It was explained why it was acceptable to start the trial in the paediatric population before the marketing authorisation for linagliptin was granted by the US Food and Drug Administration (FDA) and European Medicines Agency (EMA). A process was defined and added to the section on randomisation to ensure pre-specified proportions of treatment groups within strata for patients completing the 12-week randomised treatment period. This process was not implemented and therefore deleted from the Clinical trial protocol (CTP) with Amendment 2. Since Visit 3 was a visit generally performed by telephone, there was no possibility to check treatment compliance by pill count at this visit. It was clarified that there will be no treatment compliance check at Visit 3. It was made clear that no study medication was to be given to patients at Visit 6, since this visit was the end of treatment (EoT) visit. Laboratory parameters were added to the definition of pancreatitis. Instructions for the follow-up of patients with pancreatitis including arrangements for follow-up blood samples to be collected for repeat analysis of lipase were added. The substance to be given to the patient to perform the Boost Challenge was changed to a beverage composed of proteins, lipids and carbohydrates to limit the carbohydrates patient’s intake. Furthermore, it contained some administrative changes, additions to achieve consistency between the different sections in the CTP, further clarifications, and an update of the number of patients treated and of the summary of Adverse Events (AEs/SAEs) in the CTP introduction section to be consistent with the latest version of the Investigator Brochure.

To follow a request by the FDA, the assumptions for treatment effect as basis for the sample size determination were revised. Sample size was increased to 39 randomised patients in each treatment group leading to a total of 117 randomised patients. A para on the constant surveillance for drug-induced liver injury (DILI) was added to benefit-risk assessment section. A detailed definition of hepatic injury based on alterations in liver laboratory parameters was inserted in the section on the definition of protocol-specified significant AEs. Procedures to follow-up patients with renal events, skin lesions, or pancreatitis were described. Associated laboratory values defining such events were specified. Precise action regarding study drug was provided. A section on the monitoring for allergic reactions and the instructions to patients was added. Section on AEs to be considered as always serious was added to comply with new Boehringer Ingelheim (BI) safety guidelines. Some sections were rephrased to be in accordance with a new BI standards. A body mass index (BMI)>50th percentile for age and sex was added as inclusion criterion 7. For inclusion criteria 4 and 5, it was clarified that lab samples for GAD, ICA and C-peptide were to be taken prior to the randomisation visit but results were to be checked at Visit 2 prior to randomisation. Upper limit for lipase and amylase values was added to the definition of clinically significant hepatic disease in exclusion criterion 8. A history of pancreatitis and hereditary angioedema or of serious hypersensitivity reactions was introduced as exclusion criterion 9. Process to ensure pre-specified proportions of treatment groups within strata for patients completing the 12-week period introduced by Amendment 1 was not implemented and so deleted from the CTP with Amendment 2. It clarified that in the primary analysis the Last Observation Carried Forward (LOCF) approach will be used. An additional sensitivity analysis was introduced.

As agreed with PDCO, patients treated with metformin could be included in addition to drug naive patients with insufficient glycaemic control. All sections of the CTP affected by this change were revised. The sample size was not modified because of the change. ‘Background therapy was added as stratification factor. ‘Background therapy’ and ‘background therapy by treatment interaction’ were included in the statistical model as fixed classification effects. Metformin was regarded as a NIMP thus the investigator was to assess the causal relationship of the SAE to metformin. Number of patients to be included in the PK rich sampling was increased from 12 to 13 patients per treatment arm due to early termination there were 12 pateint with rich sample. The term ‘protocol-defined significant adverse event’ was replaced by ‘adverse event of special interest’. The number of planned trial centres was increased to around 50 and the following statement was removed: ‘At least 2 patients should be randomised at each trial centre. Investigators who fail to screen at least one patient in the first 6 months of the trial may be excluded from further participation’. It was clarified that the results for Islet cell antigen (ICA) and Glutamic acid decarboxylase (GAD) antibodies had to be checked prior to randomisation and when auto-antibodies measurement were performed. For inclusion criterion 1 the relevant visit was changed from the screening visit to the randomisation visit. Contraindications to metformin according to the local label and a medical history of cancer and/or treatment for cancer within the last 5 years were added. Information based on the annual update of the IB was incorporated into the CTP. It was clarified that patients with intermediate PK sampling at Visit 6 had be dosed at Visit 6. Rescue medication was allowed after the randomisation of the patient to allow initiation of rescue medication between Visit 6 and 7. The wording regarding follow-up period was updated.

Pancreatitis was added as a reason to stop the study medication. This change was implemented immediately to eliminate hazard.

As agreed with the PDCO and the FDA, inclusion criterion 3 was modified and patients on stable basal insulin were allowed into the trial. Basal insulin was added as background therapy permitted during the trial. ‘Insulin’ and ‘metformin in combination with insulin’ were added as levels of the stratification factor background therapy. It was defined that for patients treated with insulin, an increase in insulin total daily dose by more than 10% from the prescribed baseline total daily dose for 7 consecutive days or more is considered as initiation of rescue therapy. Patients on short-acting insulin or patients who had received short-acting insulin for more than 3 days within the 3 months prior to randomisation were excluded from the trial (exclusion criterion 3). All sections in the CTP affected by these changes including the study title and the benefit-risk assessment section were revised. The HbA1c range in inclusion criterion 3 was broadened. A rationale was added why the sample size for the final analysis was not changed although the HbA1c inclusion range was broadened. A DPP-4 interim analysis to be performed by the independent DSMB was introduced. In the final analysis DPP-4 inhibition at trough at steady state was introduced as a key secondary endpoint. Relative efficacy response was added as a further efficacy endpoint. PK endpoints were classified as further endpoints. The primary ANCOVA analysis as planned in the original CTP was replaced by an MMRM approach. The assignment of patients to treatment groups was corrected for the FAS. The approach to handle missing data was changed from LOCF to OC. The definitions for AESIs were further specified and detailed. An adjudication committee for pancreatic events was set up. Procedures for glucometer measurements and glucose monitoring were clarified for the investigational sites. Exclusion criteria 1 and 8 were reworded. It was specified that the trial medication had to be taken after the last PK blood sampling.

The following changes were performed to be in line with BI standards and applicable SOPs, the latest version of the IB, and the project requirements: Thyroid neoplasm and thyroid cancer, pancreatic cancer, and cardiac failure were added to the list of AESIs. Further criteria for hepatic events were added to consider them as AESIs. The wording regarding the SAE declaration for events occurring during the screening period was changed. It was clarified what was expected for the assessment of the causal relationship of an SAE and what was expected in terms of follow-up of SAEs. To comply with the both, the PIP agreed with the Paediatric Committee of the European Medicines Agency (PDCO) and the post-marketing requirements and commitments agreed with the FDA, the timing of the DPP-4 interim analysis was modified. It was defined that the interim analysis was to be performed as soon as the PPSDPP4 included at least 8 patients from the linagliptin 1 mg group and 8 patients from the linagliptin 5 mg group. Power calculations based on 8 evaluable patients per linagliptin arm were added to the section on the determination of the sample size for the interim analysis. The focused interim PK/PD analyses were introduced to be performed in case the DPP-4 interim analysis did not lead to premature termination of the trial. In the original CTP it was stated that the popPK analyses are described in a separate report. However, it was decided to include the results of the popPK analyses into the Clinical trial report (CTR). It was clarified that therefore it was not necessary to develop a separate popPK analysis plan.