Version 1 was submitted to improve patient recruitment and facilitation of the trial conduction. Performance of screening and baseline procedures on the same day was allowed, but start of study medication (NSAID) must be delayed until laboratory reports confirming the patient’s eligibility have been obtained. Patient will be informed about the laboratory results by the investigator and whether he/she is allowed to start treatment with NSAID. Date of the patient contact and date of treatment initiation must be recorded in the source documents. The concomitamt participation in any observational (non-therapeutic) study was allowed. Concomitant therapy with a disease modifying antirheumatic drug such as methotrexate, sulfasalazine, leflunomide, hydroxychloroquine, azathioprine) and systemic corticosteroids (≤10 md/day prednisolone equivalent) was allowed. Any NSAID can be chosen as study medication (diclofenac was supplied as the free study drug), assuming that the retardation of radiographic progression of AS is a class effect of NSAIDs which is based on several observations. This will facilitate the recruitment of patients who are treated already on demand with an NSAID. The NSAID dose in the continuous treatment arm can be varied between 50% and 100% of the maximally daily recommended dose (not 100% only). Again, this will increase the acceptance of both the patients and the rheumatologists to participate in the study. The rationale behind this is based on the assumption that the daily (continuous) therapy with NSAIDs is more relevant for inhibiting bone formation than the actualy dose. Hospitalization which has been planned prior to the screening visit due to pre-existing concomitant conditions (i.e. elective hospitalizations) will not be considered as serious adverse events (SAE). For study visits from week 16 to week 112 visit the time window for the conduct of a study visit was broadened to +/- 28-days.

Ammendment 2, Version 1,also to improve the patient recruitment and facilitation of the trial conduction as well as administrative changes. Requirement of a syndesmophyte as an inclusion criterium has been taken out. This requirement may in fact hamper the inclusion for three reasons: First, rheumatologists do not always feel confident in identifying a syndesmopyhte and rely on the radiologist. Second, a patient cannot included at the first visit because radiographs need to be reviewed first (searching for a syndesmopyhte). Third, more AS patients are potentially eligible for the study because syndesmophytes are found in 50-70% of AS patients on average, but not in all patients. Thus, skipping this inclusion requirement will facilitate the inclusion of patients. The sample size recalculation has been performed. The number of patients need to be included in the trial in order to reveal differences in radiographic progression between two treatment group is 174 now (87 patients in each arm). The coordinatingl investigator of the trial, principal investigator of the trial site 1 and legal representative of the sponosr has been changed: coordinating investigator, site 1 principal investigator and legal representative - Prof. Dr. Joachim Sieper, principal co-investigator - Dr. Martin Rudwaleit. 4 new studies sites are added, one study site is excluded.