1. Introduction
Maintaining or ameliorating the nutritional status is an important pillar in the treatment of head and neck cancer (HNC), as it has been proven that patients with HNC often suffer considerable weight loss before, during and after treatment [
1,
2,
3]. HNC affects the most fundamental activities of the upper aerodigestive tract including speaking, breathing, eating and drinking [
4]. Many patients with HNC already have a restricted nutritional status before or at the time of diagnosis. In addition, a premorbid lifestyle such as alcohol and nicotine abuse favors malnutrition [
5]. This also has an impact on the prognosis of patients with HNC [
2,
6,
7]. Furthermore, HNC is often the cause of dysphagia, odynophagia and taste disorders [
4]. The prevalence of malnutrition is associated with the location of the tumor [
6]. Malnutrition therefore plays a major role in patients with HNC [
8]. In addition, cancer patients frequently suffer from metabolic changes attributable to systemic inflammation. These include increased energy requirements, proteolysis and lipolysis [
9]. Ultimately, treatment of HNC exacerbates the problem of malnutrition [
10]. Treatment-related toxicities include progressive dysphagia due to oral mucositis which occurs primarily during radiochemotherapy treatment [
6,
11].
As soon as the oral form of nutrition is no longer sufficient, an enteral form of nutritional support is favored over the parenteral form [
12,
13,
14]. There are several ways to provide enteral nutrition. These include nasogastric (NG) and nasojejunal tubes as well as percutaneous endoscopic gastrostomy (PEG). On the one hand, the duration of the planned enteral nutrition is important. On the other hand, the patient’s wishes need to be taken into account, and the tumor’s characteristics and the type of treatment also need to be considered [
15]. PEG has a special role in this context, as enteral nutrition can be provided over a longer period of time, as in the case of HNC patients who suffer from dysphagia and experience an exacerbation of symptoms during treatment. The advantage of PEG over other forms of enteral nutrition is the bypass of the upper aerodigestive tract. In addition, PEG is also indicated for patients with HNC in a palliative situation as they suffer from stenosis in the natural path of food due to the tumor [
16]. However, there is discussion in the literature as to whether a prophylactic PEG should be placed if there is no swallowing disorder and no weight loss prior to treatment, or whether a reactive PEG should be placed if there is an existing swallowing disorder and malnutrition [
17,
18]. On the one hand, PEG placement can lead to acute (wound infection, bleeding, pneumoperitoneum, aspiration pneumonia and ileus) or chronic complications (tube leakage or obstruction, spontaneous tube removal, ulceration and necrosis of the stomach wall and recurrent aspiration), which are associated with a higher mortality rate; on the other hand, the swallowing muscles are not trained and scarring with stenosis develops more quickly, leading to more swallowing problems in the long term [
19].
The nutritional status during the course of patients with HNC has not yet been comprehensively investigated in literature, especially the nutritional status several months after the end of treatment. In order to evaluate the nutritional situation and in particular PEG in a functional and oncological context, data from patients with HNC at a tertiary university hospital were analyzed.
2. Methods
2.1. Ethical Considerations
This monocentric, retrospective study was approved by the Ethics Committee of the University Hospital of Jena (IRB No. 2023-2944). The Ethics Committee waived the requirement for informed consent from patients because the study was non-interventional, retrospective, and all data were analyzed anonymously.
2.2. Study Design
This retrospective study included all patients with primary diagnosis of head and cancer in the Department of Otorhinolaryngology, Jena University Hospital, Germany, from January 2017 to December 2019. These years were chosen to guarantee a sufficient follow-up time. In total, 457 patients were initially registered. All patients with the diagnosis of primary HNC in the period from January 2017 to December 2017 were included. Duplicates were excluded. Pathological stages of the tumors were recorded using the UICC-and TNM (8th edition) classifications [
20]. Ultimately, 289 patients were included in this study. Patients’ medical records were reviewed. All relevant parameters (patient characteristics, histopathological characteristics, treatment characteristics, type of nutrition, symptoms and follow-up) were recorded anonymously in a database. The documentation focused on PEG as a type of nutrition, weight and size parameters over time in patients with HNC. PEG insertion was primarily indicated in cases of low initial weight, severe restrictions on oral nutrition and before planned chemotherapy or radiochemotherapy treatment.
2.3. Timeline and Treatment
Four points in time were defined for the assessment of the nutrition status: the day of hospitalization/before treatment (Z0), and then one week (Z1), six weeks (Z2) and six months (Z3) after the end of head and neck cancer treatment. One of the most important and simplest ways to implement a nutritional process is anthropometry. This primarily involved measuring height and weight, and using these data to calculate the body mass index (BMI). This information on height, weight and nutritional characteristics including patient, tumor and treatment characteristics was obtained from the tumor board and operative reports. Letters from departments of ENT and Radiation Oncology contained important information about follow-up at the previously defined times (Z0–Z3).
It is also useful to assess nutritional status in combination with performance status and quality of life. The Karnofsky performance status was used for classification. The Karnofsky performance status was also taken from the tumor board reports written by the treating physician. At each time point, type of nutrition and weight were documented and the BMI and Karnofsky performance status were determined. The medical history always included questions about food intake complaints and information about the current diet.
The decision in favor of PEG or a nasogastric tube was made in accordance with the German guidelines for the nutrition of patients with different types of HNC [
21,
22,
23]. These generally state that artificial feeding, a nasogastric tube or PEG is indicated in the postoperative period. If it is to be expected that oral nutrition will no longer be possible for at least a few weeks, the PEG tube has proven to be safe and effective. If tube feeding (transnasal or transcutaneous) is necessary, the PEG tube should be preferred to a nasogastric tube if dysphagia is present or expected to persist. In uncomplicated cases, however, feeding with soft or liquid high calorie food under local and systemic analgesia is recommended, provided the patient does not aspirate. If these measures are not sufficient, nutrition must be provided via a PEG tube or nasogastric tube or parenterally. Prophylactic PEG placement is recommended if intensive radiotherapy or radiochemotherapy is planned [
24]. However, the German guidelines also point out that the available literature shows no significant benefit of prophylactic PEG placement in terms of weight progression, quality of life or oncological outcome [
25,
26,
27]. Furthermore, it is described that there is no evidence for the correct timing of PEG placement [
28,
29]. It is also recommended to replace a nasogastric tube with PEG if dysphagia and especially aspiration continue to occur after the healing phase and intensive swallowing training.
2.4. Statistical Analysis
SPSS Statistics Version 27.0 (IBM Deutschland GmbH, Ehningen, Germany) was used to carry out the statistical analyses. Nominal and ordinal data are presented as absolute numbers and percentages. Metric data were calculated as mean, standard deviation, median and range. To compare nominal or ordinal data of two subgroups, the Pearson chi-square test or the Fischer exact test was used. The Mann–Whitney U test for two independent samples was used to test for differences of metric data of two subgroups. The significance level was set to p ≤ 0.05. All significant factors from these univariate analyses were included in a time-dependent analysis of variance (ANOVA) for the four primary outcome parameters (Z0, Z1, Z2 and Z3). This was preceded by a normality test according to Kolmogorov–Smirnov for the factors used in the ANOVAs. The rate of missing data was less than 5% per parameter. Based on the assumption that the missingness that occurred are completely random, the listwise deletion approach was used to handle the missing data in the ANOVA calculations. Due to the Mauchly’s test of sphericity, the sphericity assumption was violated in all ANOVA calculations. Therefore, an adjustment with the Huynh–Feldt correction was performed. Hence, all ANOVA data presented (mean square, F, p-values) include these corrections. The significance level was set to p ≤ 0.05.
4. Discussion
In the present study, PEG was evaluated in the nutritional status of patients with HNC. PEG placement was associated with certain parameters such as alcohol or nicotine consumption at Z0, localization in oropharyngeal and hypopharyngeal carcinoma at Z0, squamous cell carcinoma at Z0, higher cancer stages (stage III/IV) at Z0, chemotherapy and impairment of food intake (swallowing disorders, pain during nutrition and globus pharyngeus) at Z0–Z3. The percentage of patients requiring PEG increased over the course of the follow-up. The percentage values were 14.9% (43 of 289 patients) before the start of treatment (Z0), 14% (40 of 286 patients) at Z1, 22.7% (58 of 255 patients) at Z2 and finally 23% (53 of 230 patients) at Z3. It has to be emphasized when interpreting the relative numbers that the number of patients alive decreased during the follow-up. In the study by Wermker et al., which compared a PEG-dependent group with a control group, the percentage of PEG was 17% (without specifying the exact date) [
30]. In a retrospective study of 196 patients with HNC, Nugent et al. showed a percentage of PEG of 22.5% (again without specifying the exact date). These results are also comparable to those of the present study [
31]. The study by Nugent et al. did not analyze different time points, but only compared prophylactic PEG placement (P-PEG) with therapeutic PEG placed during treatment (T-PEG). In the present study, 14% of patients received a prophylactic PEG. The placement of a PEG before the start of treatment as a prophylactic PEG is discussed in the literature [
32,
33,
34]. Din-Lovinescu et al. analyzed 4068 cases with HNC from the Nationwide Inpatient Sample (NIS) from 2003 to 2014 from US non-federal hospitals and reported a significant benefit of prophylactic PEG in terms of lower complication rates, shorter length of stay and lower hospital costs [
34]. In this study of NIS data, most patients had a tumor in the digastric tract, in the oral cavity (prophylactic PEG 37.6%; late PEG 35.2%) and in the oropharynx (prophylactic PEG 25.6%; late PEG 24.7%). However, only radiotherapy was taken into account for treatment. Also of note is the randomized, controlled phase III trial, the Swall-PEG study by Dragan et al. with the primary endpoint of patient-reported outcomes in terms of swallowing and quality of life after prophylactic versus reactive PEG tube placement in advanced oropharyngeal cancer patients treated with definitive chemoradiotherapy [
35]. The results have not yet been published.
In our study, the indication for PEG was increasingly given for certain localizations. These were tumors of the hypopharynx and oropharynx. The study by Zuercher et al. showed a correlation between these two localizations and also between tumors of the oral cavity and PEG. These three localizations were thus associated with more than three-quarters of PEG placement [
36]. The study by Sieron et al. also confirmed these two localizations as predisposing to prophylactic PEG [
37]. In Riera et al., hypo- and oropharyngeal carcinomas were among the most common sites for PEG placement, but laryngeal carcinomas showed a stronger association with PEG use than HNC of hypopharynx and oropharynx [
38]. It is possible that the smaller number of laryngeal cancer cases in the present study had an influence on the fact that there was no significance between laryngeal cancer and PEG use. There was still a dependency on the cancer stage. Patients with a higher stage required PEG significantly more often than patients with a lower stage. This was also found to be the case in the studies by Zuercher et al., Riera et al. and Wermker et al. [
30,
36,
38]. Riera et al. also demonstrated that squamous cell carcinomas were more frequently associated with PEG use [
38]. This was also found by Sieron et al. for prophylactic PEG use in squamous cell carcinoma, which is consistent with our results [
37]. Current or previous nicotine and/or alcohol consumption were also predictive factors for the placement of a PEG tube and for the placement of a prophylactic PEG tube which is confirmed by Wermker et al. [
30]. PEG placement was often indicated for all types of complaints relating to oral feeding. Riera et al. found dysphagia to be the most common indication for PEG insertion. Wermker et al. showed similar results.
With regard to the association between the treatment and PEG placement, Zuercher et al. described that patients with chemotherapy, radiotherapy or radiochemotherapy primarily required a PEG. In our study, we were able to show that chemotherapy or radiochemotherapy is associated with PEG placement. The study by Sieron et al. showed similar results. Van der Linden et al. also added that patients who received radiotherapy alone were more likely to receive a prophylactic PEG. It was argued that radiotherapy may worsen existing poor dental status and that poor dental status was associated with PEG placement. Wermker et al. were also able to establish this. Furthermore, patients who were primarily treated with surgery were significantly less likely to require a PEG tube before starting treatment [
39]. The present study was also able to show the same results. Ehrsson et al. stated an average PEG duration of 3 to 12 months. Accordingly, it was confirmed that patients with a PEG tube at the previous follow-up were very likely to still have it at the next follow-up [
40], which is consistent with our results. Patients with a prophylactic PEG used it for longer than patients who only received the PEG tube after completion of treatment [
41]. Nguyen et al. stated a mean duration of 8 months until a prophylactic PEG was removed [
42]. This also explains that patients with prophylactic PEG were still using the PEG more frequently at the follow-up (Z1, Z2 and Z3) than patients with subsequently placed PEG. It is possible that the PEG dependency is the reason why patients with a prophylactic PEG were significantly less able to feed themselves orally at the individual follow-up than patients without a prophylactic PEG tube.
There were significant differences in weight and BMI between patients with and without PEG. Patients with PEG showed a lower weight and thus also a lower BMI than patients without PEG. Wermker et al. were also able to demonstrate the same results, but the Karnofsky performance status was not affected by these significant differences, as was the case in our study. In their retrospective study of 186 patients with HNC who received a PEG placement prior to treatment and underwent radiotherapy or chemoradiotherapy, Lang et al. categorized all patients into three groups according to their weight and BMI [
43]. Group 2 (BMI/weight range 12–18/33 kg–63 kg), which is most comparable to the patients in our study, showed only a slight change in weight (initial weight: 66.5 kg; weight at the end of treatment: 65 kg) and BMI (initial BMI: 22.1, BMI at the end of treatment: 21.6). The ANOVA in our study showed that with regard to the group differences between patients with and without prophylactic PEG, there were no differences in weight loss and in reduction in Karnofsky performance status between the two groups. It was only recognizable that patients with PEG had a lower initial weight than patients without PEG and had higher initial values in Karnofsky performance status. Patients with PEG had an average weight loss of 5.48 kg. This corresponds to a percentage value of 7.6%. In the study by Nugent et al., the total weight loss of patients with PEG was only 5.2%. However, it should be noted that in Nugent et al., the patients with prophylactic PEG showed a much lower weight loss than the patients with subsequently placed PEG. The values are slightly distorted due to these differences in weight loss. In Nugent et al., the percentage of weight loss would be 7.5% if only PEG tubes placed during or after therapy were considered. This value is also comparable with the results of our study [
31]. In summary, the percentage weight loss of 7.6% is within the expected 6–12% for patients with HNC. Therefore, PEG feeding is considered acceptable. Regarding the quality of life, it was found that patients with PEG tubes reported more frequent discomfort with food intake. In addition, these patients showed greater weight loss. It could therefore be assumed that patients with a PEG tube had a lower quality of life than patients with oral nutrition. Low weight loss also had a positive effect on quality of life. As patients with oral nutrition showed the least weight loss in this study, it can be assumed that these patients had the highest quality of life. Prevost et al. also found that a better nutritional status correlates with a better quality of life [
44]. Ehrsson et al. described that patients with enteral nutrition had a lower quality of life than patients with oral nutrition. The patients found the enteral feeding tubes irritating and missed oral nutrition [
45].
The decision of who indicates/decides to insert a PEG tube varies considerably depending on the institution. To this end, the management of tube feeding in an interdisciplinary context in Germany was analyzed in a web-based survey by seventy participants (forty-two radiation oncologists, twelve medical oncologists, fourteen head and neck surgeons and two physicians from several specialties) in a study by Löser et al. [
46]. The study showed that a significant proportion of participants (70%) do not perform standardized nutritional screening prior to a planned chemoradiation and that there are significant differences between institutions and specialties in Germany. However, several studies have already confirmed the benefits of pre-therapeutic nutritional screening [
47,
48]. Löser et al. cited work overload and staff shortages in the German healthcare system as possible reasons for not carrying out pre-therapeutic nutritional screening. Furthermore, Löser et al. stated that 80% of respondents were in favor of prophylactic PEG tube insertion. However, it should be noted that radiation oncologists are overrepresented in this web-based study.
The present study was limited by a retrospective monocentric study design. As a result, decisions regarding nutritional treatment could not be retraced retrospectively. Thus, associations could be analyzed on the basis of the available data, but no causal relationships could be determined. Furthermore, missing values, although this was rare, could have led to a selection bias. In addition, there were also incomplete data sets for instances like socioeconomic factors or the time of death. Furthermore, the impact of PEG tube use on overall survival is an important factor but was not the focus of the present study. We are currently conducting a study to investigate the impact of PEG tube use on survival. Although nutritional risk screening scores typically do not include blood values due to high variability [
49], some factors like albumin seem to be surrogate markers for the nutritional status of a head and neck cancer patient [
50]. Hence, it would be worthwhile to include metabolic blood values in further investigations.
Overall, the present study was able to analyze PEG for the nutrition of patients with HNC before treatment and after several follow-ups. This is in contrast to many studies in the literature. There is often an analysis of just one point time before and after treatment. The patient data were collected from a tertiary university hospital. A prospective study in a larger multicenter setting could provide further and more precise information on the nutritional status in HNC. In any case, nutritional aspects should be included in the therapeutic measures for HNC.