Diacerein Versus Chondroitin Sulfate in Treatment of Adult Kashin-Beck Disease: An Intervention Trial in Heilongjiang Province, China

Xianhao Wu Luyang Cheng Second A liated Hospital of Harbin Medical University Jinjing Jia Jiaxing University Shili Song Tongkun Shi Jiaxing University Fenghua Wang Jiaxing University Xin Zhang Chinese Center for Disease Control and Prevention, Harbin Medical University, Education Bureau of Heilongjiang Province & Ministry of Health (23618504) Zhe Jiao Chinese Center for Disease Control and Prevention, Harbin Medical University, Education Bureau of Heilongjiang Province & Ministry of Health (23618504) Liyan Sun (  sunliyanlily@zjxu.edu.cn ) Jiaxing University


Abstract
Background: Kashin-Beck disease (KBD) is a special type of osteoarthritis and has disabled and stunted the growth of hundreds of people in China. It also can affect patients' ability to work and live. So, how to conduct effective treatment for adult KBD patients has become a major public health problem in current KBD endemic areas. In this trial, therapeutic effects of diacerein and chondroitin sulfate on adult KBD was to evaluate and compare so that we can screen out more suitable drug.
Methods: 308 KBD patients were divided into two groups and received chondroitin sulphate (Group A) and diacerein (Group B) for 24 weeks, respectively.
Data were collected at 0 week (baseline), 12 weeks (primary end point) and 24 weeks (secondary end point) to calculate the proportion of patients with effective therapeutic effect and overall improvement rate (primary e cacy parameters), WOMAC pain and stiffness scores (secondary e cacy parameters).
Blood sample was collected to measure liver and renal function indexes. All indexes and parameters were analysed with SPSS software and intent-to-treat (ITT) analysis was applied. Results: Two primary e cacy parameters in group B at primary end point were signi cantly higher than those in group A (P=0.021, P=0.007), but no statistical differences were seen in these two primary e cacy parameters between two groups at secondary end point or between primary and secondary end points in same group (all P>0.05). In both groups, with the prolongation of treatment time, WOMAC pain and stiffness scores decreased signi cantly (all P<0.001), but no signi cant differences were seen between primary and secondary end points (all P>0.05). In addition, in both groups, the occurrences of total adverse events were relatively low and no side effect on liver function was seen. Diacerein also had no side effect on renal function.
Conclusion: For treatment of adult KBD, both diacerein and chondroitin sulfate were effective, and diacerein might work stronger than chondroitin sulfate.
Taking into account both e cacy and safety, the optimal intervention time of diacerein was 12 weeks.
Trial registration: The trial was registered complementally on 31/10/2020, and the registration number in the Chinese Clinical Trial Registry is ChiCTR2000039600 (http://www.chictr.org.cn).

Background
Kashin-Beck disease (KBD) is an endemic, deformed and special type of osteoarthritis (OA) involving with growth plate and articular cartilage, and it mainly occurs in children aged 5-15 years old [1]. The main clinical manifestations of KBD include joint pain, enlarged and deformed joints and limb muscle atrophy. Severe cases present as short stature, even dwar sm [2]. KBD spread across 13 provinces from southwest to northeast China, including Tibet, Qinghai, Sichuan, Shaanxi, Heilongjiang and so on, and has crippled and stunted the growth of hundreds of people in China [3].
KBD can be prevented but treated di cultly. So, Chinese government has launched a massive effort and series of etiotropic measures to prevent and control KBD [4], and recent monitoring data showed that except in Tibet, no new cases occurred in other 12 KBD endemic provinces and autonomous regions [3]. In 1960s and 1970s, the incidence of KBD was about 60%-80%, and a large number of adult KBD patients aged from 40 to 70 years old were left behind from that period. KBD is a teratogenic and disabling disease, and can affect patients' ability to work and live. So, how to conduct effective treatment for adult KBD patients has become one of major public health problems in current KBD endemic areas.
Refer to OA, currently available treatments for KBD include nonspeci c interventions, pharmacologic management of symptoms and surgical treatment [4].
Among these treatment methods, oral drug therapy is widely used and two main categories of drugs were usually chosen: disease-modifying OA drugs (DMOADs) and non-steroidal anti-in ammatory drugs (NSAIDs). NSAIDs are heavily used to alleviate KBD patients' pain because these agents are inexpensive and easy to get, although they are well known to accompany by a high risk of adverse events, such as serious gastro-intestinal and vascular injuries [5]. Unlike NSAIDs, the DMOADs (such as chondroitin sulfate) have been advocated as safe and effective choice for treatment of KBD [6]. .
There is strong evidence that interleukin-1 (IL-1) plays an important role in OA cartilage degradation [9]. As an IL-1 inhibitor, diacerein can lower the activity and reduce the content of IL-1, achieved certain treatment effects in OA. Many studies have shown that diacerein could relieve pain and slow down the progression of knee OA in terms of joint structure changes and symptom [10,11]. However, as far as we know, no study has reported the therapeutic effect of diacerein on adult KBD.
Based on current treatment status of KBD detailed above, a 24-week study of continuous medicine intervention was conducted. The primary aim of this trial was to verify and evaluate therapeutic effect of diacerein on KBD. At the same time, the therapeutic effect of diacerein and chondroitin sulfate on adult KBD patients was compared.

Study design
This study was conducted as a prospective and contrastive intervention trial with a follow-up period of 24 weeks. According to the e cacy, dose and safety of diacerein in the treatment of hand and knee OA reported in literatures [12][13][14], although patients were followed up 24 weeks, 12 weeks was the primary end point and 24 weeks was the secondary end point. Questionnaire and physical examinations were conducted once every 12 weeks. The primary e cacy parameters were assessed at primary and secondary end points, and the secondary e cacy parameters were assessed at baseline, 12weeks and 24 weeks.
Subjects in this trial were grouped according to their own health situation, medication history and contraindications of drug use. No placebo was included because the constraint of Administration Village Committee. Subjects were assigned to receive chondroitin sulfate or diacerein for 24 weeks.

The selections of test sites
Heilongjiang Province is a historically severe KBD endemic area and located in northwestern China. In accordance with principles of matching natural and social factors, such as geologic environments, economic situations and educational standards, towns were included if they met the following eligibility criteria: (1) the monitoring data of KBD in recent 30 years were complete; (2) historically severe KBD endemic areas; (3) the X-ray detective rate of KBD in children was < 3% in recent 3 years; (4) the number of long-term residents above 45 years old were more than 50. Ultimately, ve towns were chosen as our test sites, they were Fanrong Town, Fulu Town, Longanqiao Town, Shaowen Town of Fuyu County and Lianghe Town of Shangzhi City.

Patients
A simple questionnaire was performed and data on demographic variables (age and sex) for all residents who were living in above-mentioned towns were gained rstly. Next, radiologic and clinical examinations of both hands and/or feet were performed for villagers aged 45-70 years in order to screen adult KBD patients. Adult KBD patients were diagnosed and those in degree I and II were selected in according to the Diagnosis of Kashin-Beck Disease (WS/T 207-2010 [15]).
After then, standard questionnaire and physical examinations were conducted for all selected KBD patients in degree I and II. Standard questionnaire mainly included health status, previous medication history, WOMAC pain and stiffness subscales or Assessment for Therapeutic E cacy on Kashin-Beck Disease scale (WS/T 79-2011 [16]). Physical examinations included measurements of patients' height, weight, liver and renal function, etc. The heights and weights of patients were measured using height and weight scales. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of the individual's height in meters. According to World Health Organization (WHO) criteria, the BMIs were categorized into four grades: underweight (BMI<18.50 kg/m 2 ), normal (18.5≤BMI<25 kg/m 2 ), overweight (25≤BMI<30 kg/m 2 ) and obese (BMI≥30.00 kg/m 2 ).
KBD patients aged 45-70 years in degree I and II would be enrolled in this trial if they had lived in our test sites since they were children (6 years old).
Individuals with other joint diseases, such as joint in ammation, metabolic bone diseases, neoplasia, osteoporosis, etc., or coexisting conditions such as a history of diabetes mellitus, stroke, emaciation, long-term use of drugs and cardiovascular, gastrointestinal, kidney, liver, respiratory diseases, or used drugs for OA/KBD in past three months and those who declined to participate in the study were excluded. Finally, 308 patients entered the study cohort.
Group setting and interventions 308 KBD patients were divided into two different groups (154 in group A and 154 in group B). On the basis of the literatures [12][13][14] and drug instructions, subjects in group A and group B were assigned to receive chondroitin sulfate tablet (300mg once, three times daily) and diacerein capsule (50mg once, twice daily), respectively. Drinking was not allowed during medical treatment. Patients should stop taking drugs immediately as soon as suspicious side effects appeared.
The collection of blood sample At 0, 12 and 24 weeks, the morning fasting blood samples of subjects (not less than 3 mL/person) were collected in 5-mL tubes with or without 10% ethylene diamine tetraacetic acid, then left at room temperature for 1-2 h. After that, the blood samples were centrifuged at 3000 rpm for 10 min to separate plasma or serum. The plasma and serum were dispensed into 120 μL aliquots into microcentrifuge tubes that were stored at -80 °C until assay.

Follow-up
The subjects were followed up every 4 weeks by doctors from township hospital. The main jobs of each follow-up were to record amount of drug use and occurrence of AEs or SAEs. Meanwhile, drugs of next phase were distributed to subjects with good compliance. The compliance of subjects was mainly assessed by following up their medication use, including whether they taken drugs according to the instructions, the amount of drugs left over from this cycle and whether they taken other drugs at the same time.

E cacy assessment
In this trial, Assessment for Therapeutic E cacy on Kashin-Beck Disease (WS/T 79-2011) [16] was applied to evaluate therapeutic effect of drugs, and the proportion of patients with effective therapeutic effect and overall improvement rate were calculated and set as primary e cacy parameters. Firstly, joint dysfunctions including joint rest pain (0 to 2 points), joint kinesthetic pain (0 to 2 points), morning stiffness (0 to 2 points), maximum walking distance (0 to 2 points) and lower limb mobility (0 to 2 points) before and after treatment were evaluated. Then, the recovery rates of joint functions after interventions were calculated using above joint dysfunctions scores. The therapeutic effects of drugs were judged to be signi cant, effective or invalid when the recovery rate of joint functions ≥70%, 30%-70% and <30%, respectively. Afterwards, the proportion of patients with effective therapeutic effect was calculated and compared. Finally, the proportion of patients with both signi cant and effective therapeutic effects (i.e. overall improvement rate) after interventions in different groups was gained and compared.
The WOMAC scale is a validated questionnaire addressing severity of joint pain ( ve questions), stiffness (two questions) and limitation of physical function (seventeen questions) of OA, researchers can use the whole system or choose parts of it. In this trial, the mean changes in WOMAC pain and stiffness subscales were chosen as the secondary e cacy parameters.

Safety evaluations
Safety evaluations in this trial included occurrences of AEs, SAEs at each visit and changes of liver and renal functions at 0, 12 and 24 weeks. AEs were mainly listed in the package inserts of drugs, including diarrhea, stomach upset, nausea, dry mouth and dizziness. SAEs were de ned as any fatal or lifethreatening clinical experience or disabling event, or requiring long-term hospitalization, whether or not it is judged to be related to treatments.
Reagents for all indexes were purchased from Ningbo Meikang Biotechnology Co., Ltd., Ningbo, China.

Quality control
Our study was carried out by a well-trained research staff along with doctors, and they were trained before the trial. The training module contents included purpose of this trial, study procedures and how to implement the questionnaire, etc. The instruments used in trial were standardized by local technical supervision bureau. Additionally, to unify various instrument operation methods and formulate standard operation procedures so that the team could complete all diagnosis and follow-up tasks according to uni ed standard.
Three exports or technologists read same X-ray image and provide their own evaluations based on the Diagnosis of Kashin-Beck Disease (WS/T207-2010) [15] to our team. The diagnosis of KBD patients was nalized by our team according to views of these experts.
A health education program was implemented to our subjects. The purpose and signi cance of treatment were clari ed three times so that subjects could actively cooperate with treatment. Patients changed groups without permission and terminated trial unauthorized were recorded in order to control the rate of loss of follow-up strictly. The corresponding data of subjects withdrew from the trial should be handled statistically.

Statistical analysis
Data were analyzed by using SPSS software (version 22.0 SPSS Inc., Chicago IL). Basic characters, BMI, X-ray diagnosis of patients, primary e cacy parameters and other categorical variables were analyzed with Chi-square test. Data of WOMAC pain and stiffness scores was expressed as mean ± standard deviation and analyzed by use of Two-way Repeated Measures Analysis of Variance. Intent-to-treat (ITT) analysis was applied in this study. All tests were two-tailed, and P<0.05 was considered statistically signi cant.

Results
Basic characteristics, BMI and X-ray diagnosis of patients Among 308 subjects, 24 (15.58%) in group A and 38 (24.68%) in group B withdrew from the study at primary end point; 49 (31.82%) in group A and 75 (48.70%) in group B failed to complete the trial at secondary end point, and the reasons for loss of follow-up were shown in Figure 1.
Basic characteristics of subjects including age and sex were described in Table 1. All basic characteristics were balanced and comparable between two groups at baseline (all P>0.05, Table 1). At the same time, there were no signi cant differences in BMI and X-ray diagnosis of subjects between two groups at the beginning of trial (all P>0.05, Table 1).

Evaluations of primary e cacy parameters
At primary end point, the proportion of patients with effective therapeutic effect (P=0.021) and overall improvement rate in group B (P=0.007) were signi cantly higher than those in group A. However, there were no statistical differences in the proportion of patients with effective therapeutic effects (P=0.151) and overall improvement rates (P=0.301) between two groups at secondary end point. In both groups, no statistical differences were seen in the proportions of patients with effective therapeutic effects (P Group A =0.121, P Group B =0.729) and overall improvement rates (P Group A =0.400, P Group B =0.509) between primary and secondary end points. The data was present in Figure 2. between two groups. But 95% con dence intervals (CI) for differentials in WOMAC pain and stiffness scores between two groups (group B minus group A) were basically zero axisymmetric at baseline, and were mostly present in the left of axis of symmetry, showing left-biased distribution at primary and secondary end points. The data was shown in Table 2.
In group A, compared to baseline, WOMAC pain and stiffness scores at primary (P pain =0.0015, P stiffness =0.0023) and secondary end points (P pain =0.0003, P stiffness =0.0082) decreased obviously. The differentials of WOMAC pain scores were 5.79 (95%CI: 1.90 to 9.68, baseline VS primary end point) and 6.49 (95%CI: 2.60 to 10.38, baseline VS secondary end point), and the differentials of WOMAC stiffness scores were 2.29 (95%CI: 0.70 to 3.88, baseline VS primary end point) and 2.03 (95%CI: 0.44 to 3.62, baseline VS secondary end point). In group B, WOMAC pain and stiffness scores at primary (P pain <0.0001, P stiffness =0.003) and secondary end points (P pain <0.0001, P stiffness =0.0021) also decreased clearly. The differentials of WOMAC pain scores were 7.75 (95%CI: 3.43 to 12.06, baseline VS primary end point) and 8.53 (95%CI: 4.22 to 12.85, baseline VS secondary end point), and the differentials of WOMAC stiffness scores were 2.76 (95%CI: 0.80 to 4.72, baseline VS primary end point) and 2.85 (95%CI: 0.89 to 4.81, baseline VS secondary end point). There were no signi cant differences in WOMAC pain and stiffness scores between primary and secondary end points in both groups. The detail was shown in Table 2 and Table 3.

Safety assessment
No severe adverse events (SAEs) were seen throughout the trial according to our follow-up records. Total adverse events (AEs) were reported in 11 patients over rst 12 weeks, 4 in group A and 7 in group B, and no signi cant difference was seen between two groups (P=0.353). Total AEs were reported in 21 patients over second 12 weeks, 9 in group A and 12 in group B, and no signi cant difference was seen between two groups (P=0.646). Among all AEs, stomach upset accounted for the highest proportion, and there was signi cant difference in stomach upset between two groups at secondary end point (P=0.022). No signi cant differences were seen in total AEs in both groups between primary and secondary end points (P group A =0.152, P group B =0.336). The data was shown in Table 4.
At primary end point, the proportion of patients with abnormal DBIL in group A was signi cantly higher than that in group B (P=0.012). Besides DBIL at primary end point, no signi cant differences were seen in other liver function indexes between two groups at other time points. At same time, there were no signi cant differences in all 8 liver function indexes in both groups before and after interventions (Figure 3).
At baseline, the proportion of patient with abnormal mALB in group A (38.2%) was signi cantly higher than that in group B (19.0%, P=0.004). At secondary end point, the proportion of patients with abnormal UREA in group A (3.8%) was signi cantly lower than that in group B (11.4%, P=0.047). Besides above two indexes, no signi cant differences were seen in the proportions of abnormal patients with other renal function indexes between two groups at other time points (all P>0.05, Figure 4).
In group A: the proportions of patients with abnormal UREA and CREA at secondary end point were signi cantly lower than those at primary time point (P UREA =0.014, P CREA =0.003) and baseline (P UREA =0.008, P CREA =0.005); the proportions of patients with abnormal mALB at primary (P=0.002) and secondary end points (P<0.001) were signi cantly lower than that at baseline; the proportion of patient with abnormal NAG at secondary end point was signi cantly higher than those at primary end point (P=0.031) and baseline (P=0.031). Differently, in group B, no signi cant differences were seen in the proportions of abnormal patients with four renal function indexes throughout the trial (all P>0.05, Figure 4). NSAIDs, DMOADs, vitamins, proprietary Chinese medicines were adopted alone or jointly according to actual situation of each endemic area. Unfortunately, the therapeutic effect of drugs used in this program was di cult to accurately evaluate due to different drug selection and treatment cycle. Meanwhile, the results showed that oral drugs were better for adult KBD patients in degree I and II and not ideal for adult KBD patients in degree III. Based on this program, adult KBD patients in degree I and II were selected as our subjects in this study.

Discussion
Assessment for Therapeutic E cacy on Kashin-Beck Disease (WS/T 79-2011) [16] stipulates principle and method of determining clinical treatment effect, which can assess joints status by evaluating daily living and working functions of KBD patients. The standard was speci c and reliable, and has been used widely to evaluate the treatment of KBD [17][18][19]. Previous results of our team in 2010 indicated that chondroitin sulfate was effective in treating adult KBD [7], and the result was veri ed once again in this trial. During 24-week interventions, nearly 20% of patients in group A showed effective and/or signi cant therapeutic effects. Encouragingly, overall improvement rates of patients in group B were consistently higher than 30% at primary and secondary end points.
In rst 12-week, the therapeutic effects (both the proportion of patients with effective therapeutic effect and overall improvement rate) of diacerein were much better than chondroitin sulfate. In second 12-week, the therapeutic advantage of diacerein remained, but became less pronounced. And from 12 to 24 weeks, there was little increase in the proportions of patients with effective therapeutic effects and overall improvement rates in two groups, namely the differences in therapeutic effects of two drugs between two end points were not obviously. On the one hand, above results suggested that therapeutic effect of diacarin for KBD was stronger than that of chondroitin sulfate. On the other hand, therapeutic effects of both two drugs could be achieved better after 12week interventions, and could be maintained but not enhanced signi cantly with the extension of treatment time.
The WOMAC scale is used widely to assess the severity of OA or KBD and the e cacy of treatment based on associated symptoms and signs [8,12]. In scale, the structure and function of hip and knee were evaluated by pain, stiffness and joint function, and the functional description was mainly for lower limbs [13,20]. However, in KBD patients, the interphalangeal, wrist and ankle were most frequently affected among all involved joints because the disease occurs in childhood and mainly harmed epiphysis and metaphysis of phalanges [21]. So, depending on patients' main symptoms and involved joints, WOMAC pain and stiffness subscales were used as secondary e cacy parameters in this trial.
There were no statistical differences in WOMAC pain and stiffness scores between two groups at all three time points. But from baseline to primary and secondary end points, 95% CI for differentials in WOMAC pain and stiffness scores between two groups went from basically zero axisymmetric to left-biased distribution, indicating WOMAC pain and stiffness scores of two groups changed from essentially equal to that of group B being less than group A. Therefore, above results implied that therapeutic effects of diacarin for pain and stiffness in KBD were slightly better than chondroitin sulfate. In both groups, with the prolongation of treatment time, WOMAC pain and stiffness scores decreased signi cantly, but there were no signi cant differences in WOMAC pain and stiffness scores between primary and secondary end points. The change trends of WOMAC pain and stiffness scores were consistent with therapeutic effects of drugs in above section. These two results corroborated each other and proved together that 12 weeks was a suitable intervention time for diacerein and chondroitin sulfate. Besides, in 95% CI terms, both drugs improved pain better.
The occurrences of AEs, SAEs and changes of liver and renal functions at each visit were monitored to evaluate safety of drugs used in this study. Firstly, no SAEs were found in this study. Secondly, the occurrences of total AEs in both groups were relatively low, and no signi cant differences were seen between two groups at same time point or between two end points at same group. But after 24-week treatment, the incidence of stomach upset caused by diacerein was clearly higher than that caused by chondroitin sulfate. Thirdly, diacerein had no side effects on liver and renal functions at any end point. Finally, chondroitin sulfate had no side effects on liver function and could the proportion of patients with abnormal UREA, CREA and mALB decrease apparently, but the proportion of patients with abnormal NAG increased. NAG is a sensitive indicator of kidney injury, the level of NAG can be more than 10 times higher than the upper limit of normal value when renal function is impaired. Among 7 patients with abnormal NAG, 2 patients had high NAG levels and which were slightly higher than the upper limit of normal value; remaining 5 patients had lower NAG levels and which were lower than the lower limit of normal value (data was not provided in this article). In other words, none of above changes of NAG had clinical signi cance. In total, all above results showed that the safety of drugs used in this study were better, especially at primary end point.
The rate of patients who lost to follow-up at secondary end point was relatively high and this would reduce the reliability of results. In order to analyze possible impact of withdraw on results, a sensitivity analysis was conducted. According to information obtained in the eld and results of sensitivity analysis, possible reasons for loss of follow-up including following three points: rstly, at baseline of the trial, it was busy farming season and most villagers were farming at home, so subjects were easy to recruit. However, at secondary end point of the trial, it was slack farming season, and some relatively younger subjects went to work in other places, so they could not return to their own villages for follow-up. This was consistent with result of sensitivity analysis that the age of subjects who lost to follow-up in group B (56.55±7.13) was younger than that in group A (59.53±5.60, P=0.0148, S- Table 1). Secondly, some KBD patients did not know two drugs used in this trial and feel its e cacies, especially lessening pain were not obvious, so they give up taking the drug and withdrew from the trial, and data supporting this hypothesis were shown in S- Table 2. In both groups, the WOMAC pain scores of patients who were lost to follow-up were signi cantly lower than those of patients who completed the trial (P Group A =0.0465, P Group B =0.0058). Finally, the last follow-up was completed in mid-December 2019, the COVID-19 outbreak prevented our team from making any further investigation of lost subjects and evaluating the subsequent effects of DC after 6-week and 12-week treatment-free observation period.
Besides the rate of subjects who lost to follow-up was relatively high, this trial still has other limitations. First, no placebo group was set due to the restrictions of Administration Village Committee, which would reduce the accuracy of the study. Second, blind grouping was not applied because subjects' health status. The statistical e ciency would improve if blind grouping is used. Last, all subjects were from Heilongjiang Province and there might be regional limitations. Therefore, we will continue to conduct further study in future to verify the authenticity of this study.

Conclusions
In summary, the results showed that both diacerein and chondroitin sulfate had certain therapeutic effect on adult KBD patients and could improve pain and stiffness. 12-week interventions were suitable, drugs could work effectively and avoid side effects minimally. Compared with chondroitin sulfate, the therapeutic effect of diacerein on adult KBD might work stronger. The results of this trial suggested that diacerein could be used to treat adult KBD in KBD endemic areas in future.   Figure 1 Flow chart of the trial.