The signal knee was chosen prior to randomization based on which knee met the inclusion and exclusion criteria. The JSW is the space measured in mm between the 2 bones in the knee joint and this is assessed by x-ray. The JSW decreases with disease progression. The lower limit for participation in the trial were 2 mm JSW. There were no upper limit as long as inclusion and exclusion criteria were met. The outcome was measured as a change in JSW from baseline to month 24.

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication. There were 2 patients in the randomized population who did not receive any study drug and where therefore excluded from the ITT analysis.

WOMAC is a self-administered set of standardized questionnaires to evaluate the condition of patients with osteoarthritis of the knee. The subject marks on a scale (1-100) the pain associated with performing each daily activity listed in the questionnaire. 0 is no pain (best), 100 is extreme pain (Worst). The total pain sub score for the questions are then calculated. Total possible minimum sub score is 0, maximum is 500. The final outcome is the absolute change from baseline to 24 months. If the outcome is less that 0 there is improvement (less pain).

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication. There were 2 patients in the randomized population who did not receive any study drug and where therefore excluded from the ITT analysis.

The central laboratory analyzed serum CTX-I (S-Crosslaps, Elecsys) and osteocalcin as well as urine CTX I/creatinine and CTX-II/creatinine. These biomarkers were analysed in order to assess the cartilage and bone turonver ratio to baseline at month 24.

The number of participants analysed for these outcomes were a total of 150 subjects from the intent-to-treat (ITT) population evenly distributed across sites, all with completed baseline as well as all follow-up visits.

Knee cartilage volume and thickness was assessed by MRI in patients from the sites in Ballerup, Denmark, the Czech Republic, and Romania using a quality controlled low-field 0.18T C-Span scanner from Esaote dedicated to the imaging of extremities. The same solenoid coil was used for all patients at a given site.

Subset of patients from the sites in Ballerup (Denmark), Czech Republic and Romania

Function and physical activity were assessed by assessed by WOMAC function sub-score in the signal knee. The criteria for assessment of the functional classification according to the American Rheumatism Association (ARA) were as follows (Hochberg et al 1992):

I. Completely able to perform usual activities of daily living (self-care, vocational and avocational) II. Able to perform usual self-care and vocational activities, but limited in avocational activities.

III. Able to perform usual self-care activities, but limited in vocational and avocational activities.

IV. Limited ability to perform usual self-care activities, vocational and avocational activities.

Self-care activities included dressing, feeding, bathing, grooming, and toileting. Avocational (recreational and/or leisure) and vocational (work, school, homemaking) activities were patient-desired and age- and sex-specific.

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication.

WOMAC's stiffness subscore was used to assess the stiffness in the signal knee. WOMAC is a self-administered set of standardized questionnaires to evaluate the condition of patients with osteoarthritis of the knee. The subject marks on a scale (1-100) the degree of joint stiffness for when performing each daily function listed in the questionnaire. 0 is no stiffness (best), 100 is extreme stiffness (worst). The total function sub score for the questions are then calculated. Total possible minimum sub score is 0, maximum is 200. The final outcome is the absolute change from baseline to 24 months. If the outcome is less that 0 there is improvement (less stiffness).

Patients were instructed not to take analgesics for 3 days prior to the WOMAC test.

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication.

Health-related quality of life was assessed by the EQ-5D questionnaire, which is a patient questionnaire for measure of health, developed by the EuroQol Group.The EQ-5D questionnaire was administered to patients in order to measure change in health-related quality of life (HRQoL) over 2 years.

The subjects marks on a VAS scale from 0 to 100 the number that best describes their health today (0 is worst imaginable health state and 100 is best imaginable health state).

The change from baseline in the EQ-5D VAS was calculated.

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication.

Pain was assessed by the WOMAC subscore in the signal knee by visit. Patients assessed their current pain level using a 100 mm visual analogue scales (VAS) by placing an X on the line that best describes his/her pain, where 0 equaled "No Pain" and 100 equaled "Worst Pain Imaginable". Patients were instructed not to take analgesics for 3 days prior to the VAS.

The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication.