A multicentre non-blinded randomised intervention trial with a parallel cluster design. The multicentre study will be performed at 12 hospitals in the Netherlands. A parallel cluster design per hospital was selected to prevent contact between participants in the same hospital who are randomised into different study arms. Such contact may result in bias as it may allow for patients in the control arm to be informed about elements of taste and smell training.

To examine the effect of at-home taste and smell training versus standard care on taste function and other outcome parameters, measurements will take place at baseline (before the training starts), and after 12 weeks. The intervention will take place at home. The control group will receive usual care. Questionnaires will be filled in online at home, while taste and smell tests and saliva collection will be conducted either at home or in the hospital during regular visits. Patients in both study arms are contacted by their dietitian every 3 weeks.

Inclusion Criteria:

• Age ≥18 and ≤70 years
• Intention of being treated with tyrosine kinase inhibitors for next 12 weeks.
• Experiencing taste alterations after starting treatment with tyrosine kinase inhibitors
• Consuming solid foods and drinks is possible
• ≤50% of recommended daily intake in kcal consisting of oral nutritional supplements
• Oral mucositis grade ≤2 (Common Terminology Criteria for Adverse Events, version 5.0)
• Ability to comply with all protocol-required actions
• Written informed consent

Exclusion Criteria:

• Pregnancy
• History of taste, smell, or saliva production dysfunction before treatment with tyrosine kinase inhibitors
• Previous or current radiotherapy of head and neck region
• Enteral feeding through tube or parenteral feeding
• Participation in another clinical trial aimed at preventing or treating taste and/or smell alterations
• Chronic (>1 month) high dose corticosteroids (>10 mg prednisone/day or equivalent)