Evaluation of sweating responses in patients with systemic connective tissue disorders using the quantitative sudomotor axon reflex test

In systemic connective tissue disorders, eccrine sweat glands are frequently attacked by immune cells, as evidenced by pathological observations.


Introduction
Sweat glands play an essential role in seasonal heat acclimation and in maintaining systemic body homeostasis 1 . Therefore, to defend eccrine sweat glands from an autoimmune response, sweat glands are immune privileged 2 . Eccrine sweat glands in patients with systemic connective tissue disorders (SCTDs) exhibit histopathological abnormalities. In particular, lymphocyte in ltration and epithelialmesenchymal transition were observed around the eccrine sweat glands of patients with systemic sclerosis (SSc), systemic lupus erythematosus (SLE), and Sjogren's syndrome (SS) [3][4][5] . Possibly related to the above ndings, fatal heat stroke in a patient with diffuse scleroderma has been reported 6 . This led us to hypothesize that patients with SCTDs might have impaired sweating ability as well as an impaired capacity to respond to seasonal changes in temperature. Although there are several reports on sweating ability in patients with SS and atopic dermatitis 7,8 , no studies have been published focusing on seasonal changes in the sweating response and skin symptoms in patients diagnosed with SCTDs.
Eccrine sweat glands receive both cholinergic and adrenergic innervation and are controlled by the autonomic nervous system 9 . Aside from sweating, sympathetic nerve activity also regulates changes in body temperature according to the environment via vasoconstrictor nerves primarily responsive to noradrenaline, with the result that autonomic failure can affect both sweating ability and vasoactivity.
The hyperactivation of the sympathetic nervous system in response to cold results in peripheral vasospasm and vasoconstriction in patients with SCTDs, known as Raynaud's phenomenon 10 . These responses cause pain, digital ulceration, and necrosis, resulting in a signi cantly reduced quality of life for those patients, especially in winter. Given the physiological functions of the sympathetic nervous system, autonomic abnormalities affect both sweating ability and vascular activity. Therefore, we speculate that there may be a relationship between sweating and Raynaud's phenomenon. In recent years, botulinum toxin, which is widely used as a treatment for hyperhidrosis 11 , has been used to treat Raynaud's symptoms by blocking the sympathetic nervous system, and several systematic reviews and follow-up studies have validated its therapeutic effects [12][13][14] . However, it is not clear whether the antiperspirant effect of botulinum toxin contributes to the improvements in Raynaud's symptoms or digital ulcers in patients with SCTDs.
The study aimed to evaluate sweating ability in patients with SCTDs by using the quantitative sudomotor axon re ex test (QSART) and to identify clinical pro les that include seasonal variations, disease-related differences, and associations with clinical factors such as Raynaud's phenomenon. This is the rst study to provide a basis for understanding sweating ability in patients with SCTDs and contribute to developing treatment strategies for patients with autonomic peripheral circulatory disorders.

Results
The characteristics of patients and healthy participants are summarized in Table 1. Among 19 patients with SS, 12 patients had a comorbidity: eight and ve patients had SSc and SLE, respectively, one of whom had both SSc and SLE. Among 86 patients, seven and 11 patients were unable to perform the QSART in the summer and winter, respectively. Consequently, the number of seasonal paired data was 67.
Systemic corticosteroids, prostacyclin derivatives, serotonin receptor antagonists, and cholinergic agents were used in 25, 22, 14, and 3 patients, respectively. No parasympathomimetic agents were used in the study participants.  Axon re ex sweat volume

Comparison by disease types
We investigated axon re ex sweat volumes (ARSVs) in patients with SSc, mixed connective tissue disease (MCTD), SLE, SS, and dermatomyositis (DM) and healthy controls (see Supplementary Fig. S1a online). All disease groups showed as much or more sweating than healthy participants in both summer and winter. We analysed the mean differences in ARSVs between participants with each disease type and healthy participants (see Supplementary Table S1 online). We did not observe a signi cant difference in axonal re ex sweat volume in any disease group compared to healthy controls.

Seasonal comparisons
We investigated ARSVs as participant-wise ratios of the volume in summer to that in winter (Fig. 1a). For all patients, the geometric mean of the seasonal ratios (95% Con dence Interval: 95%CI) was 1.49 (1.20-1.87), indicating that ARSVs were larger in summer than in winter. This trend was also observed in the healthy control participants, with a geometric mean of 1.90 (0.99-3.66). However, some patients showed higher sweat volume in winter than in summer. There were 23 patients, corresponding to 34% in the seasonal paired data (Fig. 1b).
Patients who exhibited more perspiration in winter than in summer were found among all disease groups except for patients with DM (see Supplementary Fig. S1b online). In a comparison with the healthy participants, we observed a slight association of this phenomenon with having a diagnosis of SSc (OR [95% CI], 6.14 [0.40-94.72]). However, we did not observe an association between increased axial re ex sweat volume in winter compared with summer and smoking history, illness duration, nger temperature, nailfold capillary changes, skin sclerosis severity, skin symptoms (including digit ulcers, chilblains, subcutaneous calci cations, and telangiectasia), or disease complications (see Supplementary Table S2 and Figure S2 online).

Relationship with skin sclerosis
We investigated the relationship between the ARSV and the degree of skin sclerosis, de ned by the modi ed Rodnan total skin thickness score (MRSS), in patients with SSc ( Fig. 2). MRSS scores were dichotomized as low scores (≤ 10) and high scores (> 10), which represents weak disease and moderate to severe disease, respectively 15 . The groups with MRSSs of ≤ 10 and > 10 included patients with scores of 0 to 10 and 13 to 26 in summer and 0 to 9 and 13 to 25 in winter. We observed lower ARSVs in the group with an MRSS > 10 than in the group with an MRSS ≤ 10.

Relationship with Raynaud's phenomenon
We analysed the association of the ARSV with the severity of Raynaud's activity, de ned by the Raynaud's condition score (RCS). The total scores were dichotomized as low scores (0-7) and high scores (8-16) (Fig. 3a). Values for the subcomponents of attack, pain, colour, and duration were dichotomized into those without symptoms (0) and those with symptoms (≥ 1) (Fig. 3b). We did not observe a clear relationship between the dichotomized total RCS and the sweat volume. However, there was a clear relationship between the pain score and the sweat volume. In summer, the geometric mean of the ratio

Sweat latency Comparison by disease types or speci c autoantibodies
We investigated the sweat latencies for each disease (Fig. 4a). In the summer, all patients with MCTD, SLE, and DM began to sweat within 160 s. In contrast, three of the patients with SSc (7%) and two of the patients with SS (11%) did not begin to sweat, even after 300 s (the time limit for this measurement). In the winter, patients with SSc and SS showed the same trends as seen in the summer, and two of the patients with MCTD (29%) also showed this trend.
Since the sweat latency of SSc patients was prolonged compared with that of patients with other diseases, we compared the sweat latencies between groups of patients with each speci c antibody for SSc (Fig. 4b).
In Analysis regarding an attenuated response to acetylcholine In Fig. 4, patients who did not sweat during the observation period were considered to have an attenuated response to acetylcholine. Therefore, we further analysed the details of these patients (

Discussion
We employed the QSART to assess sweating ability. A low ARSV and/or prolonged sweat latency in the QSART can be used to diagnose abnormalities of the postganglionic sympathetic bres or eccrine sweat glands associated with poor acetylcholine-induced sudomotor responses 16,17 . The sweating responses of individuals living in Japan are more pronounced in summer than in winter 18 . The changes in sweating activity measured by the QSART con rmed the involvement of the peripheral nervous system in altering sudomotor activity during seasonal acclimation 19 .
In the present study, we assessed both sweating ability and its association with the clinical severity of Raynaud's symptoms in patients with SCTDs. We found that none of the disease groups showed an apparent decrease in sweat volume compared to healthy participants (Supplementary Table S1 online). However, it was a novel nding that approximately one in three patients (34%) showed less sweating in summer than in winter (Fig. 1). This phenomenon was more common in patients with SSc than in healthy participants.
Because of the unique seasonal changes in sweating ability shown in the present study, we anticipated that patients with SCTDs have a dysregulated sweating ability due to abnormal peripheral nerve responses. While problems associated with heat adaptation are major factors for heat stroke 20 , there are no speci c data available on the risk of heat stroke in patients with SCTDs. Li et al 21 noted that left uncontrolled, recent trends in global warming will lead to an increased risk of heat stroke in 1.2 billion people by the year 2100. According to this assumption, we should pay attention to the relationship between global warming and seasonal perspiration in patients with SCTDs.
We focused on RNAP-positive SSc patients because they may have characteristic sweating abnormalities that are not present in other patient groups. Patients with RNAP-positive SSc had prolonged sweat latencies (Fig. 4), with 44% of them showing a poor response to acetylcholine (Table 2). Furthermore, RNAP-positive patients showed both less sweating and smaller seasonal differences than ACA-positive or Topo1-positive patients (see Supplementary Fig. S4a, Fig. S4b online). Patients with a high degree of skin stiffness showed less sweating than patients with a low degree of skin stiffness or no skin stiffness (Fig. 2); 57% of patients with an MRSS > 10 were RNAP positive.
Autoantibodies reactive with RNA polymerase (RNAP) III are con rmed to be strongly associated with diffuse or extensive cutaneous involvement and renal crisis 22,23 . Severe and rapidly progressive cutaneous brosis may attenuate the response to acetylcholine by disrupting and reducing the number of eccrine sweat glands and nerve bres. In some patients with SS, eccrine sweat gland dysfunction is associated with autoimmune mechanisms mediated by CD8 T cells 24 or M3 receptor-speci c autoantibodies 25 . As we did not perform pathological assessment of eccrine glands, we cannot exclude the possibility that RNAP is directly associated with eccrine gland dysfunction. Further research on sweat gland impairment and the autonomic nervous system in RNAP-positive SSc patients may lead to a better understanding of peripheral circulation in patients with SCTDs.
Our results also indicated that patients with a higher degree of the pain in the RCS evaluation had a higher sweat volume ( Fig. 3 and Supplementary Table S3). Regarding this phenomenon, we anticipated that the neuronal transmitters that convey pain signals might be involved in sweating ability. It has been reported that patients with Raynaud's symptoms exhibit abnormal responses to pain-associated neurotransmitters, including substance P, glutamate, and calcitonin gene-related peptides 26 , which may contribute to Raynaud-related pain. On the other hand, substance P and calcitonin gene-related peptide are expressed in normal sweat gland secretory cells or around the sweat glands 27,28 and contribute to gland secretion in response to harmful stimuli. Taken together, these ndings suggest that the response to neurotransmitters might link the pain in Raynaud's phenomenon and increased sweating in winter.
Increased winter sweating with severe pain in Raynaud's phenomenon might explain the phenomenon of increased winter sweating in some SCTD patients shown in Fig. 1. Tabata et al. studied sweating in SSc patients by using capillaroscopy and reported that 7 out of 21 patients developed increased sweating, although they did not perform a seasonal analysis 29 . The mechanism by which the overactivity of the sympathetic nerves that causes Raynaud's phenomenon affects sweating remains to be explored in additional studies involving a larger patient cohort, autonomic function tests, and pathological examination.
In conclusion, most patients did not show decreased sweating compared to healthy participants, but RNAP-positive patients with SSc had impaired sweating. One in three patients with an SCTD showed more sweating activity in winter than in summer, which is the opposite of the regular change. Although sweat volume was not associated with the total RCS, the pain of Raynaud's phenomenon increased the volume of sweating.
A limitation of this study was the small sample size for each disease. Our study did not consider the effects of regularly used drugs, including external agents such as moisturizers, the obscurity of patients' answers about Raynaud's symptoms, the practice of sports, the living environment, and patients' physical constitutions. In SSc patients, the reduced permeability of acetylcholine due to the hardness of the skin should be considered. Further study in combination with other autonomic nervous system assessments and more detailed patient backgrounds can provide a better understanding of the signs and biomarkers associated with peripheral nerve disorders and contribute to the development of treatment strategies for patients with autonomic peripheral circulatory disorders.

Materials And Methods
This study was conducted according to the study protocol 30 , which is available at the Japan registry of clinical trials.

Participants
The study population comprised 85 Japanese patients in the dermatology department of Nagasaki University Hospital with established diagnoses of SCTDs associated with Raynaud's phenomenon, including those diagnosed with SSc, MCTD, SLE, SS, and DM. Patients with SSc, MCTD, SLE, and SS who met the diagnostic criteria for those diseases were included in this study [31][32][33][34] . The study included 11 healthy individuals as controls. The normal use of either oral or topical medications was not restricted for any patient.

QSART
To assess the effect of seasonal changes on the sweating response, the QSART was conducted during the summer (June 2019-September 2019, mean, minimum, and maximum temperatures of 25.6°C, 17.5°C, and 37.3°C, respectively) and during the winter (December 2019-March 2020, mean, minimum, and maximum temperatures of 10.6°C, 0.8°C, and 21.2°C, respectively).
The QSART was developed by Low et al 17 and involves the ow of dry air into a capsule followed by the measurement of the sweat volume. This is performed by quantifying the moisture levels in the out owing air with a high-sensitivity hygrometer. In this study, the participants rested in a thermostatic chamber (room temperature, 23-26°C; room humidity, 40-60%) for at least 30 min prior to the physiological examination. The QSART was performed using a SKN-2000 (Skinos Co., Ltd., Nagano, Japan), and acetylcholine was delivered to the dermis of each participant's forearm by iontophoresis for 5 min with a current of 5 mA. The amount of sweat measured for 5 min was recorded as the ARSV, and the time until the start of sweating was recorded as the sweat latency. There are no de ned standard reference values for the QSART. Therefore, in this study, the QSART values of healthy volunteers were used as the reference values. The results for sweat latency (the interval of time required for sweating) were evaluated by graphic representation with a Kaplan-Meier method.  Figure 1 The axonal re ex sweat volume (ARSV) and its seasonal differences. (a) Plots of changes in the ARSV measured by the quantitative sudomotor axon re ex test in summer and winter. The graph on the left and the right shows data for patients with systemic connective tissue disorders (SCTDs) (67 paired sets) and

Figures
healthy controls (11 paired sets), respectively. Most patients with SCTDs sweated as much or more than healthy controls, but some patients sweated more in winter than in summer. (b) The number of patients with SCTDs and healthy controls in the groups with more sweating in winter than in summer (summer < winter) and summer than in winter (summer > winter). The MRSS was dichotomized as ≤10 and >10. The group with an MRSS >10 showed lower ARSVs than the group with an MRSS ≤10 in both summer and winter. The boxes indicate the interquartile ranges. The middle lines within the boxes represent the medians. The whiskers above/below the boxes are the ranges of the observed values from the 75% quartiles/25% quartiles to the maximums/minimums within 1.5 times the interquartile ranges.

Figure 3
Plots of the axonal re ex sweat volume (ARSV) according to the Raynaud's condition score (RCS). (a) and (b) represent the distributions of ARSVs by the total RCS and the score for the pain subcomponent, respectively. The horizontal axis shows the dichotomized score. The total RCS was dichotomized as 0-7 and 8-16. The scores for pain were dichotomized as 0 and ≥1. The top and bottom panels represent summer data (n = 78) and winter data (n = 73), respectively. A relationship was not found between the dichotomized total RCS and the sweat volume. However, a clear relationship was found between the pain score and the sweat volume. The p values were calculated via Fisher's exact method. The boxes and whiskers are drawn by the same rule in Figure 2. Supplementary Files