Randomized Controlled Trial

A - Click here for explanation of classification scheme.

POSITIVE: See Quality Criteria Checklist below.

The researchers' purpose was to determine if a cause-and-effect relationship exists between nutrition interventions, including dietary counseling or the use of oral nutritional supplements, and function and clinical outcomes in head and neck cancer patients receiving radiotherapy (RT).

• All consecutive, ambulatory patients receiving RT for head and neck cancer at the study center between July 2000 and March 2003 were considered eligible
• No distinction was made between patients who received primary RT, RT adjuvant to surgery, RT combined with chemotherapy or RT received for palliation
• Patient had to be receiving RT of 70 Gy in 35 fractions. 

• Renal disease
• Diabetes mellitus.

Recruitment

Not specifically mentioned; all consecutive study center patients who met the inclusion criteria during the study period were considered eligible.

Design

• All patients were referred for pre-operative RT
  • All patients had previously received chemotherapy: 5-fluorouracil + cisplatin + folinic acid
• Patients were randomized to one of three groups:
  • Group 1: Dietary counseling with regular foods
  • Group 2: Usual diet plus supplements
  • Group 3: Ad lib intake
• Randomly assigned patients received identical scheduled visits and contact time with the research dietitian
  • The goal of nutrition therapy was to help each patient achieve his or her estimated energy and protein requirements
  • The dietitian prescribed a therapeutic diet using regular foods
    • Diet was modified to meet individual requirements based on:
      • Adequate intake
      • Alleviation of symptoms
      • Psychological factors
      • Functional capacity of the GI tract
    • Diet was designed to be as close to each patient's usual eating pattern as possible
    • The diet prescription identified the type, amount and frequency of feedings
      • Goal calorie and protein level was specified
      • Restrictions, limitations or increases in dietary intake were specified
  • Patients randomly assigned to the nutrition supplement group received oral nutrition supplements as an adjunct to the patients' regular diet
    • Ready to use, high-protein, energy-dense commercial supplements were used
    • All supplements were the same brand (brand not mentioned)
    • Patients could select their preferred flavors
    • Patients received two cans of supplement per day
    • Each 200ml can of supplement contained 200kcal per 20g protein
    • Supplement consumption records were kept daily by patients and verified by a caretaker
  • Patients assigned to the ad lib intake group were instructed to maintain their regular diet
• Study measures were assessed and results were calculated.

Blinding Used

• Patients were stratified by cancer stage prior to randomization:
  • Patients were randomized in permutation blocks of three
  • Sequential, numbered, opaque sealed envelopes contained computer-generated random assignments
• Study personnel were blinded to the randomization sequence; a copy was kept separate from study personnel
• Randomization envelopes were opened prior to each patient's first appointment by a person blinded to the study procedures. 

Intervention

• Individualized nutrition counseling by the study dietitian
• Provision of two oral supplements per day in addition to regular diet; supplements provided a total of 400kcals per 40g protein per day
• Ad lib intake.

Statistical Analysis

• Power calculations were completed for the study:
  • Minimum sample size of 40 patients was required to detect, with a significance level of 0.01 and power of 0.85:
    • Difference in body weight of 1.9kg
    • Difference in nutritional intake of 25%
    • Difference in QOL score of 20%
  • Power calculations were based on changes observed in a previous pilot study of 36 patients
• Statistical analysis completed with SPSS 11.5 (SPSS Inc., Chicago, IL), and EPI-Info 2000 (Centers for Disease Control, Atlanta, Georgia)
• All analysis were conducted on an intent-to-treat basis; any missing data was replaced with the average of the study group, which would have no effect on the estimators 
• Continuous variables were analyzed using one-way analysis of variance (ANOVA) or Wilcoxen rank sum tests
• Categorical variables, incidence, prevalence and frequency were analyzed with the chi-square test
• Univariate or multiple correlations were analyzed with two-tail non-parametric Spearman tests
• Two-tailed multiple correlation analysis was performed to determine the relationship of dietary intake and RT-induced symptoms on worsening nutritional status
• Statistical significance level was P<0.05.

Timing of Measurements

Data Collection by Days
		Baseline	RT Treatment Period						End RT	Three Months Following End RT
	Day		7	14	21	28	35	42	49	139
Measure		x
Demography		x
Medical history		x
Informed consent		x
Randomizations		x
Medications*		x	x	x	x	x	x	x	x	x
PG-SGA nutritional status		x	x	x	x	x	x	x	x	x
Weight		x	x	x	x	x	x	x	x	x
Diet history		x								x
24-hour recall			x	x	x	x	x	x	x	x
RT-induced morbidity			x	x	x	x	x	x	x	x
QOL s/EORTC QLQ-C30		x							x	x
Nutritional intervention (Gp 1, Gp 2)		x	x	x	x	x	x	x	x
Acceptability and compliance		x	x	x	x	x	x	x	x

*Note: Use of dietary supplements was also monitored.

Dependent Variables

• Variable 1:
  • Nutrition outcomes:
    • Nutritional intake (energy and protein intake)
      • Basal energy requirements estimated using the World Health Organization formula for patients 60 years or less, or the Owen et al formula for patients older than 60 years
      • To estimate daily energy requirements (EER), basal requirements were multiplied by a 1.5 activity factor
      • Protein requirements were estimated from age and sex standardized reference values, and ranged from 0.8g per kg per day to 1.0g per kg per day
      • Burke's diet history, along with multiple and sequential 24-hour recall evaluations (two weekdays and one weekend day), were performed at each
    • Nutritional status [measured using Ottery's Patient Generated Subjective Global Assessment (PG-SGA)]
      • Validated tool designed for patients with cancer
      • Measures three areas:
        • Weight changes, symptoms, alterations in food intake and functional capacity
        • Components of metabolic stress
        • Physical examination
      • Patients were classified as normal, moderately malnourished or severely malnourished
• Variable 2: Symtom-induced morbidity
  • Measured using the PG-SGA
    • Addresses symptoms: Anorexia, nausea, constipation, mucositis, vomiting, diarrhea, xerostomia, pain
• Variable 3: Quality of life (QOL)
  •  Assessed with the EORTC Quality of Life questionnaire version 3.0
    • Includes six function scales (physical, emotional, cognitive, social, role and global health/QOL
    • Includes three symptom scales (fatigue, pain, nausea/vomiting)
    • Includes six single items (symptoms, financial impact)

Independent Variables

• Dietary counseling
• Oral nutrition supplements.

Control Variables

• Presence of head and neck cancer
• Receiving radiation therapy.

• Initial N: N=75 (60 males, 15 females)
  • Group 1 (individualized dietary counseling based on regular foods): N=25
  • Group 2 (supplement group): N=25
  • Group 3 (control group; ad lib intake): N=25
• Attrition (final N): 
  • All data was analyzed on an intent-to-treat basis; all enrolled patients accounted for; no patients lost to follow-up
  • All patients completed the study
  • No patients took any other dietary supplements during the study period
• Age: 60±11 (36 to 79) years
• Other relevant demographics:
  • Patients had cancers of the base of the tongue, nasopharynx, oropharynx or larynx
  • 30 patients with Stage I/II disease; 45 patients with Stage III/IV disease
    • 16 patients in Group 1 were malnourished; 14 patients in Group 2 were malnourished; and 15 patients in Group 3 were malnourished
    • 56% of the malnourished patients had Stage III/IV disease; 4% of the malnourished patients had Stage I/II disease 
  • All patients had received prior chemotherapy with an identical regimen (5-flourouracil + cisplatin + folinic acid)
  • All patients were receiving radiotherapy in preparation for surgery
• Anthropometrics: At baseline, there were no significant differences between groups for energy and protein intake, prevalence of malnutrition or presence of symptom-induced morbidity
• Location: Radiotherapy Department, University Hospital of Santa Maria, Lisbon, Portugal. 

 

Nutritional Outcomes

• Energy intake:
  • Energy intake at end of RT compared with initiation of RT:
    • Group 1: Net increase, +521kcal per day (358 to 732), P=0.002
    • Group 2: Net increase, +322kcal per day (286 to 412), P=0.005
    • Group 3: Net decrease, -400kcal per day (201 to 502), P<0.01
  • Energy intake at three-month follow-up:
    • Group 1: Maintained increase in energy intake
    • Group 2: Energy intake decreased to either baseline or below baseline, P=0.005
    • Group 3: Energy intake decreased to either baseline or below baseline, P=0.005
• Protein intake:
  • Protein intake at end of RT compared with initiation of RT:
    • Group 1: Net increase, +26g per day (20 to 34), P=0.006
    • Group 2: Net increase, +35g per day (20 to 44), P=0.001
      • Group 2 protein intake was higher than Group 1 protein intake, P=0.06
    • Group 3: Net decrease, -15g per day (9 to 21), P<0.01
  • Protein intake at three-month follow-up:
    • Group 1: Maintained increased protein intake
    • Group 2: Protein intake decreased to either baseline or below baseline, P<0.05
    • Group 3: Protein intake decreased to either baseline or below baseline, P<0.05
• Nutritional status:
  • Group 1: Eight of 16 malnourished patients at initiation of RT improved their nutrition status [net average recovery at three-months follow-up: 4kg (two to six)]
  • Group 2: No patient showed improved nutrition status
  • Group 3: No patient showed improved nutrition status.

Changes in Nutritional Status during RT and at Three Months 

Group 1

Group 2

Group 3

Decline

Maintaind/Improved

Decline

Maintained/Improved

Decline

Maintained/Improvedd

Method

End RT

Three Months

End RT

Three Months

End RT

Three Months

End RT

Three Months

End RT

Three Months

End RT

Three Months

P1*

P2*

PG-SGA

5

3

20

22

19

24

6

1

24

25

1

0

<0.002

<0.001

*P¹ describes the significance of the statistical difference between the intervention groups regarding nutritional decline, at the end of RT and at three months follow-up.

*P² describes the significance of the statistical difference between the intervention groups regarding maintenance or improvement of nutritional status, at the end of RT and at three months' follow-up.

Symptom-induced Morbidity Outcomes

• By end of RT, 90% of the study participants experienced radiation-induced toxicity
  • The incidence did not differ significantly between groups (P<0.08)
  • A trend for reduced symptoms was found in Group 1 vs. Groups 2 and 3 (P<0.07)
• At three-month follow-up, a significant difference in Grade 1 and 2 anorexia, nausea/vomiting, xerostomia and dysgeusia was found between Group 1 patients (90% showed improvement), Group 2 patients (67% showed improvement) and Group 3 patients (51% showed improvement); P<0.0001. Reduction of Grade 1 + 2 dysphagia/odynophagia was not significant between groups.
• During RT, medications (oral antifungal solutions, local anesthetics and anti-inflammatories) were prescribed as appropriate for symptom control in 57% of Group 1 patients, 61% of Group 2 patients and 68% of Group 3 patients; however, the difference was not significant
• Dietary intake was correlated significantly with nutritional status in all groups (R≤-0.59, P≤0.002).

RT-induced Morbidity According to Severity Grade

	Group 1				Group 2				Group 3
	Grade 1		Grade 2		Grade 1		Grade 2		Grade 1		Grade 2
Symptoms	End RT	Three Months	End RT	Three Months	End RT	Three Months	End RT	Three Months	End RT	Three Months	End RT	Three Months	P1*	P2*	P3*
Anorexia	10	1	2	0	9	4	5	3	9	5	7	3	<0.05	<0.12	<0.001
Nausea/Vomiting	4	0	1	0	3	2	2	1	3	2	2	1	<0.001	<0.10	<0.05
Xerostomia	12	2	3	0	10	6	6	3	10	5	7	3	<0.04	<0.05	<0.0003
Dysgueusia	10	1	7	2	10	6	11	5	11	5	12	6	<0.04	<0.008	<0.0004
Odynophagia/dysphagia	14	2	8	1	12	3	10	3	12	6	12	6	<0.0001	<0.05	<0.0002

*P¹ describes the significance of the statistical difference between the intervention groups regarding the reduction of Grade 1 symptoms, at the end of RT and at three months' follow-up.

*P² describes the significance of the statistical difference between the intervention groups regarding the reduction of Grade 2 symptoms, at the end of RT and at three months' follow-up.

*P³ describes the significance of the statistical difference between the intervention groups regarding the reduction of Grade 1 + 2 symptoms, at the end of RT and at three months' follow-up.

Quality of Life Outcomes

• At end of RT:

  • For Group 1, all QOL scores significantly improved (P<0.003)

    • QOL scores were correlated with increases in energy and protein intake (R<0.83; P<0.001)
    • QOL scores were related to improvement in nutritional status (P<0.05)
  • For Group 2, all function scores improved (P<0.009); function scores were correlated with improvement in protein intake (R<0.58; P<0.05)
  • For Group 3, all QOL scores worsened as nutritional intake and nutritional status worsened (P<0.0001, P<0.002, respectively)
•   At three-month follow-up: 
  • For Group 1, all patients maintained or improved overall QOL
    • QOL was positively correlated with maintenance or improvement of nutritional status (P<0.008) and adequate energy and protein intake (P<0.01)
    • Function scores were improved or maintained
    • Symptom scales were significantly improved over baseline (P<0.002)
  • For Group 2, patients either maintained or experienced a decline in QOL (P<0.03). Decline in QOL was associated with poor dietary intake (P<0.003) and worsened nutritional status (P<0.002)
  • For Group 3, function scores deteriorated further when compared with baseline and end of RT
    • Decline in function scores was associated with inadequate energy and protein intake (P<0.001) and worsened nutritional status (P<0.002)
    • Symptom scores were worse than at baseline (P<0.003). The worst symptom scores were associated with poor energy and protein intake (P<0.005).

• Nutrition is a key determinant of quality of life in patients with cancer
• Both at end of RT and at three months' follow-up, dietary counseling resulted in significantly improved QOL function scores, in conjunction with sufficient intake and adequate nutritional status
• Patients in Group 2 showed an improvement in QOL function scores as long as supplementation continued, but when supplementation was discontinued, QOL scores declined
• Patients in Group 3, who received no nutrition counseling or supplementation, showed a significant decrease in QOL function scores throughout the entire study period
• The benefits of dietary counseling for QOL scores included improved physiological and clinical outcomes
• Nutrition intervention, including nutrition assessment, dietary counseling, diet education, monitoring compliance and symptom management, was central to the improvement of patients' outcomes, both nutritional and non-nutritional
• Both dietary counseling and oral supplementation improved patient outcomes in the short term, with dietary counseling showing an equal or greater benefit than supplementation; however, at the three-month follow-up, only dietary counseling resulted in improved patient outcomes
• The authors urge early nutrition intervention in this patient population.

University/Hospital:

Instituto de Medicina Molecular Faculdade de Medicina da Universidadede Lisboa (Portugal), Santa Maria University Hospital (Portuagal)

• Well-designed RTC; included power calculation, randomization with blinding of investigator assigning patients to groups; used validated instruments to measure outcomes; used concurrent controls
• Patients receiving RT combined with chemotherapy and RT combined with surgery were included in the study
• Records of supplement consumption were kept by patients and patient caregivers 
• Basal energy requirements, EER and protein requirements were calculated from formulas and reference values; 1.5 activity factor was used for all patients
• Many more men than women in study; no information provided for number of males and females in each study group.