Issued before inclusion of patients after review from ethics committee and additional feedback from investigators to further specify inclusion and exclusion criteria and to implement changes to the schedule of assessments. The changes were as follows: - The inclusion/exclusion criteria were modified to add a diagnosis of depression according to ICD-10 criteria to further characterize the patient population having a definite diagnosis of depression. - Exclusion criterion number 10 was revised to specify the exclusion of patients with arrhythmia receiving antiarrhythmic drugs. - An additional study drug discontinuation criterion in terms of QTc-prolongation was added to the protocol. - In order to assess possible QTc-prolongations after uptitration of antidepressants and during the clinical assessment additional ECG recordings were included at visit 3.1 and visit 4. Visit 3.1 was changed to be a site visit for patients in the ECG-subgroup. - The correct Modified Fatigue Impact Scale was added to the Appendix; the Fatigue Impact Scale (39 items) was incorrectly inserted in the original protocol.

Issued before inclusion of patients based on recommendations from ethics committee: - Exclusion criterion number 10 is revised to specify the exclusion of patients with an increased QT-interval. - In section 7.5.6 the 3-lead ECG was replaced by a 12-lead ECG.

Issued before inclusion of patients after additional feedback from investigators to eliminate the sertraline arm. It introduced the following changes: - The study design was amended and the sertraline arm was eliminated. - The patient number of the ECG group was increased to 125 patients.

Issued to implement recommendations of the CHMP to strengthen cardiovascular monitoring during treatment initiation of fingolimod. The EMA was reviewing the benefits and risks of fingolimod. The review had been started after Novartis had informed EMA of the unexplained sudden death of a 59-year-old female patient with multiple sclerosis in the United States of America within 24 hours of taking the first dose of fingolimod in November. The patient was being treated with metoprolol and amlodipine for hypertension. 6 other cases of unexplained death had been reported, 3 of which were sudden. In addition, other reports included 3 deaths due to heart attack and 1 due to disruption of the heart rhythm. At the time of its authorisation, no cases of sudden or unexplained death had been reported in studies with fingolimod. However, fingolimod was known to cause a transient bradycardia and might be associated with atrioventricular block after the first dose.

In agreement with the EMA, the following recommendations are effective immediately for patients treated with fingolimod as described in “Dear Health Care Professional Letter” as follows: ‘For all patients starting treatment, monitoring during the first 6 hours after dosing should include:- a 12-lead ECG prior and 6 hours after the first dose; - continuous 6-hour ECG monitoring; -lood pressure and heart rate measurement every hour. -In those patients with evidence of clinically important cardiac effects, monitoring should be extended until resolution. The following criteria for extended monitoring are recommended: - The presence at the 6-hour time-point after first dose of: o heart rate less than 40 beats per minute; -decrease in heart rate of more than 20 beats per minute compared with baseline; - persistent new-onset 2nd degree atrioventricular block, Mobitz Type I (Wenckebach). - The occurrence at anytime during the 6-hour-monitoring of: - symptomatic bradycardia; - new onset 2nd degree atrioventricular block, Mobitz Type II; - new onset 3rd degree atrioventricular block. When fingolimod treatment is interrupted for more than 14 days for re-commencing the above recommendations for first dose monitoring will also apply.’ On January 27, 2012 the “Dear Health Care Professional Letter” was sent to all FTY720 study sites. These recommendations were effective immediately. Sites were required to inform the patients about these new recommendations and patients were required to reconsent their participation.

The citalopram arm was being closed due to ethics committee request to increase the frequency of ECG recordings in the citalopram arm. This request could not be fulfilled as a higher frequency of ECG controls on a weekly basis would extremely minimize acceptance of investigators and patients and therefore strongly affect recruitment. Treatment with the combination of fingolimod and citalopram therapy was available without any additional ECG in clinical routine. Therefore the sample size was reduced to 250 patients. The inclusion of patients with cardiac risk factors was revised according to the new recommendations of CHMP and EMA. - Patients who had already been on fingolimod for at least 7 days could be enrolled or patients where the first dose observation took place in the study CFTY720DDE17 and completed this study were also eligible for inclusion. The study CFTY720DDE17 was implemented to collect additional safety data on bradyarrythmic events after the first dose of fingolimod. Patients were monitored by ECG (pre-dose 12-lead ECG, 6h continuous ECG and a 12-lead ECG 6h post dose) following the first dose of fingolimod. - The uptitration period was prolonged to max. 28 days and the wash-out period for prior antidepressant therapy was shortened to 14 days prior visit 2. - The most recent recommendations from EMA in terms of cardiac response monitoring following the first dose of fingolimod were included. Sites were required to inform the patient about these new recommendations and patients were required to re-consent their participation.

Issued to implement corrective changes: - A well-known mode of action of fingolimod is a reduction of peripheral lymphocyte count to values which are approximately 70-75% lower than baseline counts. If patients are pretreated with fingolimod lymphocyte counts can be lower than 800 cells/mm3. Exclusion criterion #12 was modified to reflect the inclusion of pretreated patients. - The number of dispensed packs of antidepressant medication at visit 3 was corrected according to the longer titration period implemented with Amendment 5. - Treatment initiation recommendations in cardiac risk patients were removed as these patients could not be included in the trial.

Issued to stop recruitment due to slow enrollment. The first patient was recruited on 11 Nov 2011 and the necessary patient number had still not been reached, even though the recruitment phase had been extended by 30 weeks. 54 out of the 250 anticipated patients had entered the study and it was considered unlikely that the planned number of patients could be achieved in a reasonable period of time. Therefore, the recruitment for this study was stopped prematurely. The analysis was performed as planned. Centers were notified to stop recruitment of patients on 30 Apr 2013. Patients, who had already been recruited into the study performed all planned visits according to the protocol.

Issued to implement an update of the Gilenya (fingolimod) label in the EU approved by the Committee for Medicinal Products for Human Use (CHMP): - The label update provides refined guidance on when existing first dose monitoring procedures should be repeated. In patients, who are re-initiated after a certain treatment interruption, a repetition of first-dose-monitoring strategy is necessary. - In patients, who require pharmacological intervention during the first dose monitoring and who are monitored overnight in a medical facility, monitoring should be repeated after the second dose of fingolimod.