• follow-up visits at 48 weeks (remission criteria assessment, for phase III non-remitters only) and 72 (remission criteria assessment) were added to enable long term follow-up of participants. • contraceptive use as inclusion criteria was added on request of regulatory authorities. • inclusion criteria regarding the onset of illness was changed from ‘a maximum of 2 years since onset of positive symptoms’ to ‘a maximum of 2 years since onset of psychosis’. Many patients have experienced vague symptoms in their childhood, but the onset of psychosis was regarded a more relevant starting point. • For patients using clozapine, leucocyte checks need to be continued for at least 4 weeks when a patient discontinues clozapine. • Several reasons for withdrawal of participants were added: 1) The nature of the patients treatment is changed to coercive treatment (based on judicial ruling); 2) In contrast to the patient's status at enrollment, the patient is now represented by a legal guardian or under legal custody; 3) Emergence of one or more contraindications against any of the study drugs as mentioned in the Summaries of Product Characteristics (refer to Appendix B). In particular, clozapine use needs to be discontinued when one or more of the following adverse events occur: severe leucopenia (leucocyte count <3000/mm3 or 3.0x109/l) or neutropenia (count <1500/mm3 or 1.5x109/l), myocarditis or cardiac arrhythmias; 4) Patient becomes pregnant or initiates lactation

• the recommended tapering schedule of study medication was adjusted to slow down the dose increase in order to decrease the chance on and severity of extrapyramidal symptoms, as the first patient who entered the study suffered from severe EPS. • on request of the participating centers, a titration recommendation for the transfer from phase I to phase II, and from phase II to phase III study medication was included. • target dose of 400 mg/day amisulpride was added, in line with findings from the EUFEST study. However, clinicians could deviate from the target dose as well as the titration scheme if deemed necessary.

• the eligibility for entering the Psychosocial Intervention component after the pharmaco-therapeutic component was no longer limited to patients meeting remission criteria, but also for drop outs and patients not meeting remission criteria, as they could also benefit from this intervention. • It was found that ‘Schizophreniform disorder’ could not be completely assessed through the M.I.N.I. diagnostic interview. Therefore the confirmation of this diagnosis was rephrased as follows: Schizophreniform disorder is assessed through a M.I.N.I. diagnosis of psychosis NOS complemented by a diagnosis of schizophreniform disorder according to DSM-IV criteria. • a clinical diagnosis was added to the long term f/u visit 22 (74 weeks) to gain insight into the stability of the diagnosis of participants at baseline.

• closure of one participating center, addition of a new participating center. • increase of patient sample from 350 to 500, due to the high remission rate in phase I. • a third MRS scan was added, 10 weeks after baseline, providing a longer term follow up of the timing of any glutamate changes, and investigating any differential effects of amisulpride versus olanzapine on glutamate changes. • changes in Serious Adverse Event reporting were implemented: 1) pregnancy is no longer reported as SAE but rather an AE; 2) hospitalisation due to psychiatric exacerbation is reported only in the annual line listings, due to the high frequency of occurrence at this early stage of the illness and the fact that immediate reporting does not have added value.

• recalculation of power analyses for MRS. • following changes in the amisulpride SPC, a safety procedure was added: if female patients have a history of breast cancer, and/or a first degree relative with a (history of) breast cancer, prolactin levels should be assessed at the local lab, at visit 2 and visit 5. • a blood count assessment was added to the biomarker blood draws, in order to support epigenetic analyses.

The generic amisulpride used for the study thus far, was no longer commercially available, therefore a switch to another generic amisulpride was required.