. The PET findings were superimposed on a magnetic resonance imaging (MRI) template to ensure accurate identification of the structures of interest.

3. RESULTS

3.1. Clinical characteristics

There were no significant differences in sex, age, weight, height, blood pressure, or fasting blood glucose between the rosacea group and the control group (Table S1).

The main rosacea phenotype among the patients was the erythematotelangiectatic subtype, with an average Investigator Global Assessment score of 2.75 ± 0.97 (Figure S1). Most of the patients had mild to moderate erythema, papules, and telangiectasis. Most of the patients experienced a stinging/burning sensation and easy flushing.

3.2. Visual examination of rosacea and SPM analysis

On visual examination, the between‐group difference in the ¹⁸F‐FDG uptake (Figure 1) was not found to be significant. No hypermetabolic or hypometabolic lesions were found in the PET images for the two patients who underwent whole‐body scanning (Figure S2).

FIGURE 1.

The ¹⁸F‐fluorodeoxyglucose (¹⁸F‐FDG) positron emission tomography/computed tomography brain images from a healthy control and a patient with rosacea. (A) An image from a 33‐year‐old woman showing relatively symmetrical cerebral ¹⁸F‐FDG metabolism. (B) No obvious abnormal metabolism is visible in this image from a 30‐year‐old female patient with rosacea

The SPM analysis showed that compared to the control group, the patients in the rosacea group had focal hypometabolism of glucose in the right superior temporal gyrus (BA22), right superior parietal lobule (BA7), right postcentral gyrus (BA7), right parahippocampal gyrus, left superior frontal gyrus (BA6), and posterolateral thalamic nucleus compared with the participants in the control group, as well as higher glucose metabolism in the right precentral gyrus (BA44), left inferior frontal gyrus (BA44, 46, and 47), and cerebellar tonsil (Figure 2). The CGM changes in the patients with rosacea are shown in detail in Table 1.

FIGURE 2.

Cerebral glucose metabolism changes in the patients with rosacea compared with the healthy controls. In these three‐dimensional renderings, the (A) cerebral hypometabolism and (B) hypermetabolism regions are indicated by the red‐to‐yellow color (K = 25). The results of the statistical parametric mapping analysis of the two groups’ positron emission tomography images are superimposed on a magnetic resonance imaging template to facilitate the accurate identification of the affected structures. The increased metabolism regions are indicated by the red‐to‐yellow color, and the decreased regions are indicated by the blue‐to‐green color (K = 25, the color bar indicates t‐values) (C and D)

TABLE 1.

Comparison of the cerebral glucose metabolism between the patients with rosacea and the healthy controls

Glucose metabolism	Brain region	Peak coordinates x, y, z (mm)	Brodmann areas	t	p‐Value
Decreased	R. superior temporal gyrus	50, −24, 1	22	7.80	<0.001
	R. superior parietal lobule	22, −67, 64	7	6.35	<0.001
	R. postcentral gyrus	24, −55, 69	7	4.56	<0.001
	R. parahippocampal gyrus	24, −12, −13		5.22	<0.001
	L. superior frontal gyrus	−12, 24, 64	6	3.79	<0.001
	L. lateral posterior nucleus thalamic nucleolus	−20, −19, 12		5.70	<0.001
Increased	R. precentral gyrus	57, 16, 7	44	5.44	<0.001
	L. inferior frontal gyrus	−53, 12, 16	44	5.04	<0.001
		−44, 15, −7	47	5.00
		−48, 43, 13	46	4.17
	L. cerebellar tonsil	−30, −39, −33		4.17	<0.001

Abbreviations: L., left; R., right.

4. DISCUSSION

Rosacea is a common chronic skin disorder characterized by erythema, telangiectasias, flushing, pustules, and fibrosis affecting the central aspect of the face. In recent years, studies of the pathogenesis of rosacea have focused on the role of neurovascular homeostasis. Some systemic diseases, including cardiovascular disease, gastrointestinal and autoimmune disorders, depression, and neurologic conditions, ⁹ are believed to be associated with rosacea. It has been reported that rosacea is related to multiple neurologic disorders, such as migraines, ¹⁰ Parkinson's disease, ¹¹ dementia (particularly Alzheimer's disease), ¹² , ¹³ glioma, ¹⁴ and facial dystonia, ¹⁵ as well as to various psychiatric conditions or disorders, such as anxiety, ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ depression, and schizophrenia. ¹⁷ Therefore, an increased focus on neurologic changes in patients with rosacea may be warranted.

¹⁸F‐FDG PET/CT provides complementary information on CGM, and it is a helpful adjunctive diagnostic tool in clinical practice. There has been some research on the cerebral activity associated with several neurological and psychological conditions, including some of the above‐mentioned comorbidities of rosacea. ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶

In dermatological clinical practice, ¹⁸F‐FDG PET/CT is used primarily for neoplasms, ²⁷ , ²⁸ leukemia cutis, ²⁹ sarcoidosis, ³⁰ and plasma cell disorders. ³¹ To date, no neuroimaging studies have been conducted on rosacea. The present study is the first to use ¹⁸F‐FDG PET/CT to evaluate both cerebral and whole‐body glucose metabolism in patients with rosacea.

The major brain areas that showed hypometabolism in patients with rosacea as compared to the healthy controls were the parietal and temporal lobes and the postcentral gyrus, while the areas that showed hypermetabolism were the frontal lobes and the precentral gyrus, along with parts of the cerebellum. Our findings show some similarities and differences, with various reports of changes in brain metabolism in rosacea‐related diseases.

For patients with Parkinson's disease, the metabolism pattern is characterized by increased activity in the basal ganglia, pons, and cerebellum, with concomitant reductions in glucose metabolism in the premotor, pre‐supplementary motor, and posterior parietal cortices. ²¹ For patients with Alzheimer's disease, ¹⁸F‐FDG PET/CT imaging shows hypometabolism first in the precuneus and posterior cingulate cortex, followed by the temporoparietal lobes, and finally in the frontal lobes, with relative sparing of the motor and visual cortices. ²² In the present study, the metabolic changes in patients with rosacea were different from those observed in patients with Parkinson's and Alzheimer's disease. In patients with rosacea, the metabolic changes in the cerebral cortex were characterized by hypermetabolism in the areas in front of the central sulcus (e.g., the frontal lobe and anterior central gyrus) and hypometabolism in the area posterior to the central sulcus (e.g., the posterior central gyrus and parietal lobe).

Some of the CGM changes exhibited by the patients with rosacea are similar to those in patients with migraines and sensitive skin (SS). For patients with migraines, the changes were reported to be related to weakened functional connections from the right amygdala to the middle cingulate gyrus, the right anterior central gyrus to the right parahippocampal gyrus, and the left posterior central gyrus to the right anterior central gyrus. ²³ In our study, the patients with rosacea also showed decreased glucose metabolism in the postcentral and the parahippocampal gyri.

Querleux et al. ³² reported the use of functional MRI for the evaluation of brain function in patients with SS. Compared with the controls in their study, the patients with SS showed specific cerebral activation during a skin irritation test, which manifested as skin discomfort due to increased lactic acid activity in the primary sensorimotor cortex contralateral to the application site and bilaterally in the frontoparietal network. This includes the parietal cortex, prefrontal areas around the superior frontal sulcus, and the supplementary motor area, whose activity also spread into the ipsilateral primary sensorimotor cortex and bilateral peri‐insular secondary somatosensory area. In our patients with rosacea, hypermetabolism was evident in the inferior frontal and precentral gyri. Querleux et al. ³² also found that for patients with SS, lactic acid test revealed increased activity in the contralateral and bilateral anterior parietal skin networks of the primary sensorimotor cortex, which is consistent with the findings in our study, indicating that there are some neural functional changes common between rosacea and SS.

In our study, we observed extensive changes in the CGM signals in the cerebral cortex and limbic system of the patients with rosacea. There were some changes common with migraines and SS, which manifested in the parahippocampal gyrus and thalamus. We found decreased glucose metabolism in the left thalamus (primarily the posterolateral nucleus), which is consistent with the abnormal thalamic function and reduced volume of the thalamic nuclei, including the central nucleus complex, anterior nucleus, and lateral dorsal nucleus in patients with migraines.

Migraines and rosacea have some common neurovascular pathophysiological features, including changes in both the blood flow to the facial skin and the levels of perivascular neuropeptide, ³³ which may explain the similar changes in the CGM. Furthermore, there are some overlapping neurovascular mechanisms, including certain trigger factors (such as psychological stress and alcohol consumption) that activate the sensory component of the trigeminal nerve in the skin or brain. This results in the release of various neuropeptides, such as calcitonin gene‐related peptide, pituitary adenylate cyclase‐activating peptide, substance P, and neuropeptide Y, which can induce vasodilation, increased vascular permeability, and neurogenic inflammation, leading to aggravation of migraines or rosacea. ³³

According to the classification devised by Baumann, ³⁴ there are four discrete SS subtypes, including acne, stinging, allergic, and rosacea. Patients with the rosacea subtype show a tendency toward recurrent flushing, facial redness, and hot sensations. The clinical characteristics common to both the rosacea subtype of SS and rosacea might explain the similarity between the neural functional changes observed in our study and those reported by Querleux et al. ³² Functional neuroimaging has consistently highlighted hyperactivity in the limbic and paralimbic structures and the frontal gyrus in anxiety. ²⁴ , ²⁵ , ²⁶

The rates of embarrassment, low self‐esteem, social anxiety, and decreased Dermatology Life Quality Index scores are particularly high in patients with rosacea. ³⁵ In addition to feelings of stigmatization, patients with rosacea have been found to have a greater risk of depression or anxiety disorders, ³⁶ including generalized anxiety disorder. ³⁷ Consequently, our findings regarding the metabolic changes in patients with rosacea might be influenced by their anxiety status. Further research on the neurological mechanism of rosacea is needed to confirm this possibility.

This study has several limitations, the primary one being the limited sample size of the patients with rosacea. Injection of a radio‐labeled tracer like ¹⁸F‐FDG for research purposes can be an obstacle to recruiting volunteers for studies of this type. Further research is needed to assess the correlation between anxiety scores and the clinical evaluation of specific diseases and changes in brain metabolism.

5. CONCLUSION

In conclusion, we demonstrated that extensive changes occur in the CGM signals in the cerebral cortex and limbic system in patients with rosacea, making this the first study to provide a neural functional basis for this condition.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

Supporting information

Supporting Information

Liu Y, Xu Y, Guo Z, Wang X, Xu Y, Tang L. Identifying the neural basis for rosacea using positron emission tomography‐computed tomography cerebral functional imaging analysis: A cross‐sectional study. Skin Res Technol. 2022;28:708–713. 10.1111/srt.13171

DATA AVAILABILITY STATEMENT

The data that supports the findings of this study are available in the supplementary material of this article.