Evaluating the impact on quality of life of chemoradiation in gastric cancer

Thirty-three patients receiving radiotherapy (45 Gy in 25 fractions), together with 12 weeks of infusional 5-fluorouacil and escalating doses of cisplatin every 2 weeks, completed the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire C30 at five time points: baseline, completion of radiation, 4 weeks after completion of radiation, 6–12 months after completion of chemoradiation, and 2–3 years after completion of chemoradiation.


INTRODUCTION
Quality of life (qol) is increasingly recognized as an important endpoint in clinical trials of cancer therapies.However few data are available on qol in patients who have received abdominal radiation therapy 1,2 .
In 2001, Macdonald et al. published a landmark study, the Intergroup 0116 trial 3 .This randomized controlled trial showed a significant survival advantage for adjuvant chemoradiation (crt), as compared with surgery alone, in patients with completely resected high-risk gastric and gastroesophageal adenocarcinoma 3 .Adjuvant crt treatment using the regimen of the Intergroup 0116 trial is now the standard of care in North America and in certain centres in Europe.This prolonged treatment course (taking 5 months) is associated with a 40% or worse rate of acute toxicity.Alternative chemotherapy regimens are being explored in this setting to improve survival outcomes and toxicity 4 .
Few data are available on the qol of patients undergoing adjuvant therapy for gastric adenocarcinoma 2 .Here, we present a prospective evaluation of qol for patients with completely resected gastric or gastroesophageal adenocarcinoma undergoing adjuvant abdominal radiation therapy with chemotherapy using infusional 5-fluorouracil (5fu) and escalating doses of cisplatin in a single-institution phase i study.
Evaluating the impact on quality of life of chemoradiation in gastric cancer Based on the results of that study, a phase ii study is currently underway 5 .
We hypothesised that qol would be reduced during treatment, but that it would return to baseline levels at 6-12 months after completion of treatment.We also hypothesised that the experience of greater toxicity and the higher doses of chemotherapy would both be associated with poorer qol.

METHODS
Approval to conduct this study was granted by the institutional research ethics board.All study patients provided informed consent.

Patient Population
This phase i study enrolled 34 patients with histologically confirmed Union Internationale Contre le Cancer stage ib to iv (M0) completely resected gastric or gastroesophageal junction adenocarcinoma.Patients had an Eastern Cooperative Oncology Group performance status of 2 or less.An upper age limit was not initially defined for the study, but because of significant therapy-related adverse events, the protocol was later amended to exclude patients over the age of 70.Participants were required to have a pre-treatment nutritional intake of at least 1500 calories daily.Patients were excluded if they had previously received radiation therapy or if they had a contraindication to radiation therapy.Tests for complete blood count, electrolytes (sodium, potassium), creatinine, bilirubin, aspartate transaminase, alanine transaminase, and alkaline phosphatase were obtained, as was a computed tomography scan of the chest, abdomen, and pelvis (to exclude patients with progressive disease or comorbidity that would preclude adjuvant therapy).

Chemotherapy
All patients were prescribed 12 weeks of infusional 5fu 200 mg/m 2 daily.Patients were recruited in 5 consecutive cohorts.Three patients without doselimiting toxicity (dlt) were treated per cohort, in a standard 3+3 dose-escalation design: If 1 of 3 patients experienced a dlt, 3 further patients were treated at the same dose level.Cisplatin every 2 weeks was added to the treatment plan in escalating dose intensity 5 .In cohorts 1-4, the cisplatin dose was 0 mg/m 2 , 20 mg/m 2 , 30 mg/m 2 , and 40 mg/m 2 respectively, given 4 times starting at week 1 (that is, weeks 1, 3, 5, and 7).In cohort 5 (the final cohort), an additional dose of 40 mg/m 2 was given in week 9 (see Table i) 5 ."Dose-limiting toxicity" was defined as grade 3 or greater toxicity requiring treatment or hospitalization."Maximal tolerated dose" was the highest dose cohort in which no more than 1 of 6 patients experienced a dlt.

Radiation Therapy
Participants were treated according to institutional protocol with three-dimensional conformal planning and delivery 6,7 .Radiation dose was 45 Gy in 25 fractions starting on day 16 after commencement of the chemotherapy.

Toxicity
Acute toxicity was scored according to the Common Toxicity Criteria, version 2.0 (ctc v2.0).Late toxicity was scored using the Radiation Therapy Oncology Group late toxicity score 8,9 .All toxicities occurring 90 days or more after completion of crt were considered late toxicities.

QOL Assessment
The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire C-30 (eortc qlq-C30) was used to assess qol.The qlq-C30 was completed at 5 time points: enrolment (before crt, but after definitive surgery), at completion of radiation therapy, and then at 4 weeks, 6-12 months, and 2-3 years after completion of chemotherapy.The eortc qlq-C30 is a self-administered, cancer-specific questionnaire 10 .It has 5 functional scales (each scale consisting of questions reflecting the physical, role, social, emotional, and cognitive functioning of the patient), 3 symptom scales (fatigue, pain, and nausea and vomiting), and global health status and overall qol scales.A number of single items assess additional symptoms of dyspnea, sleep disturbance, constipation, diarrhea, and perceived financial impact.This measure has been rigorously developed and has demonstrated reliability, validity, and responsiveness in a variety of cancer populations, both for discriminative and for evaluative purposes 1,10-15 .

Analyses
Survival and follow-up were determined from the date of diagnosis.Overall and relapse-free survival were estimated using the Kaplan-Meier method.
Each qol domain was scored and reported separately using previously described standard methods 10 .Raw scores were linearly transformed to give values between 1 and 100.Higher scores in the functional domains and in global qol indicate better functioning; For this study, a 10-point difference in qol score was determined a priori to be a clinically significant difference.This difference was determined to be clinically meaningful in the "moderate" category (10-20 points) by a previous prospective study that correlated changes in qol score with a patient's perceived difference in qol.In that study, a 5-to 10-point change in qol score corresponded to a "little" change in qol as perceived by the patient; a 10-to 20-point change, to a "moderate" change; and a greater than 20-point change, to a "very much" change 16 .
For analysis of compliance with questionnaires, patients that on follow-up had not reached the analysis time point and those that had died or relapsed before that time point were excluded from the numerator and the denominator 17 .In other words, compliance is reported as the proportion of expected questionnaires actually able to be scored.
Mean and median scores were calculated for each qol domain at each time point.Nonparametric testing (signed rank test) was used to determine the significance of changes in qol during follow-up as compared with baseline.For each domain at each time point, the percentages of patients with qol scores that had improved (10-point or more improvement), remained unchanged (less than a 10-point change), or worsened (10-point or more deterioration) were calculated.
In an exploratory analysis, the qol scores at completion of therapy were individually examined for patients who experienced a dlt and were compared with the mean scores for the rest of the cohort.Also, the potential associations of qol domains with dose cohort were assessed using the linear mixed-model across all time points.

Baseline Characteristics and Compliance
Our phase i study included 34 patients treated between September 2002 and March 2007.One patient was not eligible for the qol assessment because of a language barrier.Of the remaining 33 eligible patients, 1 did not complete treatment because of social issues and therefore only had a baseline qol assessment; another patient completed follow-up qol assessments, but lacked a baseline assessment.Those 2 patients were included in the analysis.Table ii shows patient and tumour characteristics.At the 5 time points, the number of qol forms that could be scored were 33 (baseline), 32 (completion of radiation), 32 (4 weeks after crt), 28 (6-12 months after crt), and 19 (2-3 years after crt), for completion rates of 97%, 91%, 84%, 75%, and 74% respectively.

Relapse and Survival
Median follow-up for the 28 patients alive at the time of analysis was 2.7 years (range: 0.3-5 years).Relapse had occurred in 9 patients (26%), and 6 patients (18%) had died of their disease.At 3 years, overall survival and relapse-free survival were 83% and 71% respectively, with median survival not yet reached 5 .

QOL Over Time
Figure 1 shows the mean scores for global qol and for the functional and symptom scales at each time point.Table iii shows median differences from baseline in the scores for each domain at completion of radiation and at 4 weeks, 6-12 months, and 2-3 years after crt completion.Median scores for several domains-global qol and the social, role, emotional, nausea and vomiting, and fatigue scales-showed clinically and statistically significant worsening at the completion of radiation.In addition, statistically significant worsening of scores was seen for the physical and appetite scales.At 4 weeks after completion of crt, global qol and fatigue remained significantly worse than at baseline.By 6-12 months, no subscale showed a difference from baseline on average.However, at that time point, a proportion of patients did show worse scores as compared with baseline.Figure 2 shows the proportion of patients with improved, unchanged, and worsened qol scores for each domain at completion of radiation therapy and at 6-12 months after crt.

QOL and Dose Cohort
Global qol (p = 0.0007) and the nausea and vomiting scale (p = 0.04) were both significantly worse with higher chemotherapy doses.Figure 3 shows the mean scores for these domains across all time points and dose levels.

QOL and Toxicity
Compared with the other patients, the 5 patients who experienced dlts scored worse at completion of radiation on the emotional (p = 0.05) and the nausea and vomiting (p = 0.02) scales.Figure 4 shows the scores in those domains across all time points for each patient with a dlt and the mean scores for the remaining cohort.

DISCUSSION
Very few data are available on the qol of patients undergoing adjuvant therapy for gastric and gastroesophageal adenocarcinoma, and no long-term data are available 2 .Our sample size is small, but the results show that qol is impaired during adjuvant crt treatment.Our results suggest that higher chemotherapy doses are associated with poorer scores for the nausea and vomiting scale and for global qol.Moreover, the small number of patients experiencing a dlt had poorer scores on the emotional and the nausea and vomiting scales at completion of radiation.In most patients, qol scores return to baseline, but recovery may take 6-12 months.Importantly, subscale scores at 6-12 months remain below baseline in a significant proportion of patients.This study was primarily a dose-finding study for an alternative chemotherapy regimen in the   A B adjuvant crt therapy of gastric cancer; qol was a secondary endpoint.Increasingly, qol is being recognized as an important endpoint in clinical studies.
In the setting of adjuvant treatment, the survival benefits of additional therapies are significantly smaller than they are in the definitive treatment setting.Consequently, the long-term adverse impact on qol is an important issue in patient counselling and in the decision-making process.Quality of life is also an important endpoint in the overall evaluation of alternative therapies as compared with standard treatments.All patients in our study had undergone radical surgery before they commenced adjuvant treatment.Several research groups, using a variety of instruments [Short Form (SF-36) Health Survey, D15, Spitzer Quality of Life Index, eortc qlq-C30 and -STO22], have reported a reduction in qol in similar populations [18][19][20][21] .
Tyrvainen et al. 19 used the SF-36 and D15 to compare qol in 25 long-term survivors after total gastrectomy for gastric cancer with qol in a normal population.Those authors found that mental health, physical and social functioning, energy, and vitality were similar in both groups.However certain dimensions, such as sleeping, eating, and distress, scored worse in cancer survivors.Lee et al. 22 assessed 408 gastric cancer survivors and found that they had more difficulties in performing housework and in maintaining employment than did a general population, with cancer survivors working fewer hours and having less work-related ability.A proportion of the patients in that study also received radiation therapy (12%) and chemotherapy (26%), but compared with patients treated by surgery alone, the patients receiving additional treatment did not show any additional effect with regard to ability to work.
The qlq-C30 addresses general qol issues.Blazeby et al. 2 reported favourable psychometric properties of the new gastric cancer qol module, the qlq-STO22, in 219 gastric cancer patients treated with palliative and radical intent.Thirty-three patients included in that study were treated radically with surgery and neoadjuvant or adjuvant chemotherapy with or without radiation therapy.The qlq-STO22 assesses specific symptoms (reflux, eating, pain, hair loss, dry mouth), anxiety, and body image.This additional measure will be valuable for assessing qol specific to gastric cancer patients in future studies of that population.

CONCLUSIONS
The factors associated with poorer qol outcomes in the setting studied here are not yet clear.The complex QOL AND CHEMORADIATION IN GASTRIC CANCER relationship between treatment-induced toxicity and qol has been studied for various cancer sites.Crosssectional and longitudinal studies have provided important information on the relationship between adverse events and qol in prostate, head-and-neck, breast, and lung cancers [23][24][25][26][27][28][29][30][31][32] .However, data are lacking on the relationship between long-term toxicity and qol after radiation treatment for abdominal malignancies.Toxicity and other potential patient and therapy factors associated with poorer qol outcomes in the short and long term are important areas for future study; these factors may contribute to the tailoring of therapy to the individual patient.

CONFLICT OF INTEREST DISCLOSURES
All the authors declare that there are no financial conflicts of interest.

figure 1
figure 1 Mean scores of (A) global and functional scales across all time points and (B) symptom scales across all time points.

figure 2 figure 3 figure 4
figure 2 Percentage change in scores for global quality of life (glObal) and for functional scales (A) from baseline to completion of radiation and (B) from baseline to 6-12 months.Percentage change in the scores for symptom scales (C) from baseline to completion of radiation and (D) from baseline to 6-12 months.
Some of this information was presented at the Canadian Association of Radiation Oncology 2008 Annual Meeting, the American Society of Therapeutic Radiology and Oncology 2008 Annual Meeting, and the International Society for Quality of Life Research 2008 Annual Meeting.

table i
Cisplatin administration by cohort

table ii
Baseline characteristics for all patients and for patients by dose cohort

table iii
Score difference a between baseline and follow-up md p Value md p Value md p Value a Boldface type indicates median difference values (md) and p values associated with pre determined clinical or statistical significance.