Amendment No. 1 • The CT scan imaging technique only was allowed at inclusion to confirm diagnosis and at 48 hours to evaluate the haematoma volume. The same technique had to be used at inclusion and 48 hours. If CT scan and MRI are commonly used to confirm diagnosis, CT scan is preferred at 48 hours for control imaging. • Modification of the exclusion criteria: Some exclusion criteria needed an explanation for a better understanding. General seizure has been deleted as they are frequently associated with intracranial haemorrhage. • Modification of Octaplex administration: According to Scientific Committee members request (Pr P. Coriat, principal investigator and Pr B. Riou, investigator, Dr B. Vigué and Pr. C Négrier), an additional Octaplex dose could be administered at 10 ± 5 minutes and at 6 hours after the first Octaplex infusion, following precise recommendation and not at the discretion of the investigator. The recommendation was 1.5 as target INR for the first rescue dose. The table has been simplified to facilitate the additional dose calculation by the investigators. • Modification of Octaplex infusion rate: The context of very high emergency implies to shorten all the delays in patient care including treatment administration. • Update of information on Octaplex storage (new Octaplex variation approved by reference member state via Mutual Recognition Procedure in 30-Nov-2007). • New laboratory parameters: Digital glucose was performed instead of being analysed in the laboratory. Additional laboratory tests were done (liver function, haematology) in order to follow the overall safety data.

Amendment No. 2 • Modification of the informed consent procedure: The previous procedure was not adapted to the emergency situation. According to the law, in emergency situations, if prior informed consent of the patient is not possible and the patient’s legally acceptable representative is not available, the patient can be enrolled without prior informed consent. The informed consent procedure was modified in consequence: if the patient’s legally acceptable representative was not available, the patient could be included by the investigator in presence of a witness independent from the investigator and the Sponsor. • Because of the change of the informed consent procedure, the patient concomitant disability could not be assessed by the physician if the patient’s legally acceptable representative was absent. This exclusion criterion was thus suppressed. • Population of the study better defined: The type of intracranial haemorrhage has been precisely defined for a better understanding.

Amendment No. 3 • Modification of biological analyses: Some of the centralised analyses (fibrin degradation products and thrombin-prothrombin complexes) have been cancelled with the agreement of the central laboratory, in order to lighten plasma tubes preparation for local laboratories. Moreover, central laboratory analyses were only to be performed in the centres that possessed laboratory facilities (adequate devices and personnel during night and day) to prepare plasma tubes. The laboratory parameters analysis was performed locally or centrally upon local laboratory facilities. • ECG surveillance: ECG is monitored continuously during the whole stay in the investigator’s ward, thus the surveillance at 48 hours was added. • Modification of the determination of sample size based on the results obtained from previous studies with Octaplex. • Precision of the anticoagulant therapy: The type of anticoagulant therapy has been added. Only patients under vitamin K antagonist therapy were admitted. • Precision of the type of subdural haematoma: Only patients with acute subdural haematoma were admitted. This criterion was assessed by CT scan.

Amendment No. 5 • Change of Sponsor: Octapharma France SAS (Boulogne-Billancourt, France) replaced Octapharma AG (Lachen, Switzerland). • Modification of the sample labels for investigational medicinal products: Addition of FIX content in Octaplex and change of Sponsor. • Modification of the study duration: Due to the late clinical study start, the study duration was prolonged accordingly.

Amendment No. 6 • In the entire text, thrombin generation assay (TGA) analysis was suppressed. The results of TGA analysis in the first patients included in the study were not interpretable due to a bad plasma preparation. As this test was still experimental, it was decided to stop this analysis in agreement with Pr Negrier (Head of Central Laboratory). • In the entire text, a time window of 15 minutes before and 15 minutes after the fixed time points 1 hour, 3 hours, 6 hours and 24 hours was added. As this study was carried out in a context of emergency, it appeared that greater flexibility in the times of blood sample collection was appropriate to avoid protocol deviations and give more comfort to the investigator during the care of patient. • In the entire text, the National Institute of Health (NIH) stroke scale was suppressed. As the NIH stroke scale was not usually used by most study investigators and requires a specific training, it was decided to not perform this scale anymore. Alternatively, the neurological worsening was assessed by a routine neurological examination. • An English translation, certified by a sworn translator, of the Informed consent and the Patient information approved by the IEC on 01-Jul-2008 was added to Appendix 2 of the protocol. This translation was supplied, for information, to two patients speaking English.

Amendment No. 7 • Modification of the study duration: Due to a recruitment rate below what was expected, the duration of the study had to be extended accordingly.

Amendment No. 8 • Modification of the study duration: Due to a recruitment rate below what was expected, the duration of the study had to be extended accordingly.