Guidelines for nutrition in adults with head and neck cancer: The American Society for Parenteral and Enteral Nutrition.

PURPOSE
Head and neck cancer is defined as tumors or cancerous cells arising from the mucosa of the oral cavity, lips, larynx, pharynx, cervical esophagus, nose, sinuses, skin, and salivary glands.
1
,
2
Globally, in 2021, head and neck cancers were the seventh most common type of cancer worldwide, accounting for 792,280 new cancer diagnoses and 424,066 deaths.
3
Patients with head and neck cancer present special nutrition‐related challenges and are at high risk for malnutrition because of difficulties chewing and swallowing, loss of appetite, and other symptoms caused by tumor location and symptomatic response to treatment.
4
For the purpose of this guideline, malnutrition refers to undernutrition. Guidance is crucial to help interdisciplinary teams administering nutrition care to this key population.
Evidence‐based guidelines for the nutrition management of adult patients with head and neck cancer were first published by the Clinical Oncology Society of Australia (COSA) in 2011 and have since been updated and maintained as a living guideline.
5
The COSA guidelines are a comprehensive synthesis of the literature addressing the continuum of care from diagnosis through survivorship and palliative care and encompassing all study designs from randomized control trials (RCTs) to cross‐sectional designs, as well as studies that included other high nutrition risk cancer populations. This American Society for Parenteral and Enteral Nutrition (ASPEN) guideline departs from the COSA effort by restricting to study designs capable of providing causal inference (RCTs and quasi‐experimental designs) and by tightening population restrictions to exclusively include head and neck cancer populations to form recommendations under these strict criteria. Additionally, this guideline is expanded to include new questions reflecting areas of current clinical interest and multidisciplinary team care.
This clinical guideline's objective is to provide practice guidance for all clinicians on the nutrition care of adult patients with head and neck cancer via a systematic review and grading of the scientific literature, and a modified Delphi approach to facilitate expert consensus. The complete guideline recommendations are presented in Table 1.
Recommendations in this guideline do not constitute medical or other professional advice and should not be taken as such. To the extent that the information published herein may be used to assist in the care of patients, the primary component of quality medical care is the result of the professional judgment of the healthcare professionals providing care. The information presented here is not a substitute or replacement for the exercise of professional judgment by healthcare professionals; rather, it is intended to supplement professional training and judgment. Circumstances and patient specifics in clinical settings may require actions different from those recommended in this document; in those cases, the judgment of the treating professionals should prevail. Use of this information does not in any way guarantee any specific benefit in outcome or survival. This paper was approved by the ASPEN Board of Directors.

Target Population: The target population is adult patients ≥16 years with head and neck cancer receiving any treatment modality (eg, chemotherapy, radiation, surgery, or a combination of therapies).

Target Audience: This clinical guideline is intended for use by any medical health professional involved in the nutrition care of adult patients with head and neck cancer (eg, dietitians, nurses, pharmacists, physicians, and speech‐language pathologists).

METHODS

The panels
The guideline was comprised of four panels: a clinical panel, a bias panel, a guidelines relief panel, and an external validation panel. The clinical panel was chaired by Nicole Kiss, PhD, APD. The clinical panel included dietitians, an epidemiologist/methodologist, a pharmacist, a radiation oncologist, and a speech pathologist. A nurse and an ear, nose, and throat physician were included in the early phases of the project, and although they were not included in the later activities of the clinical panel, they participated in the external validation panel to generate expert consensus and as independent journal reviewers before publication.
The role of the clinical panel was to create the guideline key questions using the Population, Intervention, Comparison, Outcome, and Timeframe (PICOT) framework, assist in the creation of search terms, perform the screening and data extraction, create the initial recommendations, and edit the final manuscript. The bias panel was composed of doctoral‐level researchers with a background in nutrition. They were trained and overseen by the epidemiologist/methodologist to perform bias analysis of the included studies. The guidelines relief panel was composed of an interdisciplinary team of dietitians, nurses, pharmacists, and physicians, and their role was to assist in screening and data extraction. Finally, the external validation panel was an interdisciplinary group of dietitians, a nurse, pharmacists, and an ear, nose, and throat physician (Figure 1). Their role was to review the recommendations approved by the clinical panel and provide external validation to the wording of the recommendations via a blinded vote.

The protocol
An “a priori” protocol for this guideline was published and made available for public comment for 6 weeks.
6
Emails were sent to members of ASPEN to solicit feedback from clinicians and researchers. Two patients with head and neck cancer were also solicited for feedback to ensure all key stakeholders had an equal opportunity to provide input. The protocol was adjusted based on the feedback received.

Study inclusion/exclusion criteria
To be included, an article had to be a study of patients aged ≥16 years with head and neck cancer, published after January 1, 2001, whose primary or secondary objective was directly relevant to at least one of our questions and outcomes of interest. To include European studies that use a younger cutoff to define adults, the age of 16 years rather than 18 years was chosen to define an adult. For each question, we restricted to the study design most able to answer that specific question. The decision‐making process for this is outlined in the published protocol for this guideline.
6
The result was that studies for all guideline questions were restricted to RCTs and quasi‐experimental designs. Study designs that were vulnerable to known unmanaged confounders were excluded. When such confounding is present, the resulting statistic is altered in magnitude and/or direction, losing its ability to meaningfully reflect the relationship under study.

The search strategy
A comprehensive search was performed in the PubMed database from 2001 to July 27, 2025. The decision to restrict the search to studies published after 2001 reflects modern radiation techniques, treatment modalities, and protocols. The PubMed search strategy may be viewed in Figure 2. Analogous search strategies were composed and run for the EMBASE, CINAHL, and Cochrane Central databases. The results of these searches were uploaded into Covidence for data collection.

Data collection and presentation
All data were screened and extracted in duplicate in Covidence by researchers blinded to each other's decisions, except for the catch‐up search, which collects citations at the end of the guidelines process before journal submission. This was screened in duplicate but extracted directly by the guideline methodologist. Practice tests were performed before data collection to improve internal agreement.
6
Any discrepancies among panel members’ study selections were adjudicated by a third party. Data were presented as narrative tables to account for the considerable heterogeneity in study design, outcome types, and timepoints explored. Forest plots were created wherever three or more studies were highly similar in population, intervention, comparator, outcome, and time point of measurement.

Bias analysis
Bias of included studies was assessed in duplicate by the bias panel and checked by the methodologist. Each study received two bias assessments to account for differences in blinding bias between outcomes. The Cochrane Risk of Bias 2 tool
7
was used to assess RCTs and the Cochrane Risk of Bias in Non‐Randomized Interventions tool
8
was used to assess quasi‐experimental designs. Bias assessments are available in the Supplementary Materials.

Statistical analysis
Wherever three or more studies were able to be meaningfully conflated, a summary statistic was run using a random‐effects model with a Knapp Hartung adjustment. Small study numbers (<10) tend to destabilize a summary statistic leading to statistical false positive results. The Knapp Hartung adjustment adjusts the error to account for this issue.
9
,
10
Publication bias was assessed through funnel plots and Egger tests wherever ≥10 studies are available for conflation into a forest plot.

Recommendations
The Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) tool was used to facilitate transparency in the creation of the guideline recommendations. The clinical panel assembled multiple times to discuss the recommendations for every key question and wrote the recommendations together. The GRADE process generates two major outputs: it generates a certainty level that the current evidence can answer the key questions, which ranges from “very low” to “high,” and it also generates a strength of the recommendation. For this guideline, where certainty existed that the potential benefits of following the recommendation outweighed any potential harms, the strength was listed as “strong.” Where we were less certain but still felt that the benefits outweighed the harms, the strength is listed as “weak” (Figure 3). New terminologies have recently been introduced by the GRADE group that have replaced the term “weak” with the term “conditional.” We have chosen not to adopt that alteration, but the meaning of the two terms may be considered equivalent for this guideline. The separation of certainty of evidence from recommendation strength makes it is possible to have a strong recommendation despite very low certainty of evidence if the clinical panel is certain that the potential benefits outweigh the potential harms.
A modified Delphi approach to generate expert consensus was performed, permitting a blinded vote by the clinical panel members. This continued until ≥70% agreement was met for each recommendation. Then, the recommendations were sent to the external validation panel, which repeated the modified Delphi. This was a larger panel, and so an 80% agreement was required. Whenever the requisite percent agreement was not met, the chair modified the recommendation, and the vote was rerun for that recommendation until sufficient agreement was reached.

RESULTS
The search strategies yielded 10,581 citations. Of these, 2240 were duplicates and were removed. After review, 8249 were removed for not meeting the inclusion criteria. This left 92 studies for data extraction and inclusion in this guideline (Figure 4).

Question 1a
In adult patients with head and neck cancer receiving chemoradiation or radiation, does earlier enteral nutrition (EN) vs later EN change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicities, or unplanned hospital admission?

Recommendation: In adults with head and neck cancer planned for or receiving radiotherapy with or without chemotherapy or other systemic therapy, we suggest initiation of enteral feeding when there is clinical evidence that nutrition intake or status is compromised, despite other strategies (oral nutrition supplements or food fortification) having been attempted. This should be individualized to oral nutrition intake, symptom burden, and nutrition status while considering individual treatment plans and clinical, psychosocial, and socioeconomic status.

Certainty of Evidence: Moderate

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Six studies met the inclusion criteria for this PICOT question.
11
,
12
,
13
,
14
,
15
,
16
Four were RCTs and two were secondary analyses of RCTs examining different outcomes. Three studies demonstrated a significant benefit in the intervention vs the control. One study reported improved dysphagia at 1 year,
15
another reported improved quality of life,
13
and a third reported decreased length of hospital stay.
16
The remainder found no significant differences between the two groups for any outcome (Tables 2 and 3). No studies favored the control. On average, patients in the intervention group had a 3% reduction in 1‐year mortality compared with the control, but this was not statistically significant (Risk Difference (RD) = −0.03, 95% CI = −0.1 to 0.05; P = 0.35) (Figure 5). Brown et al
11
and its secondary analysis were found to be low risk of bias, whereas the other five studies were rated as having some concerns for bias (Tables S1 and S2).

Rationale and discussion
Patients with head and neck cancer have a high prevalence of malnutrition and muscle loss, which has been associated with worse outcomes.
17
,
18
,
19
Proactive nutrition intervention may minimize or prevent the decline in nutrition status for these patients.
20
Early commencement of EN may, therefore, be an appropriate strategy to meet energy and protein requirements while preventing nutrition decline. The current recommendation is based on this potential benefit, along with some evidence of benefit from the literature, and low risk of harm and aligns with a similar recommendation in the COSA guideline.
5

The timing and amount of nutrition received varied considerably between the studies. Two studies placed their percutaneous endoscopic gastrostomy (PEG) tubes prophylactically in both groups and commenced feeding the intervention group immediately, whereas the control group did not commence until weight loss or reduced intake were noted. The third study restricted PEG placement to the intervention group. Feeding in the intervention group commenced only if inadequate intake was noted, whereas the control group was left to standard care. It is important to note that, in all studies, adherence to EN was poor and patients were unable to achieve their recommended energy intake. So, aiming to commence early EN does not necessarily equate to achieving energy goals, and the one study that tracked energy intake reported no differences between the groups.
14

Future research
The true effect of achieving the interventions in the studies for this PICOT question is unknown because of poor adherence to the interventions and variability in implementation. Randomized trials of interventions using structured protocols to achieve energy and protein requirements are needed, with the ability to tailor the intervention to individual patients. These protocols should use strategies that are common in clinical practice (eg, food fortification or modification initially) and should describe criteria for when oral nutrition supplements or EN should be initiated. They should assess the effect on nutrition status, treatment completion/interruptions, treatment toxicities, unplanned hospital admissions, and survival. Similar interventions in nonrandomized studies have shown benefit for some of these outcomes.
21
,
22
Codesigning the protocol with patients and other stakeholders is suggested to ensure patient and healthcare provider perspectives and the barriers to early EN are considered. This is consistent with suggestions in a recently published roadmap for nutrition research on head and neck cancer care.
23

Question 1b
In adult patients with head and neck cancer, does longer postoperative nutrition (EN or oral nutrition supplements) vs shorter duration of nutrition support change progression‐free survival, overall survival, nutrition intake, time to transition to full oral diet, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, length of stay, surgical complications, or hospital readmissions?

Recommendation: In adults with head and neck cancer, we suggest early commencement (within 24 h) of postoperative nutrition intake (EN, oral, or otherwise) to meet estimated nutrition requirements. Early initiation of oral intake should occur in consultation with the surgical team, dietitian, and speech pathologist. To meet nutrition requirements, oral intake should be increased gradually, and supplemental enteral feeding should be maintained until sufficient oral intake is established. Early initiation of oral intake may be inappropriate in certain situations, making consultation with surgical, speech, and nutrition teams critical.

Certainty of Evidence: Very low

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Two studies met the inclusion criteria for this question.
24
,
25
No studies found significant differences between the intervention and control for any outcome (Tables 4 and 5). One study reported a longer length of hospital stay in the control group, but did not provide statistics.
24
No studies favored the control. One study was found to have “some concerns” for bias, and the other, a quasi‐experimental design, was found to be a “serious” risk for bias (Tables S3–S5).

Rationale and discussion
Patients with head and neck cancer are at risk of malnutrition and muscle loss preoperatively and postoperatively.
26
Malnutrition and low muscle mass are associated with adverse postoperative outcomes, such as surgical complications, delayed wound healing, longer hospital stay, and poorer survival.
19
,
26
The timing of initiating nutrition support, including oral nutrition support, may be important in preventing decline in nutrition status and, therefore, adverse postoperative outcomes. The current recommendation is based on this potential benefit, along with no evidence of harm from early oral nutrition in the two included studies, and aligns with similar recommendations in the COSA guideline and other guidelines.
5
,
27

Whether or not commencing oral intake is appropriate depends on factors such as the extent of surgical resection and/or reconstruction, severity of dysphagia, risk of aspiration, and other symptoms. Some patients will require long‐term EN and may be unable to tolerate oral intake. Therefore, the timing of the introduction of oral intake and the duration of EN should be determined collaboratively through consultation among the surgical team, speech and language pathologist, and dietitian.

Future research
No studies specifically addressed the optimal duration of postoperative EN or oral nutrition supplements. In clinical practice, these therapies may stop before postoperative discharge regardless of the patient's nutrition intake and/or nutrition status. Randomized trials are needed to compare continuation of EN or oral nutrition supplements until oral intake is adequate to meet energy and protein requirements to standard care and assess the effect on nutrition status, time to transition to full oral diet, surgical complications, postoperative length of stay, hospital readmissions, and survival. Such studies should stratify participants by the site and extent of surgical resection and/or reconstruction.

Question 1c
In adult patients with head and neck cancer, does increasing the frequency of dietetic intervention vs standard care during chemoradiation or radiation and up to 3 months posttreatment change progression‐free survival, overall survival, nutrition intake, time to transition to full oral diet, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicities, or unplanned hospital admissions?

Recommendation: In adults with head and neck cancer receiving radiotherapy with or without chemotherapy or other systemic therapy, we recommend weekly consultation with a dietitian during treatment and every 2 weeks for up to 6 weeks after treatment to maintain nutrition status and quality of life while preventing unplanned hospital admissions and early cessation of treatment.

Certainty of Evidence: High

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Thirteen studies met the inclusion criteria for this question.
28
,
29
,
30
,
31
,
32
,
33
,
34
,
35
,
36
,
37
,
38
,
39
,
40
Nine studies demonstrated a significant benefit in the intervention compared with the control.
28
,
29
,
33
,
34
,
36
,
37
,
38
,
39
,
40
Four studies reported lower rates of malnutrition,
28
,
29
,
34
,
36
seven found less weight loss,
28
,
33
,
36
,
37
,
38
,
39
,
40
two found fewer treatment interruptions,
28
,
33
five found improved quality of life,
28
,
34
,
37
,
38
,
39
four reported higher energy intake,
29
,
34
,
38
,
39
two reported higher protein intake,
34
,
38
one reported reduced incidence of symptoms at 3‐months,
34
and one reported fewer hospital admissions due to mucositis.
33
The remainder found no significant differences between the two groups. No studies favored the control. Two studies were rated as having “low” risk of bias. Seven had “some concerns” for bias. One RCT was rated at “high” risk for bias, and three quasi‐experimental designs were rated at “serious” risk for bias (Tables 6, 7, and S6–S9).

Rationale and discussion
Patients with head and neck cancer are at high risk of malnutrition and muscle loss during and following radiotherapy with or without chemotherapy.
17
Observational studies show that malnutrition and low muscle mass are associated with unplanned hospital admission, poorer quality of life, and reduced survival.
17
,
41
,
42
The frequency of dietitian intervention may be important to minimize this decline in nutrition status and prevent adverse outcomes. The current recommendation is based on consistent evidence of benefit from the literature and no evidence of harm other than the possible inconvenience of additional appointments and aligns with a similar recommendation in the COSA guideline.
5

The frequency of dietitian interventions varied throughout the studies, with the most common schedule being weekly during treatment and every second week for 6 weeks following treatment. Cancer services should consider how they can provide this level of nutrition care while minimizing appointment fatigue for patients. Colocated interdisciplinary clinics, such as combined dietitian and speech‐language therapy clinics, may be beneficial to streamline care. Cancer services with limited dietetic resources may need to evaluate how best to provide this care to patients receiving radiotherapy with or without chemotherapy while also ensuring equitable access to dietetic care for patients undergoing surgery, which may require prioritizing patients at the highest risk of malnutrition.

Future research
The current evidence is sufficient to conclude that increased frequency of dietetic interventions improves important clinical outcomes. Future research in this area should focus on how best to deliver this care equitably, efficiently, and sustainably in routine practice. This includes testing models of care, for example, telehealth‐enabled approaches, codesigned care pathways, workforce models that optimize specialist availability, and risk‐based triage tools to support prioritization and access. Implementation‐effectiveness studies should evaluate clinical outcomes such as treatment completion, unplanned hospitalization, nutrition status, quality of life, and long‐term survival alongside system‐level outcomes including economic evaluation, resource utilization, and value‐based care measures. There is also more to discover regarding the optimal timing of nutrition consultation, which would require RCTs.

Question 1d
In adult patients with head and neck cancer, does longer preoperative and postoperative intervention by a dietitian compared with shorter intervention duration change progression‐free survival, overall survival, nutrition intake, time to transition to full oral diet nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, length of stay, surgical complications, or hospital readmissions?

Recommendation: In adults with head and neck cancer, we recommend that the duration of dietitian intervention preoperatively and postoperatively should be individualized according to the patient's nutrition status, swallowing function, symptom burden, and clinical, psychosocial, and socioeconomic status. Intervention should continue while there remains a risk to nutrition status from nutrition‐impact symptoms, reduced dietary intake, continued indication for nutrition support, and/or other factors relating to nutrition risk. Patients undergoing certain surgical procedures, such as those with extensive surgical resections (eg, free flap reconstructions) or with dysphagia, may require dietetic support for a longer duration.

Certainty of Evidence: Very low/expert opinion

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
No studies met the inclusion criteria for this question.

Rationale and discussion
In patients treated with surgery for head and neck cancer, malnutrition and low muscle mass are associated with prolonged length of hospital stay, more surgical complications, delayed wound healing, and poorer survival.
19
,
26
The duration of nutrition intervention before and after surgery may be important to minimize or prevent the decline in nutrition status and the associated adverse outcomes. Although no literature was found to answer this question, the current recommendation is based on this potential benefit and no known harm other than the possible inconvenience of additional appointments. This confidence that the potential benefits outweigh the potential harms necessitated a strong recommendation.

Future research
RCTs examining the effect of longer durations of dietetic intervention before and after surgery are needed. Relevant outcomes of importance are nutrition status, hospital length of stay, surgical complications, unplanned readmission, quality of life, and survival. Studies should also stratify patients by nutrition status to account for the known adverse impact of malnutrition on these outcomes.

Question 2a
In adult patients with head and neck cancer receiving any treatment modality, does nutrition screening vs not screening impact progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion treatment interruptions, treatment toxicities, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer, we recommend that all patients be screened for malnutrition using a validated tool at their first presentation to the healthcare facility and regularly throughout treatment and recovery to facilitate timely referral for nutrition intervention. In certain patient subgroups who are at higher risk of malnutrition, screening may be bypassed if processes are established for automatic referral to a dietitian.

Certainty of Evidence: Very low/expert opinion

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
No studies met the inclusion criteria for this question.

Rationale and discussion
Malnutrition is prevalent in patients with head and neck cancer.
18
Early identification of malnutrition risk may facilitate prompt referral of at‐risk patients to a dietitian for comprehensive nutrition assessment and intervention. Therefore, screening for malnutrition risk may improve patient outcomes. Although no literature was found to answer this question, the current recommendation is based on this potential benefit and no known harm other than the possible inconvenience of participating in screening and aligns with a similar recommendation in the COSA guideline.
5

In health services where access to a dietitian is dependent on a referral, or where dietitian resources are limited, screening using validated tools is an appropriate strategy to triage patients at risk of malnutrition to a dietitian. However, some patients with head and neck cancer (eg, those planned for combined chemoradiotherapy) are already considered at high risk of malnutrition because of treatment‐related factors. These patients may not be identified as being at risk of malnutrition through malnutrition screening processes but are at high risk of becoming malnourished. For these patients, automatic referral to a dietitian may be more appropriate, where resources permit.

Future research
The value of screening as a triage tool is widely accepted. Randomized trials to prove that screening identifies risk of malnutrition or prompts referral to a dietitian are likely unnecessary. Instead, future research should investigate the impact of implementing nutrition screening on referral to a dietitian, specifically in health services where patients are not automatically referred to a dietitian. Screening usually involves the implementation of a new model of care, policy, or procedure. Therefore, randomization to screening vs no screening is likely to be difficult. Stepped‐wedge randomized trials or interrupted time series analysis may be appropriate study designs to assess the effect of malnutrition screening on the rate of referral to a dietitian, nutrition status, treatment completion/interruptions, surgical complications, unplanned admission or readmission, survival, cost effectiveness, and impact on the workforce.

Question 2b
In adult patients with head and neck cancer receiving any treatment modality, does nutrition assessment vs no nutrition assessment change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicities, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer, we recommend that patients undergo a comprehensive nutrition assessment if they have been screened and found to be at risk of malnutrition, if they are automatically referred through established protocols because of high malnutrition risk, or if they present with an enteral access device, either planned or already in situ. We further recommend this assessment be performed by a dietitian or other qualified nutrition professional using a tool that has been validated in the oncology population (eg, Patient‐Generated Subjective Global Assessment (PG‐SGA) or SGA).

Certainty of Evidence: Very low/expert opinion

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
One quasi‐experimental study
43
met the inclusion criteria for this question (Tables 8, 9, S10, and S11).

Rationale and discussion
One nonrandomized matched‐pair study (n = 82) comparing nutrition assessment via SGA with dietetic counseling when needed against a group receiving no active nutrition assessment reported decreased delay in adjuvant therapy (P = 0.0076), wound infection and dehiscence (P = 0.0065), flap failure (P = 0.0031), delay in oral feeding (P = 0.0115), a decrease in prolonged hospital stay >6 days (P = 0.0121), and improvement in body composition metrics.
43
Although the bias risk was rated as “serious” for this study, nutrition assessment is a key step in the nutrition care process, required to determine nutrition status and guide interventions.
44
It is also important for monitoring nutrition status and may therefore be a strategy to prevent nutrition decline. The current recommendation is based on this potential benefit and no known harms and aligns with a similar recommendation in the COSA guideline.
5

Future research
Although nutrition assessment is a key step in the nutrition care process, the optimal frequency of nutrition assessment required to minimize decline in nutrition status is unknown. RCTs are required to examine the effect of nutrition assessment frequency (before, during, and after treatment) on nutrition status in patients with head and neck cancer. Other relevant outcomes include the number of treatment completions/interruptions, surgical complications, unplanned admission/readmission, and survival.

Question 3a
In adult patients with head and neck cancer receiving any treatment modality, does intensive nutrition therapy designed to meet current recommendations for protein intake vs standard care change progression‐free survival, overall survival, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer receiving any modality of treatment, we recommend a protein intake of 1.2–1.5 g/kg/day, which would meet the needs of most patients with head and neck cancer. This may be met and maintained through one or a combination of oral intake, oral nutrition supplements, or EN to meet protein requirements and should be tailored to symptom burden and nutrition status while considering individual treatment plans and clinical, psychosocial, and socioeconomic status.

Certainty of Evidence: Very low/expert opinion

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
One RCT
45
met the inclusion criteria for this question (Tables 10, 11 S12, and S13).

Rationale and discussion
Muscle loss is common in patients with head and neck cancer, and observational studies show it is associated with poor outcomes.
17
,
18
Adequate protein is important to maintain muscle mass or minimize muscle loss. The one study that met our inclusion criteria (n = 49) examined the presurgical provision of whey protein against standard care and reported decreased postoperative complications in the whey protein group (P < 0.001), although unmanaged between‐group differences at baseline make this statistic difficult to interpret.
45
It was rated as having “some concerns” for bias risk. Further, studies that did not meet our inclusion criteria because of differences in design, time period, or rigor of intervention, but that examined protein delivery at different levels in patients with cancer, have not reported any harms in protein delivery at these levels. One meta‐analysis examined eight studies in patients with cancer with a high prevalence of sarcopenia. Five of these were in patients with head and neck cancer. One RCT and four pre‐post studies in head and neck cancer supported the benefit of protein for maintaining muscle mass. The authors found that, taking all eight studies into account, lean muscle wasting during treatment was found in patient groups that maintained an intake of <1.2 g/day of protein. Groups receiving >1.4 g/day experienced lean muscle maintenance.
46
It is important to note that the inclusion criteria for the systematic review only required protein to have been tracked between groups. Protein did not need to be the primary intervention, and co‐interventions were not considered. Therefore, we cannot be confident that protein was the cause of the outcome, but it does imply that protein at those levels was not harmful. Although the effect of supplementing protein in these patients is unknown, one meta‐analysis of 11 retrospective studies in head and neck cancer found sarcopenia to be associated with decreased overall survival and increased relapse‐free survival.
47
Although neither of these reviews can be considered definitive and no studies specifically designed to answer this question were found, the current recommendation has no reported harms, and the idea that it may be somewhat protective against severe muscle wasting is biologically plausible, supported by these meta‐analyses, and is consistent with the COSA guideline and other guidelines.
5
,
48

Future research
The effect of intensive nutrition therapy, using one or a combination of oral intake, oral nutrition supplements, or EN, on protein intake has not been established. RCTs designed to achieve and maintain protein intake using dietary counseling, escalating to oral nutrition supplements and/or EN as clinically indicated, are needed to assess the effect on muscle mass, nutrition status, treatment completion/interruptions, treatment toxicities, surgical complications, postoperative length of stay, unplanned hospital admissions, and survival. The exploration of protein intakes up to 2 g/kg/day has occurred in other cancer populations.
49

Question 3b
In adult patients with head and neck cancer receiving any treatment modality, does intensive nutrition therapy designed to meet current recommendations for energy intake vs standard care change progression‐free survival, overall survival, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer receiving any modality of treatment, we recommend an energy intake of ≥30 kcal/kg/day. This may be met and maintained through one or a combination of oral intake, oral nutrition supplements, or EN to meet energy requirements and should be tailored to symptom burden and nutrition status while considering individual treatment plans and clinical, psychosocial, and socioeconomic status. Nutrition status should be monitored regularly to determine if energy intake is sufficient, noting that sufficient energy intake is also important to ensure protein intake is used for the preservation of muscle mass.

Certainty of Evidence: High

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Seven studies met the inclusion criteria for this question.
24
,
50
,
51
,
52
,
53
,
54
,
55
Five demonstrated a significant benefit in the intervention compared with the control group,
50
,
51
,
53
,
54
,
55
with one study reporting less chemotherapy and radiotherapy interruptions in the intervention group
50
; one reporting less flap failure, flap dehiscence, and neck hematoma and decreased hospital length of stay
54
; three studies reporting better weight/body mass index (BMI) maintenance in the intervention group
51
,
53
,
55
; two reporting improved measures of body composition
51
,
55
; and one reporting improved PG‐SGA scores in the intervention group.
53
Of these, three had “some concerns.” Two were “low” risk of bias and two quasi‐experimental designs were reported at “serious” risk for bias. No studies reported worse outcomes in the intervention vs the control (Tables 12, 13, and S14–S16).

Rationale and discussion
Malnutrition is common in patients with head and neck cancer and is associated with poor outcomes.
17
,
18
Observational studies show that adequate energy intake is vital for maintaining nutrition stores and minimizing a decline in nutrition status.
56
With energy needs repleted, protein utilization strategies can shift to prioritize maintenance of lean muscle mass. One small observational study in 41 patients with head and neck cancer found that an energy intake of 30 kcal/kg/day was required to attenuate muscle loss, whereas an energy intake of 25 kcal/kg/day was associated with muscle loss.
56
The current recommendation is based on this potential benefit, along with clear evidence of benefit concerning weight maintenance and no evidence of harm. It is also consistent with the COSA guideline and other guidelines.
5
,
48

Future research
The effect of intensive nutrition therapy on energy intake, using one or a combination of oral intake, oral nutrition supplements, or EN, has not been established. Randomized trials designed to achieve and maintain energy intake using dietary counseling and escalating to oral nutrition supplements and/or EN if clinically indicated, are needed. Relevant outcomes for these trials include energy intake, muscle mass, nutrition status, treatment completion/interruptions, treatment toxicities, surgical complications, postoperative length of stay, unplanned hospital admissions, and survival.

Question 4a
In adult patients with head and neck cancer receiving any treatment modality, does estimating protein requirements based on an alternate body weight or composition vs standard care (actual weight) change progression‐free survival, overall survival, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admissions or readmission?

Recommendation: In adults with head and neck cancer receiving any treatment modality, owing to insufficient evidence at present to demonstrate a benefit from individualizing protein requirements based on body composition, we suggest estimating protein requirements based on actual body weight. However, the risk of overestimating protein requirements in patients with obesity is higher. An acceptable solution to this may be to use the higher end of ideal body weight or to use actual body weight while using clinical judgment to determine if the resulting target protein amount is achievable. We recommend ongoing monitoring of nutrition intake alongside nutrition status, muscle mass, muscle strength, and physical performance as an indication of the adequacy of protein intake.

Certainty of Evidence: Very low/expert opinion

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
No studies met the inclusion criteria for this question. See Question 4b for discussion.
The rationale, discussion, and future research directions for this recommendation are included in Question 4b because of similarities.

Question 4b
In adult patients with head and neck cancer receiving any treatment modality, does estimating energy requirements based on an alternate body weight or composition vs standard care (actual weight) change progression‐free survival, overall survival, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer receiving any treatment modality, we suggest estimating energy requirements based on actual body weight because of insufficient evidence at present to demonstrate a benefit from individualizing energy requirements based on body composition. However, the risk of overestimating energy requirements in patients with obesity is higher. In this situation, either ideal body weight or actual body weight may be used with clinical judgment to determine if the resulting target energy requirement is achievable. We recommend ongoing monitoring of nutrition intake alongside weight, nutrition status, muscle mass, muscle strength, and physical performance as an indication of adequacy of energy intake.

Certainty of Evidence: Very low/expert opinion

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
No studies met the inclusion criteria for this question.

Rationale and discussion
Current methods to estimate protein requirements are based on body weight and do not consider muscle mass or body composition. Observational studies show that protein requirements that are estimated using body weight may underestimate or overestimate protein requirements, particularly in patients with obesity.
57
Protein intake that is individualized to body composition may prevent underfeeding or overfeeding of protein. Inadequate energy intake may lead to the utilization of muscle protein as an energy source and result in muscle loss. However, a narrative review suggests excess energy intake leads to increased fat mass, which may increase the risk of cancer recurrence and other chronic health conditions.
58
Individualizing estimation of energy requirements may, therefore, be important to prevent adverse changes to body composition (eg, muscle loss or increased fat mass) or exacerbation of existing adverse body composition. However, individualizing protein and energy requirements relies on accurate assessment of body composition, in particular muscle mass, and most assessment techniques have limitations at this time.
59
Inaccurate assessment of muscle mass could result in underestimation or overestimation of protein and/or energy requirements, with underestimation having an undesirable effect on muscle loss and overestimation having an undesirable effect on fat mass. The current recommendation to use actual body weight or ideal body weight in obesity is based on an effort to avoid this potential harm.
It should be noted that the method to calculate ideal body weight for patients with obesity may vary regionally. In the United States, it is common to use the Hamwi equation.
60
In Australia, it is common to define ideal body weight as the weight at the upper end of the range for BMI within the healthy reference range, noting that ranges appropriate to ethnicity should be used. Patients with head and neck cancer who are underweight are likely to require the higher end of the recommended range for protein intake (1.5 g/kg/day). If weight gain occurs, requirements may need to be adjusted.
In patients with head and neck cancer who are underweight, we have not suggested alternatives to using actual body weight. This is because using actual body weight to estimate energy requirements is expected to lead to weight gain.

Future research
Individualizing protein and energy requirements based on body composition should be investigated in randomized trials to determine the effect on muscle mass, fat mass, nutrition status, treatment toxicities, treatment completion/interruptions, surgical complications, postoperative length of stay, unplanned hospital admissions, and survival. There is potential for findings to vary depending on which body composition assessment technique is used. Therefore, to ensure comparability across studies, a discussion needs to occur regarding which techniques should be used. Priority should be given to techniques that are both feasible in clinical practice and validated in a cancer population.

Question 5
In adult patients with head and neck cancer receiving any treatment modality, does gastrostomy feeding (via PEG or radiologically inserted gastrostomy [RIG]) vs nasogastric tube (NGT) feeding change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, dysphagia, incidence of stricture, fistula development, global quality of life, fatigue, return to work, performance status, treatment completion, feeding tube dependence, time of transition to full oral diet, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer receiving any treatment modality, we suggest that the decision to place a PEG or RIG tube vs an NGT is made through discussion among interdisciplinary team members including a dietitian or another member with nutrition training. The decision regarding the type of enteral access device should be based on the clinical situation (including tumor location and stage), symptom burden (especially preexisting dysphagia), treatment plan, psychosocial situation, and the anticipated duration of enteral feeding. If EN is indicated, feeding via PEG/RIG may be more appropriate when anticipated for longer durations (commonly >4–6 weeks); otherwise, an NGT should be considered.

Certainty of Evidence: Moderate

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Four studies met the inclusion criteria for this PICOT question. Three studies demonstrated a significant benefit in the intervention compared with the NGT feeding control group.
61
,
62
,
63
Three studies reported less weight loss,
61
,
62
,
63
two found less tube dislodgement,
61
,
62
one reported less surgical site dehiscence,
63
one reported less infections,
62
one found improved quality of life,
62
one reported less altered body image,
61
one reported less loss of mid‐upper arm circumference,
62
and one study reported higher triceps skinfold thickness.
61
The remaining study found no significant differences between the two groups for any outcome (Tables 14, 15, S17, and S18). Although no studies favored the control, one study reported a longer duration of enteral feeding, more pain, and more inconvenience in the intervention group.
61
All four studies were rated as having some concerns for risk of bias. This position is also consistent with observational studies that did not meet our inclusion criteria but found that using protocols to guide decisions regarding the route of enteral feeding improved clinical outcomes.
21
,
22
,
65

Rationale and discussion
Patients with head and neck cancer are at high risk of dysphagia, odynophagia, and mucositis owing to the tumor location and treatment.
66
Enteral feeding is required to meet energy and protein requirements when patients are unable to achieve sufficient oral intake. Enteral feeding using a PEG or RIG is typically used when enteral feeding is expected to be needed for longer durations, whereas NGT feeding is typically used when a shorter duration is anticipated. PEG and RIG tube insertion is generally safe; however, a narrative review indicates complications are reported to occur in 5%–40% of cases.
67
The most common complication is wound infection, which is typically minor, although major complications requiring surgical intervention or that are life‐threatening may occur.
67
A narrative review suggests NGT insertion is most commonly associated with minor complications, including discomfort, sinusitis, or epistaxis, and long‐term use may result in irritation of the gastric lining, nasal pressure ulcers, and gastrointestinal bleeding.
68
Confirmation of placement must occur before initiating EN.
68
The current weak recommendation is based on these considerations, along with some evidence of benefit for PEG feeding from the literature, balanced with the potential harms, and aligns with similar recommendations in the COSA guidelines.
5

Observational studies suggest that a longer duration of enteral feeding is likely to be required in patients requiring extensive surgical resection, those planned for oral or bilateral radiotherapy with concurrent chemotherapy, or patients with pre‐existing dysphagia.
21
,
69
Patients who have experienced severe unintentional weight loss (>10% in 6 months), severe malnutrition, or pre‐existing poor oral intake may also be at high risk of needing a longer duration of enteral feeding.
21
While this recommendation explicitly states PEG and RIG tubes, it applies to all types of gastrostomy tubes.

Future research
Randomizing patients to PEG compared with NGT may be challenging when clear indications exist for one vs the other, but RCTs may be warranted in patients for whom indication for an NGT or PEG is equal. Further studies should examine the pre and post effect of implementing protocols or algorithms to guide decisions on the enteral access device as treatment modalities and their toxicity profiles evolve. Outcomes that should be examined include nutrition status, treatment completion/interruptions, treatment toxicities, unplanned hospital admissions, dysphagia, and survival.

Question 6
In adult patients with head and neck cancer receiving any treatment modality, does more frequent speech pathology intervention compared with the standard of care change the time to transition to full oral diet, progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, dysphagia, global quality of life, fatigue, return to work, performance status, treatment completion feeding tube dependence, incidence of stricture, fistula development, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer receiving any treatment modality, we recommend consultation by a speech pathologist before treatment (surgery or radiotherapy with or without chemotherapy) for baseline assessment and education if the treatment is likely to affect swallowing function or in the case of pre‐existing dysphagia. We recommend that the frequency of consultation by a speech pathologist during and after radiotherapy (with or without chemotherapy) and after surgery be guided by the treatment plan as well as the severity of dysphagia and other treatment toxicities. Additional considerations include clinical, psychosocial, and socioeconomic status. Interventions should be tailored to reduce dysphagia risk, minimize malnutrition, and improve quality of life. These interventions may include the maintenance of oral intake throughout radiotherapy (if safe to do so), prophylactic or therapeutic swallowing exercises, texture modification, swallowing maneuvers, compensatory strategies, and education.

Certainty of Evidence: Moderate

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Eight studies met the inclusion criteria for this question.
70
,
71
,
72
,
73
,
74
,
75
,
76
,
77
Five studies demonstrated a significant benefit in the intervention group compared with the control.
71
,
72
,
73
,
74
,
77
Three reported better swallow function,
71
,
72
,
74
three found improved mouth opening,
72
,
74
,
77
one reported improved normalcy of diet at 3 and 6 months,
73
and one reported improved self‐reported functional oral intake in the intervention group.
71
One study favored the control, finding improved ability to swallow
70
in the control group. The remaining study found no significant differences between the two groups.
75
One study was rated as having a “low” risk of bias. One quasi‐experimental design was rated as having “serious” risk of bias. The remaining studies were rated as having some concerns or high risk of bias (Tables 16, 17, and S19–S21).

Rationale and discussion
Dysphagia is prevalent in patients with head and neck cancer, which affects patients’ ability to achieve an adequate oral intake, and observational studies suggest this may lead to malnutrition.
66
Speech pathologist intervention may improve or prevent decline in dysphagia. More frequent interventions by a speech pathologist may, therefore, be a strategy to improve oral intake and prevent decline in nutrition status. The current recommendation is based on this potential benefit, along with some evidence of benefit from the literature, and no known harms.
There was significant variation across the studies in the type of intervention provided, frequency of the intervention, and outcome measures reported (type and timing). In one study,
71
the intervention was provided at least 3 months after surgery, whereas in the remaining studies, exercises were provided as a prophylactic approach during radiotherapy. It is important to note that compliance with prophylactic exercises during radiotherapy is often poor, and in two studies, the swallowing intervention was not provided by a speech pathologist. As such, many of these studies did not investigate the comprehensive and individualized approach generally used by speech pathologists in clinical practice.
There has been increasing interest in alternate service delivery models to support the provision of prophylactic exercises during radiotherapy. A study by Wall et al did not find differences in outcomes between clinician‐directed telepractice therapy and self‐directed therapy; however, patients significantly preferred the first two approaches over the patient‐directed approach.
78

Future research
Future research on the impact of speech pathology interventions for patients with head and neck cancer needs to address the effectiveness of specific therapy/rehabilitation strategies, as well as investigate different delivery and methodological approaches to such interventions. Such research would generally be appropriate as either an RCT or a pre‐post implementation approach, depending on the current practices of the center(s) involved. Consideration should be given to the outcome measures used, with comprehensive research ideally including all the following: patient‐reported outcomes, clinician‐rated function, and instrumental swallow assessments. Consideration should also be given to the specific patient population, time frames of the intervention, and data collection and therapy regimens (where relevant), to maximize the ability to make comparisons between studies and implement outcomes into clinical practice.

Question 7
In adult patients with head and neck cancer undergoing any treatment modality, does an interdisciplinary approach to nutrition management vs standard care change progression‐free survival, overall survival, nutrition intake, time to transition to full oral diet, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion feeding tube dependence, treatment interruptions, treatment toxicity, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer, we recommend an interdisciplinary approach to nutrition management. An interdisciplinary approach should involve collaboration between health professionals with the expertise to manage any symptom or issue that is affecting or anticipated to affect the patient's nutrition intake or nutrition status. We recommend that the core team for nutrition management include dietitians, nurses, pharmacists, physicians, and speech pathologists. Additional members may include dental professionals, physical therapists, psychologists, and social workers.

Certainty of Evidence: Moderate

Strength: Strong

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
Four studies met the inclusion criteria for this question.
79
,
80
,
81
,
82
All four studies favored an interdisciplinary approach to nutrition management. Three studies reported less weight loss.
79
,
81
,
82
Two reported less mucositis.
79
,
81
One reported a shorter hospital length of stay.
79
Two reported decreased risk of malnutrition,
80
,
81
with one reporting improved nutrition status.
82
One study reported fewer speech problems, improved handgrip strength, improved role functioning, and improved pain on a quality of life questionnaire in the intervention group but favored the control group for physical function, feeling more ill, and higher nutrition supplement use, which may be related to detection bias.
80
This RCT was found to be low risk of bias,
80
whereas the other studies were at “high” risk of “some concerns” for bias (Tables 18, 19, and S22–S24).
79

Rationale and discussion
Malnutrition is common in patients with head and neck cancer.
18
The factors contributing to malnutrition are multifactorial, including symptoms affecting nutrition intake as well as psychosocial and socioeconomic status. An interdisciplinary approach to nutrition management may, therefore, be an appropriate strategy to manage these complex factors and prevent nutrition decline. The current recommendation is based on this potential conceptual benefit, along with some evidence of benefit from the literature, especially the low‐bias RCT. Although more studies are needed, the balance between potential benefits and harms were considered, so it was given a rating of “strong.”
Dietitians, nurses, pharmacists, physicians, and speech pathologists are expected to be required for the nutrition management of all patients with head and neck cancer. Some patients may require a social worker to organize meal delivery or shopping and address financial hardship if it impacts food security. If dentition is affecting a patient's nutrition intake, a dental professional may be required. Mental health concerns, such as anxiety and depression, may affect nutrition intake and require support from a mental health worker (eg, psychologist, social worker, or psychiatrist). In patients with muscle loss, a physical therapist may be needed to maintain or improve muscle mass.

Future research
RCTs of an interdisciplinary approach to nutrition management during treatment are required. If more than one site is planning to initiate an interdisciplinary team approach, a stepped‐wedge randomized design where implementation is staggered and compared between sites may be particularly suitable. If not, a quasi‐experimental design would be suitable. This could be a pre‐post design comparing findings before and after a policy change. Alternatively, one hospital without an interdisciplinary nutrition team could be compared with one without. Outcomes of importance for such studies include nutrition status, muscle mass, dysphagia, time to transition to a full oral diet, treatment interruptions, hospital length of stay, return to work, and survival.

Question 8
In adult patients with head and neck cancer receiving any treatment modality, does a pharmaceutical appetite stimulant compared with no pharmaceutical appetite stimulant change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myeosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion treatment interruptions, treatment toxicities, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation: In adults with head and neck cancer who are experiencing anorexia and receiving any treatment modality, we suggest dietary counseling (including oral nutrition supplements or enteral feeding) and management of other symptoms that are affecting oral intake as first‐line strategies to address anorexia and improve nutrition intake. Otherwise, a pharmaceutical appetite stimulant may be considered for short‐term use where clinically appropriate. In conjunction with the medical team and dietitian, this decision should ideally include discussion with a pharmacist specializing in oncology.

Certainty of Evidence: Very low/expert opinion

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
One quasi‐experimental study was found that addressed this question (Tables 20, 21, S25, and S26).

Rationale and discussion
Anorexia and cachexia occur in some patients with head and neck cancer, and observational studies indicate they are associated with adverse outcomes.
84
In patients for whom dietary counseling and other symptom management approaches have not improved nutrition intake, a pharmaceutical appetite stimulant (eg, corticosteroids, progesterone analogs, or olanzapine) may be an appropriate strategy to increase nutrition intake and prevent nutrition decline. Pharmaceutical appetite stimulants are not recommended for long‐term use. This is because the agents that may be most beneficial, corticosteroids or progesterone analogs, may cause other, more serious, issues or toxicities with chronic use. Ghrelin receptor agonists, such as anamorelin, have been shown to improve body weight, muscle mass, and quality of life in patients with advanced lung cancer,
85
with no studies in patients with head and neck cancer. Ghrelin receptor agonists may have side effects including gastrointestinal upset, hyperglycemia, and, rarely, fatal arrhythmia. One quasi‐experimental trial (n = 104) comparing megesterol acetate (400 mg/day) with no megesterol acetate unless >5% weight loss occurred met the inclusion criteria for this question.
83
Although this study was found to be at “serious” risk for bias, it reported decreased weight loss, duration of neutropenia and Grade 3–4 mucositis, and improved overall survival in the megesterol group vs the control. The current recommendation is based on the recognition that although some potential benefit may be realized by pharmaceutical appetite stimulants, this must be balanced against the potential harmful side effects, which can be determined through discussion with an oncology clinical pharmacist.

Future research
Adequately powered, double blind RCTs examining the effect of appetite stimulants compared with standard care are needed. Relevant outcomes include measures of appetite, nutrition intake, nutrition status, muscle mass, quality of life, and survival. Such trials should specifically recruit participants with pre‐existing anorexia and stratify by nutrition status.

Question 9
In adult patients with head and neck cancer receiving chemoradiation or radiation, does continuing oral intake (if tolerated) after the initiation of EN compared with not continuing oral intake change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia (skeletal muscle mass + strength), myosteatosis, dysphagia, incidence of stricture, global quality of life, fatigue, return to work, performance status, treatment completion, feeding tube dependence, time of transition to a full oral diet, surgical complications, length of stay, or hospital readmission?

Recommendation: In adults with head and neck cancer who have commenced EN and who can safely continue oral intake per consult with a speech pathologist, we suggest that continuing any degree of oral intake may be beneficial for maintaining swallow function. The amount, type, and texture of the oral intake will be dependent on swallow safety and treatment toxicities. The volume and timing of enteral feeding should be adjusted according to what is consumed orally to optimize the opportunity for the patient to continue oral intake while also ensuring that nutrition requirements are met.

Certainty of Evidence: Very low/expert opinion

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%
No studies met the inclusion criteria for this question.

Rationale and discussion
Observational studies demonstrate that swallow function may deteriorate in the absence of any oral intake,
86
of which exclusive EN during treatment may be one contributing factor. Reduced swallow function leads to difficulties in resuming oral intake.
86
Maintaining some degree of oral intake throughout enteral feeding may, therefore, be an appropriate strategy for maintaining swallow function, which has the potential to improve quality of life. The current recommendation is based on this potential benefit and no known harms, assuming this occurs in consultation with a speech‐language therapist to minimize the risk of aspiration or other dysphagia‐related complications. This recommendation is consistent with findings from a large retrospective observational study, where maintenance of oral intake during treatment was independently associated with better long‐term diet after treatment and shorter duration of gastrostomy dependence, when adjusted for tumor and treatment burden.
86

It is important to note that this may not be appropriate for all patients. Continuing oral intake throughout enteral feeding in some patients may be very challenging and possibly inappropriate because of the severity of dysphagia and/or other treatment‐related side effects. Decisions to maintain oral intake should, therefore, be made on an individualized basis while considering the potential for a negative psychological impact from being asked to perform a task they are unable to complete.

Future research
Traditional RCTs to investigate maintaining any degree of oral intake alongside enteral feeding would be unethical. This is because patients who are otherwise capable of eating and drinking (albeit in small volumes) would be required to remain on nothing by mouth status if randomized to the control group. Instead, quasi‐experimental studies investigating the pre‐post effect of implementing a protocol to guide clinical decisions regarding maintaining oral intake are required. Such protocols should prompt clinicians to consider the appropriateness of maintaining oral intake based on pre‐existing dysphagia and toxicities such as pain, taste, and appetite. A stepped‐wedge randomized trial comparing the intervention with usual care may be an appropriate study design if more than one hospital is initiating a new protocol for oral intake alongside enteral feeding. Regardless of design, outcomes such as swallow function, consistency, amount of oral intake, ongoing need for EN, and dysphagia‐related complications should be assessed.

Question 10
In adult patients with head and neck cancer receiving any treatment modality, does use of special‐purpose nutrients (eg, arginine or glutamine) compared with not using special‐purpose nutrients change progression‐free survival, overall survival, nutrition intake, nutrition status, weight, muscle mass, sarcopenia, myosteatosis, global quality of life, fatigue, return to work, performance status, treatment completion treatment interruptions, treatment toxicities, surgical complications, length of stay, or unplanned hospital admission or readmission?

Recommendation (arginine): Given the limited evidence on progression‐free and overall survival and some evidence of benefit for decreased fistula development and length of stay in adults with head and neck cancer, we suggest that using arginine‐supplemented nutrition may be acceptable at the discretion of the interdisciplinary team.

Certainty of Evidence: Low

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%

Recommendation (glutamine): Oral/enteral glutamine has been shown to reduce the severity of oral mucositis, with the potential to reduce other treatment toxicities and hospitalization, and improve treatment completion. We therefore suggest that the use of oral/enteral glutamine in patients with head and neck cancer may be acceptable at the discretion of the interdisciplinary team. Intravenous glutamine is more controversial because of one small study that reported increased mortality in the patients receiving intravenous glutamine and recent preclinical trials suggesting mechanisms through which glutamine may contribute to tumor growth and treatment resistance. For this reason, we suggest not adding parenteral glutamine to standard nutrition therapy in patients with head and neck cancer until further research becomes available to confirm its safety.

Certainty of Evidence: High

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%

Recommendation (ω‐3s): Given the inconsistent evidence for benefit but no evidence of significant harms in patients with head and neck cancer, we suggest that ω‐3–supplemented nutrition is unlikely to be harmful and may be used or not at the discretion of the interdisciplinary team.

Certainty of Evidence: Low

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%

Recommendation (combined nutrients): In patients with head and neck cancer, given the inconsistent evidence for benefit but no evidence of significant harms, we suggest that the use of combined special‐purpose nutrient or immunonutrition‐supplemented formulas is unlikely to be harmful and may be used or not at the discretion of the interdisciplinary team.

Certainty of Evidence: Very low

Strength: Weak

Clinical Panel Delphi Agreement: 100%

External Validation Panel Delphi Agreement: 100%

Arginine
Seven studies that met the inclusion criteria for this question examined the use of arginine.
87
,
88
,
89
,
90
,
91
,
92
,
93
Six studies demonstrated a significant benefit from arginine supplementation compared with the control.
87
,
88
,
89
,
90
,
91
,
93
Three studies reported lower development of fistulas,
88
,
89
,
90
two studies reported shorter length of stay,
89
,
90
,
91
and one study reported improved disease‐free and overall survival
87
in the intervention group. One study reported better BMI retention, improved quality of life, and lower pain and oral toxicity scores in the intervention group relative to the control.
93
The remaining two studies found no significant difference between the two groups for any outcome (Tables 22, 23, S27, and S28). Five studies were rated as having some concerns for risk of bias and two were rated at low risk of bias. Three studies were meta‐analyzable (Figure 6) and found no significant differences in general infections (RD = 0.00, 95% CI = −0.14 to 0.14; P = 0.41), fistula of wound (RD = −0.09, 95% CI = −0.29 to 0.12; P = 0.28), diarrhea (RD = 0.10, 95% CI = −0.28 to 0.48; P = 0.39), and length of stay (RD = −1.38, 95% CI = −18.24 to 15.47; P = 0.76).

Glutamine
Fifteen studies that met the inclusion criteria for this question examined the use of glutamine.
93
,
94
,
95
,
96
,
97
,
98
,
99
,
100
,
101
,
102
,
103
,
104
,
105
,
106
,
129
Twelve studies demonstrated a significant benefit from glutamine supplementation compared with the control.
93
,
94
,
95
,
96
,
98
,
100
,
101
,
102
,
103
,
104
,
105
,
129
Seven studies reported reduced severity of mucositis
93
,
95
,
96
,
98
,
100
,
101
,
105
; one reported lower incidence of mucositis
103
; one reported increased time to mucositis onset
98
; two reported less severe dysphagia
103
,
104
; four reported less pain
93
,
95
,
100
,
105
; one reported less severe odynophagia
129
; one reported less hospitalization for treatment toxicities, less weight loss, and fewer treatment interruptions
103
; one reported less nausea and less edema
104
; two reported improved quality of life
93
,
94
; one reported improved fat‐free mass
94
; and one reported less dermatitis
102
in the intervention group. Two studies found no significant difference between the two groups for any outcome (Tables 22, 23, and S27–S29). Importantly, one study reported on overall and progression‐free survival for parenteral glutamine and found no difference between groups.
106
Six studies were rated as having low risk of bias.
93
,
102
,
103
,
104
,
105
,
106
Two were rated as having high risk of bias,
96
,
100
and the remaining seven studies were rated as having some concerns.
Meta‐analysis was possible for two outcomes: mucositis and mucositis severity Grade 3–4 (Figure 7). On average, patients were reported to have 5% less mucositis in the glutamine intervention groups vs the controls, but this was not statistically significant (RD = −0.05, 95% CI = −0.12 to 0.03; P = 0.16). Risk of having mucositis Grade 3–4 was 28% reduced in the glutamine intervention groups vs the control (RD = −0.28, 95% CI = −0.43 to −0.13; P < 0.001).

ω‐3
Seven studies that met the inclusion criteria for this question examined the use of ω‐3 fatty acids.
107
,
108
,
109
,
110
,
111
,
112
,
119
Three studies demonstrated a significant benefit from ω‐3 supplementation.
107
,
112
,
119
One study reported improved nutrition status and improved appetite at 6‐ to 7‐week follow‐up,
119
one reported less fatigue,
112
and one reported higher energy intake and higher body weight at baseline, end of radiotherapy, and at 1 and 3 months posttreatment, improved quality of life, and higher treatment completion
107
in the intervention group. The remaining three studies found no significant difference between the two groups for any outcome (Tables 22, 23, S27, and S28). Two studies were found to be at “low” bias risk,
107
,
111
and the rest were at “some concerns” for bias. Cereda et al and Pottel et al were rated as having low risk of bias, Solís‐Martínez was rated as having high risk of bias, and the remaining three studies were rated as having some concerns.
107
,
111
Meta‐analysis was possible for absolute weight loss (Figure 8). On average, patients in the ω‐3 intervention group had 1.2 kg less weight loss, but this was not statistically significant (MD = 1.20 kg, 95% CI = −1.61 to 4.01; P = 0.21).

Combined special‐purpose nutrient formulas
Sixteen studies that met the inclusion criteria for this question examined the use of combined special‐purpose nutrient formulas.
114
,
115
,
117
,
120
,
121
,
122
,
125
,
126
,
127
,
128
,
130
Twelve studies demonstrated a significant benefit from combined immunonutrition supplementation.
113
,
115
,
116
,
117
,
118
,
121
,
122
,
123
,
125
,
126
,
130
Nine studies reported fewer complications.
113
,
115
,
117
,
118
,
121
,
122
,
125
,
126
,
130
One reported lower incidence of Grade 3 mucositis.
123
Two reported reduced postoperative or total length of hospital stay.
125
,
126
One reported higher energy and protein intake.
130
Two reported increased 3‐year survival in the intervention group,
114
,
116
with one reporting improved overall survival
116
and the other reporting better progression‐free survival in a subanalysis of compliant participants.
114
One study found better weight retention in the intervention group vs the control.
123
Three studies found no significant difference between the two groups for any outcome (Tables 22, 23, and S27–S29). Four studies were rated as having a “low” risk of bias.
114
,
116
,
117
,
127
Six studies were rated as having a “high” risk of bias,
113
,
115
,
123
,
124
,
125
,
130
and the remaining studies were rated as having “some concerns.”

Rationale and discussion
As described in a narrative review, special‐purpose nutrients have a role in modulating reactions within immune, inflammatory, and lean muscle–producing pathways.
132
Through such mechanisms, special‐purpose nutrients may have an effect on outcomes such as nutrition status, muscle mass, wound healing, and other complications.
133
,
134
To determine whether a specialty nutrient should be administered, a careful weighing of the benefits vs harms is needed.
The benefits of arginine were clear in the literature, with its ability to decrease incidence of fistula, decrease length of stay, and improved progression‐free and overall survival. Arginine has been associated with some negative outcomes such as headache, infection, gastrointestinal disturbances, and rash. However, these only occur in approximately 3% of the population.
135
For this reason, we felt the benefits of taking arginine outweighed any potential harms.
Glutamine is a more complex issue. Glutamine, not to be confused with glutamate (eg, L‐glutamic acid), is not usually present in pharmacologic doses in standard care enteral or parenteral products. This is because of its stability issues in an aqueous environment. It is available as a powdered modular product. Outside the United States, it is also available as an intravenous parenteral supplement in its more stable dipeptide form. The literature found glutamine to be very effective in delaying, reversing, and decreasing the severity of mucositis. However, certain concerns have arisen in the past few years that glutamine could itself promote tumor growth and progression as well as resistance to cancer treatment. In 2022, a compelling narrative review was published by Alden et al demonstrating the current state of evidence concerning the potential for glutamine to feed tumor growth and induce resistance to cancer therapy.
136
The mechanisms were biologically plausible. They were supported entirely by mechanistic studies showing how glutamine supports cancer cell survival, preclinical studies showing radiosensitization and improved immunotherapy efficacy when glutamine metabolism is blocked, and early clinical trials assessing the combination of glutamine inhibitors with standard treatments.
136
Furthermore, an RCT in patients undergoing autologous stem cell transplant found worse mortality in the intervention group who received 30 g/day of parenteral glutamine, although this study was very small (n = 40).
137
Conversely, a recent oral glutamine intervention study in patients with head and neck cancer reported no differences in survival. Although this study was also very small (n = 38), it directionally favored the oral glutamine group,
106
and similar findings have been reported in other cancer populations. One glutamine study in non–small cell lung cancer (n = 101) found no significant difference in overall or progression‐free survival, with the glutamine group directionally favored at most time points of survival analysis.
138
Another study (n = 122) in a similar population found the glutamine group did not have worse overall or disease‐free survival.
139
After adjusting for weight loss, low hemoglobin level, and nodal stage, both overall (P = 0.05) and disease‐free survival (P = 0.035) were significantly higher than those of the control. These findings are corroborated in a smaller study in 60 women with breast cancer. Oral glutamine did not negatively impact tumor shrinkage or immunohistochemistry.
140
Although this literature is still in its infancy and more research is needed, there is little to suggest that oral glutamine is problematic and much to suggest it is helpful in treating treatment toxicities. For this reason, we suggest the use of glutamine is acceptable with the caveat that caution is needed concerning parenteral glutamine unless future findings demonstrate its safety. This caution is consistent with the recommendation in a clinical guideline for the management of mucositis secondary to cancer therapy.
141

The reported benefits of ω‐3 supplementation were minor in the literature, encompassing decreased fatigue, improved dietary intake, and weight gain. Potential ω‐3 side effects are also minor and include gastrointestinal upset, abdominal pain, rash, infection, and taste changes.
142
The current recommendation for ω‐3 is based on the potential benefit, along with some evidence of benefit from the literature, and no serious harms.
The studies on combined immunonutrition examined various blends of immunonutrients. The data imply these formulas may be beneficial, although it is difficult to know which ingredient is performing the benefit, and different formulations were used across the different studies. Currently, all the premade immunonutritions formulations are enteral. If parenteral immunonutrition is formulated, we recommend against including dipepetide glutamine parenteral formulations until further research demonstrates its safety.

Future research
Data regarding the impact of special‐purpose nutrients on survival are limited. In addition, current studies vary in their quality and the outcomes assessed. Adequately powered, double blind randomized trials examining the effect of supplementation with special‐purpose nutrients compared with standard care on treatment complications, hospital length of stay, treatment completion, and progression‐free and overall survival are required. This is particularly crucial for glutamine supplementation because of its potential effects on tumor growth, cancer progression, and treatment resistance. Large RCTs are needed to assess the safety of glutamine but especially of parental dipeptide glutamine formulations.

CONCLUSION
This clinical guideline has built on current guidelines and is intended to support clinicians in delivering optimal nutrition care to patients with head and neck cancer. The recommendations are based on evidence from RCTs, quasi‐experimental designs, and expert opinion. The strongest evidence was found for the frequency of dietitian intervention, intensity of nutrition therapy, timing of EN, frequency of speech pathology intervention, and an interdisciplinary approach to nutrition management. It is anticipated these guidelines will be used to advocate for sufficient resources within health services to provide evidence‐based care to people with head and neck cancer. During the synthesis of data from the included trials, it was evident that standardized outcome reporting is lacking in this field, meaning meta‐analysis was only possible for a limited number of questions. We call on researchers in the field of head and neck cancer nutrition to work toward a standardized set of outcome measures to improve the quality of evidence and ability to synthesize the data underpinning future recommendations. Furthermore, suggestions for further interdisciplinary research, appropriate study designs, and outcomes were identified within each rationale above to encourage ongoing research in the nutrition care for this vulnerable population.

AUTHOR CONTRIBUTIONS

Nicole Kiss: Conceptualization; Writing—original draft; Data curation; Supervision; Writing—review and editing. Merran Findlay: Conceptualization; Writing—review and editing. Jacqui Frowen: Conceptualization; Writing—original draft; Writing—review and editing; Data curation. Whitney E. Lewis: Conceptualization; Writing—review and editing; Data curation. Jeannine Mills: Conceptualization; Data curation; Writing—review and editing. Anurag K. Singh: Conceptualization; Writing—review and editing; Data curation. David D. Church: Data curation; Conceptualization; Writing—review and editing. Jacob T. Mey: Conceptualization; Data curation; Writing—review and editing. Sarah Peterson: Data curation; Writing—review and editing; Conceptualization. Kathleen Aguzzi: Data curation; Writing—review and editing. Sarah Bellini: Data curation; Writing—review and editing. Maria Paula Villela Coelho: Writing—review and editing; Data curation. Laura Cordwin: Writing—review and editing; Data curation. Michael Duffy: Writing—review and editing; Data curation. Shanna Hager: Writing—review and editing; Data curation. Manpreet S. Mundi: Writing—review and editing; Data curation. Michael Owen‐Michaane: Writing—review and editing; Data curation. Kathleen Price: Writing—review and editing; Data curation. Heather Stanner: Writing—review and editing; Data curation. Bridget Storm: Writing—review and editing; Data curation. Malika Udagedara: Writing—review and editing; Data curation. Liam McKeever: Conceptualization; Writing—original draft; Methodology; Validation; Software; Formal analysis; Project administration; Data curation; Supervision.

CONFLICT OF INTEREST STATEMENT
Nicole Kiss has received honoraria from Nutricia Australia and Abbott Australasia for consultancy and presentations. Merran Findlay has received honoraria from Fresenius Kabi Australia and Nutricia Australia for invited presentations and was supported by a Maridulu Budyari Gumal (SPHERE) Cancer Clinical Academic Group Senior Research Fellowship funded by a Cancer Institute NSW Research Capacity Building Grant (2021/CBG003). Whitney E. Lewis accepted a position as a medical liason at Bristol Myers Squibb, USA, near the end of the guideline creation period. She was consulted for her expertise but did not provide a Delphi vote for any recommendation. Manpreet S. Mundi reports no conflict of interest pertinent to this topic; there are research grants to his institution from Nestlé and NorthSea. The remaining authors declare no conflict of interests.

Supporting information
Supplemental Appendix.