• The main purpose of this amendment was to allow for the inclusion of subjects with a platelet count ≤ 50 x 10^9/L, with a history of bleeding (rather than a history of clinically significant bleeding). Revising this inclusion criterion was intended to enhance enrollment. • To ensure consistency among romiplostim MDS study protocols, the definition of disease progression was revised. Previously, disease progression was defined as an IPSS low or INT-1 risk score that increased to an intermediate-2 or higher risk. With this amendment, disease progression was newly defined as transformation to AML. • Other administrative and typographical errors were corrected including the revision of the informed consent form (ICF) (Platelet Function Addendum) to accurately reflect the requirements of the protocol regarding the times of sampling.

• The original protocol inadvertently excluded subjects with a normal total bilirubin. The main purpose of this amendment was to allow for the inclusion of subjects with a normal total bilirubin (total bilirubin ≤ 2.0 × the laboratory normal range). • The eligibility criteria pertaining to written informed consent (IC) was revised to allow a subject’s legally acceptable representative to sign the ICF. If legally acceptable representatives were not permitted to sign the ICF per local law, the ICF and patient information sheet were to be locally revised appropriately. • Data captured in the long term follow up (LTFU) were clarified to inform that subjects were requested to allow Sponsor continued access to medical records, so that information related to subjects’ health condition may be obtained during the LTFU period. • Bone marrow biopsy with aspirate and cytogenetics could be sent to a central laboratory if a site’s local laboratory cytogenetics certification does not meet ICH GCP requirements. • Clarification of dose adjustment rules: for subjects receiving 250 μg romiplostim or volume-matched placebo weekly, the dose of IP would be withheld any time the platelet count was > 450 x 10^9/L and then reinitiated once every 2 weeks at the next scheduled visit the platelet count was < 200 x 10^9/L; subjects who had a previous dose reduction could increase the dose of IP if their platelet count was < 50 x 10^9/L for 3 consecutive weeks, beginning on the fourth week after the platelet count first fell to < 50 x 10^9/L. • To ensure consistency among romiplostim MDS studies, if the subject’s marrow was inaspirable, a cytogenetic analysis was performed on a peripheral blood sample. • Other administrative and typographical errors were corrected including the revision of the ICF to accurately reflect the purpose of the study, regions in which the study was conducted and to remove the reference to exploratory research and discontinuation of medications for ITP.

• The definition of disease progression and transformation to AML was clarified. Disease progression to AML was now to be assessed per WHO guidelines requiring confirmation of marrow or peripheral blast cells ≥ 20% that exists in the absence of romiplostim (4 weeks off dosing) and other hematopoietic growth factors (2 weeks off dosing). A pathology report confirming other leukemias such as chloroma (granulocytic sarcoma, myeloid sarcoma) or leukemia cutis would also constitute disease progression to AML. • For the purpose of this study, the definition of transformation to AML was expanded to also include any subject that clinically required the initiation of antileukemic treatment based on physician judgment and clinical diagnosis. • Due to the new requirement for subjects to be removed from IP for 4 weeks to study the bone marrow, the interim washout period duration was revised from 2 to 4 weeks to 4 weeks. • The timing of entrance criteria was further clarified to provide investigative sites with more detailed instruction of when disease related assessments should take place relative to the screening period. • The platelet function testing sub-study was removed from the protocol as a result of the primary investigator who was leading the sub-study resigned from the investigative site and declined further study participation. • The consent form was removed as a protocol appendix and was now to be provided separately to investigative sites. The consent form was also updated to reflect changes in study procedures and the most current safety information.