- Email: [email protected]
The impact of HPV status on weight loss and feeding tube use in oropharyngeal carcinoma
Oral Oncology 79 (2018) 33–39
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The impact of HPV status on weight loss and feeding tube use in oropharyngeal carcinoma
T
⁎
Belinda Vangelova, , Damian P. Kotevskia, Janet R. Williamsa,b, Robert I. Smeea,b,c a
Department of Radiation Oncology, The Nelune Comprehensive Cancer Centre, The Prince of Wales Hospital, Randwick, New South Wales, Australia University of New South Wales Clinical Teaching School, The Prince of Wales Hospital, Randwick, New South Wales, Australia c Department of Radiation Oncology, Tamworth Base Hospital, Tamworth, NSW, Australia b
A R T I C L E I N F O
A B S T R A C T
Keywords: Human papillomavirus HPV Oropharynx cancer Weight loss Radiotherapy Head and neck cancer Prophylactic feeding tubes Reactive feeding tubes Chemotherapy Nutrition
Objectives: It has been well established that patients with oropharyngeal carcinoma are at high nutritional risk, with significant weight loss and tube feeding common. Human papillomavirus (HPV)-associated disease has led to a change in the “typical” presentation and nutritional profile of this population. The aim of our study was to determine whether the need for a feeding tube, and weight loss during radiotherapy (RT) in patients with oropharyngeal carcinoma differed with HPV status. Materials and methods: Patients who received curative RT ± chemotherapy from January 2011 to January 2016 were included (n = 100). We retrospectively evaluated feeding tube use and timing of insertion (prophylactic vs reactive), percentage weight loss during RT and the prevalence of critical weight loss (CWL) ≥5%. Results: HPV-positive patients had significantly higher weight loss during RT compared to the rest of the cohort (8.4% vs 6.1%, 95%CI 0.8–3.9, p = 0.003). CWL was observed in 86% and in a higher proportion with HPVpositive disease (93%, p = 0.011). Conditional probability modelling analysis revealed, with 74% accuracy, concurrent chemoradiotherapy and HPV-positive status were predictors of CWL when comparing HPV-positive patients to HPV-negative (96%, p = 0.001 and 98%, p = 0.012 respectively). More HPV-positive patients required feeding tubes (n = 43, 63%, p = 0.05), most being reactive (n = 27, 63%). All patients with reactive tubes experienced CWL. Conclusion: The high incidence of CWL in patients with HPV-positive oropharyngeal carcinoma is of concern. Tube feeding continues to be a necessary nutritional intervention in this population and predicting who will require a tube is challenging. Larger, prospective cohort studies are required.
Introduction Patients with oropharyngeal carcinoma (OPC) are at nutritional risk due to the effects of the tumour itself and the toxicities of the treatment modalities used [1–3]. It has been well established that malnutrition and weight loss is common in head and neck cancer (HNC), and critical weight loss (CWL) (defined as ≥5% loss in one month [4,5]), has been shown to affect outcomes and survival [6–8]. The epidemiology of OPC has changed in the last decade with the increasing incidence of human papillomavirus (HPV)-associated disease [9]. HPV-positive patients often present at a younger age,[10] are more likely to be life nonesmokers [11,12] and are less likely to present with tumour-related dysphagia or odynophagia affecting oral intake [13]. As a result, the identification of patients who are more likely to experience the toxicities of treatment that potentially lead to malnutrition and the need for
a feeding tube, is an increasing challenge for clinicians. Controversy remains in regards to the optimal timing of feeding tube insertion in HNC [14], with some studies advocating for prophylactic placement [15–20], and others indicating insertion only as needed [16,21–23]. Although de-intensification of radiotherapy (RT) in the treatment of HPV-positive OPC may reduce the level of treatment toxicities in the future, the current use of high dose RT and concurrent chemoradiotherapy (CRT) in this population continues to result in acute and late toxicities that compromise nutritional status, often necessitating the need for a feeding tube. To our knowledge, there are no studies to date that have investigated the timing of feeding tube insertions and associated weight loss, in relation to HPV status. Therefore, the purpose of this study was to determine whether HPV status influences the need for, and timing of, feeding tube insertion, the incidence of CWL, and whether there is a
⁎ Corresponding author at: Department of Radiation Oncology Research Precinct, Nelune Comprehensive Cancer Centre, Level 1, Bright Building, Avoca Street, Randwick, NSW 2031, Australia. E-mail address: [email protected] (B. Vangelov).
https://doi.org/10.1016/j.oraloncology.2018.02.012 Received 6 December 2017; Received in revised form 5 February 2018; Accepted 11 February 2018 Available online 22 February 2018 1368-8375/ © 2018 Elsevier Ltd. All rights reserved.
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
investigation, these patients have been included in the PFT group. Insertion of a PFT was decided by a multidisciplinary team (MDT) at the time of patient presentation, based on clinical presentation, perceived expected need for enteral feeding during RT, and treating Oncologist preferences. PFTs were inserted prior to, or within the first 2 weeks of RT, and were either a percutaneous endoscopic gastrostomy (PEG) or surgically inserted balloon gastrostomy. RFTs were inserted on a case by case basis in response to poor oral intake and subsequent weight loss during RT or within 2 weeks of completion. RFTs were predominantly nasogastric tubes (NGTs) and in some cases, proceeded to gastrostomy insertion. Duration of tube feeding was calculated from the date feeding commenced via the tube, to the date ceased.
difference in weight loss during radiotherapy ( ± chemotherapy), in patients with HPV-positive disease. Materials and methods Study design This was a single-institution retrospective study of all adult (≥18 years) patients with newly diagnosed OPC who received definitive RT either as a primary, adjuvant, or combined modality treatment at The Nelune Comprehensive Cancer Centre in Sydney, Australia between January 2011 and January 2016. This is a South Eastern Sydney Local Health District Area Health Ethics approved research project (HREC11/070).
Statistical analysis
Patient eligibility
Continuous data are presented as means for normally distributed data and medians for non-normally distributed data. All categorical data are presented as frequencies (%). One-way ANOVA was used to analyse the differences in weight change during RT across 3 tube groups (ie: PFT vs RFT vs no tube) in the whole cohort and by HPV status. Data was then dichotomised by HPV status, and an Independent samples Ttest was used to assess differences in percentage weight change between the two groups. Kaplan-Meier analysis using the log-rank test was performed to investigate the difference between prophylactic and reactive gastrostomy tube feeding duration in HPV-positive patients. CWL was investigated dichotomously by HPV status in univariate analysis using the Fisher’s exact test. The probability of CWL with HPV status was also analysed using binary logistic regression (using the backward likelihood ratio method) for both univariate and multivariate models. Prediction analysis was performed using SPSS Modeler 18.0 (IBM, Armonk, NY). Data was partitioned into training data to build a Chi-square automatic interaction detection (CHAID) model, and testing data to evaluate the accuracy of the model using Receiver Operating Characteristic curves. Two models were constructed using CHAID to predict variables that increased the probability of a patient: (1) experiencing CWL, and (2) needing a RFT. Patient characteristics and treatment specifics were used as predictor variables. These included: HPV status, age, gender, stage, treatment modality, RT dose, neck node irradiation and pre-treatment weight loss. Models were performed twice, once including the HPV-unknown status patients and again with only those confirmed HPV-negative cases. Prediction of the need for a RFT was analysed comparing those with a RFT to those with no tube. Significance level was set at p < 0.05 for all analyses. All other statistical analysis was conducted using SPSS Statistics 24.0 (IBM, Armonk, NY).
Eligibility criteria included: pathology confirmed oropharynx carcinoma (newly diagnosed, previously untreated), treatment with curative intent, and entire RT completion within our facility. Patients who underwent surgical resection (namely wide local excision, diagnostic tonsillectomy, excisional biopsy or neck node dissection) prior to RT +/− chemotherapy were also included. Patients were excluded if referred for recurrent or progressive disease, distant metastases or treated with palliative intent. HPV status was collected from confirmed histopathology reports where p16 immunostaining had been completed. Positive p16 results were associated with HPV-positive disease status as previously described [24]. Patients were not routinely tested for HPV in the earlier years of the study if they did not exhibit the common characteristics for the disease. These patients were grouped with confirmed HPV-negative patients for the purpose of comparison to those with HPV-positive disease. Patients were dichotomized into two groups: HPV-positive vs HPV-negative and unknown status. Those with unknown status were then removed for additional analysis with prediction modelling. All data were sourced from medical and dietetic documentation in both electronic (MOSAIQ) and paper Oncology Department records, and hospital medical records. Treatment All patients received RT administered daily as 5 fractions per week over 6 weeks, delivered as either Intensity-Modulated Radiation Therapy (IMRT), or conformal 3D RT to the primary tumour site, and to neck nodes, either unilaterally or bilaterally when indicated. RT included concomitant boosts where required and all treatment was delivered via external beam using a 6MV Linear Accelerator. Concurrent chemotherapy was administered where appropriate using either cisplatin (weekly or third-weekly) or cetuximab.
Results Patient characteristics
Weight loss One hundred patients were eligible for inclusion, the majority being male (85%). Sixty-eight percent had HPV-positive disease, with the remainder either HPV-negative (n = 10) or of unknown status (n = 22). The highest proportion of oropharynx sub-site were tonsil tumours (49%), however, in the HPV-positive cohort, the base of tongue tumours were slightly more prevalent (50%). Concurrent CRT was the most commonly used treatment modality in the whole cohort (78%), and the HPV-positive cohort (87%), with concurrent cisplatin administered in 86% of these. Over two-thirds of patients received the median RT dose of 68 Gy in 34 fractions over 6 weeks. Thirty-one percent of the whole cohort presented with unintentional weight loss, with 29% of the HPV-positive group having lost weight prior to diagnosis. Of these HPV-positive patients, 60% (n = 12) presented with weight loss ≥ 5%. Demographic and clinical characteristics are summarised in Table 1.
Percentage weight change during treatment was calculated using weight at week one of RT as baseline and weight in the last week of RT. Mean percentage weight change of the whole cohort was calculated weekly during treatment, and presented as total weight change percentage at the end of RT. CWL was defined as ≥5% weight loss during this period. Tube feeding Data were collected on the timing and type of feeding tube inserted. Feeding tubes were defined as either: (a) prophylactic feeding tubes (PFT), or (b) reactive feeding tubes (RFT). A third type of tube – the “therapeutic” tube, was defined as a tube inserted during surgery in anticipation of dysphagia post operatively. For the purposes of this 34
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Table 1 Patient demographics and clinical characteristics. Whole cohort = (n = 100) N (%)
HPV-positive (n = 68) N (%)
HPV-negative (n = 10) N (%)
Unknown HPV status (n = 22) N (%)
PFT (n = 28)
RFT (n = 33)
N (%)
N (%)
No Tube (n = 39) N (%)
Gender Male Female
85 (85) 15 (15)
59 (87) 9 (13)
6 (60) 4 (40)
20 (91) 2 (9)
25 (89) 3 (11)
29 (88) 4 (12)
31 (80) 8 (20)
Age (years) Mean (range)
61 (32–90)
59 (32–85)
67 (55–90)
64 (46–87)
62 (46–85)
58 (32–69)
62 (36–90)
Primary site Tonsil Base of tongue Other
49 (49) 46 (46) 5 (5)
33 (49) 34 (50) 1 (1)
5 (50) 4 (40) 1 (10)
11 (50) 8 (36) 3 (14)
10 (36) 17 (61) 1 (3)
18 (55) 14 (42) 1 (3)
21 (54) 15 (39) 3 (7)
T- stage T1 T2 T3 T4 Tx
29 (29) 34 (34) 24 (24) 6 (6) 7 (7)
23 (34) 22 (32) 13 (19) 4 (6) 6 (9)
0 5 (50) 3 (30) 2 (20) 0
6 7 8 0 1
(5)
7 8 8 3 2
(25) (29) (29) (11) (7)
8 (24) 11 (33) 9 (27) 3 (9) 2 (6)
14 (36) 15 (38) 7 (18) 0 3 (8)
N-stage N0 N1 N2 N3
12 (12) 27 (27) 53 (53) 8 (8)
4 (6) 18 (27) 39 (57) 7 (10)
3 (30) 2 (20) 5 (50) 0
5 7 9 1
(22) (32) (41) (5)
3 (11) 2 (7) 22 (79) 1 (3)
2 (6) 12 (36) 15 (46) 4 (12)
7 (18) 13 (33) 16 (41) 3 (8)
Modality RT only Sx + RT CRT ( ± Sx)
15 (15) 7 (7) 78 (78)
7 (10) 2 (3) 59 (87)
3 (30) 4 (40) 3 (30)
5 (23) 1 (4) 16 (73)
2 (7) 3 (11) 23 (82)
0 0 33
13 (33) 4 (11) 22 (56)
Chemotherapy agent Cisplatin Induction + cisplatin Cetuximab
62 (62) 1 (1) 15 (15)
51 (75) 1 (2) 7 (10)
2 (20) 0 1 (10)
9 (41) 0 7 (32)
15 (54) 0 8 (29)
28 (85) 1 (3) 4 (12)
19 (53) 0 3 (8)
Radiotherapy IMRT Conformal 3D
73 (73) 27 (27)
52 (77) 16 (23)
6 (60) 4 (40)
15 (68) 7 (32)
27 (97) 1 (3)
24 (73) 9 (37)
22 (56) 17 (44)
Radiation Dose (Gy) Median (IQ range)
68
68
62 (6)
68
68
68
68
Neck nodes treated None Unilateral Bilateral
1 (1) 48 (48) 51 (51)
0 31 (46) 37 (54)
0 5 (50) 5 (50)
1 (4) 12 (55) 9 (41)
0 4 (14) 24 (86)
0 13 (39) 20 (61)
1 (3) 31 (79) 7 (18)
Pre-Tx wt loss Yes No
31 (31) 69 (69)
20 (29) 48 (71)
5 (50) 5 (50)
6 (27) 16 (73)
11 (39) 17 (61)
13 (39) 20 (61)
7 (18) 32 (82)
CWL Yes No
86 (86) 14 (14)
63 (93) 5 (7)
6 (60) 4 (40)
17 (77) 5 (23)
22 (79) 6 (21)
33 (100) 0
31 (80) 8 (20)
Feeding tubes Prophylactic Reactive No tube
28 (28) 33 (33) 39 (39)
16 (24) 27 (40) 25 (36)
5 (50) 1 (10) 4 (40)
7 (32) 5 (23) 10 (45)
– – –
– – –
– – –
(27) (32) (36)
Abbreviations: HPV, human papillomavirus; RT, radiotherapy; IMRT, intensity modulated radiotherapy; IQ, inter-quartile; Sx, surgery; Pre-Tx, pre-treatment; wt, weight; CWL, critical weight loss; PFT, prophylactic feeding tube; RFT, reactive feeding tube.
rest of the cohort (n = 23/32, 72%) (p = 0.011). On prediction modelling, HPV-positive status and CRT were found to be predictors of CWL, with our model having 83% prediction accuracy (Fig. 2a). Gender also featured as a predictor, with a high male to female ratio, consistent with OPC demographics. Other characteristics included in the model were not found to be predictive. When prediction modelling was repeated without the “unknown” group, the same predictors were found: HPV-positive, CRT and gender (model accuracy −74%) (Fig. 2b). Logistic regression analysis revealed HPV-positive status and CRT to be predictors of CWL at univariate level. In multivariate analysis, CRT (OR 4.9, 95% CI 1.4–17.3, p = 0.014) retained significance, however HPV status did not (OR 3.3, 95% CI 0.9–11.7, p = 0.068). More significant predictors of CWL were identified using CHIAD.
Weight change The mean percentage weight change during treatment across the whole cohort was −7.7 ± 3.7%. All but one HPV-positive patient lost weight during RT, with the range of weight loss 0–17%. HPV-positive patients had significantly higher mean percentage weight loss during RT compared to the rest of the cohort (8.4% vs 6.1%) (mean difference 2.3%, 95% CI 0.8–3.9, p = 0.003). Fig. 1 illustrates the distribution of weight change by HPV status, with the largest proportion of HPV-positive patients losing between 5% and 10%. Highest mean percentage weight loss occurred in the final week of RT. CWL was observed in 86% of patients, and in a significantly higher number of HPV-positive patients (n = 63/68, 93%), compared with the 35
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Fig. 1. Percentage weight change during radiotherapy.
Fig. 2. Predictive modelling for Critical Weight Loss (CWL) using CHAID (a) Including unknowns (prediction accuracy 83%) (b) Excluding unknowns (prediction accuracy 74%).
56% others) (p = 0.050), most being RFTs (n = 27/43, 63%), and only 4/12 patients presenting with ≥5% loss at diagnosis actually had a PFT inserted. No HPV-positive patients had a therapeutic tube. Twenty-eight patients had PFTs inserted. Sixteen HPV-positive patients received a PFT (24%). Twenty-eight RFTs were inserted during
Feeding tube use Sixty-one patients required a feeding tube, with the majority inserted reactively (n = 33/61, 54%). A larger proportion of HPV-positive patients required feeding tubes (n = 43/68, 63% vs n = 18/32, 36
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Of note is that the mean percentage weight change in both HPV status groups was at critical level (≥5%). When analysing the HPVpositive group alone, the only significant difference in mean percentage weight loss was observed between those who had RFTs and those with no tubes (the former losing more, 9.6% vs 7.1%) (p = 0.023), again, both means at critical level. Discussion This study aimed to investigate whether feeding tube use and subsequent weight loss during RT ± chemotherapy, varied with HPV status. Determining which patients with OPC will require intensive nutrition support is clinically challenging, especially in light of the changing epidemiology of the patient with HPV-positive disease. Patients may not present with malnutrition or unintentional weight loss at diagnosis, and tumour-related symptoms such as odynophagia may not impact on oral intake at that stage. Deciding whether to advocate for the insertion of a PFT in these patients is a cause for debate within our MDT. It has been well established that weight loss during treatment for HNC is common, and unintentional weight lost prior to diagnosis is a predictor of poorer outcomes [6,25,26]. The number of patients in our study presenting with unintentional weight loss at diagnosis is low (31%) and we were unable to demonstrate that this is a predictor of further losses during treatment, or the need for a feeding tube. In fact, 84% of patients who had no pre-treatment weight loss, experienced CWL during treatment and almost a third required reactive feeding tubes. The proportion of HPV-positive patients presenting with unintentional weight loss was even lower at 29% and despite more than half presenting with ≥5% loss, only 4 had a PFT. In our facility, there is a reluctance to insert PFTs in patients who are younger, fitter and “motivated” to continue with oral intake for as long as possible during treatment. Despite a perceived level of “motivation” amongst patients with HPV-positive disease, we have demonstrated that an alarming number experience CWL, especially in the last week of treatment when toxicities are reaching their peak. The prevalence of CWL was higher than has been previously reported in HNC. Langius et al. found a 50% prevalence of CWL in their heterogenous HNC cohort, however, HPV status was not included [8]. Other studies have looked at weight loss during RT in heterogenous HNC groups with lower prevalence, likely due to the inclusion of other HNC sites, eg larynx, where weight loss may be lower than in OPC [1,26,27]. The percentage of HPV-positive patients with CWL in our cohort is considerably high at 93% and we have shown that CRT had a significant impact on this. The reason for HPV-positive patients experiencing higher levels of CWL needs further investigation to better understand if there is a difference in the toxicity profile dependent on HPV status [28]. Bledsoe et al. found similar acute toxicity rates between HPV-positive and negative status, and similar rates of tube feeding in patients with OPC, however, weight loss as a result of toxicities was not investigated [29]. A study by Vatca et al. identified HPV status as an independent predictor of grade 3–4 mucositis during CRT and HPV-positive status was associated with higher percentage change in body weight when compared to HPV-negative patients, however, only 15 HPV-negative patients were included and CWL or weight loss associated with the timing of feeding tube insertion was not investigated [30]. Feeding tubes are used frequently in HNC patients during RT/CRT, however, the perceived optimal timing of insertions vary institutionally and remains controversial due to a lack of high quality evidence to guide practice [14]. In our cohort, a higher number of RFTs were required, the majority in the HPV-positive group, indicating that there is a clinically significant proportion of patients unable to meet nutritional requirements orally during treatment, and those with HPV-positive disease appear to be most at risk. Kramer et al. investigated prophylactic and reactive PEG insertion
Fig. 3. Duration of gastrostomy tube feeding in HPV-positive patients. Log Rank p = 0.093.
RT, most being NGTs (n = 21). Five patients had RFT insertion post RT. Eighty-two percent of PFT were being used by week 4 of RT, with 72% of RFTs inserted and used in week 5 and 6. Thirty-nine patients did not have a feeding tube (64% of these patients were HPV-positive, n = 25) (Table 1). When comparing prophylactic gastrostomies with reactive gastrostomies, there was no significant difference in the mean length of time tubes were used for feeding (89 vs 78 days respectively, p = 0.60), and no difference was observed specifically in the HPV-positive group (71 vs 83 days respectively, p = 0.93) (Fig. 3). No HPV-positive patients were using PFTs six months post RT and only one HPV-positive patient was still using a reactively inserted gastrostomy at this time point. Prediction modelling was unable to provide predictive variables for the type of tube required in HPV-positive patients, with equal predictive numbers in prophylactic and reactive groups. Weight loss and tube use All patients with RFTs experienced CWL and of those HPV-positive patients with a PFT, 94% also had CWL. The incidence of CWL amongst the tube groups in relation to HPV status is shown in Table 2. A greater proportion of patients with RFTs had CWL compared to those with PFTs and no tube (p = 0.006). HPV-positive patients with PFTs had significantly higher mean percentage weight loss when compared to those with PFTs in the remaining cohort (8.6% vs 3.9% respectively) (mean difference 4.7%, 95% CI 1.7–7.6, p = 0.003). This was not the case when comparing RFTs and the no tube group (Fig. 4). Patients with RFTs in the HPV-negative and unknown group lost significantly more weight compared to patients with a PFT or no tube (p = 0.038). No difference was observed in the HPV-positive group when comparing mean percentage weight loss by tube type (p = 0.889) (Fig. 4). Table 2 Critical weight loss incidence with tube type.
Feeding tubes Prophylactic Reactive No tube
Whole cohort N = 86
HPV-positive N = 63
HPV-negative N=6
Unknown status N = 17
22 (79%) 33 (100%) 31 (80%)
15 (94%) 27 (100%) 21 (84%)
3 (60%) 1 (100%) 2 (50%)
4 (57%) 5 (100%) 8 (80%)
Note: Percentages display the proportion of patients (by HPV status) who experienced critical weight loss.
37
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Fig. 4. Mean percentage weight change: HPV status and tube type. Note: Bars represent the mean percentage weight change (value included) with 95% confidence intervals.
discussed, HPV-positive patients are often a younger, fitter demographic, and using a tube whilst they are still able to consume some food orally, is often not seen as a priority, despite recommendations from the MDT. Patients mostly commenced using their PFTs in weeks 3 and 4 of RT and most reactive feeding commenced in week 5 and 6. The largest weight losses occurred in the last weeks of RT and are likely the main reason RFTs were inserted. It could be argued that patients with PFTs should commence using tubes earlier to prevent CWL. Wiggenraad et al. demonstrated that weight loss may be limited with the early use of a PFT in HNC patients who receive CRT, with mean weight loss of only 2.8% [20] compared to 8.6% in our HPV-positive cohort. This, however, was not shown to be the case in a mixed HNC study by Brown et al. where there was no difference in weight change when patients commenced early feeding via PFT [36]. Therefore, it remains unclear whether earlier nutritional intervention would impact on weight loss and ultimately nutritional status specifically in HPV-positive patients. PFTs being used in reaction to poor oral intake and weight loss, potentially adds to the incidence of CWL in our study. The rate of CWL was also high in patients with no tube, indicating that although 39% of patients completed treatment without enteral feeding, some may have benefitted from the additional nutritional support and possible CWL prevention. Many patients refuse the insertion of a tube despite weight loss and malnutrition, choosing instead to persevere for as long as possible. In light of the findings of our study, methods and timing of nutritional interventions may need to be tailored differently for the HPV-positive population, specifically in our MDT decision-making processes for feeding tube insertion. This is a single institution, retrospective study and as such has several limitations. Our cohort is relatively small, and lacking a complete set of HPV status results. Some patients in the “unknown” group may have in fact been HPV-positive, which may impact on the final results. We attempted to address this potential bias by running the prediction model again without the unknown status group and found the same predictors, even with small numbers in the HPV-negative arm.
in a mixed HNC group and found that PFTs remained in place for significantly longer than RFTs, however, there was no distinction between how long tubes were actually used for feeding and when they were removed [23]. Our results indicate that in the HPV-positive population, there is no difference between the length of time prophylactic gastrostomy and reactive gastrostomy tubes were used, and in fact, all PFT patients had stopped using their tubes 6 months post treatment. PFTs have been described in the literature as barriers to patients returning to normal oral intake and can even impact on dysphagia long term, creating “tube dependence” [31]. This may be the case in HPV-negative patients as shown by Ward et al., where those with prophylactic PEGs had a higher incidence of severe late dysphagia [32]. In another study, HPV-positive patients had reduced swallowing dysfunction after CRT when compared to HPV-negative patients, however, rarely required long term feeding tubes [33]. Setton et al. demonstrated low rates of long term tube dependence in a large cohort of non-surgically treated OPC patients across three institutions, although HPV status was not included [34]. Other studies compare reactive NG feeding to gastrostomy feeding, showing the former are used for significantly shorter periods of time [31,35]. However, NGs are specifically designed for short term use and often removed when some oral intake is resumed, which may not necessarily be the patients’ full nutritional requirement. Gastrostomies are used for longer periods to supplement oral intake, often until almost all nutritional requirements are met orally. Many studies lack the documentation of volumes administered via a tube and whether a patient is solely reliant on nutrition via that tube, and hence the actual level of “dependence”. Therefore the concern of tube dependence needs better defining in OPC patients and may not be a valid reason to avoid gastrostomies in the HPV-positive population, warranting further investigation. The majority of patients with PFTs experienced CWL in our study, and HPV-positive patients lost significantly more weight on average than the rest of the cohort. This could in part be due to the reluctance of patients to commence using their tubes for feeding. As previously
38
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
The rising incidence of HPV-positive tumours and their proportionately higher numbers, make comparisons with equal numbers of HPV-negative patients challenging. There were too few patients in our HPV-negative group for gastrostomy feeding duration comparisons. Full nutritional assessments were not available for all patients prior to treatment commencement and this may have been beneficial when looking at feeding tube prediction. Larger numbers are needed for future research, with specific attention to the nutritional management of patients with HPV-positive disease. CWL in the HPV-positive population is of concern, considering improved survival rates, and prevention of long-term treatment-related toxicities is becoming more important. Tube feeding remains an important and appropriate clinical intervention, as oral intake is almost certainly compromised with the current treatment modalities utilised. Future changes in treatments for this patient group may reduce toxicities, however, until such changes occur, HPV-positive patients with OPC remain nutritionally vulnerable and further research specific to this population is needed to guide practice.
[14]
[15]
[16]
[17]
[18]
[19]
[20]
Conflict of interest and source of funding [21]
As with all material emanating from the Head & Neck Database (H& N Database) this is an Ethics approved study, number ∗HREC Ref: (11/ 070). This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. All contributing authors agree to the content and publishing of this manuscript. All work conforms to the provisions of the Declaration of Helsinki.
[22] [23] [24]
Acknowledgement
[25]
The authors wish to acknowledge the entire Head and Neck Multidisciplinary Team at The Prince of Wales Hospital involved in the ongoing management and care of the patients included in this study.
[26]
[27]
References [28]
[1] Beaver ME, Matheny KE, Roberts DB, Myers JN. Predictors of weight loss during radiation therapy. Otolaryngology - Head Neck Surg 2001;125:645–8. [2] Kubrak C, Olson K, Jha N, Jensen L, McCargar L, Seikaly H, et al. Nutrition impact symptoms: key determinants of reduced dietary intake, weight loss, and reduced functional capacity of patients with head and neck cancer before treatment. Head Neck 2009;32:290–300. [3] van Bokhorst-de van der Schueren MAE, van Leeuwen PAM, Kuik DJ, Klop WMC, Sauerwein HP, Snow GB, et al. The impact of nutritional status on the prognoses of patients with advanced head and neck cancer. Cancer 1999;86:519–527. [4] Ottery FD. Definition of standardized nutritional assessment and interventional pathways in oncology. Nutrition 1996;12:S15–9. [5] Bauer J, Capra S, Ferguson M. Use of the scored patient-generated subjective global assessment (PG-SGA) as a nutrition assessment tool in patients with cancer. Eur J Clin Nutr 2002;56:779–85. [6] Langius JA, Bakker S, Rietveld DH, Kruizenga HM, Langendijk JA, Weijs PJ. Critical weight loss is a major prognostic indicator for disease-specific survival in patients with head and neck cancer receiving radiotherapy. Br J Cancer 2013;109. [7] Cho Y, Roh J, Jung J, Kim S, Lee S, Choi S, et al. Prediction of posttreatment significant body weight loss and its correlation with disease-free survival in patients with oral squamous cell carcinomas. Nutr Cancer 2013;65:417–23. [8] Langius JA, Twisk J, Kampman M, Doornaert P, Kramer MH, Weijs PJ, et al. Prediction model to predict critical weight loss in patients with head and neck cancer during (chemo)radiotherapy. Oral Oncol 2016;52:91–6. [9] Gillison M, Chaturvedi A, Anderson W, Fakhry C. Epidemiology of human papillomavirus–positive head and neck squamous cell carcinoma. J Clin Oncol 2015;33:3235–42. [10] Benson E, Li R, Eisele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014;50:565–74. [11] Vokes EE, Agrawal N, Seiwert TY. HPV-Associated Head and Neck Cancer. J Natl Cancer Inst 2015;107:djv344-djv344. [12] Okami K. Clinical features and treatment strategy for HPV-related oropharyngeal cancer. Int J Clin Oncol 2016;21:827–35. [13] McIlwain WR, Sood AJ, Nguyen SA, Day TA. Initial symptoms in patients with HPV-
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
39
positive and HPV-negative oropharyngeal cancer. JAMA Otolaryngol Head Neck Surg; 140:441–447. Nugent B, Lewis S, O'Sullivan JM. Enteral feeding methods for nutritional management in patients with head and neck cancers being treated with radiotherapy and/or chemotherapy. Cochrane Database of Syst Rev 2013;1:CD007904. Baschnagel AM, Yadav S, Marina O, Parzuchowski A, Lanni Jr. TB, Warner JN, et al. Toxicities and costs of placing prophylactic and reactive percutaneous gastrostomy tubes in patients with locally advanced head and neck cancers treated with chemoradiotherapy. Head Neck 2014;36:1155–61. Chen A, Li B, Lau D, Farwell D, Luu Q, Stuart K, et al. Evaluating the role of prophylactic gastrostomy tube placement prior to definitive chemoradiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 2010;78:1026–32. Rutter CE, Yovino S, Taylor R, Wolf J, Cullen KJ, Ord R, et al. Impact of early percutaneous endoscopic gastrostomy tube placement on nutritional status and hospitalization in patients with head and neck cancer receiving definitive chemoradiation therapy. Head Neck 2011;33:1441–7. Morton RP, Crowder VL, Mawdsley R, Ong E, Izzard M. Elective gastrostomy, nutritional status and quality of life in advanced head and neck cancer patients receiving chemoradiotherapy. ANZ J Surg 2009;79:713–8. Silander E, Nyman J, Bove M, Johansson L, Larsson S, Hammerlid E. Impact of prophylactic percutaneous endoscopic gastrostomy on malnutrition and quality of life in patients with head and neck cancer: a randomized study. Head Neck 2012;34:1–9. Wiggenraad R, Flierman L, Goossens ABR, Verschuur H, Croll G, Moser L, et al. Prophylactic gastrostomy placement and early tube feeding may limit loss of weight during chemoradiotherapy for advanced head and neck cancer, a preliminary study. Clin Otolaryngol 2007;32:384–90. Clavel S, Fortin B, Despres P, Donath D, Soulieres D, Khaouam N, et al. Enteral feeding during chemoradiotherapy for advanced head and neck cancer: a singleinstitution experience using a reactive approach. Int J Radiat Oncol Biol Phys 2011;79:763–9. Sheth C, Sharp S, Walters E. Enteral feeding in head and neck cancer patients at a UK cancer centre. J Hum Nutr Diet 2013;26:421–8. Kramer S, Newcomb M, Hessler J, Siddiqui F. Prophylactic versus reactive PEG tube placement in head and neck cancer. Otolaryngol Head Neck Surg 2014;150:407–12. Klussmann JP, Gültekin E, Weissenborn SJ, Wieland U, Dries V, Dienes HP, et al. Expression of p16 protein identifies a distinct entity of tonsillar carcinomas associated with human papillomavirus. Am J Pathol 2003;162:747–53. Capuano G, Grosso A, Gentile P, Battista M, Bianciardi F, Di Palma A, et al. Influence of weight loss on outcomes in patients with head and neck cancer undergoing concomitant chemoradiotherapy. Head Neck 2008;30:503–8. Ghadjar P, Hayoz S, Zimmermann F, Bodis S, Kaul D, Badakhshi H, et al. Impact of weight loss on survival after chemoradiation for locally advanced head and neck Cancer: secondary results of a randomized phase III trial (SAKK 10/94). Radiat Oncol 2015;10:21. Lees J. Incidence of weight loss in head and neck cancer patients on commencing radiotherapy treatment at a regional oncology centre. Eur J Cancer Care 1999;8:133–6. Koyfman S, Adelstein D. Enteral feeding tubes in patients undergoing definitive chemoradiation therapy for head-and-neck cancer: a critical review. Int J Radiat Oncol Biol Phys 2011;84:581–9. Bledsoe T, Noble A, Hunter G, Rybicki L, Hoschar A, Chute D, et al. Oropharyngeal squamous cell carcinoma with human papillomavirus status treated with definitive chemoradiotherapy: patterns of failure and toxicity outcomes. Radiat Oncol 2013;8:1–7. Vatca M, Lucas Jr JT, Laudadio J, D’Agostino RB, Waltonen JD, Sullivan CA, et al. Retrospective analysis of the impact of HPV status and smoking on mucositis in patients with oropharyngeal squamous cell carcinoma treated with concurrent chemotherapy and radiotherapy. Oral Oncol 2014;50:869–76. Williams G, Teo M, Sen M, Dyker K, Coyle C, Prestwich R. Enteral feeding outcomes after chemoradiotherapy for oropharynx cancer: a role for prophylactic gastrostomy. Oral Oncol 2012;48:434–40. Ward MC, Bhateja P, Nwizu T, Kmiecik J, Reddy CA, Scharpf J, et al. Impact of feeding tube choice on severe late dysphagia after definitive chemoradiotherapy for human papillomavirus–negative head and neck cancer. Head Neck 2016;38:E1054–60. Naik M, Ward MC, Bledsoe TJ, Kumar AMS, Rybicki LA, Saxton JP, et al. It is not just IMRT: human papillomavirus related oropharynx squamous cell carcinoma is associated with better swallowing outcomes after definitive chemoradiotherapy. Oral Oncol 2015;51:800–4. Setton J, Lee N, Riaz N, Huang S, Waldron J, O'Sullivan B, et al. A multi-institution pooled analysis of gastrostomy tube dependence in patients with oropharyngeal cancer treated with definitive intensity-modulated radiotherapy. Cancer 2015;121:294–301. Corry J, Poon W, McPhee N, Milner A, Cruickshank D, Porceddu S, et al. Prospective study of percutaneous endoscopic gastrostomy tubes versus nasogastric tubes for enteral feeding in patients with head and neck cancer undergoing (chemo)radiation. Head Neck 2009;31:867–76. Brown TE, Banks MD, Hughes BGM, Lin CY, Kenny LM, Bauer JD. Randomised controlled trial of early prophylactic feeding vs standard care in patients with head and neck cancer. Br J Cancer 2017;117:15–24.
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The impact of HPV status on weight loss and feeding tube use in oropharyngeal carcinoma
T
⁎
Belinda Vangelova, , Damian P. Kotevskia, Janet R. Williamsa,b, Robert I. Smeea,b,c a
Department of Radiation Oncology, The Nelune Comprehensive Cancer Centre, The Prince of Wales Hospital, Randwick, New South Wales, Australia University of New South Wales Clinical Teaching School, The Prince of Wales Hospital, Randwick, New South Wales, Australia c Department of Radiation Oncology, Tamworth Base Hospital, Tamworth, NSW, Australia b
A R T I C L E I N F O
A B S T R A C T
Keywords: Human papillomavirus HPV Oropharynx cancer Weight loss Radiotherapy Head and neck cancer Prophylactic feeding tubes Reactive feeding tubes Chemotherapy Nutrition
Objectives: It has been well established that patients with oropharyngeal carcinoma are at high nutritional risk, with significant weight loss and tube feeding common. Human papillomavirus (HPV)-associated disease has led to a change in the “typical” presentation and nutritional profile of this population. The aim of our study was to determine whether the need for a feeding tube, and weight loss during radiotherapy (RT) in patients with oropharyngeal carcinoma differed with HPV status. Materials and methods: Patients who received curative RT ± chemotherapy from January 2011 to January 2016 were included (n = 100). We retrospectively evaluated feeding tube use and timing of insertion (prophylactic vs reactive), percentage weight loss during RT and the prevalence of critical weight loss (CWL) ≥5%. Results: HPV-positive patients had significantly higher weight loss during RT compared to the rest of the cohort (8.4% vs 6.1%, 95%CI 0.8–3.9, p = 0.003). CWL was observed in 86% and in a higher proportion with HPVpositive disease (93%, p = 0.011). Conditional probability modelling analysis revealed, with 74% accuracy, concurrent chemoradiotherapy and HPV-positive status were predictors of CWL when comparing HPV-positive patients to HPV-negative (96%, p = 0.001 and 98%, p = 0.012 respectively). More HPV-positive patients required feeding tubes (n = 43, 63%, p = 0.05), most being reactive (n = 27, 63%). All patients with reactive tubes experienced CWL. Conclusion: The high incidence of CWL in patients with HPV-positive oropharyngeal carcinoma is of concern. Tube feeding continues to be a necessary nutritional intervention in this population and predicting who will require a tube is challenging. Larger, prospective cohort studies are required.
Introduction Patients with oropharyngeal carcinoma (OPC) are at nutritional risk due to the effects of the tumour itself and the toxicities of the treatment modalities used [1–3]. It has been well established that malnutrition and weight loss is common in head and neck cancer (HNC), and critical weight loss (CWL) (defined as ≥5% loss in one month [4,5]), has been shown to affect outcomes and survival [6–8]. The epidemiology of OPC has changed in the last decade with the increasing incidence of human papillomavirus (HPV)-associated disease [9]. HPV-positive patients often present at a younger age,[10] are more likely to be life nonesmokers [11,12] and are less likely to present with tumour-related dysphagia or odynophagia affecting oral intake [13]. As a result, the identification of patients who are more likely to experience the toxicities of treatment that potentially lead to malnutrition and the need for
a feeding tube, is an increasing challenge for clinicians. Controversy remains in regards to the optimal timing of feeding tube insertion in HNC [14], with some studies advocating for prophylactic placement [15–20], and others indicating insertion only as needed [16,21–23]. Although de-intensification of radiotherapy (RT) in the treatment of HPV-positive OPC may reduce the level of treatment toxicities in the future, the current use of high dose RT and concurrent chemoradiotherapy (CRT) in this population continues to result in acute and late toxicities that compromise nutritional status, often necessitating the need for a feeding tube. To our knowledge, there are no studies to date that have investigated the timing of feeding tube insertions and associated weight loss, in relation to HPV status. Therefore, the purpose of this study was to determine whether HPV status influences the need for, and timing of, feeding tube insertion, the incidence of CWL, and whether there is a
⁎ Corresponding author at: Department of Radiation Oncology Research Precinct, Nelune Comprehensive Cancer Centre, Level 1, Bright Building, Avoca Street, Randwick, NSW 2031, Australia. E-mail address: [email protected] (B. Vangelov).
https://doi.org/10.1016/j.oraloncology.2018.02.012 Received 6 December 2017; Received in revised form 5 February 2018; Accepted 11 February 2018 Available online 22 February 2018 1368-8375/ © 2018 Elsevier Ltd. All rights reserved.
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
investigation, these patients have been included in the PFT group. Insertion of a PFT was decided by a multidisciplinary team (MDT) at the time of patient presentation, based on clinical presentation, perceived expected need for enteral feeding during RT, and treating Oncologist preferences. PFTs were inserted prior to, or within the first 2 weeks of RT, and were either a percutaneous endoscopic gastrostomy (PEG) or surgically inserted balloon gastrostomy. RFTs were inserted on a case by case basis in response to poor oral intake and subsequent weight loss during RT or within 2 weeks of completion. RFTs were predominantly nasogastric tubes (NGTs) and in some cases, proceeded to gastrostomy insertion. Duration of tube feeding was calculated from the date feeding commenced via the tube, to the date ceased.
difference in weight loss during radiotherapy ( ± chemotherapy), in patients with HPV-positive disease. Materials and methods Study design This was a single-institution retrospective study of all adult (≥18 years) patients with newly diagnosed OPC who received definitive RT either as a primary, adjuvant, or combined modality treatment at The Nelune Comprehensive Cancer Centre in Sydney, Australia between January 2011 and January 2016. This is a South Eastern Sydney Local Health District Area Health Ethics approved research project (HREC11/070).
Statistical analysis
Patient eligibility
Continuous data are presented as means for normally distributed data and medians for non-normally distributed data. All categorical data are presented as frequencies (%). One-way ANOVA was used to analyse the differences in weight change during RT across 3 tube groups (ie: PFT vs RFT vs no tube) in the whole cohort and by HPV status. Data was then dichotomised by HPV status, and an Independent samples Ttest was used to assess differences in percentage weight change between the two groups. Kaplan-Meier analysis using the log-rank test was performed to investigate the difference between prophylactic and reactive gastrostomy tube feeding duration in HPV-positive patients. CWL was investigated dichotomously by HPV status in univariate analysis using the Fisher’s exact test. The probability of CWL with HPV status was also analysed using binary logistic regression (using the backward likelihood ratio method) for both univariate and multivariate models. Prediction analysis was performed using SPSS Modeler 18.0 (IBM, Armonk, NY). Data was partitioned into training data to build a Chi-square automatic interaction detection (CHAID) model, and testing data to evaluate the accuracy of the model using Receiver Operating Characteristic curves. Two models were constructed using CHAID to predict variables that increased the probability of a patient: (1) experiencing CWL, and (2) needing a RFT. Patient characteristics and treatment specifics were used as predictor variables. These included: HPV status, age, gender, stage, treatment modality, RT dose, neck node irradiation and pre-treatment weight loss. Models were performed twice, once including the HPV-unknown status patients and again with only those confirmed HPV-negative cases. Prediction of the need for a RFT was analysed comparing those with a RFT to those with no tube. Significance level was set at p < 0.05 for all analyses. All other statistical analysis was conducted using SPSS Statistics 24.0 (IBM, Armonk, NY).
Eligibility criteria included: pathology confirmed oropharynx carcinoma (newly diagnosed, previously untreated), treatment with curative intent, and entire RT completion within our facility. Patients who underwent surgical resection (namely wide local excision, diagnostic tonsillectomy, excisional biopsy or neck node dissection) prior to RT +/− chemotherapy were also included. Patients were excluded if referred for recurrent or progressive disease, distant metastases or treated with palliative intent. HPV status was collected from confirmed histopathology reports where p16 immunostaining had been completed. Positive p16 results were associated with HPV-positive disease status as previously described [24]. Patients were not routinely tested for HPV in the earlier years of the study if they did not exhibit the common characteristics for the disease. These patients were grouped with confirmed HPV-negative patients for the purpose of comparison to those with HPV-positive disease. Patients were dichotomized into two groups: HPV-positive vs HPV-negative and unknown status. Those with unknown status were then removed for additional analysis with prediction modelling. All data were sourced from medical and dietetic documentation in both electronic (MOSAIQ) and paper Oncology Department records, and hospital medical records. Treatment All patients received RT administered daily as 5 fractions per week over 6 weeks, delivered as either Intensity-Modulated Radiation Therapy (IMRT), or conformal 3D RT to the primary tumour site, and to neck nodes, either unilaterally or bilaterally when indicated. RT included concomitant boosts where required and all treatment was delivered via external beam using a 6MV Linear Accelerator. Concurrent chemotherapy was administered where appropriate using either cisplatin (weekly or third-weekly) or cetuximab.
Results Patient characteristics
Weight loss One hundred patients were eligible for inclusion, the majority being male (85%). Sixty-eight percent had HPV-positive disease, with the remainder either HPV-negative (n = 10) or of unknown status (n = 22). The highest proportion of oropharynx sub-site were tonsil tumours (49%), however, in the HPV-positive cohort, the base of tongue tumours were slightly more prevalent (50%). Concurrent CRT was the most commonly used treatment modality in the whole cohort (78%), and the HPV-positive cohort (87%), with concurrent cisplatin administered in 86% of these. Over two-thirds of patients received the median RT dose of 68 Gy in 34 fractions over 6 weeks. Thirty-one percent of the whole cohort presented with unintentional weight loss, with 29% of the HPV-positive group having lost weight prior to diagnosis. Of these HPV-positive patients, 60% (n = 12) presented with weight loss ≥ 5%. Demographic and clinical characteristics are summarised in Table 1.
Percentage weight change during treatment was calculated using weight at week one of RT as baseline and weight in the last week of RT. Mean percentage weight change of the whole cohort was calculated weekly during treatment, and presented as total weight change percentage at the end of RT. CWL was defined as ≥5% weight loss during this period. Tube feeding Data were collected on the timing and type of feeding tube inserted. Feeding tubes were defined as either: (a) prophylactic feeding tubes (PFT), or (b) reactive feeding tubes (RFT). A third type of tube – the “therapeutic” tube, was defined as a tube inserted during surgery in anticipation of dysphagia post operatively. For the purposes of this 34
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Table 1 Patient demographics and clinical characteristics. Whole cohort = (n = 100) N (%)
HPV-positive (n = 68) N (%)
HPV-negative (n = 10) N (%)
Unknown HPV status (n = 22) N (%)
PFT (n = 28)
RFT (n = 33)
N (%)
N (%)
No Tube (n = 39) N (%)
Gender Male Female
85 (85) 15 (15)
59 (87) 9 (13)
6 (60) 4 (40)
20 (91) 2 (9)
25 (89) 3 (11)
29 (88) 4 (12)
31 (80) 8 (20)
Age (years) Mean (range)
61 (32–90)
59 (32–85)
67 (55–90)
64 (46–87)
62 (46–85)
58 (32–69)
62 (36–90)
Primary site Tonsil Base of tongue Other
49 (49) 46 (46) 5 (5)
33 (49) 34 (50) 1 (1)
5 (50) 4 (40) 1 (10)
11 (50) 8 (36) 3 (14)
10 (36) 17 (61) 1 (3)
18 (55) 14 (42) 1 (3)
21 (54) 15 (39) 3 (7)
T- stage T1 T2 T3 T4 Tx
29 (29) 34 (34) 24 (24) 6 (6) 7 (7)
23 (34) 22 (32) 13 (19) 4 (6) 6 (9)
0 5 (50) 3 (30) 2 (20) 0
6 7 8 0 1
(5)
7 8 8 3 2
(25) (29) (29) (11) (7)
8 (24) 11 (33) 9 (27) 3 (9) 2 (6)
14 (36) 15 (38) 7 (18) 0 3 (8)
N-stage N0 N1 N2 N3
12 (12) 27 (27) 53 (53) 8 (8)
4 (6) 18 (27) 39 (57) 7 (10)
3 (30) 2 (20) 5 (50) 0
5 7 9 1
(22) (32) (41) (5)
3 (11) 2 (7) 22 (79) 1 (3)
2 (6) 12 (36) 15 (46) 4 (12)
7 (18) 13 (33) 16 (41) 3 (8)
Modality RT only Sx + RT CRT ( ± Sx)
15 (15) 7 (7) 78 (78)
7 (10) 2 (3) 59 (87)
3 (30) 4 (40) 3 (30)
5 (23) 1 (4) 16 (73)
2 (7) 3 (11) 23 (82)
0 0 33
13 (33) 4 (11) 22 (56)
Chemotherapy agent Cisplatin Induction + cisplatin Cetuximab
62 (62) 1 (1) 15 (15)
51 (75) 1 (2) 7 (10)
2 (20) 0 1 (10)
9 (41) 0 7 (32)
15 (54) 0 8 (29)
28 (85) 1 (3) 4 (12)
19 (53) 0 3 (8)
Radiotherapy IMRT Conformal 3D
73 (73) 27 (27)
52 (77) 16 (23)
6 (60) 4 (40)
15 (68) 7 (32)
27 (97) 1 (3)
24 (73) 9 (37)
22 (56) 17 (44)
Radiation Dose (Gy) Median (IQ range)
68
68
62 (6)
68
68
68
68
Neck nodes treated None Unilateral Bilateral
1 (1) 48 (48) 51 (51)
0 31 (46) 37 (54)
0 5 (50) 5 (50)
1 (4) 12 (55) 9 (41)
0 4 (14) 24 (86)
0 13 (39) 20 (61)
1 (3) 31 (79) 7 (18)
Pre-Tx wt loss Yes No
31 (31) 69 (69)
20 (29) 48 (71)
5 (50) 5 (50)
6 (27) 16 (73)
11 (39) 17 (61)
13 (39) 20 (61)
7 (18) 32 (82)
CWL Yes No
86 (86) 14 (14)
63 (93) 5 (7)
6 (60) 4 (40)
17 (77) 5 (23)
22 (79) 6 (21)
33 (100) 0
31 (80) 8 (20)
Feeding tubes Prophylactic Reactive No tube
28 (28) 33 (33) 39 (39)
16 (24) 27 (40) 25 (36)
5 (50) 1 (10) 4 (40)
7 (32) 5 (23) 10 (45)
– – –
– – –
– – –
(27) (32) (36)
Abbreviations: HPV, human papillomavirus; RT, radiotherapy; IMRT, intensity modulated radiotherapy; IQ, inter-quartile; Sx, surgery; Pre-Tx, pre-treatment; wt, weight; CWL, critical weight loss; PFT, prophylactic feeding tube; RFT, reactive feeding tube.
rest of the cohort (n = 23/32, 72%) (p = 0.011). On prediction modelling, HPV-positive status and CRT were found to be predictors of CWL, with our model having 83% prediction accuracy (Fig. 2a). Gender also featured as a predictor, with a high male to female ratio, consistent with OPC demographics. Other characteristics included in the model were not found to be predictive. When prediction modelling was repeated without the “unknown” group, the same predictors were found: HPV-positive, CRT and gender (model accuracy −74%) (Fig. 2b). Logistic regression analysis revealed HPV-positive status and CRT to be predictors of CWL at univariate level. In multivariate analysis, CRT (OR 4.9, 95% CI 1.4–17.3, p = 0.014) retained significance, however HPV status did not (OR 3.3, 95% CI 0.9–11.7, p = 0.068). More significant predictors of CWL were identified using CHIAD.
Weight change The mean percentage weight change during treatment across the whole cohort was −7.7 ± 3.7%. All but one HPV-positive patient lost weight during RT, with the range of weight loss 0–17%. HPV-positive patients had significantly higher mean percentage weight loss during RT compared to the rest of the cohort (8.4% vs 6.1%) (mean difference 2.3%, 95% CI 0.8–3.9, p = 0.003). Fig. 1 illustrates the distribution of weight change by HPV status, with the largest proportion of HPV-positive patients losing between 5% and 10%. Highest mean percentage weight loss occurred in the final week of RT. CWL was observed in 86% of patients, and in a significantly higher number of HPV-positive patients (n = 63/68, 93%), compared with the 35
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Fig. 1. Percentage weight change during radiotherapy.
Fig. 2. Predictive modelling for Critical Weight Loss (CWL) using CHAID (a) Including unknowns (prediction accuracy 83%) (b) Excluding unknowns (prediction accuracy 74%).
56% others) (p = 0.050), most being RFTs (n = 27/43, 63%), and only 4/12 patients presenting with ≥5% loss at diagnosis actually had a PFT inserted. No HPV-positive patients had a therapeutic tube. Twenty-eight patients had PFTs inserted. Sixteen HPV-positive patients received a PFT (24%). Twenty-eight RFTs were inserted during
Feeding tube use Sixty-one patients required a feeding tube, with the majority inserted reactively (n = 33/61, 54%). A larger proportion of HPV-positive patients required feeding tubes (n = 43/68, 63% vs n = 18/32, 36
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Of note is that the mean percentage weight change in both HPV status groups was at critical level (≥5%). When analysing the HPVpositive group alone, the only significant difference in mean percentage weight loss was observed between those who had RFTs and those with no tubes (the former losing more, 9.6% vs 7.1%) (p = 0.023), again, both means at critical level. Discussion This study aimed to investigate whether feeding tube use and subsequent weight loss during RT ± chemotherapy, varied with HPV status. Determining which patients with OPC will require intensive nutrition support is clinically challenging, especially in light of the changing epidemiology of the patient with HPV-positive disease. Patients may not present with malnutrition or unintentional weight loss at diagnosis, and tumour-related symptoms such as odynophagia may not impact on oral intake at that stage. Deciding whether to advocate for the insertion of a PFT in these patients is a cause for debate within our MDT. It has been well established that weight loss during treatment for HNC is common, and unintentional weight lost prior to diagnosis is a predictor of poorer outcomes [6,25,26]. The number of patients in our study presenting with unintentional weight loss at diagnosis is low (31%) and we were unable to demonstrate that this is a predictor of further losses during treatment, or the need for a feeding tube. In fact, 84% of patients who had no pre-treatment weight loss, experienced CWL during treatment and almost a third required reactive feeding tubes. The proportion of HPV-positive patients presenting with unintentional weight loss was even lower at 29% and despite more than half presenting with ≥5% loss, only 4 had a PFT. In our facility, there is a reluctance to insert PFTs in patients who are younger, fitter and “motivated” to continue with oral intake for as long as possible during treatment. Despite a perceived level of “motivation” amongst patients with HPV-positive disease, we have demonstrated that an alarming number experience CWL, especially in the last week of treatment when toxicities are reaching their peak. The prevalence of CWL was higher than has been previously reported in HNC. Langius et al. found a 50% prevalence of CWL in their heterogenous HNC cohort, however, HPV status was not included [8]. Other studies have looked at weight loss during RT in heterogenous HNC groups with lower prevalence, likely due to the inclusion of other HNC sites, eg larynx, where weight loss may be lower than in OPC [1,26,27]. The percentage of HPV-positive patients with CWL in our cohort is considerably high at 93% and we have shown that CRT had a significant impact on this. The reason for HPV-positive patients experiencing higher levels of CWL needs further investigation to better understand if there is a difference in the toxicity profile dependent on HPV status [28]. Bledsoe et al. found similar acute toxicity rates between HPV-positive and negative status, and similar rates of tube feeding in patients with OPC, however, weight loss as a result of toxicities was not investigated [29]. A study by Vatca et al. identified HPV status as an independent predictor of grade 3–4 mucositis during CRT and HPV-positive status was associated with higher percentage change in body weight when compared to HPV-negative patients, however, only 15 HPV-negative patients were included and CWL or weight loss associated with the timing of feeding tube insertion was not investigated [30]. Feeding tubes are used frequently in HNC patients during RT/CRT, however, the perceived optimal timing of insertions vary institutionally and remains controversial due to a lack of high quality evidence to guide practice [14]. In our cohort, a higher number of RFTs were required, the majority in the HPV-positive group, indicating that there is a clinically significant proportion of patients unable to meet nutritional requirements orally during treatment, and those with HPV-positive disease appear to be most at risk. Kramer et al. investigated prophylactic and reactive PEG insertion
Fig. 3. Duration of gastrostomy tube feeding in HPV-positive patients. Log Rank p = 0.093.
RT, most being NGTs (n = 21). Five patients had RFT insertion post RT. Eighty-two percent of PFT were being used by week 4 of RT, with 72% of RFTs inserted and used in week 5 and 6. Thirty-nine patients did not have a feeding tube (64% of these patients were HPV-positive, n = 25) (Table 1). When comparing prophylactic gastrostomies with reactive gastrostomies, there was no significant difference in the mean length of time tubes were used for feeding (89 vs 78 days respectively, p = 0.60), and no difference was observed specifically in the HPV-positive group (71 vs 83 days respectively, p = 0.93) (Fig. 3). No HPV-positive patients were using PFTs six months post RT and only one HPV-positive patient was still using a reactively inserted gastrostomy at this time point. Prediction modelling was unable to provide predictive variables for the type of tube required in HPV-positive patients, with equal predictive numbers in prophylactic and reactive groups. Weight loss and tube use All patients with RFTs experienced CWL and of those HPV-positive patients with a PFT, 94% also had CWL. The incidence of CWL amongst the tube groups in relation to HPV status is shown in Table 2. A greater proportion of patients with RFTs had CWL compared to those with PFTs and no tube (p = 0.006). HPV-positive patients with PFTs had significantly higher mean percentage weight loss when compared to those with PFTs in the remaining cohort (8.6% vs 3.9% respectively) (mean difference 4.7%, 95% CI 1.7–7.6, p = 0.003). This was not the case when comparing RFTs and the no tube group (Fig. 4). Patients with RFTs in the HPV-negative and unknown group lost significantly more weight compared to patients with a PFT or no tube (p = 0.038). No difference was observed in the HPV-positive group when comparing mean percentage weight loss by tube type (p = 0.889) (Fig. 4). Table 2 Critical weight loss incidence with tube type.
Feeding tubes Prophylactic Reactive No tube
Whole cohort N = 86
HPV-positive N = 63
HPV-negative N=6
Unknown status N = 17
22 (79%) 33 (100%) 31 (80%)
15 (94%) 27 (100%) 21 (84%)
3 (60%) 1 (100%) 2 (50%)
4 (57%) 5 (100%) 8 (80%)
Note: Percentages display the proportion of patients (by HPV status) who experienced critical weight loss.
37
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
Fig. 4. Mean percentage weight change: HPV status and tube type. Note: Bars represent the mean percentage weight change (value included) with 95% confidence intervals.
discussed, HPV-positive patients are often a younger, fitter demographic, and using a tube whilst they are still able to consume some food orally, is often not seen as a priority, despite recommendations from the MDT. Patients mostly commenced using their PFTs in weeks 3 and 4 of RT and most reactive feeding commenced in week 5 and 6. The largest weight losses occurred in the last weeks of RT and are likely the main reason RFTs were inserted. It could be argued that patients with PFTs should commence using tubes earlier to prevent CWL. Wiggenraad et al. demonstrated that weight loss may be limited with the early use of a PFT in HNC patients who receive CRT, with mean weight loss of only 2.8% [20] compared to 8.6% in our HPV-positive cohort. This, however, was not shown to be the case in a mixed HNC study by Brown et al. where there was no difference in weight change when patients commenced early feeding via PFT [36]. Therefore, it remains unclear whether earlier nutritional intervention would impact on weight loss and ultimately nutritional status specifically in HPV-positive patients. PFTs being used in reaction to poor oral intake and weight loss, potentially adds to the incidence of CWL in our study. The rate of CWL was also high in patients with no tube, indicating that although 39% of patients completed treatment without enteral feeding, some may have benefitted from the additional nutritional support and possible CWL prevention. Many patients refuse the insertion of a tube despite weight loss and malnutrition, choosing instead to persevere for as long as possible. In light of the findings of our study, methods and timing of nutritional interventions may need to be tailored differently for the HPV-positive population, specifically in our MDT decision-making processes for feeding tube insertion. This is a single institution, retrospective study and as such has several limitations. Our cohort is relatively small, and lacking a complete set of HPV status results. Some patients in the “unknown” group may have in fact been HPV-positive, which may impact on the final results. We attempted to address this potential bias by running the prediction model again without the unknown status group and found the same predictors, even with small numbers in the HPV-negative arm.
in a mixed HNC group and found that PFTs remained in place for significantly longer than RFTs, however, there was no distinction between how long tubes were actually used for feeding and when they were removed [23]. Our results indicate that in the HPV-positive population, there is no difference between the length of time prophylactic gastrostomy and reactive gastrostomy tubes were used, and in fact, all PFT patients had stopped using their tubes 6 months post treatment. PFTs have been described in the literature as barriers to patients returning to normal oral intake and can even impact on dysphagia long term, creating “tube dependence” [31]. This may be the case in HPV-negative patients as shown by Ward et al., where those with prophylactic PEGs had a higher incidence of severe late dysphagia [32]. In another study, HPV-positive patients had reduced swallowing dysfunction after CRT when compared to HPV-negative patients, however, rarely required long term feeding tubes [33]. Setton et al. demonstrated low rates of long term tube dependence in a large cohort of non-surgically treated OPC patients across three institutions, although HPV status was not included [34]. Other studies compare reactive NG feeding to gastrostomy feeding, showing the former are used for significantly shorter periods of time [31,35]. However, NGs are specifically designed for short term use and often removed when some oral intake is resumed, which may not necessarily be the patients’ full nutritional requirement. Gastrostomies are used for longer periods to supplement oral intake, often until almost all nutritional requirements are met orally. Many studies lack the documentation of volumes administered via a tube and whether a patient is solely reliant on nutrition via that tube, and hence the actual level of “dependence”. Therefore the concern of tube dependence needs better defining in OPC patients and may not be a valid reason to avoid gastrostomies in the HPV-positive population, warranting further investigation. The majority of patients with PFTs experienced CWL in our study, and HPV-positive patients lost significantly more weight on average than the rest of the cohort. This could in part be due to the reluctance of patients to commence using their tubes for feeding. As previously
38
Oral Oncology 79 (2018) 33–39
B. Vangelov et al.
The rising incidence of HPV-positive tumours and their proportionately higher numbers, make comparisons with equal numbers of HPV-negative patients challenging. There were too few patients in our HPV-negative group for gastrostomy feeding duration comparisons. Full nutritional assessments were not available for all patients prior to treatment commencement and this may have been beneficial when looking at feeding tube prediction. Larger numbers are needed for future research, with specific attention to the nutritional management of patients with HPV-positive disease. CWL in the HPV-positive population is of concern, considering improved survival rates, and prevention of long-term treatment-related toxicities is becoming more important. Tube feeding remains an important and appropriate clinical intervention, as oral intake is almost certainly compromised with the current treatment modalities utilised. Future changes in treatments for this patient group may reduce toxicities, however, until such changes occur, HPV-positive patients with OPC remain nutritionally vulnerable and further research specific to this population is needed to guide practice.
[14]
[15]
[16]
[17]
[18]
[19]
[20]
Conflict of interest and source of funding [21]
As with all material emanating from the Head & Neck Database (H& N Database) this is an Ethics approved study, number ∗HREC Ref: (11/ 070). This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. All contributing authors agree to the content and publishing of this manuscript. All work conforms to the provisions of the Declaration of Helsinki.
[22] [23] [24]
Acknowledgement
[25]
The authors wish to acknowledge the entire Head and Neck Multidisciplinary Team at The Prince of Wales Hospital involved in the ongoing management and care of the patients included in this study.
[26]
[27]
References [28]
[1] Beaver ME, Matheny KE, Roberts DB, Myers JN. Predictors of weight loss during radiation therapy. Otolaryngology - Head Neck Surg 2001;125:645–8. [2] Kubrak C, Olson K, Jha N, Jensen L, McCargar L, Seikaly H, et al. Nutrition impact symptoms: key determinants of reduced dietary intake, weight loss, and reduced functional capacity of patients with head and neck cancer before treatment. Head Neck 2009;32:290–300. [3] van Bokhorst-de van der Schueren MAE, van Leeuwen PAM, Kuik DJ, Klop WMC, Sauerwein HP, Snow GB, et al. The impact of nutritional status on the prognoses of patients with advanced head and neck cancer. Cancer 1999;86:519–527. [4] Ottery FD. Definition of standardized nutritional assessment and interventional pathways in oncology. Nutrition 1996;12:S15–9. [5] Bauer J, Capra S, Ferguson M. Use of the scored patient-generated subjective global assessment (PG-SGA) as a nutrition assessment tool in patients with cancer. Eur J Clin Nutr 2002;56:779–85. [6] Langius JA, Bakker S, Rietveld DH, Kruizenga HM, Langendijk JA, Weijs PJ. Critical weight loss is a major prognostic indicator for disease-specific survival in patients with head and neck cancer receiving radiotherapy. Br J Cancer 2013;109. [7] Cho Y, Roh J, Jung J, Kim S, Lee S, Choi S, et al. Prediction of posttreatment significant body weight loss and its correlation with disease-free survival in patients with oral squamous cell carcinomas. Nutr Cancer 2013;65:417–23. [8] Langius JA, Twisk J, Kampman M, Doornaert P, Kramer MH, Weijs PJ, et al. Prediction model to predict critical weight loss in patients with head and neck cancer during (chemo)radiotherapy. Oral Oncol 2016;52:91–6. [9] Gillison M, Chaturvedi A, Anderson W, Fakhry C. Epidemiology of human papillomavirus–positive head and neck squamous cell carcinoma. J Clin Oncol 2015;33:3235–42. [10] Benson E, Li R, Eisele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014;50:565–74. [11] Vokes EE, Agrawal N, Seiwert TY. HPV-Associated Head and Neck Cancer. J Natl Cancer Inst 2015;107:djv344-djv344. [12] Okami K. Clinical features and treatment strategy for HPV-related oropharyngeal cancer. Int J Clin Oncol 2016;21:827–35. [13] McIlwain WR, Sood AJ, Nguyen SA, Day TA. Initial symptoms in patients with HPV-
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
39
positive and HPV-negative oropharyngeal cancer. JAMA Otolaryngol Head Neck Surg; 140:441–447. Nugent B, Lewis S, O'Sullivan JM. Enteral feeding methods for nutritional management in patients with head and neck cancers being treated with radiotherapy and/or chemotherapy. Cochrane Database of Syst Rev 2013;1:CD007904. Baschnagel AM, Yadav S, Marina O, Parzuchowski A, Lanni Jr. TB, Warner JN, et al. Toxicities and costs of placing prophylactic and reactive percutaneous gastrostomy tubes in patients with locally advanced head and neck cancers treated with chemoradiotherapy. Head Neck 2014;36:1155–61. Chen A, Li B, Lau D, Farwell D, Luu Q, Stuart K, et al. Evaluating the role of prophylactic gastrostomy tube placement prior to definitive chemoradiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 2010;78:1026–32. Rutter CE, Yovino S, Taylor R, Wolf J, Cullen KJ, Ord R, et al. Impact of early percutaneous endoscopic gastrostomy tube placement on nutritional status and hospitalization in patients with head and neck cancer receiving definitive chemoradiation therapy. Head Neck 2011;33:1441–7. Morton RP, Crowder VL, Mawdsley R, Ong E, Izzard M. Elective gastrostomy, nutritional status and quality of life in advanced head and neck cancer patients receiving chemoradiotherapy. ANZ J Surg 2009;79:713–8. Silander E, Nyman J, Bove M, Johansson L, Larsson S, Hammerlid E. Impact of prophylactic percutaneous endoscopic gastrostomy on malnutrition and quality of life in patients with head and neck cancer: a randomized study. Head Neck 2012;34:1–9. Wiggenraad R, Flierman L, Goossens ABR, Verschuur H, Croll G, Moser L, et al. Prophylactic gastrostomy placement and early tube feeding may limit loss of weight during chemoradiotherapy for advanced head and neck cancer, a preliminary study. Clin Otolaryngol 2007;32:384–90. Clavel S, Fortin B, Despres P, Donath D, Soulieres D, Khaouam N, et al. Enteral feeding during chemoradiotherapy for advanced head and neck cancer: a singleinstitution experience using a reactive approach. Int J Radiat Oncol Biol Phys 2011;79:763–9. Sheth C, Sharp S, Walters E. Enteral feeding in head and neck cancer patients at a UK cancer centre. J Hum Nutr Diet 2013;26:421–8. Kramer S, Newcomb M, Hessler J, Siddiqui F. Prophylactic versus reactive PEG tube placement in head and neck cancer. Otolaryngol Head Neck Surg 2014;150:407–12. Klussmann JP, Gültekin E, Weissenborn SJ, Wieland U, Dries V, Dienes HP, et al. Expression of p16 protein identifies a distinct entity of tonsillar carcinomas associated with human papillomavirus. Am J Pathol 2003;162:747–53. Capuano G, Grosso A, Gentile P, Battista M, Bianciardi F, Di Palma A, et al. Influence of weight loss on outcomes in patients with head and neck cancer undergoing concomitant chemoradiotherapy. Head Neck 2008;30:503–8. Ghadjar P, Hayoz S, Zimmermann F, Bodis S, Kaul D, Badakhshi H, et al. Impact of weight loss on survival after chemoradiation for locally advanced head and neck Cancer: secondary results of a randomized phase III trial (SAKK 10/94). Radiat Oncol 2015;10:21. Lees J. Incidence of weight loss in head and neck cancer patients on commencing radiotherapy treatment at a regional oncology centre. Eur J Cancer Care 1999;8:133–6. Koyfman S, Adelstein D. Enteral feeding tubes in patients undergoing definitive chemoradiation therapy for head-and-neck cancer: a critical review. Int J Radiat Oncol Biol Phys 2011;84:581–9. Bledsoe T, Noble A, Hunter G, Rybicki L, Hoschar A, Chute D, et al. Oropharyngeal squamous cell carcinoma with human papillomavirus status treated with definitive chemoradiotherapy: patterns of failure and toxicity outcomes. Radiat Oncol 2013;8:1–7. Vatca M, Lucas Jr JT, Laudadio J, D’Agostino RB, Waltonen JD, Sullivan CA, et al. Retrospective analysis of the impact of HPV status and smoking on mucositis in patients with oropharyngeal squamous cell carcinoma treated with concurrent chemotherapy and radiotherapy. Oral Oncol 2014;50:869–76. Williams G, Teo M, Sen M, Dyker K, Coyle C, Prestwich R. Enteral feeding outcomes after chemoradiotherapy for oropharynx cancer: a role for prophylactic gastrostomy. Oral Oncol 2012;48:434–40. Ward MC, Bhateja P, Nwizu T, Kmiecik J, Reddy CA, Scharpf J, et al. Impact of feeding tube choice on severe late dysphagia after definitive chemoradiotherapy for human papillomavirus–negative head and neck cancer. Head Neck 2016;38:E1054–60. Naik M, Ward MC, Bledsoe TJ, Kumar AMS, Rybicki LA, Saxton JP, et al. It is not just IMRT: human papillomavirus related oropharynx squamous cell carcinoma is associated with better swallowing outcomes after definitive chemoradiotherapy. Oral Oncol 2015;51:800–4. Setton J, Lee N, Riaz N, Huang S, Waldron J, O'Sullivan B, et al. A multi-institution pooled analysis of gastrostomy tube dependence in patients with oropharyngeal cancer treated with definitive intensity-modulated radiotherapy. Cancer 2015;121:294–301. Corry J, Poon W, McPhee N, Milner A, Cruickshank D, Porceddu S, et al. Prospective study of percutaneous endoscopic gastrostomy tubes versus nasogastric tubes for enteral feeding in patients with head and neck cancer undergoing (chemo)radiation. Head Neck 2009;31:867–76. Brown TE, Banks MD, Hughes BGM, Lin CY, Kenny LM, Bauer JD. Randomised controlled trial of early prophylactic feeding vs standard care in patients with head and neck cancer. Br J Cancer 2017;117:15–24.