Page 173 - Functional impairment and cues for rehabilitation of head and neck cancer patients -

INTRODUCTION
Chemoradiotherapy (CRT) is widely used in advanced stage head and neck cancer (HNC). Common side effects of radiotherapy are dysphagia and weight loss (1), and concomitant chemotherapy increases the prevalence of these toxicities (2). Malnourishment and/or dehydration resulting from mucositis, loss of taste, xerostomia, and impaired swallowing function can cause feeding tube dependency in a proportion of patients during treatment (3). However, numerous studies have shown that still a considerable proportion of patients maintain their oral intake during CRT (4, 5).
Reactive feeding tube (RFT) placement, placement of a feeding tube (i.e., nasogastric tube (NGT) or a percutaneous radiological gastrostomy (PRG)) reactive to excessive weight loss (> 5% over three months or > 10% over six months), dehydration or aspiration, has a role in decreasing the incidence of (long-term) functional problems. Maintaining oral intake, along with targeted preventive exercises, prevents non-use atrophy of the swallowing muscles (6-9). This is therefore standard of care in HNC patients treated with CRT in the Netherlands Cancer Institute. Frequent monitoring of patients’ oral intake is thereby mandatory to timely identify patients requiring a feeding tube to lower the risk of weight loss, dehydration and treatment interruption associated with RFT (10-13). On the other hand, prophylactic feeding tube placement may prevent this (14-16), but at the same time convicts all patients tube feeding, whereas this would be unnecessary in a substantial proportion.
Both protocols thus have advantages and disadvantages and it would be beneficial if one could predict whether a reactive or prophylactic approach would be most appropriate for a given patient (i.e., personalized medicine) (17). Predictive factors for tube placement and (prolonged) dependency have been identified (16, 18-27). These factors include radiotherapy variables, tumor and nodal stage, and weight loss prior treatment. However, a clinically applicable prediction model to select HNC patients treated with CRT for proactive tube feeding in high risk patients is, to our knowledge, still lacking.
Therefore, we hypothesized that clinical decision-making on proactive tube placement could be aided by a prediction model based on the known predictive factors. The model should enable accurate identification of patients at risk of prolonged (> 90 days) tube dependency during primary CRT. This would allow for a personalized strategy regarding proactive/reactive tube placement, feeding and supportive care (e.g., swallowing exercises).
7
Prediction model for tube feeding during CRT
171