TGF-β1 in plasma and cerebrospinal fluid can be used as a biological indicator of chronic pain in patients with osteoarthritis.

INTRODUCTION: Previous studies have demonstrated that cytokines, transforming growth factor (TGF-β1), and brain-derived neurotrophic factor (BDNF) can impact the intensity of pain in rodents. However, the roles of cytokines, TGF-β1 and BDNF in humans with chronic pain in osteoarthritis remains unclear, and no comparison between plasma and central cerebral spinal fluid (CSF) has been conducted.
METHODS: Patients with osteoarthritis who were scheduled to receive spinal anesthesia were enrolled. The intensity of pain was evaluated with a visual analogue scale (VAS). In addition, patients with genitourinary system (GU) diseases and without obvious pain (VAS 0-1) were included as a comparison (control) group. The levels of TGF-β1, BDNF, tumor necrosis factor-α (TNF-α), and interleukin (IL)-8 within the CSF and plasma were collected and evaluated before surgery.
RESULTS: The plasma and CSF TGF-β1 levels were significantly lower in the osteoarthritis patients with pain (VAS ≥ 3) than in the GU control patients. Downregulation of plasma BDNF was also found in osteoarthritis patients with pain. The Spearman correlation analysis showed that the VAS pain scores were significantly negatively correlated with the levels of TGF-β1 in the CSF of patients with osteoarthritis. However, there was no significant correlations between the pain scores and the levels of BDNF, TNF-α, and IL-8 in either the CSF or plasma.
CONCLUSIONS: TGF-β1 but not BDNF, TNF-α, or IL-8 may be an important biological indicator in the CSF of osteoarthritis patients with chronic pain.

1. Introduction

Pain is an unpleasant stressful sensation that is caused by an acute injury or chronic disease [1]. Persistent chronic pain has been suggested to be a chronic disease state. For example, underlying diseases such as arthritis are usually associated with chronic pain for weeks or even months [2]. In patients with osteoarthritis (OA), chronic pain persists due to the progression of the disease and becomes a disabling symptom of OA [3]. Therefore, selection of pain control agents and finding the relevant biological indicators are important in the treatment of OA. Up to the present time, many studies have focused on treatments for the chronic pain caused by OA, in the hope that a pain-relieving strategy will improve the symptoms and the daily functioning in these patients [4-6]. In addition, an increasing number of studies are focusing on related biological or proteomic changes in pain conditions to determine either the pain indicators or pain control targets. Thus, in this study, we assumed that gaining an understanding of the molecular basis of pain perception may provide new insights into the treatment of chronic pain caused by OA.

Although the mechanisms of chronic pain in patients with OA is not well understood [7], the inflammatory process in the joint and its surrounding tissue is assumed to play a role in the developing of pain in patients with arthritis [8]. In patients with inflammatory arthritis, treatment with anti-rheumatic drugs is effective in reducing inflammatory pain symptoms [9]. Previous studies have found that inflammatory cytokines in the blood are associated with chronic low back pain [10]. Furthermore, serum cytokines, including tumor necrosis factor-α (TNF-α) and interleukin 8 (IL-8), have been found to be higher in patients with sciatica than in healthy controls [10]. In animal pain models, TNF-α [11, 12] and IL-8 [13, 14] have been shown to be significantly increased and to contribute to chronic neuropathic pain. Furthermore, transforming growth factor (TGF-β1), as a neuronal protective factor with an anti-inflammatory effect, has been demonstrated to inhibit the pain process in chronic constriction injury (CCI) models [15] and in neuropathic pain models [16]. These studies present strong evidence suggesting that cytokines and TGF-β1 may serve as potential pain indicators. However, the role of cytokines in chronic pain perception in the peripheral blood and central nervous system (CNS) of human subjects remains unclear. Evaluating the expression of cytokines and TGF-β1 in the blood and CNS may provide a better reference for the evaluation or treatment of chronic pain in clinical settings.

On the other hand, brain-derived neurotrophic factor (BDNF) can increase the intensity of pain in rodent inflammatory and neuropathic pain models [17, 18]. BDNF has been suggested to be the driving force for neuroplasticity in central sensitization in neuropathic pain models [17]. Conditional BDNF gene knockout mice have also demonstrated its involvement in chronic inflammatory pain [18]. These reports indicated that BDNF plays an important role in pain. However, the controversy surrounding the effects of inflammatory/anti-inflammatory cytokines and neurotrophic factor in pain remains. Thus, understanding the relationship between these molecules and pain perceptions are important, especially in the CNS and peripheral blood.

Therefore, in this study TGF-β1, TNF-α, IL-8 and BDNF were simultaneously collected from the cerebrospinal fluid (CSF) and plasma samples taken from patients with OA with (VAS score ≥ 3) or without (VAS score ≤ 2) significant clinical pain. The relationship between pain perceptions and molecular expression in CSF and plasma was evaluated and compared. We hypothesized that cytokines, TGF-β1 and the neurotrophic factors in CSF and plasma are correlated to the intensity of pain and that these molecules in subjects with chronic OA experiencing pain may serve as pain indicators of chronic pain in OA.

2. Materials and methods

2.1 Ethics approval and consent

The trial was registered prior to patient enrollment at clinicaltrials.gov (NCT03606915, Principal investigator: Yen-Chin Liu, Date of registration: 2015-05-22). This study was approved by the University’s Institutional Review Board (IRB #B-ER-104-070) for the Protection of Human Subjects at National Cheng Kung University Hospital and Tainan Hospital of Ministry of Health and Welfare, Taiwan. The study procedures were performed according to the rules of the IRB and fully explained to the participants. After the study had been completely described to the participants, informed consent was obtained from all patients in the study. This study was performed in accordance with the standards of ethics outlined in the Declaration of Helsinki.

2.2 The design of the study

To evaluate the changes in the biological molecules in patients at different levels of pain, patients with or without chronic pain who had received spinal anesthesia for lower body surgery were enrolled in this study. The patients’ pain intensity was evaluated based on the pain score prior to surgery. The levels of TNF-α, IL-8, TGF-β1, and BDNF both in CSF and plasma were also evaluated.

2.3 The selection of the patients

We collected patients with a pain history from 07/2015 to 02/2017 before they received spinal anesthesia for knee/hip OA surgery. Moreover, patients with genitourinary system (GU) diseases without obvious current pain at rest served as a comparison (control) group. None of the patients were pregnant; they had an American Society of Anesthesiologists physical status of 1 to 3 and were scheduled for lower body surgery. The inclusion criteria included 1) patients undergoing spinal anesthesia who were over 20 years of age, 2) patients who were able to understand the purpose of the study and had provided written informed consent. The exclusion criteria were 1) patients with contraindications for spinal anesthesia (refused spinal anesthesia, coagulopathy, or severe aortic stenosis) or refusal to participate in the study, 2) patients with pre-existing neurological diseases, 3) patients using immunomodulatory drugs, and 4) patients over 70 years old or with cancer. All patient medical records were reviewed.

2.4 Pain evaluation and analgesics used

We evaluated the intensity of pain before surgery. All patients were asked to evaluate their current pain intensity using the Visual Analogue Scale (VAS, ranged from 0–10, no moving or standing), based on their native language Mandarin or Taiwanese a day before the operation and biosample collection. The pain intensity was recorded by experienced nurses and was believed to be their current pain intensity as impacted by their disease. To confirm the situation of chronic pain, the duration of pain (days) was collected. Patients were asked to recall how long the pain or discomfort caused by the disease has persisted. Complete blood count, liver and renal function, chest X-ray, and electrocardiogram were also routinely checked to exclude other medical conditions. The patients typically received their surgery and CFS sampling with the next few days. According to the patients’ statements and their medical records, they were then grouped into 3 different groups: GU without obvious pain (control, VAS 0–1), OA with very mild pain (VAS ≤ 2), and OA with pain (VAS ≥ 3) based on their pain level a day before the surgery. The analgesic drugs used by the patients were also recorded based on their medical records within one month prior to sampling.

2.5 Blood and CSF collection

All patients were sampled in the operating room when they received spinal anesthesia for their scheduled surgery. All patients were collected in the daytime (8:30~16:30) and had no food and water intake for at least 8 hours. In the operating room, patients were in the lateral decubitus knee-chest position. After sterilizing and dressing the patient’s lower back, we introduced a 25G spinal needle into the spinal canal and CSF (1 ml) was drained, collected, and then replaced with 0.5% bupivacaine (2–3 ml) as the spinal anesthesia. We also collected whole blood simultaneously before surgery began via venipuncture. Whole blood (10 ml) was drawn into a test tube that contained the anticoagulant ethylenediaminetetraacetic acid. The blood and CSF samples were ice bathed and then transported to a laboratory for a biological molecule analysis. The CSF and whole blood samples were then centrifuged (3000 × g at 4°C for 10 min), and the CSF and plasma was isolated. The CSF and plasma were immediately stored at −80°C.

2.6 Measuring of plasma and CSF IL-8, TNF-α, TGF-β1 and BDNF, levels

The levels of IL-8, TNF-α, TGF-β1, and BDNF were quantified using enzyme-linked immunosorbent assays (ELISA) (HSTA00D for TNF-α, D8000C for IL-8, DB100B for TGF-β1 and DBD00 for free form BDNF, Quantikine Human Cytokine Kit; R&D Systems, Minneapolis, MN, USA) as before [19]. These kits were validated to measure human cytokines, BDNF in plasma [19] and supernatant of cell culture. In brief, the standard of molecules was established using recombinant human IL-8, TNF-α, BDNF and TGF-β1 (R&D) proteins diluted in buffer. In order to measure TGF-β1 and BDNF in the plasma, the plasma was diluted (1:10) with an assay diluent buffer before measurement. The relevant dilution factors were calculated back to the exact concentration of TGF-β1 and BDNF in the plasma. A total of 50 μl of either standard or diluted samples (plasma or CSF) was added to the plate. All laboratory assays were performed in duplicate. The ELISA plate was then incubated for 3 hours at room temperature on a rotary shaker and washed. After incubation with the conjugated buffer and substrate solution, the concentrations of the biological molecules were evaluated with an ELISA reader (SpectraMax-M2; Molecular Devices, Sunnyvale, CA, USA) and calculated by the concentration of standard.

2.7 Statistical analysis

Chi-square (χ2) tests, a one-way ANOVA with a Bonferroni post-hoc multiple comparison the demographic data. A multivariate analysis of covariance (MANCOVA) (covariates: age, gender, body mass index [BMI], hypertension, diabetes mellitus [DM] and the use of pain-relieving drugs) was used for the statistical evaluations. For the biological molecules are often non-normality, a non-parametric statistical test (Kruskal-Wallis test) was also chosen to analysis the data. To determine the correlation between the VAS pain scores and the plasma and CSF molecule, the Spearman correlations were calculated. SPSS 22.0 for Windows was used for the statistical computations. Significance was set at p < 0.05.

3. Results

We collected and analyzed 29 patients with end-stage OA (X-ray presented articular cartilage has worn away almost completely; bone on bone contact is occurring and passed another objective orthpedicst’s check) when they received spinal anesthesia for total knee or hip replacement surgery. Our data demonstrated that some patients with chronic OA do not feel obvious pain when resting (VAS score 0–2, no moving or standing). Therefore, in order to understand the cause of different expressions of pain between patients, the OA patients were then divided into groups according to their VAS score as very mild pain (VAS score is ≤ 2, n = 8) and pain (VAS score ≥ 3, n = 21) groups. Moreover, 14 patients with GU diseases without obvious pain (VAS score 0–1 at rest) were included as a control group (Table 1).

Table 1

Comparison of the general clinical parameters for the patients.

	GU patients	Chronic OA patients			p
	VAS score 0–1 (control)	VAS score 0–2 (very mild pain)	VAS score ≥ 3 (pain)
N	14	8	21
Age (years, mean ± SD)	57.3 ± 8.7	59.9 ± 6.6	62.2 ± 7.6	F = 1.661a	0.203
Female sex, n (%)	6 (42.9)	7 (87.5)	15 (71.4)	χ2 = 5.187b	0.075
Body weight (kg, mean ± SD)	66.5 ± 6.9	73.1 ± 15.9	69.9 ± 13.1	F = 0.793a	0.459
BMI (kg/m2, mean ± SD)	25.5 ± 2.8	28.6 ± 4.8	28.6 ± 4.6	F = 2.737a	0.077
Hypertension, n (%)	3 (21.4)	2 (25.0)	12 (57.1)	χ2 = 5.351b	0.069
DM, n (%)	1 (7.1)	1 (12.5)	6 (28.6)	χ2 = 2.789b	0.248
Mean VAS pain score (mean ± SD)	0.3 ± 0.6	1.5 ± 0.8**	3.4 ± 0.7***###	F = 85.907a	< 0.0001
Duration of pain (day, mean ± SD)	75.1 ± 186.6	1665.0 ± 1200.0***	1060.0 ± 921.9**	F = 10.635a	< 0.0001
Pain-relieving drugs used, n (%)	5 (35.7)	1 (12.5)	7 (31.8)	χ2 = 1.488b	0.475
Opioid, n	0	0	0
NSAID, n	2	1	6
Acetaminophen, n	3	0	1
Operation (n)	Hernia (2)	OA hip replacement (2)	OA hip replacement (3)
	Hydrocele (1)	OA knee replacement (6)	OA knee replacement (17)
	Ureterscope (11)		Knee arthroscopy (1)

GU, genitourinary examination or operation. OA, osteoarthritis. VAS, Visual Analogue Scale. BMI, body mass index. DM, diabetes mellitus. NSAIDS, nonsteroidal anti-inflammatory drugs (etoricoxib, naproxen). Patients with GU diseases without obvious pain (VAS score 0–1 at rest) were included as a comparison (control) group. Patients with OA were divided into very mild pain (VAS score 0–2) and pain (VAS score ≥ 3) groups according to their level of pain. Data were presented as means ± standard deviation (SD), n, or n (%).

aOne-way ANOVA with a Bonferroni post-hoc multiple comparison

**p < 0.001

***p < 0.0001 vs. GU control patients

###p < 0.0001 vs. OA patients with very mild pain.

bAnalysis of χ2 test.

There were no significant between-group differences in age, body weight, or body mass index (BMI) (Table 1). There were no statistical differences in the distribution of gender, hypertension, diabetes mellitus (DM), and pain-relieving drugs used for the groups (Table 1). The mean VAS pain scores for the GU control, OA with very mild pain and OA with pain groups were 0.3, 1.5 and 3.4 (p < 0.0001) respectively, which indicated the different degrees of pain intensity in these patients at rest. Moreover, to better understand the situation of chronic pain, the duration of pain (days) were recorded according to the recollection of the patients. The significant different duration of pain experience caused by the disease were found among GU control patients (mean duration of pain is 75.1 days), OA patients with very mild pain (mean duration of pain is 1,665 days), and OA patients with pain (mean duration of pain is 1,060 days) (p < 0.0001).

3.1 Plasma levels of TGF-β1 and BDNF were significantly down regulated in patients with chronic painful osteoarthritis

A multivariate analysis of covariance adjusted for the covariates (age, gender, BMI, hypertension, DM and pain-relieving drugs used) was used. And because of the non-normal distribution of these biological molecules, a non-parametric method (Kruskal- Wallis test) was used to analyze the data. OA patients with pain (VAS ≥ 3), but not those with VAS ≤ 2, had plasma TGF-β1 and BDNF levels that were significantly lower compared to the GU control patients (Table 2 and Fig 1A, p < 0.05).

Table 2

Comparison of laboratory parameters in GU and chronic osteoarthritis patients.

	GU patients	Chronic OA patients		MANCOVA		Kruskal- Wallis Test
	VAS score 0–1 (control)	VAS score 0–2 (very mild pain)	VAS score ≥ 3 (pain)
				F	p	p
N	14	8	21
Plasma parameters (pg/mL)
TGF-β1	5108.3 ± 2094.6	1579.7 ± 661.1	739.3 ± 46.5*†	4.790	0.015	0.035
BDNF	3952.0 ± 1106.5	2494.8 ± 1027.4	721.1 ± 211.6*††	4.896	0.014	0.001
TNF-α	1.87 ± 0.41	1.51 ± 0.58	1.21 ± 0.11	0.993	0.381	0.488
IL-8	1.34 ± 0.32	1.39 ± 0.57	2.26 ± 0.45	0.286	0.753	0.251
CSF parameters (pg/mL)
TGF-β1	38.41 ± 7.85	36.05 ± 10.63	12.16 ± 1.26*†#	5.351	0.010	0.008
BDNF	12.11 ± 4.94	4.18 ± 2.97	1.74 ± 0.55	3.198	0.053	0.104
TNF-α	0.91 ± 0.34	0.27 ± 0.14	0.83 ± 0.30	0.175	0.840	0.395
IL-8	74.28 ± 27.46	58.53 ± 28.39	68.05 ± 19.27	0.156	0.856	0.987

GU, genitourinary examination or operation. OA, osteoarthritis. VAS, Visual Analogue Scale. Patients with GU diseases without obvious pain (VAS score 0–1 at rest) were included as a comparison (control) group. Patients with OA were divided into very mild pain (VAS score 0–2) and pain (VAS score ≥ 3) groups according to their level of pain. Data were mean ± S.E. A multivariate analysis of covariance (MANCOVA) adjusted for covariates (age, gender, body mass index, hypertension, diabetes mellitus and pain-relieving drugs used) was used to analyze the data. Data are mean ± SE.

*p < 0.05 vs. GU control patients. And a non-parametric method (Kruskal- Wallis test) was used to analyze the data.

†p < 0.05

††p < 0.01 vs. GU control patients.

#p < 0.05 OA with very mild pain vs OA with pain.

Fig 1

The levels of cytokines and BDNF in plasma and cerebral fluid (CSF) in patients with genitourinary examination or operation (GU, control) and patients with osteoarthritis (OA).

OA patients were divided into very mild pain (Visual Analogue Scale, VAS score 0–2) and pain (VAS score ≥ 3) groups according to their level of pain. The non-parametric method (Kruskal- Wallis test) was used to analyze the data. Data are mean ± SE. *p < 0.05, ** p < 0.01 vs. GU control patients. #p < 0.05 OA with very mild pain vs OA with pain.

3.2 The levels of TGF-β1 in CSF were significantly downregulated in the patients with chronic painful osteoarthritis

In the CSF, a multivariate analysis of covariance adjusted for the covariates (age, gender, BMI, hypertension, DM and pain-relieving drugs used) was used. And a non-parametric method (Kruskal- Wallis test) was also used to analyze the data. We demonstrated that the level of TGF-β1 (Table 2 and Fig 1B, p < 0.05) was significantly down regulated in the OA patients with pain compared to the GU control group. The levels of BDNF, TNF-α or IL-8 in CSF did not change in these patients with chronic OA compared to the GU control group (Table 2 and Fig 1B).

3.3 The levels of TGF-β1 in plasma and CSF were significantly correlated with the VAS scores

The Spearman correlation with control of age, gender, body mass index and pain drug showed that in patients with chronic OA, there were no significant correlations between age, gender, BMI, pain-relieving drugs used and the VAS scores. Their CSF TGF-β1 levels were significantly negatively correlated with their VAS pain scores (Table 3 and Fig 2, p < 0.05). However, there were no significant correlations between plasma and CSF BDNF, TNF-α and IL-8 levels and the VAS scores (Table 3).

Table 3

Comparison of blood and CSF parameters with pain scores in patients with chronic osteoarthritis.

Plasma molecule	VAS score		CSF molecule	VAS score
	Spearman r	p		Spearman r	p
Plasma TGF-β1	0.038	0.855	CSFTGF-β1	–0.498	0.011
Plasma BDNF	–0.324	0.114	CSFBDNF	0.187	0.370
Plasma TNF-α	–0.006	0.975	CSFTNF-α	–0.016	0.938
Plasma IL-8	0.151	0.471	CSFIL-8	–0.110	0.602

The correlation of plasma and cerebral spinal fluid (CSF) molecules with Visual Analogue Scale (VAS) scores in patients with chronic osteoarthritis (OA, n = 29). BDNF, brain derived neurotrophic factor. The Spearman correlation analysis with control of age, gender, body mass index and pain drug were used. p < 0.05 represented a significant correlation of variables with the VAS scores.

Fig 2

Scatter plot of the Visual Analogue Scale (VAS) scores with the concentration of cerebral spinal fluid (CSF) and plasma TGF-β1 of patients with OA (n = 29).

Spearman correlation analyses with control of age, gender, body mass index and pain drug were used. p < 0.05 represented a significant correlation of variables with the VAS scores.

4. Discussion

Up to the present time, treatment of pain has been a challenge because its pathogenesis and progression are very complex. Thus, an analysis of the underlying mechanism of pain in patients with osteoarthritis will help the treatment of chronic pain during the progression of osteoarthritis. This is the first study to simultaneously investigate the roles of cytokines and neurotrophic factors in human plasma and CSF for osteoarthritis patients with chronic pain. Our data demonstrated a significant decrease in the plasma and CSF levels of TGF-β1 in patients with osteoarthritis who were experiencing pain, but this was not found in those with very mild pain. In the osteoarthritis patients, the Spearman correlation analyses further revealed that the VAS pain scores were negatively correlated with the levels of TGF-β1 in CSF. However, there were no significant correlations between the pain scores and the levels of BDNF, TNF-α, and IL-8 in either the CSF or the plasma samples. These results indicate the important role of TGF-β1 in pain suffered by patients with chronic osteoarthritis.

Several studies have reported that serum TNF-α is significantly higher in humans experiencing chronic pain, such as people with sciatica [10], fibromyalgia [20, 21], lumbar disc herniation [22], and low back pain [23]. Previous studies also have shown that the levels of inflammatory cytokines in the plasma of osteoarthritis patients [24] or synovial [25] samples are up-regulated. The downregulation of serum TNF-α has been correlated with improvements in the VAS scores of patients with osteoarthritis after treatment [5]. In this study, we did not find any significant differences in the levels of TNF-α and IL-8 in the plasma and CSF in osteoarthritis patients in the chronic pain group. However, our finding was partially supported by previous studies [21, 26–28] suggesting that the level of TNF-α in CSF do not increase in patients with complex regional pain syndrome [26] and post-traumatic osteoarthritis [28]. For patients with disc herniation and sciatica, the levels of TNF-α in their CSF and serum have also not been found to be significantly increased [27]. Our data suggest that the discrepancy in cytokines expression in different diseases may lead to variations in observations of pain. Here, it was assumed that inflammatory cytokines, which can be measured in the synovium, histological samples of osteoarthritis, blood and CSF, may differ depending on the stage of osteoarthritis. In this study, we collected patients with osteoarthritis and with chronic pain before they underwent joint replacement surgery. These patients may have suffered from pain for many years and are in the end stage of osteoarthritis. Our data represented that at the end stage of osteoarthritis, the levels of TNF-α and IL-8 in the plasma and CSF may not be the significant physiological indicators of pain.

Although the role of inflammatory cytokine expression in peripheral blood or the CNS remains controversial in terms of chronic pain, studies have suggested that intrathecal administration of anti-inflammatory cytokines can reverse chronic constriction injury-induced mechanical allodynia and thermal hyperalgesia in rats [29]. The significant decrease in the anti-inflammatory cytokine TGF-β1 in CSF and plasma found in the present study also hint at possible therapeutic targets for patients in pain. TGF-β1, a 25-kDa protein, expresses in different types of tissue and can be induced by activated glia cells [30]. It is classified as an anti-inflammatory cytokine [30]. The expression and production of TGF-β1 both in peripheral blood and brain have been found [30-33]. In blood, TGF-β1 has been found in high amounts within α granules of platelets and in platelet-rich plasma [31-33]. In peripheral tissue, platelets store large amounts of TGF-β1, which immediately after wounding facilitates the formation of a hemostatic plug [31] and accelerates wound healing in vivo [34]. In addition, platelet-rich plasma (PRP) has been intensively used for treatment of OA [35, 36], and patients with OA treated with PRP showed improvement in terms of pain level and joint function compared to a control group [35]. Furthermore, numerous studies have shown that TGF-β1 is required for the formation of articular cartilage at the early stages of joint development. In mice, postnatal TGF-β1 signaling transduction molecule deletion in chondrocytes results in OA-like pathologies [37]. Activating the TGF-β1 signaling maintain the articular cartilage homeostasis and against the progression of OA [37]. Thus, the downregulation of plasma TGF-β1 in our patients with OA may decrease protection to the joint and accelerate disease progression. However, other investigations have indicated that TGF-β1 may, in fact, be a factor in joint destruction [38]. In an OA mouse model, high concentrations of active TGF–β1 in the subchondral bone initiated pathological changes in OA [38]. The conflicting roles of TGF-β1 signaling in joints in the development of OA may be due to the different stages of bone development and the concentration of TGF-β1 in local tissue.

In the central nervous system, astrocytes express and produce TGF-β1 [39]. TGF-β1 has dominant neuronal protective effects. It can promote the survival of dopaminergic neurons [40] and protect hippocampal neurons against transient global ischemia-induced damage [41]. In addition, TGF-β1 has been demonstrated to inhibit the intensity of pain in chronic constriction injury models [15, 42] and neuropathic pain [15] in rodents. Intrathecal injections of bone marrow stromal cells may increase TGF-β1 expression and inhibit neuropathic pain in mice [15]. The distinctive positive effects of TGF-β1 in terms of anti-inflammation, neuronal protection and the spinal sensitized pain clarifies its importance in chronic pain in central nervous system.

This is the first study to compare the levels of TGF-β1 in CSF and plasma in human subjects with chronic OA and with pain. Our data reveals that patients with end-stage OA experience long duration of pain (> 1000 days). In addition, patients with end-stage OA who experience very mild pain (VAS ≤ 2) at rest had a higher CSF TGF-β1 than those patients who experience pain at rest. Thus, we hypothesize that the levels of TGF-β1 in CSF in humans may modulate pain in OA subjects. Although the mechanism of TGF-β1 down-regulation in OA patients is still unclear, the findings of our study indicate that in patients with end stage of osteoarthritis chronic pain, the downregulation of TGF-β1 in CSF may be a significant indicator of the pain intensity associated with osteoarthritis.

With the exception of the downregulation of plasma and CSF TGF-β1, we also found a significant decrease in the level of plasma BDNF in patients with chronic pain. However, we did not find a significant downregulation of CSF BDNF in those experiencing chronic pain. In the CNS, BDNF is produced from neurons and the glia cells [43, 44] and also serves as a protective neuronal factor. BDNF has quite different effects from those of TGF-β1 in pain studies [17, 18, 45, 46]. In rodent inflammatory or neuropathic pain models, BDNF was found to be potentiated and involved in the generation of pain [17, 18]. Serum levels of BDNF were found to be increased in older women with low back pain [45], and plasma levels of BDNF were positively correlated with the severity of pain in women with pelvic pain [46]. However, for patients with post-herpetic neuralgia, lower levels of BDNF in CSF have been reported [47]. Thus, the roles of BDNF in pain still require further study.

There are some limitations in this study. First of all, there was no CSF from the healthy subjects. We collected it from a disease reference group without obvious pain (VAS score 0–1). We did not exclude the molecular differences in circulation or CSF between unhealthy and healthy control subjects. Second, the use of analgesics (acetaminophen and NSAIDs) may impact the expression of cytokines. Although we attempted to exclude their impact through the use of a partial correlation with adjusted the used of pain-relieving drugs, further studies still need to be conducted to confirm the effects of analgesics on cytokines in CSF and blood in patients with pain. Third, although the ELISA kits were validated to measure these molecules in human plasma and supernatants of cell culture. And the recombined human IL-8, TNF-α, BDNF and TGF-β1 diluted in buffer were used as a calibration. However, this is the first time ELISA kits is used to detect these molecules in CSF, and the interpretation of CSF data should be more cautious.

5. Conclusions

Our findings revealed that the levels of TGF-β1 in plasma and CSF may be biological indicators in patients with end stage osteoarthritis experiencing chronic pain. Therapeutics that promote TGF-β1 expression may serve as a therapy for pain control in patients with osteoarthritis. The levels of BDNF, TNF-α, or IL8, were not shown to be suitable biological indicators of chronic pain.

22 Jun 2021

PONE-D-20-38558

TGF-β1 in plasma and cerebrospinal fluid can be used as a biological indicator of chronic pain sensation in patients with osteoarthritis

PLOS ONE

Dear Dr. Chen,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 06 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Firas H Kobeissy, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. Thank you for stating the following financial disclosure:

No.

At this time, please address the following queries:

a) Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

b) State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

c) If any authors received a salary from any of your funders, please state which authors and which funders.

d) If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

5. Thank you for stating the following in the Acknowledgments Section of your manuscript:

This study was supported in part by grant NCKUH-10509003 from National Cheng Kung University Hospital, Taiwan and 105-2314-B-006-007-, 106-2314-B-006-026-MY2 (to YCL), 105-2628-B-037 -003 -MY3 (to SLC), from the Ministry of Science and Technology, Taiwan.

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

This study was supported in part by grant NCKUH-10509003 from National Cheng Kung University Hospital, Taiwan and 105-2314-B-006-007-, 106-2314-B-006-026-MY2 (to YCL), 105-2628-B-037 -003 -MY3 (to SLC), from the Ministry of Science and Technology, Taiwan.

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

6. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

7. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This interesting manuscript presents data on cytokine-immunoreactivity in paired plasma and CSF from people with OA and urological controls. The findings are of interest, not least because they are not what one might expect. As such, the manuscript raises more questions than it answers, although much of this might be addressed by more detailed presentation of methods and data.

Specific points:

What was the lead question and recall timeframe for the pain VAS?

Please refer to molecule detection by ELISA to indicate that what was measured was immunoreactivity (e.g. BDNF-LI), or otherwise provide evidence of assay validation with this sample type that confirms molecular identity with e.g. BDNF.

Please specify diagnoses in control group in more detail, as, in the absence of a healthy matched control group, it might be that controls are abnormal, rather than the patients with OA. How do the authors explain that 5 in the control group were using pain-relieving medications if they did not have pain?

The mean VAS for those with OA pain (3.4) is very low (Table 1) considering that all these people had at least VAS=3. What was the range/distribution of VAS scores in the groups?

Were cytokine levels and VAS scores normally distributed? Cytokine levels will usually be positively skewed, requiring logarithmic transformation prior to parametric analysis. It seems that the range of VAS scores is very low (mild pain). Descriptives (table 1) should normally be presented with SD or interquartile range rather than SEM, to enable the reader to gain an impression of the distribution of data, rather than just the precision of the mean estimate. Additionally, the regression analyses might be clearer if presented through scatterplots, to demonstrate that the significant Pearson correlation coefficients are not driven by a small number of extreme values.

Please include diagnostic group as a covariate (ie control or OA) in regression analyses in case observed differences are attributable to other aspects of diagnosis rather than pain.

The origin of TGF beta in the current population is unclear, however, differences in plasma levels as well as CSF levels might well point to a peripheral source. TGF beta has been associated with OA pathology and it would be important to relate the current findings to previous data on circulating TGF beta in OA. Furthermore, please describe the `stage’ of OA in more detail (mentioned in discussion as end stage). Given that knee structural changes in OA are also associated with pain, might TGF beta be a marker for structural severity rather than pain severity?

The authors should be commended for not emphasising conclusions from their negative findings, which might have resulted from power limitations.

Pain is, by definition, a sensation. Please avoid the term `pain sensation’ which is easily misread for `pain sensitisation’ which is not studied here.

Reviewer #2: PLos ONE

TGF-β1 in plasma and cerebrospinal fluid can be used as a biological indicator of chronic pain sensation in patients with osteoarthritis.

Yen-Chin Liu, Hung-Tsung Hsiao, Jeffrey Chi-Fei Wang, Tzu-Cheng Wen, Shiou-Lan Chen.

Cytokines can affect pain in osteoarthritis (OA), as well as in many other conditions, in various ways e.g. by enhancing or depressing inflammatory, and various other tissue changes, but also by direct effects on the nervous system. The authors present their data on the plasma and CSF levels of 4 cytokines in patients with OA of the hip, and/or knee. The cytokines they studied were TGF-β1, BDNF, TNF-α, and IL-8. Plasma and CSF samples were obtained from patients scheduled to have spinal anesthesia for surgery. Plasma and CSF samples were also obtained from patients with genitourinary system disease (GU), scheduled for surgery under spinal anesthesia. The samples were assessed for levels of the above 4 cytokines. The authors compared the levels of each of the 4 cytokines between the group of OA patients with little or no pain, and OA patients with more severe pain. The authors also compared cytokine levels between OA patients and GU patients. Their findings showed a significant difference in TGF-β1 levels in OA patients with severe pain and OA patients with mild or no pain; CSF and plasma TGF-β1 levels were significantly lower in the OA patients with severe pain compared to those levels in OA patients with no or mild pain. Similarly, the OA patients with severe pain also had significantly lower TGF-β1 levels (CSF and plasma) than the GU patients. BDNF plasma levels were lower in patients with OA associated with severe pain, but there was no significant lowering of BDNF levels in CSF. The lower plasma BDNF levels represented an unexpected finding for which there does not appear to be a good explanation. There were no significant differences between the patients with severe OA, mild OA, and GU with respect to IL-8 and TNF-α levels in either plasma or CSF. In OA patients with severe pain there was a significant negative correlation between pain severity (assessed by VAS) and TGF-β1 levels (both plasma and CSF).

Assessment.

Major Points

1.These results are interesting, but it is uncertain whether they are generalizable because of some questions with respect to choices in the patient populations and their clinical assessments.

*The authors included 36 patients with OA; they excluded 5 for “a chronic pain history with acute pain surgery”. What does this mean?

*Two more were excluded because they had neural disease “when they received spinal anesthesia for the osteoarthritis of knee or hip surgery” What does this mean?

*It is not clear what surgery the other 29 OA patients had. I assume that it would have been either knee or hip arthroplasties but this has to be specified.

*If the above is correct, it appears that such OA patients differ from the usual population selected for arthroplasties in Europe or North America where such operations are performed in a somewhat older population with a male predominance.

*Eight out of the 26 patients with OA had a VAS of 0-2. Patients with OA of hips or knees proceeding

to an arthroplasty have severe pain. A VAS of 0-2 would be unusual in such patients.

2. The VAS was assessed at the time the patients entered the hospital. I am concerned about the validity of such an assessment. An average VAS for the preceding week would be more informative.

3. The GU comparison group was selected as “having no pain”. Yet in Table 1 it appears that 5/14 of these patients were taking “pain relief drugs”. Somewhat surprisingly, only 7/21 OA patients with VAS scores of > were taking such medications.

4. The low levels of TGF-β1 in patients with more severe pain are an interesting finding. The authors should discuss the possible reasons for this. Could this be due to inadequate production or excessive consumption? What do animal experiments suggest?

Minor Points

1. The paper requires extensive corrections with respect to the English language. Examples : Page 2 (Introduction) line 2 : “The persisted chronic pain has been considered”; Page 11, lines 240-241 ( first and second from the top) in the “Discussion” part : “Our data suggest that then discrepancy of cytokines expression in different disease may provide the multiple aspect of observation in pain sensation”.

2. Each table should be presented on a separate page.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

5 Aug 2021

Reviewers' comments:

Reviewer #1: This interesting manuscript presents data on cytokine-immunoreactivity in paired plasma and CSF from people with OA and urological controls. The findings are of interest, not least because they are not what one might expect. As such, the manuscript raises more questions than it answers, although much of this might be addressed by more detailed presentation of methods and data.

Response: We thank the reviewer for the positive words. As recommended by the reviewer, we have added the revision in the manuscript.

Specific points:

What was the lead question and recall timeframe for the pain VAS?

Response: This is a cross-sectional study of CSF and plasma molecules in human subjects with chronic pain. We evaluated the intensity of pain before surgery. So even the GU patients were evaluated. About the OA patients, all patients fit the OA X ray criteria (including joint space loss and osteophyte formation) and passed another objective orthpedicst's check, their chronic OA diagnosis was confirmed. About patients' VAS recall timeframe, we asked patient recall their VAS when admission. We also added the duration of pain to confirm their chronic pain situation. The related descriptions were revised in method (Page 6, last paragraph), Table 1 and results (Page 8, last paragraph).

Please refer to molecule detection by ELISA to indicate that what was measured was immunoreactivity (e.g. BDNF-LI), or otherwise provide evidence of assay validation with this sample type that confirms molecular identity with e.g. BDNF.

Response: As we mentioned in method, the commercial ELISA kit (Quantikine Human Cytokine Kit; R&D Systems, Minneapolis, MN, USA) were used for the measurement of protein levels of TNF-α, IL-8, TGF-�  1 and mature form of BDNF (Page 7, last paragraph).

Please specify diagnoses in control group in more detail, as, in the absence of a healthy matched control group, it might be that controls are abnormal, rather than the patients with OA. How do the authors explain that 5 in the control group were using pain-relieving medications if they did not have pain?

Response: We agree with the comment, thus this point was described in the limitation of this study (Page 21, lines 326-328). We also revised our Table 1 to present more details of each groups. We used the GU patients with no pain as a control group and presented no significant differences of age, gender, body weight, or body mass index (BMI) between-groups. However, the significant difference of VAS scores and pain duration were found between groups (Table 1).

About the 5 GU patients that using pain-relieving medications that we founded from their outpatient clinic record, we think it is the ethic issue that doctors prescribed the pain -relieving medication because of the complain of pain of patients in the outpatient clinic. Thus, only weak pain killers were used in this group. And as we revised in Table 1 that 3 of those GU patients used acetaminophen and 2 of those GU patients used NSAIDs (etoricoxib, naproxen) which were adjusted as covariates in Table 2 and figure 1. We also described this point in the limitation of this study (Page 21, lines 328-332).

The mean VAS for those with OA pain (3.4) is very low (Table 1) considering that all these people had at least VAS=3. What was the range/distribution of VAS scores in the groups?

Response: In this study, we found that OA patients with very mild pain (n=8), their VAS score are 0 to 2. And in OA subjects with chronic pain (n=21), the range of VAS scores are 3 to 6.

Were cytokine levels and VAS scores normally distributed? Cytokine levels will usually be positively skewed, requiring logarithmic transformation prior to parametric analysis. It seems that the range of VAS scores is very low (mild pain). Descriptives (table 1) should normally be presented with SD or interquartile range rather than SEM, to enable the reader to gain an impression of the distribution of data, rather than just the precision of the mean estimate. Additionally, the regression analyses might be clearer if presented through scatterplots, to demonstrate that the significant Pearson correlation coefficients are not driven by a small number of extreme values.

Response: We had revised the Table 1 as the suggestion. And we thanks the reviewer's comments that the distribution of cytokine levels are indeed positively skewed. Thus, we also transformed the levels of cytokines logarithmically prior to the analysis of Pearson correlation in Table 3 (Page 16). The scatterplots of VAS with log CSF TGF-β1 and log plasma TGF-β1 are also shown in Fig 2 (Page 17).

Please include diagnostic group as a covariate (ie control or OA) in regression analyses in case observed differences are attributable to other aspects of diagnosis rather than pain.

Response: In Table 3 (Page 16), we only included the patients with same diagnosis (OA patients, n=29). And we transformed the levels of cytokines logarithmically prior to the analysis of Pearson correlation. We also added the scatterplots of VAS with log CSF TGF-β1 and log plasma TGF-β1 are also shown in Fig 2 (Page 17).

The origin of TGF beta in the current population is unclear, however, differences in plasma levels as well as CSF levels might well point to a peripheral source. TGF beta has been associated with OA pathology and it would be important to relate the current findings to previous data on circulating TGF beta in OA.

Response: We thank the reviewer for the excellent comments. We had added the description of origin of TGF beta in blood and brain. And the possible relationship of TGF-β1 in OA (Page 19, last paragraph and Page 20, lines 299-315)

Furthermore, please describe the `stage’ of OA in more detail (mentioned in discussion as end stage). Given that knee structural changes in OA are also associated with pain, might TGF beta be a marker for structural severity rather than pain severity?

Response: Thanks the comments. We added the describe in result (Page 8, lines 172-174) that the definition of end-stage OA was that X-ray presented articular cartilage wears away almost completely and bone on bone contact occurs and passed another objective orthpedicst’s check. We also added the duration of pain of each groups in Table 1.

In addition, we agree with the comments of reviewer, that the levels of plasma TGF-β1 may as a marker for structural severity of OA. However, in this study we collect the CSF TGF-β1 that may explain the intensity of central pain in OA subjects. However, the expression of TGF-β1 in different stage of OA need more study in the future.

The authors should be commended for not emphasising conclusions from their negative findings, which might have resulted from power limitations.

Response: Thank you for this comment.

Pain is, by definition, a sensation. Please avoid the term `pain sensation’ which is easily misread for `pain sensitisation’ which is not studied here.

Response: Thank you for this comment. In this revision, we revised VAS as the measurement of pain intensity. We revised most of the “pain sensation” description as “pain intensity” or “the intensity of pain”.

Reviewer #2: PLos ONE

Cytokines can affect pain in osteoarthritis (OA), as well as in many other conditions, in various ways e.g. by enhancing or depressing inflammatory, and various other tissue changes, but also by direct effects on the nervous system. The authors present their data on the plasma and CSF levels of 4 cytokines in patients with OA of the hip, and/or knee. The cytokines they studied were TGF-β1, BDNF, TNF-α, and IL-8. Plasma and CSF samples were obtained from patients scheduled to have spinal anesthesia for surgery. Plasma and CSF samples were also obtained from patients with genitourinary system disease (GU), scheduled for surgery under spinal anesthesia. The samples were assessed for levels of the above 4 cytokines. The authors compared the levels of each of the 4 cytokines between the group of OA patients with little or no pain, and OA patients with more severe pain. The authors also compared cytokine levels between OA patients and GU patients. Their findings showed a significant difference in TGF-β1 levels in OA patients with severe pain and OA patients with mild or no pain; CSF and plasma TGF-β1 levels were significantly lower in the OA patients with severe pain compared to those levels in OA patients with no or mild pain. Similarly, the OA patients with severe pain also had significantly lower TGF-β1 levels (CSF and plasma) than the GU patients. BDNF plasma levels were lower in patients with OA associated with severe pain, but there was no significant lowering of BDNF levels in CSF. The lower plasma BDNF levels represented an unexpected finding for which there does not appear to be a good explanation. There were no significant differences between the patients with severe OA, mild OA, and GU with respect to IL-8 and TNF-α levels in either plasma or CSF. In OA patients with severe pain there was a significant negative correlation between pain severity (assessed by VAS) and TGF-β1 levels (both plasma and CSF).

Assessment.

Major Points

1.These results are interesting, but it is uncertain whether they are generalizable because of some questions with respect to choices in the patient populations and their clinical assessments.

*The authors included 36 patients with OA; they excluded 5 for “a chronic pain history with acute pain surgery”. What does this mean?

Response: Sorry for the misunderstanding. Actually, we collected 5 OA patients with low-leg trauma and with acute pain. Thus, these patients with chronic and acute pain were excluded. We deleted this sentence (Page 8) for possible misunderstanding.

*Two more were excluded because they had neural disease “when they received spinal anesthesia for the osteoarthritis of knee or hip surgery” What does this mean?

Response: We deleted this sentence (Page 8). Actually, we had mentioned in the exclusion criteria 2) we excluded those patients having pre-existing neurological diseases. The pre-existing neurological disease may have influence on CSF data.

*It is not clear what surgery the other 29 OA patients had. I assume that it would have been either knee or hip arthroplasties but this has to be specified.

Response: Thank you for your comment. We revised the Table 1 and presented the surgery type which was OA hip (total hip replacement), OA knee (total knee replacement) and knee arthroscopy.

*If the above is correct, it appears that such OA patients differ from the usual population selected for arthroplasties in Europe or North America where such operations are performed in a somewhat older population with a male predominance.

Response: Thank you for your comment. We agree that generally, male may receive more OA surgery than female. However, some data still claim that OA is more common in women than in men. Although we have the limitation in the selection of case, however, the comparison of VAS scores with CSF and plasma molecules may provide a reference in the observation of OA with chronic pain.

*Eight out of the 26 patients with OA had a VAS of 0-2. Patients with OA of hips or knees proceeding to an arthroplasty have severe pain. A VAS of 0-2 would be unusual in such patients.

Response: Thank you for your comment. We agree that OA surgery caused a lot of pain. However, we had described clearly that patients’ VAS was recorded when they administered to the hospital before surgery, not after surgery. And patients were asked to present their VAS before operation. It is also reasonable that OA patients have low VAS once they reduced activity or taking drugs. We thought their low VAS may present some clinical meanings and separated as another groups

2. The VAS was assessed at the time the patients entered the hospital. I am concerned about the validity of such an assessment. An average VAS for the preceding week would be more informative.

Response: Thank you for your comment. We agreed that the VAS validity is an important issue. However, we also concerned the recall validity if we asked patient recall the preceding week. Furthermore, there is no data showed that recalling preceding week VAS is a better presentation than VAS at administration. So we used the VAS at administration as grouping category and we also noticed that those with low VAS at administration also used less pain drug than those with high VAS at administration which also proved that our grouping using VAS at administration is reasonable.

3. The GU comparison group was selected as “having no pain”. Yet in Table 1 it appears that 5/14 of these patients were taking “pain relief drugs”. Somewhat surprisingly, only 7/21 OA patients with VAS scores of > were taking such medications.

Response: Thank you for your comment. About the 5 GU patients that using pain killers, that we founded from their outpatient clinic record, we think it is the ethic issue that doctors prescribed the pain killer because of the complain of pain of patients in the outpatient clinic. Thus, only weak pain killers were used in this group. And as we presented in the revised Table 1 that 3 of those GU patients used acetaminophen and 2 of those GU patients used NSAIDs (etoricoxib, naproxen) which were adjusted as covariates in Table 2 and figure 1. We also described this point in the limitation of this study (Page 21, lines 328-332)

For OA patients, we still find that less pain group patient used less pain drugs (1/8 vs.7/21) We also think that, in our hospital, many patients receive surgery because they don’t want taking pain drugs because of many reasons (afraid GI/renal side effect, already taken many drug, still pain even taking drug….). So even the surgeon prescribed less drugs if patients claimed they don’t want take medicine. The surgeon just arrange operation as soon as possible. So there are less ratio for patients receiving drugs.

4. The low levels of TGF-β1 in patients with more severe pain are an interesting finding. The authors should discuss the possible reasons for this. Could this be due to inadequate production or excessive consumption? What do animal experiments suggest?

Response: This is a first study that compared the levels of TGF-β1, BDNF and cytokines in CSF and plasma of human subjects with no pain, very mild pain and chronic pain condition. As we discussed in page 19-20 that TGF-β1 may involve in the articular cartilage homeostasis and inhibit pain sensation in rodents. Thus we hypothesize that the levels of TGF-β1 in CSF/plasma in humans may associated with the progression of disease and modulate pain. Although the mechanism of TGF-β1 down-regulation in OA patients is still unclear. However, our study represented that patients with end stage of osteoarthritis and with chronic pain, the downregulation of TGF-β1 in CSF may be a significant indicator of the pain intensity associated with osteoarthritis. The related description of TGF-β1 with OA progression and chronic pain were added in Page 19 last paragraph and Page 20 lines 299-315.

Minor Points

1. The paper requires extensive corrections with respect to the English language. Examples : Page 2 (Introduction) line 2 : “The persisted chronic pain has been considered”; Page 11, lines 240-241 ( first and second from the top) in the “Discussion” part : “Our data suggest that then discrepancy of cytokines expression in different disease may provide the multiple aspect of observation in pain sensation”.

Response: Thank you for your comment. We revised this newest version again and make it easy to read. We also corrected the grammar errors with the native speaker's help.

2. Each table should be presented on a separate page.

Response: We had presented each table on a separate page.

Submitted filename: 2021-0805-PONE-reply to the reviewers comments.docx

Click here for additional data file.

13 Sep 2021

26 Oct 2021

Reviewers' comments:

Response: We thank the reviewers for the positive words. As recommended by the reviewers, we have added the revision in the manuscript.

Reviewer #1: I appreciate the authors’ efforts in revising their manuscript, which goes some way to address my original concerns. I suspect that my original comments (and those of the second reviewer where similar) might have been incompletely understood by the authors and I attempt to explain my residual concerns more clearly below.

What was the lead question and recall timeframe for the pain VAS? (?p6) I note also that this point was raised by the second reviewer, who made additional pertinent points. Having read the authors’ response, I think that I might understand the source of apparent confusion and discrepancies, but this is not yet clear to the general reader of the manuscript. Am I correct to think that participants were recruited on the basis of clinical evidence of chronic pain (undergoing arthroplasty), but that the pain VAS used in analyses asked about `recent pain’. As in my original comments, please give the exact text and anchors for the VAS used. I expect this was not in English language, and so the exact text in the original language, plus a verbatim translation might suffice. It makes a very big difference what time-frame defined `recent’ if the original text used the word `recent’. Does `recent’ mean since admission, over the past few minutes while you have not been standing, or over the past few months? I suspect from the results that participants rated their `current’ pain whilst not mobilising or standing. Our experience also is that patients listed for arthroplasty might not have severe pain at rest, but as soon as they try to do something, then their pain becomes severe. Chronic pain (e.g. as measured by WOMAC or clinically used VAS scores) typically refers to pain over the past week; either average pain, or worst pain, thereby reflecting the intermittent nature often of OA pain. On reflection, I can understand that `current pain’ might be the most appropriate index here, because molecules in CSF might change over short time periods, and the authors might wish to measure the pain experienced at a time point closest to biosample collection.

Response: Thank you for your comments. Correctly, all patients were asked to evaluate their current pain intensity using the Visual Analogue Scale (VAS, ranged from 0-10, no moving or standing), based on their native language Mandarin or Taiwanese a day before the operation and biosample collection. All of these were revised in method (Page 6, last paragraph) and results (Page 8, last paragraph).

However, if this is the case, they might wish to speculate as to how relevant this might be to clinical pain. At least, they should explain the apparent inconsistency commented on by both reviewers that people with very low pain scores were undergoing joint surgery.

Response: Thank you for your comments. And, as you recognized, our data support that patients with end-stage OA experienced with chronic pain (> 1000 days, Table 1). In addition, our data also revealed that some patients with chronic OA do not feel obvious pain when resting (VAS ≤ 2, in OA with very mild pain group). And this patients (OA with very mild pain) at rest had higher CSF TGF-β1 than those OA patients with pain. Thus, we hypothesize that the levels of TGF-β1 in CSF in humans may modulate pain in OA subjects. All of these were revised in result (Page 8, last paragraph; Page 9, last paragraph) and discussion (Page 20, line 328-330).

Also I note that reviewer 2, like myself, commented on the GU comparison group was selected as “having no pain”; so am I correct to understand that they had no `current pain on admission to hospital’ but that they might have had chronic pain prior to admission? This again should be clarified, and might be part of the same explanation as the clarification of what the VAS score represents.

Response: Thank you for your comments. We apology that we did not described the definition and condition of the GU-control group clearly. In this study, patients with GU diseases without obvious current pain at rest (VAS score 0-1, mean VAS score is 0.3 ± 0.6 in GU control patients, Table 1) before surgery were included as a comparison (control) group. Furthermore, to access that they might have had chronic pain, we also checked the experience of pain of GU patients according to their recollection (GU control, mean duration of pain, 75.1 days in Table 1). All of these were revised in method (Page 6, lines 116-117) and result (Page 9).

The authors have chosen to not `indicate that what was measured was immunoreactivity (e.g. BDNF-LI)’. I do not feels strongly about this, but the fact that ELISA measurements, especially competitive ELISAs, might measure factors other than the specific molecular species targeted, particularly when they are applied to biofluids in which the manufacturer and investigator have not validated them. I would still suggest that the authors provide data on validation in plasma and CSF, rathe than assume that these kits do what the manufacturers say they do when they sell them for profit.

Response: We agree that although the ELISA kit were validated to measure these molecules in human plasma and supernatants of cell culture. And the recombined human IL-8, TNF-α, BDNF and TGF-β1 diluted in buffer were used as a calibration. However, this is the first time that using of ELISA kits to detect these molecules in CSF, and the interpretation of CSF data should be more cautious. We have added and revised such description in method (Page 7, last paragraph; Page 8, 1st paragraph) and limitations (Page 21).

Please provide statistical evidence that log transformed data are normally distributed. If, as stated in the authors’ response, untransformed data were significantly positively skewed, then please use log transformed (or otherwise normalised) data for all parametric analyses, including between group comparisons. If log transformation converts non-normal data to normal data, then please only present analyses on log transformed data (e.g. in Table 2). In other words, only present the statistically appropriate analyses.

Response: Thanks for the suggestion of statistics. Because of the log transformed data are still not normally distributed. For the biological molecules are often non-normality, thus, a non-parametric statistical test (Kruskal-Wallis test) was added to analysis the original data (Table 2). To determine the correlation between the VAS pain scores and the plasma and CSF molecule, the Spearman correlations with control of age, gender, body mass index and pain drug were performed (Table 3). All of these were added in method (Page 8, 2nd paragraph), Table 2, Table 3 and results (Page 12, Page 15).

I’m afraid that I don’t understand the data in the statement; ` The mean duration of pain was 75.1, 1,665, and 1,060 days with no pain, OA patients with very mild pain, and OA patients experiencing pain according to the recall of the patients, respectively (p < 0.0001).’ Are the authors really stating that participants were asked how long they had had pain to the nearest 0.1 days??? Given my above comments about variability of pain in OA, this seems meaningless if I am interpreting correctly. What is the method for this?

Response: We apology that we did not clearly described the definition of this data. In this study, to confirm the situation of chronic pain, the duration of pain (days) was collected. Patients were asked to recall how long the pain or discomfort caused by the disease has persisted. All of these were added and revised in Table 1, method (Page 7, 1st paragraph) and in result (Page 9, last paragraph).

Reviewer #2: My concerns have been answered. The authors' findings are interesting although their significance is uncertain in view of conflicting evidence in thre literature on the roles of TGFbeta1 and BDNF.

Response: Thank you for you insightful feedback and comments.

Submitted filename: 2021-1026-Reply to the Reviewers comment.docx

Click here for additional data file.

19 Dec 2021

TGF-β1 in plasma and cerebrospinal fluid can be used as a biological indicator of chronic pain in patients with osteoarthritis

PONE-D-20-38558R2

Dear Dr. Chen,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Firas H Kobeissy, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors have addressed my concerns. The patient population of osteoarthritis patients is somewhat atypical, but I am willing to accept that. The results with respect to the low levels of TGF-beta are very interesting.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: ManfredHarth

22 Dec 2021

PONE-D-20-38558R2

TGF-β1 in plasma and cerebrospinal fluid can be used as a biological indicator of chronic pain in patients with osteoarthritis

Dear Dr. Chen:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Firas H Kobeissy

Academic Editor

PLOS ONE