Pembrolizumab (41) – Keytruda®

• Clinical trials
  • The KEYNOTE 689 trial is an ongoing, multicentre, open-label, randomised, controlled Phase III trial investigating pembrolizumab as neoadjuvant therapy and as postoperative adjuvant therapy in combination with radiotherapy, with or without concomitant cisplatintherapy followed by pembrolizumab monotherapy against postoperative adjuvant radiotherapy with or without concomitant cisplatin therapy in adult patients with resectable locally advanced squamous cell carcinoma of the head and neck.

a) Adults with resectable locally advanced squamous cell carcinoma of the head and neck whose tumours express PD-L1 with a CPS ≥ 1 and who are suitable for cisplatin-based therapy; neoadjuvant and adjuvant therapy

• Hint for a minor additional benefit
• Overall, the limitations arising from the available data do not call into question the advantage in terms of overall survival. Overall, the G-BA concludes that pembrolizumab, as monotherapy for neoadjuvant treatment and subsequently for adjuvant treatment in combination with radiotherapy, with or without concomitant cisplatintherapy, and subsequently as monotherapy, offers a minor additional benefit compared with the corresponding treatment regimen without pembrolizumab.
• mortality
  • Overall survival is a secondary endpoint in the KEYNOTE 689 study and is defined as the time from randomisation to death from any cause.
  • For the endpoint of overall survival, a statistically significant difference in favour of pembrolizumab was observed; the extent of this difference is assessed as a relevant improvement, though not exceeding a minor level.
• Morbidity – Failure of curative treatment
  • Patients in this therapeutic indication are treated with a curative therapeutic approach. The failure of a curative therapeutic approach is, in principle, relevant to patients.
  • Due to the uncertainties described regarding qualifying events and the lack of data on patients who did not commence treatment after randomisation, the results for the EFS endpoint according to BICR are not suitable and are not used for the assessment.
  • Taken together, the analyses presented for the post-hoc adapted EFS, as well as the associated sensitivity analyses 1 and 3, cannot be cannot be meaningfully interpreted and are therefore not included in the assessment.
  • By contrast, sensitivity analysis 2 is considered to be interpretable in a meaningful way with regard to the detailed assumptions.
  • Owing to the uncertainties described in the various EFS analyses presented above, the EFS results are not suitable for demonstrating the failure of the curative treatment approach with sufficient certainty and are not included in the assessment. Consequently, there are no suitable data available regarding the failure of the curative treatment approach.
• Morbidity – Symptoms
  • Symptoms were assessed in the KEYNOTE 689 study using the symptom scales of the EORTC QLQ-C30 and EORTC QLQ-H&N35 questionnaires.
  • Overall, the results of the patient-reported endpoints on symptoms are not suitable for assessing the additional benefit due to differences in data collection between the study arms, particularly at baseline.
• Morbidity – Health status
  • Health status is assessed in the KEYNOTE 689 study using the visual analogue scale (VAS) of the EQ-5D.
  • Due to the uncertainties described in the section on symptoms, the results of the patient-reported endpoints on health status are not suitable and are not used for the assessment.
• morbidity
  • The results for both the EFS endpoint and the patient-reported endpoints on symptoms and health status are not interpretable and are not used for the assessment. Consequently, no assessable data are available for the morbidity category as a whole.
• Health-related quality of life
  • Quality of life was assessed in the KEYNOTE 689 study using the functional scales of the EORTC QLQ-C30 and EORTC QLQ-H&N35 questionnaires.
  • Due to the uncertainties described in the section on symptoms, the results of the patient-reported endpoints on health-related quality of life are not suitable.
• Side effects – Total adverse events (AEs)
  • In the KEYNOTE 689 study, an AE occurred in almost all patients in both study arms. The results are presented here for supplementary information only.
• Overall assessment
  • For the overall survival endpoint, there is a statistically significant difference in favour of pembrolizumab, with an extent that is assessed as a relevant improvement, though not exceeding a minor degree.
  • In the morbidity endpoint category, no suitable data are available for the endpoint ‘symptoms’, as assessed using the EORTC QLQ-C30, the EORTC QLQ-H&N35 and health status as assessed using the EQ-5D VAS, due to major differences in the assessment methods between the study arms, particularly at baseline.
  • The results for the event-free survival (EFS) endpoint are not suitable for this analysis to demonstrate with sufficient certainty the failure of the curative treatment approach.
  • Consequently, no evaluable data are available for the morbidity category as a whole.
  • With regard to health-related quality of life (assessed using the EORTC QLQ-C30 and -H&N35), no suitable data are available due to major differences in the assessment methods between the study arms, particularly at baseline.
  • For the endpoints relating to side effects, the results from the respective time-to-event analyses for the endpoints ‘serious AEs’ and ‘discontinuation due to AEs’ cannot be interpreted because the proportional-hazards assumption was not met in each case. The statistically significant difference in favour of pembrolizumab from the time-to-event analysis for the endpoint of severe AEs is based solely on a later onset of severe AEs; therefore, in the overall assessment, no advantage for pembrolizumab can be inferred on the basis of this result alone. In the category of side effects, neither an advantage nor a disadvantage is identified overall.
  • In the overall assessment, there is an advantage in terms of overall survival, the extent of which is assessed as a relevant improvement. No suitable data are available regarding the failure of the curative treatment approach. Similarly, no suitable data are available for other patient-relevant endpoints relating to morbidity or health-related quality of life. With regard to side effects, the available data can only be assessed to a limited extent. However, in the overall assessment, these limitations do not call into question the advantage in overall survival.

b) Adults with resectable locally advanced squamous cell carcinoma of the head and neck whose tumours express PD-L1 with a CPS ≥ 1 and who are not suitable for cisplatin-based therapy; neoadjuvant and adjuvant therapy

• The additional benefit is not proven.
• No data are available to enable an assessment of the additional benefit. In its dossier, the pharmaceutical manufacturer provides data only for the patient population of patients who are suitable for cisplatin-based therapy (patient population a).

Courtesy translation only, please refer to the German original.