Patient DemographicsPopulation

A total of 50 patients were assessed for eligibility and included in this post-marketing clinical follow-up (PMCF) study. One patient had to be excluded from the analysis after meeting an exclusion criterion following inclusion. Therefore 49 patients comprised the full analysis set. A total of 44 patients completed the study according to protocol and 5 patients terminated the study prematurely. For patient flow, refer to the CONSORT diagram (Fig. 1).

Fig. 1

CONSORT diagram. *One patient excluded from analysis after meeting an exclusion criterion following inclusion. n number of patients

CM-chitosan was used to treat 35 women (71.4%) and 14 men (28.6%) with a mean age of 65.6 years and a body mass index (BMI) of 30.1 kg/m2. The Kellgren–Lawrence grade was II in 23 patients (46.9%) and III in 26 patients (53.1%). All subjects underwent extended clinical examination, prior to any injection, to exclude potential contraindications. All subjects received injections under ultrasound guidance.

Patients’ baseline demographics are summarized in Table 1.

Table 1 Patients’ baseline demographicsEfficacy ResultsVAS Pain Score

There was a rapid onset of effect and pain decreased significantly (p ≤ 0.001) from a median value of 49.0 mm at baseline (n = 49) to 24 mm at week 1 (n = 49) and to 18 mm at week 12 (n = 48). VAS pain score decreased further to a median value of 11.0 mm at week 24 (n = 44), and it was 18 mm at week 36 (n = 45). Corresponding mean values were 47.3 ± 26.1 mm at baseline, 29.1 ± 26.0 mm at week 1, 29.8 ± 28.1 mm at week 12, 21.6 ± 23.0 mm at week 24, and 25.4 ± 23.9 mm at week 36. Mean results are presented in Fig. 2.

Fig. 2

Evolution of VAS pain score [mm]. Data are presented as mean value, and standard deviation (error bars), *p < 0.001 versus baseline. VAS Visual Analog Scale

Hence, at all time points VAS scores were significantly lower compared to baseline (Wilcoxon matched pairs test, p < 0.001 at week 1, p = 0.001 at week 12, p < 0.001 at week 24, and p < 0,001 at week 36). From week 1 until week 36, at least 70% of patients showed an improvement in pain, indicating a long-lasting effect.

KOOS Questionnaire

This analysis was performed according to current KOOS scoring instructions. All subscales of the KOOS (pain, symptoms, activities of daily living, sport and recreation, and quality of life) were calculated independently and were converted into a scale from 0% to 100% where 0% represents “the most severe symptoms” and 100% means “no symptoms”.

In all KOOS subscales patients showed a statistically significant improvement as presented in the following tables and figures, with a summary in Fig. 3.

Fig. 3

Summary of mean KOOS scores at the different time points. ADL activities of daily living, QoL quality of life, Sport/Rec sports and recreational activities

KOOS Pain

KOOS pain improved progressively from a median of 58.3% at baseline (mean 56.2% ± 18.8%) to 86.1% (mean 74.1% ± 24%) at week 36 compared to baseline. Median scores were 66.7%, 79.2%, and 82.3% (mean = 67.4%, 73.5%, and 76.4%) at week 1, 12, and 24, respectively (Table 2 and Fig. 4).

Table 2 Subscale KOOS pain from baseline to week 36Fig. 4

Subscale KOOS pain from baseline to week 36. Data are presented as median value, and 25th and 75th percentiles, minimum and maximum (error bars), p < 0.001 versus baseline

Improvement of KOOS pain was statistically significant at all time points (Wilcoxon matched pairs test, p < 0.001 each).

KOOS ADL

KOOS ADL (activities of daily living) increased markedly from a median of 54.4% (mean 57.4% ± 20.6%) at baseline to 82.35% (mean 77.6% ± 20.8%) at week 36. Starting with week 1 all percentages related to activities were increased to greater than 73.5% (Table 3 and Fig. 5).

Table 3 Subscale KOOS ADL from baseline to week 36Fig. 5

Subscale KOOS ADL from baseline to week 36. Data are presented as median value, and 25th and 75th percentiles, minimum and maximum (error bars), p < 0.001 versus baseline

Improvement of KOOS ADL was statistically significant at all time points (Wilcoxon matched pairs test, p < 0.001 each).

KOOS Symptoms

KOOS symptoms improved from a median of 64.3% (mean 64% ± 19.5%) at baseline to 89.3% (mean 77.4% ± 21.1%) at week 24. Thereafter, there was a small decrease to 82.1% (mean 74% ± 23.1%) at week 36 (Table 4).

Table 4 Subscale KOOS symptoms from baseline to week 36

Improvement of KOOS symptoms was statistically significant at all time points (Wilcoxon matched pairs test, p = 0.004 at week 1, p = 0.003 at week 12, p < 0.001 at week 24, and p = 0.003 at week 36).

KOOS Sport and Recreation

The KOOS Sport/Rec increased markedly from a median of 25.0% (mean 30.9% ± 26.7%) at baseline to 72.5% (mean 62.4% ± 29%) at week 24. It remained high with 65.0% (mean 58.1% ± 32%) at week 36 (Table 5).

Table 5 Subscale KOOS Sport/Rec from baseline to week 36

Improvement of KOOS Sport/Rec was statistically significant (Wilcoxon matched pairs test, p = 0.001 at week 1 and p < 0.001 at weeks 12, 24, and 36).

KOOS Quality of Life

KOOS QoL improved clearly from a median score of 25.0% (mean 29.2% ± 18.3%) at baseline to a median score of 43.8% (mean > 50%) at weeks 12, 24, and 36 (Table 6).

Table 6 Subscale KOOS QoL from baseline to week 36

Improvement of KOOS QoL was statistically significant (Wilcoxon matched pairs test, p < 0.001 each).

KOOS Improvement/Worsening

An increase in score values compared to baseline, regardless of the extent, was interpreted as improvement of the KOOS subscales. A summary of the values of the five subcategories, pain, symptoms, ADL, Sport/Rec, and QoL is provided in Table 7.

Table 7 Assessment overview on improvement, worsening, and no change over time of all five subscales, given as absolute (N) and percentage (%) frequency distribution

An improvement in all KOOS subscales was noted in more than 70% of patients at 6 months, specifically 75.0%, 70.5%, 84.1%, 88.2%, and 84.1% in KOOS pain, symptoms, ADL, Sport/Rec, and QoL, respectively. Improvement in KOOS pain continued to 9 months with 80.0% of patients improved compared to baseline.

Responder Analysis

The responder rate was calculated at week 24 follow-up visit according to the OMERACT-OARSI proposed set of responder criteria [12]. The therapy response was calculated for each patient and for each performance criterion. Six months after treatment, 32 patients (72.7%) could be classified as responder and 12 patients (27.3%) as non-responder. As a result of missing data, no response status could be calculated for five patients.

Patient’s Global Assessment and Patient’s and Investigator’s Satisfaction Scales

The PGA was based on a question asked by the physician on how the patient was affected by their knee OA and how the patient would rate their condition on the day of the visit. The answer was scored on a scale of 0 to 10, where 0 is “very good condition” and 10 is “very poor condition”.

Overall, the patients’ condition had significantly improved. An improvement was noted in at least 70% of patients at all time points compared to baseline. At week 1 follow-up, 69.4% of subjects reported subjective improvement; at week 24, the proportion had risen to 84.1%. PGA scores at each follow-up time point were statistically significant compared to baseline (Wilcoxon’s matched pairs test, p < 0.001).

There was a good match with patient’s satisfaction, since more than 75% of patients stated they were either “satisfied” or “very satisfied” with the treatment.

Satisfaction with the treatment was expressed by the investigator for at least 65% of patients and the condition of most patients had improved according to the investigator.