Relationship between quality of life and clinical outcomes in advanced non-small cell lung cancer: best supportive care (BSC) versus BSC plus chemotherapy

Lung Cancer 24 (1999) 17 – 24

Relationship between quality of life and clinical outcomes in advanced non-small cell lung cancer: best supportive care (BSC) versus BSC plus chemotherapy Sumitra Thongprasert a,*, Porntip Sanguanmitra a, Wanida Juthapan a, Jennifer Clinch b a

Department of Medicine, Faculty of Medicine, Chiang Mai Uni6ersity, Chiang Mai, Thailand b Clinch Research Consultant, Winnipeg, Manitoba, Canada

Received 16 September 1998; received in revised form 1 February 1999; accepted 11 February 1999

Abstract In a prospective randomized study, 287 patients with advanced non-small cell lung cancer (NSCLC) stage IIIb or IV with ECOG performance status (PS) 0–1 or 2 were randomly assigned to receive either best supportive care (BSC) or supportive care plus combination chemotherapy (IEP regimen: ifosfamide 3 gm/m2 IV with mesna uroprotection, epirubicin 60 mg/m2 IV on day 1 and cisplatin 60 mg/m2 IV on day 2; or MVP regimen: mitomycin-C 8 mg/m2, cisplatin 100 mg/m2 IV on day 1, vinblastine 4 mg/m2 IV on days 1 and 15). Serial assessment of Karnofsky performance status (KPS), modified Functional Living Index-Cancer (T-FLIC) and modified Quality of Life-Index (T-QLI) were used to estimate the quality of life. Interviews were done at entry, at the third month and at 2 months post complete treatment. At least two courses of chemotherapy were considered to be adequate for response evaluation. Patients were treated for a total of four to six courses or until progression of disease. Partial response rates were 40 and 41.7% in IEP and MVP arms. Median survival durations were 5.9 and 8.1 months for the IEP and MVP chemotherapy arms, and 4.1 months for BSC (log-rank test: P =0.0003). One year survival was 13, 29.8 and 39.3% for the BSC, IEP and MVP regimens, respectively. Two years survival was 7.8, 6.4 and 13.1% for the BSC, IEP and MVP regimens, respectively. Improvement in quality of life (QOL) scores at the first, second and third interview were seen in chemotherapy arms only, not in the BSC arm. We conclude that combination chemotherapy improves the quality of life as well as prolonging the survival of patients with advanced NSCLC. © 1999 Elsevier Science Ireland Ltd. All rights reserved.

1. Introduction

* Corresponding author.

Advanced non-small cell lung cancer (NSCLC) occurs in the majority of patients and is associated with an extremely poor prognosis. Multiple

0169-5002/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 9 9 ) 0 0 0 1 7 - 3

18

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

randomized studies and a meta-analysis of these studies [1–3] showed that chemotherapy prolongs survival time in advanced non-small cell lung cancer, but only by 4 – 8 weeks at the median with improved 1-year survival from 10 to 20 – 30% [3 – 6]. One of these randomized studies from Canada showed that chemotherapy with CAP cost less than supportive therapy in metastatic non-small cell lung cancer. None of these studies reported quality of life information which is extremely important given the small survival advantage and the toxicity of chemotherapy. Recently, several investigators were concerned about the real benefit of chemotherapy in the treatment of non-small cell lung cancer. The main question was regarding the prolonged survival, that is whether prolonging survival time was good enough for the patients to cope with the side effects of chemotherapy? The introduction of quality of life assesment into clinical studies was considered one of the measures to answer the question [7–11]. At our hospital, in Northern Thailand we have about 1000 new cases of NSCLC per year [12]. Most of these cases are advanced stage. Thus, we started a randomized study to assess the quality of life of NSCLC treated with best supportive care (BSC) versus best supportive care plus chemotherapy. We selected two combination chemotherapy regimens: IEP and MVP. Both of these regimens were routinely administered in our hospital. The response rate and survival time of the two regimens were not different [13].

2. Patients and methods Patients with inoperable clinical stage IIIb and IV, histologically or cytologically verified NSCLC (squamous cell carcinoma, adenocarcinoma, or large cell carcinoma), aged less than 75 years, with ECOG performance status of 0 – 2, were eligible for this study. Patients with initial symptoms and/ or signs of brain metastases, cord compression, superior vena caval syndrome or pathological bone fracture were excluded. Patients who had been previously treated with cytotoxic agents or radiation therapy were not eligible for the study.

Computed tomography (CT) of the chest, and bone and liver scintigraphy were performed only in symptomatic patients. Patients had normal hematological and renal function. Bilirubin was less than 2. Informed consent was obtained prior to randomization. The randomization procedure was block randomization using combined parameters of stage and ECOG performance status i.e.: Stage IIIb, ECOG 0–1; Stage IV, ECOG 0–1; Stage IIIb, ECOG 2; Stage IV, ECOG 2. Envelopes were used for randomization. Patients were randomized into three arms: arm I, best supportive care (BSC); arm II, BSC + chemotherapy 1 (IEP); and arm III, BSC+chemotherapy 2 (MVP). IEP chemotherapy (arm II) consisted of ifosfamide 3 gm/m2 IV infusion over 4 h on day 1, with mesna uroprotection; 4% epirubicin 60 mg/m2 IV on day 1 and cisplatin 60 mg/m2 IV infusion over 2 h on day 2. Cycles were given every 28 days. MVP chemotherapy (arm III) consisted of mitomycin-C 8 mg/m2 IV on days 1, 29 and 71; vinblastine 4 mg/m2 IV every 2 weeks and cisplatin 100 mg/m2 IV infusion over 2 h on days 1, 29 and every 4 weeks. Chemotherapy was stopped if no response was seen after two courses or after maximum response was achieved (maximum number of chemotherapy courses= 6). All patients were entitled to receive supportive care, i.e. radiation therapy against bone metastases, late brain metastases, tracheal compression, bronchial compression or superior vena caval syndrome (which developed after entry to the study). Thoracocentesis and/or tube thoracoscopy with pleurodesis was administered to patients with pleural effusion on the BSC arm. Pain medication, nutritional and psychological support were given according to physician’s judgement. Questionnaires on quality of life were carried out by an interviewer at three intervals during the treatment period: (1) at the time of randomization; (2) at the end of second course of chemotherapy (prior to the start of the third course) or at 2 months after entry for the BSC group; and (3) at 2 months after cessation of chemotherapy or at 6 months after entry for the BSC group. At the same times as QOL assessment by T-FLIC and T-QLI, the patients were assessed

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

by the physician using Karnofsky performance status scale. To measure quality of life, we used two questionnaires, one was the Functional Living IndexCancer (FLIC) which was developed and reported by Schipper and his group [7,10], the second was the Quality of Life Index (QLI) which was developed by Spitzer and his group [8]. We translated these two questionnaires into the Thai language and modified them to fit Thai culture. The initial FLIC was a self-assessed questionnaire with a visual analogue response scale for each question. Our modified FLIC called T-FLIC and modified QLI called T-QLI were administered by an interviewer who had no knowledge of the patient’s status but had experience with questionnaires. The response scale for T-FLIC was changed from the original to three grades: good, bad or intermediate. The response scale for T-QLI was unchanged. T-FLIC consisted of 22 questions each scored 0, 1 or 2 (bad, intermediate, good). Thus the highest score of the T-FLIC was 44 and the lowest was 0. T-QLI consisted of five questions also scored 0, 1 or 2. Thus the highest scores of T-QLI was 10 and the lowest was 0. Both the T-FLIC and the T-QLI were previously validated [14]. Cronbach’s a coefficient was used to test the internal consistency of the questionnaires. The correlation between scores was calculated using Pearson’s correlation coefficient. For demographic data, i.e. age, stage, tumor type, non-parametric tests were used. The Logrank test was used to assess differences in survival time between regimens. The difference in quality of life time trends was tested using the SAS procedure Proc Mixed. The standard repeated measures analysis of variance requires equality of the variances at each time of assessment and equality of the covariances between times of assessment. This requirement is almost never met. SAS Proc Mixed not only permits missing data points but also allows modelling of the variance – covariance matrix thus giving a more accurate hypothesis test. By definition, changes in quality of life can only be assessed in patients who have had at least two assessments. Hence this analysis was conducted on 172 patients who had time 2 and/or time 3 assessments in addition to the initial assessment.

19

3. Results Both FLIC and QLI questionnaires, were translated into Thai language and tested in the Thai population. The reliability and consistency of the modified questionnaires (T-FLIC and T-QLI) were reported previously [14]. Internal consistency (Cronbach’s a coefficient) was 0.78–0.83. Construct validity was supported by testing the correlation between T-FLIC and T-QLI (r= 0.62). The criterion-related validity of these questionnaires was tested by Pearson’s correlation coefficient to Karnofsky performance status (KPS) [14] and was found to have a good correlation. In this study also good correlations were found between TFLIC and KPS, T-QLI and KPS and also TFLIC and T-QLI at each time of assessment (see Table 1). The characteristics of the patients entered onto the study are presented in Table 2. Of the 287 evaluable cases, 98 cases were randomized to BSC, 96 cases were randomized to IEP and 93 cases were randomized to MVP. The stage and median Karnofsky performance status of patients in each arm were not different. The median age was 59.5, 58 and 58 years in arms I–III, respectively. Of 287 patients entered, only 281 patients completed the first quality of life (QOL) assessment, 172 patients completed the second QOL assessment and 84 patients completed the third assessment. The number of cases who completed the second and third interview were lower than the first interview due to death or loss to followup. Table 3 shows the number of patients that did not complete the questionnaires due to death or loss to follow up prior to the second and third interviews for each study group. Table 1 Correlation of T-FLIC, T-QLI and PS Correlation Correlation Correlation Correlation Correlation Correlation Correlation Correlation Correlation

of of of of of of of of of

sum sum sum sum sum sum sum sum sum

T-FLIC1 and PS1 T-QLI1 and PS1 T-FLIC1 and T-QLI1 T-FLIC2 and PS2 T-QLI2 and PS2 T-FLIC2 and T-QLI2 T-FLIC3 and PS3 T-QLI3 and PS3 T-FLIC3 and T-QLI3

0.383 0.409 0.571 0.434 0.441 0.769 0.255 0.412 0.795

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

20 Table 2 Patient characteristics

Table 4 Survival time versus regimen Treatment BSC

IEP

Treatment

Median survival time (months)

BSC IEP MVP

4.1 5.9 8.1

% Partial response

MVP

No. of patients entered Median age (years) (range, years)

98 59.5 (28–73)

96 58 (36–73)

93 58 (28–76)

Pathological diagnosis: Adenocarcinoma Squamous cell CA. Large cell CA. Unclassified CA. Cytological proof.

49 31 12 4 2

68 22 6 0 –

56 23 8 5 1

Assessment of QOL1 Assessment of QOL1+2 Assessment of QOL1+2+3

97 46 19

93 64 36

91 62 29

The response rate and median survival time of the patients by study arm are shown in Table 4. The survival of the patients treated with chemotherapy (IEP and MVP) were significantly longer than the BSC arm (5.9 and 8.1 vs. 4.1 months; P= 0.0003). One-year survival for arm I (BSC), arm II (IEP) and arm III (MVP) was 13, 29.8 and 39.3%, respectively. Two-year survival for BSC, IEP and MVP was 7.8, 6.4 and 13.1%, respectively. Fig. 1 shows the survival curves of the three treatment groups. The main toxic reactions of the two chemotherapeutic regimens (IEP and MVP) were nausea and vomiting and reversible alopecia which occurred in all patients. Higher incidence of hematological toxicity was seen in IEP than MVP. Grades 3 and 4 anemia were seen in 28.5 and 15.9% for IEP and MVP, respectively.

NA 40 41.7

Leukopenia grades 3 and 4 were seen in 7.1 and 7.4% for IEP and MVP, respectively. Thrombocytopenia grades 3 and 4 were seen in 7.1 and 7.4% for IEP and MVP, respectively. Thirty-four patients treated with IEP had pleural effusion; thoracocentesis was done in 23 patients; and tube thoracoscopy was administered to 11 patients. Twenty-one patients treated with MVP had pleural effusion; thoracocentesis was done in 15 patients; and tube thoracoscopy was administered in six patients. Twenty-seven patients treated with best supportive care (BSC) had pleural effusion; thoracocentesis was done in 18 patients; and tube thoracoscopy was administered to nine patients. Sixty-eight patients in the BSC arm received palliative radiation therapy to the primary tumor, five patients received palliative radiotherapy to bone and three patients received whole brain irradiation. Twenty patients and ten patients who received chemotherapy (IEP and MVP, respectively) received palliative radiotherapy to the chest. Six patients in the IEP arm and four patients in the MVP arm received palliative radiotherapy to bone. Five patients in the IEP arm and four patients in the MVP arm received whole brain radiotherapy.

Table 3 Number of patients dead or lost follow-up prior to second and third interviews Regimen

No. No. No. No.

of of of of

patients patients patients patients

death prior to the second interview loss to follow up prior to the second interview death prior to the third interview loss to follow up prior to the third interview

BSC (cases)

IEP (cases)

MVP (cases)

37 14 17 10

23 6 19 9

19 10 23 10

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

21

4. Association of quality of life and clinical outcomes

Table 5 Percentage of patients whose FLIC scores went up, down or remained the same between times of assessment

We correlated the initial KPS, T-FLIC, and T-QLI with survival time. All three correlations were significant (For PS1, r= 0.26770, P = 0.0001; for FLIC1, r=0.18298, P = 0.0103; and for QLI1, r = 0.15090, P =0.0348). As patients with cancer are followed with serial quality of life assessments, individuals are lost from the mean calculations due to death and loss to follow-up. Since survival is usually related to initial quality of life assessment, the means can increase simply because the patients with the highest quality of life scores survive while the low scorers die. Thus it is important to know what happens to individuals over time. Patients in this study were grouped according to whether their scores increased, decreased or remained the same from times 1 to 2 and from times 2 to 3. This data is shown in Table 5. It is clear from this descriptive analysis that a greater percentage of patients experienced a time 1 to time 2 decrease in quality of life in the best supportive care group than in either of the chemotherapy groups. The same was also true of the times 2 to 3 change. The quality of life mean scores for each regimen at each time of assessment are shown in Table 6. Patients on supportive care showed a decline in the mean T-FLIC score from times 1 to 2, whereas patients in both chemotherapy groups showed an increase. Between times 2 and 3 mean scores remained about the same for all groups.

Regimen

Fig. 1. Survival curve of the three treatment arms.

Change

Time 1–2

Time 2–3

n Down No change Up n

46 47.8 10.9 41.3 64

18 50.0 0.0 50.0 35

2 = IEP

Down No change Up n

32.8 6.3 60.9 63

45.7 0.0 54.3 30

3 =MVP

Down No change Up

36.5 4.8 58.7

36.7 13.3 50.0

1 =BSC

The mean T-QLI scores behaved similarly for the supportive care group. However, the change in T-QLI was less pronounced for the two chemotherapy groups—a slight increase followed by a decrease in the IEP group and a slight overall decline in the MVP group. These differences in trend over time were significant for both measures of quality of life (T-FLIC, P= 0.05; and T-QLI, P=0.02).

5. Discussion Quality of life is a new parameter for assessing disease outcome. Previously most clinicians believed that only objective parameters such as response, survival and toxicity were useful for the assessment of clinical outcomes in cancer treatment. Recently several investigators have changed their ideas and have been using the quality of life as a parameter for assessing tumor response [9,10,15]. Several quality of life questionnaires have been developed and widely used. This study was started in early 1991. The FLIC questionnaires and the QLI questionnaires were adopted for this study. The modified version of FLIC and QLI questionnaires (T-FLIC and T-QLI) were validated [14]. Good correlations were seen between T-FLIC, T-QLI and performance status. These findings

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

22

supported that T-FLIC and T-QLI were valid and useful tools for assessing quality of life in this study. The study reported by Woods et al. [16] showed that the survival of the NSCLC patients treated with combination chemotherapy was longer than the no chemotherapy group; however, the difference was not statistically significant. Woods had stressed that the chemotherapy arm produced severe toxicity. However, that study did not incorporate quality of life measurement and the authors felt strongly that quality of life should have been used to measure the treatment outcome in addition to the response rate and survival time. The study reported by Rapp et al. [5] also demonstrated the survival improvement in the group treated with chemotherapy. In addition, they concluded that chemotherapy was more cost effective in treating NSCLC with advanced unresectable disease, presumably due to fewer complications and hospital admissions.

The study reported by Ganz et al. [15], comparing chemotherapy (cisplatin plus vinblastine) with best supportive care, included the measurement of quality of life using the Functional Living IndexCancer but unfortunately the study was not completed. Performance status was used as a crude estimate of quality of life, but there were no differences in performance status score over the duration of the trial. Because of the small number of patients in this study and the difficulty of conducting the trial, the authors could not come to any conclusion about the benefit of using quality of life questionnaires to assess the outcome of the study. It was not easy to carry out this type of study, especially due to death and disability of patients with advanced NSCLC. From our experience, the patients’ situation, their thoughts and their hopes were understood better than in previous studies which did not include the quality of life question-

Table 6 T-FLIC scores and T-QLI scores (n=173) Scores:

Regimen BSC

Mean Leastsquares meana Median Minimum Maximum No. of cases

Mean Leastsquares meana Median Minimum Maximum No. of cases

IEP

MVP

T-FLIC1

T-FLIC2

T-FLIC3

T-FLIC1

T-FLIC2

T-FLIC3

T-FLIC1

T-FLIC2

T-FLIC3

28.4 28.4

26.1 26.2

28.4 26.4

27.7 27.7

31.3 31.3

31.6 30.9

28.3 28.3

30.5 30.5

31.0 30.0

28 13 44 47

26 5 44 46

29 3 42 19

28 8 43 65

34.5 7 44 64

35 5 44 36

28 8 39 61

34 7 44 61

35 4 44 28

T-QLI1

T-QLI2

T-QLI3

T-QLI1

T-QLI2

T-QLI3

T-QLI1

T-QLI2

T-QLI3

7.3 7.3

6.3 6.3

6.5 6.2

7.2 7.2

7.5 7.5

7.6 7.3

7.6 7.6

7.4 7.4

7.3 7.0

7 4 10 47

6 3 10 46

7 2 10 19

7 3 10 65

8 2 10 64

8 3 10 36

8 4 10 61

8 0 10 61

8 2 10 29

a These means are produced by SAS Proc Mixed. When data are missing the means are estimated based on the data that are present, i.e. both a particular individual’s other data and his/her group’s data.

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

naires. Besides, the results of this study revealed that the toxicities of chemotherapy did not ruin the quality of life of the patients. The analysis of quality of life data in cancer clinical trials is problematic when patients die during the course of the trial. In the analysis presented here data for those who died were missing. However, the mean quality of life scores for those who were alive at time 2 on chemotherapy showed no deterioration from baseline in contrast to those on best supportive care only. Both quality and quantity of life were superior in the chemotherapy arms. Our findings that chemotherapy did not compromise the patients’ quality of life is very important. This adds to the results of several investigators including the results of a multivariate analysis which reported that treatment of non-small cell lung cancer with chemotherapy could prolong survival time of the patients and cost less than the supportive care treatment alone [16 – 19].

[2]

[3]

[4]

[5]

[6]

[7] [8]

6. Conclusion Modified T-FLIC and T-QLI scales are useful for the assessment of quality of life. Patients with advanced lung cancer treated with chemotherapy had benefit both in prolonged survival and improvement in quality of life. Besides tumor response both T-FLIC and T-QLI are useful in measuring quality of life. The above two questionnaires should be considered as a parameter of the clinical response. This study confirmed the importance of chemotherapy for the treatment of advance non-small cell lung cancer.

Acknowledgements We thank Professor Paul Bunn for his kindly advice on the manuscript, Asta Medica Co. for supporting part of the study and Farmitalia Carlo Erba for supporting part of the study.

References [1] Ruckdeschel JC, Finkelstein DM, Ettinger DS, et al. A

[9] [10]

[11]

[12] [13]

[14]

[15]

[16]

23

randomized trial of the four most active regimens for metastatic non-small cell lung cancer. J Clin Oncol 1986;4:14 – 22. Gralla RJ, Kris MG. Management of non-small cell lung cancer. In: Williams CJ, Whitehouse JMA, editors. Recent Advances in Clinical Oncology. New York, NY: Churchill Livingstone, 1986:167 – 80. Non-small Cell Lung Cancer Collaborative Group. Chemotherapy in non-small cell lung cancer: a meta-analysis using updated on individual patients from 52 randomised clinical trials. Br Med J 1995;311:899 – 909. Cormier Y, Bergeron D, La Forge J, et al. Benefits of polychemotherapy in advanced non-small cell bronchogenic carcinoma. Cancer 1982;50:845 – 9. Rapp E, Pater JL, Willan A, et al. Chemotherapy can prolong survival in patients with advanced non-small cell lung cancer — report of a Canadian multicenter randomized trial. J Clin Oncol 1988;6:633 – 4. Kaasa S, Lund E, Thorud E, Hatevoll R, Host H. Symptomatic treatment versus combination chemotherapy for patients with extensive non-small cell lung cancer. Cancer 1991;67:2443 – 7. Schipper H, Levitt M. Measuring quality of life: risks and benefits. Cancer Treat Rep 1985;69(10):1115– 23. Spitzer WO, Dobson JA, Hall J, et al. Measuring the quality of life of cancer patients. A concise QL-Index for use by physicians. J Chron Dis 1981;34:585 – 97. Aaronson NK, Meyerowitz BE, Bard M, et al. Quality of life research in oncology. Cancer 1991;67:839 – 43. Schipper H, Clinch J, McMurray A, Levitt M. Measuring the quality of life of cancer patients. Functional Living Index-Cancer: development and validation. J Clin Oncol 1984;2:472 – 83. Marino P, Pampallona S, Prestoni A, et al. Chemotherapy versus supportive care in advanced non-small cell lung cancer: Results of a meta-analysis of the literature. Chest 1994;106:861 – 5. Martin NC, Srisukho S. Tumor registry. Maharaj Nakorn Chiang Mai Hospital 1992;15:17. Thongprasert S. Analysis of the four combination chemotherapies in non-small cell lung cancer treated at Maharaj Nakorn Chiang Mai. Jpn J Cancer Chemother 1992;19(8):1197– 201. Sanguanmitra P, Juthapan W, Thongprasert S. Validity and reliability of the modified Functional Living IndexCancer (T-FLIC) and the modified Quality of life (T-QLI) questionnaires. In: Proceedings of the 11th Asian Pacific Cancer Conference, Bangkok, Thailand, 1993:121. Ganz PA, Figlin RA, Haskell CM, La Soto N, Siau J. Supportive care versus supportive care and combination chemotherapy in metastatic non-small cell lung cancer. Cancer 1989;63:1271 – 8. Woods RL, Williams CJ, Levi J, et al. A randomised trial of cisplatin and vindesine versus supportive care only in advanced non-small cell lung cancer. Br J Cancer 1990;61:608 – 11.

24

S. Thongprasert et al. / Lung Cancer 24 (1999) 17–24

[17] Greenberg ER, Chute CG, Stukel T, et al. Social and economic factors in the choice of lung cancer treatment. New Engl J Med 1988;318:612–7. [18] Bakker W, Van Oosterom AT, Aaronson NK, et al. Vindesine, cisplatin, and bleomycin combination chemo-

.

therapy in non-small cell lung cancer: survival and quality of life. Eur J Cancer Clin Oncol 1986;22(8):963– 70. [19] Fernandez C, Rosell R, Abad-Esteve A, et al. Quality of life during chemotherapy in non-small cell lung cancer patients. Acta Oncol 1989;28(fasc 1):29 – 33.