Introduction

Malnutrition influences quality of life (QoL) and performance status (PS) in cancer patients [1, 2]. Although nutritional deterioration is common in head and neck cancer (HNC), its earliest indicator, represented by involuntary loss of body weight (UWL), is frequently underestimated or completely ignored before treatment [3, 4]. The aim of this study was to evaluate the prevalence of malnutrition in HNC and the possible influence of nutritional status on QoL and PS before radio or concomitant chemoradiotherapy.

Materials and methods

Sixty-one consecutive outpatients (69% males, 31% females), median age 50 (range 18–70) with locally advanced HNC (stage III and IVA disease), treated at Roma Saint Peter Hospital Oncology Department, were enrolled. Criteria for eligibility included absence of diabetes, hepatic or renal failure, alcohol dependence, and not previous surgical or oncological treatment. Characteristics of patients are described in Table 1 and cancer location and stage of disease in Table 2. All patients were evaluated for malnutrition (UWL), nutritional intake (by diet history), nutritional status (Patient Generated Subjective Global Assessment (PG-SGA) score), serum prealbumin (PREA; n.v. 20–40 mg/dL), hemoglobin level (Hb; n.v. 12.0–15.0 g/dL), C-reactive protein (CRP; n.v. 0.01–5.0 mg/dL), QoL (European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ) C-30 v. 3.0), and Eastern Cooperative Oncology Group performance status scale (ECOG PS) before treatment.

Table 1 Characteristics of patients
Full size table
Table 2 Cancer location
Full size table

Statistical analysis

For statistical analysis, we used SPSS for Window, release 13.0. The end point was to investigate the prevalence of malnutrition and the possible influence of nutritional status on QoL and PS before treatment. UWL, PG-SGA, CRP, PREA, Hb, QoL, and PS in not malnourished and malnourished patients were compared with chi-square, Fisher or T-Student tests, as appropriate. All the possible correlations were analyzed with nonparametric Spearman test. Possible influence of malnutrition on QoL and PS was tested with multivariate analysis. Significance was reported at the conventional p < 0.05 level.

Results

Baseline differences in not malnourished and malnourished patients are described in Table 3.

Table 3 Baseline differences in nonmalnourished and malnourished patients
Full size table

Malnutrition

The European Society for Clinical Nutrition and Metabolism considered malnourished a patient who involuntary lost a 5% or greater of body weight in the last 3 months [5]. Thirty-six percent of patients were malnourished before treatment. Mean value ± standard deviation (SD) of UWL was 10 ± 5 kg. Weight loss was CRP-related (p = 0.001).

Nutritional intake

Nutritional intake was derived from a diet history. Both energy and protein intakes were analyzed at enrolment. At baseline, 10% of nonmalnourished HNC referred a reduction of usual food intake vs 67% of malnourished. Only two patients reduce food intake because of odynophagia.

Nutritional status

Nutritional status was assessed with Ottery’s PG-SGA. A score of 0–1 does not require an intervention, but a reassessment on routine and regular basis during treatment. A score of 2–3 requires a patient and family education by dietician, nurse, or other clinician. A score of 4–8 requires the intervention by dietician and a score ≥9 indicates a critical need for improved symptom management and/or nutrient intervention options. In nonmalnourished HNC mean ± SD PG-SGA score was 3 ± 2 and in malnourished 9 ± 5 (p = 0.001). PG-SGA score was UWL (p < 0.001) and CRP-related (p < 0.001).

Inflammatory status

We considered significant a CRP level ≥10 mg/dL (n.v. 0.01–5.0 mg/dL). At enrolment, in 27% of HNC, CRP level was ≥10 mg/dL. In nonmalnourished patients, mean ± SD CRP level was 4.3 ± 5.0 mg/dL, whereas in malnourished was 20.3 ± 22.5 mg/dL (p = 0.004). Seventy-one percent of HNC with CRP level ≥10 mg/dL were malnourished vs 13% with CRP level <10 mg/dL.

Serum prealbumin

At baseline in all HNC, the mean value ± SD of PREA (n.v. 20–40 mg/dL) was 26 ± 5 mg/dL. In nonmalnourished patients, mean value ± SD was 28 ± 5 mg/dL, whereas in malnourished was 22 ± 3 mg/dL (p < 0.001). PREA was weight loss- (p = 0.001), CRP- (p = 0.021), and PG-SGA-related (p = 0.006).

Hemoglobin

Mean ± SD hemoglobin level in nonmalnourished was 13.7 ± 0.9 mg/dL and in malnourished 12.4 ± 1.2 mg/dL (p < 0.001). Hemoglobin level was weight loss- (p < 0.001), CRP- (p = 0.019), and PG-SGA-related (p = 0.002). In Table 4, we reported the mean concentrations of hemoglobin as measured in the subgroups of patients.

Table 4 Concentration of hemoglobin in the subgroups of patients
Full size table

Performance status

At enrolment, the median score of ECOG PS in nonmalnourished HNC was 0 (0–2), whereas in malnourished was 1 (0–2; p = 0.018). PS was weight loss- (p = 0.002), Hb- (p = 0.042), PREA- (p = 0.040), CRP- (p = 0.033), and PG-SGA-related (p = 0.007).

Quality of life

The mean ± SD of EORTC QLQ-C30 scores for nonmalnourished and malnourished patients are listed on Table 5. About function scales, physical (p = 0.043), role (p = 0.047), and social functions (p = 0.024) scores were significantly better in nonmalnourished than in malnourished HNC. About symptom scales and symptoms single items, fatigue (p < 0.001), appetite loss (p < 0.001), and nausea and vomiting (p = 0.002) scores were worse in malnourished than in nonmalnourished HNC. In the multivariate analysis, malnutrition (UWL) and Hb level independently influenced physical (p = 0.002; p = 0.005), role (p = 0.004; p = 0.001), and social functions (p = 0.024; p = 0.009).

Table 5 Mean ± SD QLQ-C30 scores in nonmalnourished and malnourished patients
Full size table

Discussion

The importance of malnutrition as comorbid condition in HNC undergoing radiotherapy or concomitant chemoradiotherapy has long been recognized. Although a nutritional deterioration is reported to affect 25–50% of HNC before any therapy [6], few studies evaluated the possible prevalence and impact of malnutrition on performance and quality of life in these cancer patients before treatment. In our study, we excluded HNC affected by alcoholism, 7% of HNC in our casuistry, because a heavy alcohol intake may influence independently the nutritional status; it has been estimated that the impact of malnutrition is 70% in alcoholic patients vs 30% in nonalcoholic [7]. Our data confirm that malnutrition is common also in nonalcoholic HNC not only during but also before treatment. In our experience, 36% of patients were already malnourished at enrolment; 13% of HNC with a CRP level <10 mg/dL were malnourished because of a food intake reduction due to psychological problem or, in two patients, to odynophagia, and 71% of patients with a CRP level ≥10 mg/dL were malnourished because food intake reduction was due to early satiety and severe anorexia. In this second group of patients, the presence of unintentional weight loss, anorexia, early satiety, fatigue, and inflammatory status demonstrated that about 30% of HNC might be already affected by cancer cachexia before treatment [8]. In cachectic HNC, we also founded a significant correlation between weight loss, systemic inflammation, and hemoglobin level. The possible mechanism involved in this relationship could be that both weight loss and low hemoglobin level might be considered a cytokine-mediated disorder, resulting from complex interactions between tumor cells and host [9]. Overexpression of proinflammatory cytokines (interleukin (IL)-6, IL-1β, tumor necrosis factor-α) and CRP, due to IL-6 increase, results in shortened survival of red blood cells, suppression of erythroid progenitor cells, impaired iron utilization, and inadequate erythropoietin production [10]. However, even if systemic inflammation is considered the major cause of cancer-related anemia before chemo- and radiotherapy, in our experience, statistical analysis showed that in the subgroup of patients, the hemoglobin significantly reduced only when inflammatory status was associated with weight loss [11]. Weight loss, anemia, and inflammatory status already may compromise functional ability of HNC before the treatment-induced toxicity. In addition to weight loss prior to the diagnosis, the patient may lose an additional 10% of pretherapy body weight during treatment. A total weight loss reduction ≥20% significantly correlates with treatment interruption, infections, early mortality, hospital readmission rate after treatment completion, and survival [12–18]. Since the optimal therapy for cancer-related malnutrition is curing the underlying cancer [19], our experience suggests that an early and concomitant nutritional support may stop or counteract further weight loss and improve outcome in HNC [20]