Adjuvant Treatment of Stage I Lung Cancer With High-Dose Vitamin A By Ugo Pastorino, Maurizio Infante, Marco Maioli, Giuseppe Chiesa, Marc Buyse, Patrick Firket, Nicole Rosmentz, Maurizia Clerici, Enzo Soresi, Maurizio Valente, Pier Angelo Belloni, and Gianni Ravasi Purpose:

Vitamin A and retinoids are strong inhibitors

of epithelial cancer promotion and progression in experimental carcinogenesis. This study examined whether they may prevent the occurrence of upper aerodigestive cancer in subjects heavily exposed to tobacco smoking, such as patients already cured of an early-stage lung cancer.

Patients and Methods: The adjuvant effect of highdose vitamin A was tested on 307 patients with stage I non-small-cell lung cancer. After curative surgery, patients were randomly assigned to either a group prescribed retinal palmitate administration (orally 300,000 IU daily for 12 months) or a control group prescribed no treatment. Results: After a median follow-up of 46 months, the number of patients with either recurrence or new primary tumors was 56 (37%) in the treated arm and 75 (48%) in the control arm. Eighteen patients in the treated

LUNG CANCER REMAINS an overwhelming social problem because of its high incidence, low surgical curability, and the failure of conventional systemic treatments to cure it. Even in the presence of major reductions in tobacco consumption, the mortality rates in Europe will not decrease during the next 20 years because of the large cohort of people already affected by long-term carcinogenic exposure.' Screening programs have proven ineffective in reducing the overall and cancer-related mortality, thus deceiving the expectations for early diagnosis. 2 On the other hand, adjuvant postsurgical radiotherapy or chemotherapy did not improve survival in advanced and early-stage lung cancer.3 Even in resected stage I lung cancer, primary tumor relapses still occur in 20% to 30% of patients, and second primary tumors are detected in 10% to 20% of cases, depending on the intensity and duration of follow-up. 4 In this situation, it seems justified to evaluate new substances, such as biologic response modifiers, in adjuvant or chemopreventive trials. The interest in vitamin A and synthetic retinoids has developed from a number of epidemiologic observations From the Istituto Nazionale Tumori, OspedaleNiguarda,and Ospedale S. Carlo, Milan, Italy, and the InternationalInstitutefor Drug Development, Brussels, Belgium. Submitted February 18, 1992; accepted March 8, 1993. Address reprint requests to Ugo Pastorino, MD, Department of Thoracic Surgery, Istituto Nazionale Tumori, Via Venezian 1, 20133 Milan, Italy. ©1993 by American Society of Clinical Oncology. 0732-183X/93/1 107-0004$3.00/0

1216

group developed a second primary tumor, and 29 patients in the control group developed 33 second primary tumors. A statistically significant difference in favor of treatment was observed concerning time to new primary tumors in the field of prevention (P = .045, log-rank test). The treatment difference in terms of disease-free interval was close to statistical significance (P = .054, log-rank test) and just significant when adjusted for primary tumor classification (P = .038, Cox regression model). Conclusion: Daily oral administration of high-dose vitamin A is effective in reducing the number of new primary tumors related to tobacco consumption and may improve the disease-free interval in patients curatively resected for stage I lung cancer. The impact of such a treatment on survival needs to be further explored. J Clin Oncol 11:1216-1222. © 1993 by American Society of Clinical Oncology. into broad experimental evidence that shows that these substances are able to prevent or revert carcinogenesis and modulate epithelial cell differentiation.8 ' 9 Until recently, the clinical evidence of an anticancer activity was based on sporadic heterogeneous reports.'0 However, the positive results of two controlled studies in which isotretinoin produced regression of oral leukoplakia" and prevented second primary tumors in resected head and neck cancer' 2 have opened a new perspective in chemoprevention. Concurrent negative results from a large study on skin cancer have limited expectations regarding 0-carotene as a safe alternative to retinoids. 3 The available experience with vitamin A and retinoids suggests that the achievement of significant results is dose-dependent, and active treatment schedules are associated with typical side effects. 4-16 Positive results seem to be time-dependent, and their maintenance requires prolonged administra1 tion." 7 Moreover, vitamin A and its derivatives differ substantially in terms of absorption, bioavailability, clinical activity, and side effects.'8-21 Two alternative strategies are being developed to test such a hypothesis in a randomized design: large population trials on heavy smokers or asbestos workers22 and smaller trials on patients cured of a prior cancer in the lung or head and neck area'2,23 who remain at high risk for development of a second malignancy. This report discusses the results of a randomized trial of adjuvant treatment of resected stage I lung cancer with high-dose vitamin A. The interim results and trial methodology have been reported previously. 23

Journal of Clinical Oncology, Vol 11, No 7 (July), 1993: pp 1216-1222

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

1217

LUNG CANCER CHEMOPREVENTION

PATIENTS AND METHODS Patients Patients with pathologic diagnosis of stage I (TI-T2NOMO) nonsmall-cell lung carcinoma after complete surgical resection were selected for this study. Pathologic classification was based on the new international staging system for lung cancer. 24 The recruitment of patients began in 1985 at the Departments of Thoracic Surgery of the Istituto Nazionale Tumori, Niguarda Hospital, and S. Carlo Hospital in Milan. The accrual was closed in 1989, and follow-up of all patients was updated in July 1992. Before randomization, the baseline status of patients was assessed by physical examination, chest x-rays, sputum cytology, ECG, and complete laboratory evaluation. Eligibility criteria included an age of less than 75 years, normal liver and kidney function, impossibility of becoming or unwillingness to become pregnant, and a short interval from surgery (within 2 months). Patients with serious liver failure, acute or chronic hepatitis, serious ischemic heart disease, glaucoma, persistent disease, or who had previous chemotherapy were considered ineligible. Patients with a prior diagnosis of invasive cancer who had undergone curative treatment before the occurrence of lung cancer were considered eligible for entry onto the trial. Each eligible patient was informed about the aims and management of the study. The study protocol was approved by the three institutional ethics boards for human research, and informed consent was obtained from all patients.

Study Design Patients were randomly assigned to be given either vitamin A or no treatment. They were stratified according to the center, cell type (squamous v nonsquamous), and previous cancer at another site (absent v cured). Randomization was made by telephone call to the data center and patients were grouped in blocks of four within the strata. A preparation of emulsified retinol palmitate (A-Mulsin; Mucos Pharma Gmbh & Co, Gerestried, Munich, Germany) was chosen in consideration of its absorption properties and supplied as a liquid preparation or capsules. The drug was administered at a daily dose of 300,000 IU orally for a minimum of 12 months, or 90,000,000 IU total dose. If it was well accepted and tolerated by the patient, and in the absence of objective signs of toxicity, the treatment was continued for 12 additional months. The clinical outcome as regards cancer relapse was evaluated every 4 months for all patients by physical examination and assessment of symptoms, chest x-rays, and sputum cytology. The patients assigned to the treatment arm were additionally checked every 6 weeks during the first 4 months, every 2 months subsequently up to 12 months, and every 4 months up to 24 months for compliance and treatmentrelated toxicity. Compliance was assessed by means of an accurate interview made up of 13 questions that investigated the frequency and timing of intake on one side and the number and frequency of noncompliant events on the other. Subjective and objective symptoms were collected by questioning and physical examination only from the treatment arm, whereas the laboratory changes potentially related to treatment were investigated in all patients (treated and controls). The following blood analyses were requested at each examination: serum bilirubin, AST, ALT, gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP), blood urea nitrogen, creatinine, blood glucose, serum electrophoresis, serum lipids and hemoglobin values, and differential leukocyte, RBC, and platelet counts. Blood levels of retinol and retinol-binding protein were tested on a limited sample

of patients at the moment of randomization and during the outpatient follow-up: a total of 109 determinations were performed on patients given vitamin A, and 75 were performed on controls.25 In case of clinical or laboratory evidence of toxicity attributable to vitamin A, the dosage was halved until normalization. In case of persistent toxicity at a lower dose, the treatment was suspended until clinical symptoms had disappeared and then was resumed at the regular dose. The main end points of this adjuvant trial were represented by primary lung cancer relapse, including local recurrence and regional or distant metastases, and occurrence of a new primary tumor in the upper aerodigestive tract. Differentiation in the diagnosis between new primary lung cancer and pulmonary relapse was based on the presence of at least one of the following criteria: biopsy-proven different cell type, single contralateral lesion, or single ipsilateral lesion occurring in a different lobe from the original resection without continuity with previous resection margins. The ultimate end point was represented by the overall survival calculated from the date of randomization.

StatisticalAnalysis The statistical significance of the difference between mean serum values was assessed by the Mann-Whitney test. The following times to failure were calculated from the date of randomization: time to death from any cause (survival), time to relapse and/or new primary (disease-free interval), time to relapse (local or distant), time to new primary, time to new primary in field of prevention (lung, head and neck, bladder). The Kaplan-Meier estimates of these times were cal26 culated for each treatment group and compared using the log-rank test.27 Multivariate analyses were performed using the Cox regression model to identify prognostic variables and to adjust treatment dif28 ferences for any variables retained as significant. The variables considered were sex, age, histology, tumor extent (pT), prior cancer, and smoking habits. The assumption of proportional hazards was tested through an interaction term between treatment and time. RESULTS Relapses and New Primary Cancers Between July 1985 and October 1989, 313 patients were entered onto the trial and 307 were assessable for the present analysis. Six patients were excluded (four in the treatment arm and two in the control arm) because of major protocol violations related to pathologic stage: two cases of stage III disease and four cases of incomplete resection. Patients characteristics are listed in Table 1. The two groups were well balanced with respect to the factors that may potentially affect the probability of relapse and survival, namely, resection volume, histologic type, tumor extent (pT according to International Union Against Cancer-tumor-node-metastasis [UICC-TNM] classification), and tobacco consumption. Squamous cell carcinoma represented 47% of the cases in both arms, and pT1 represented 37% to 38%. A slight, nonsignificant excess of prior cancers occurred in the control arm (4% v 8%). The site of prior cancer was larynx (four), oral cavity (three), lymphoma (three), bladder (two), stomach (two), colon (one), lung (one), and other (two). The average time

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

1218

PASTORINO ET AL Table 1. Patient Characteristics

Characteristic

Age, years 0-59 60+

Table 3. Site of New Primary Tumors

Retinal Patients

Control Patients

No.

No.

%

%

72 78

48 52

72 85

46 54

144 163

140

93

142

90

282

Female

10

7

15

10

25

Type Squamous Other

70 80

47 53

74 83

47 53

144 163

Absent

144

96

145

92

289

Present

6

4

12

8

18

Sex

Male

Resection Lobar Pneumo pT 1

2 Smoking history Current smoker Former smoker Total

11

7

13

8

131 8

87 5

133 11

85 7

55

37

60

38

115

95

63

97

62

192

92 46

61 31

98 40

62 25

190 86

150

157

24

264 19

307

from pulmonary resection to treatment start was 21 days, and median time was 13 days (13 for treated cases and 12 for controls). None of the patients received adjuvant radiotherapy or chemotherapy. With a median follow-up of 46 months, a total of 131 (42%) patients have failed after primary treatment: 56 (37%) in the treated arm and 75 (48%) in the control arm. The pattern of failure is indicated in Table 2: pure locoregional recurrence was observed in 15 (10%) and 11 (7%) of retinol and control patients, respectively, distant recurrence in 23 (15%) and 35 (22%) patients, and new primary cancer in 18 (12%) and 29 (18%) patients. A total of 51 second primary cancers were detected in 47 patients. Eighteen patients developed a second primary tumor in the treated arm, and Table 2. Patterns of Failure Retinal Patients

Control Patients

No.

%

No.

%

All cancer failures

56

37

75

48

Recurrence Locoregional Distant New primary Deaths Recurrence

38 15 23 18 55 37

25 10 15 12 37 25

46 11 35 29 64 39

29 7 22 18* 41 25

7

5

14

9

7

11

7

New primary Other cause

Control Patients

No.

Months*

No.

Months

Lung Larynx Bladder Colorectal Prostate Skin melanoma Pancreas

11 2 0 3 0 1 1

28 42

26 27 16 12 16

35 61

21 3 3 3 3 0 0

Total

18

29

33

24

10

*Average months to detection of second primary tumors.

Previous cancer

Sublobar

Retinal Patients

Total No.

11

*Three patients had more than 1 second primary tumor.

29 patients developed 33 second primary tumors in the control arm. The site of new primary malignancies and the average time to second primary in months are listed in Table 3. In the control arm, two patients developed two second primary tumors and one patient developed three tumors. The number of patients with new primary malignancies related to tobacco smoking or malignancies in field of prevention (lung, head and neck, bladder) was 13 versus 25, respectively. Salvage surgery for new primary tumors was performed with eight of 18 patients (44%) in the retinol arm and with 20 of 29 patients (69%) in the control arm. Toxicity and Compliance The majority of treated patients experienced typical side effects related to the administration of retinol palmitate. Dryness of the skin or mucous membranes occurred in 91 (61%) patients, desquamation in 81 (54%), itching in 35 (23%), transient liver enlargement in 21 (14%), nausea or dyspepsia in 17 (11%), light nosebleeds in 13 (9%), headache in 11 (7%), and mild hair loss in five (3%). These symptoms usually subsided without interruption of the treatment, or they disappeared soon after discontinuation of treatment. Table 4 lists some of the biochemical and metabolic parameters used to monitor liver toxicity calculated as average values and prevalence of abnormal findings observed in the two arms at the time of randomization and during the follow-up. GGT values were significantly higher in patients treated with vitamin A only at 24 months from the beginning of treatment. Triglyceridemia increased constantly over the period of vitamin A intake. However, after treatment completion, patients given vitamin A resumed average values similar to those observed in the control group. Cholesterol levels were less strikingly modified, and the differences in mean values were not significant between the two groups. Renal function tests, glucose fasting levels, hemoglobin values, and RBC, WBC, and lymphocyte counts were not affected by

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

1219

LUNG CANCER CHEMOPREVENTION Table 4. Average Laboratory Values ± SD and Prevalence of Patients With Values Above the Normal Range (in parentheses) 0 Months

12 Months

Retinol AST (IU/dL) GGT (IU/dL) Triglycerides (mg/dl)

Control

Retinoal

28 ± 22(20) 45 ± 57 (15)

22 + 11(9) 38 + 35 (15)

23 ± 15(7) 81 ± 125 (49)

174 + 86 (68)

144 + 48 (50)

283 ± 161 (73)*

24 Months Control 27 + 50 +

Retinol

Control

30 ± 15(15) 149 ± 177 (69)*

24(17) 39 (32)

179 ± 141 (43)

185 ±

25 + 58 +

93 (66)

15(12) 58 (39)

193 + 172 (40)

NOTE. Abnormal values: AST > 40 IU/dL; GGT > 50 IU/dL; triglycerides > 150 mg/dL. *P < .05.

vitamin A supplementation. Average values of retinolemia, monitored in a sample of patients, at the time of randomization were 72 gg/dL for the treatment arm and 62 gg/dL for the control arm; the corresponding values at 12 months were 98 versus 66 pg/dL (P < .05). Retinolbinding protein values for the treatment versus control arms were 3.7 pg/dL versus 3.2 tg/dL, respectively, at the time of randomization and 5.9 versus 4.2 gg/dL, respectively, at 12 months (P < .05). The proportion of patients with estimated regular drug intake was more than 80% during the first 12 months. At the time of analysis, 119 patients (79%) had completed the first 12 months' schedule, and 59 (39%) had reached 2 years of treatment. In 29 patients (19%), the treatment was interrupted before the 12th month because of early recurrence (16 cases, I11%), objective signs of toxicity (five cases, 3%), or noncompliance (eight cases, 5%); in two cases the treatment was never begun. As far as toxicity is concerned, treatment has been discontinued in two cases because of recurrence of a preexisting dermatitis, in one case because of exacerbation of chronic proctitis, in one

case because of serious hypertriglyceridemia, and in one case because of acute type B viral hepatitis. Survival The time to relapse and/or new primary (disease-free interval) is shown in Fig 1. The estimated proportion of patients disease-free at 5 years was 64% versus 51%, in favor of the treatment arm (P = .054). The time to new primary cancer at any site is shown in Fig 2. The estimated proportion of patients without new primary at 5 years was 86% versus 76%, in favor of the treatment arm (P = .09). Excluding second primary tumors outside of the area targeted for chemoprevention (colon, prostate, melanoma), the time to primary tumor in the field of prevention (lung, head and neck, or bladder only) is shown in Fig 3. The estimated proportion free of new primary in field was 89% versus 80%, respectively (P = .045). At the end of follow-up, a total of 55 deaths were observed in the treatment arm compared with 64 deaths observed in the control arm (Table 2). The number of patients who died of cardiovascular disease or other causes unrelated to cancer was identical in each arm (11 cases).

1.0

1.0

0.9

0.9

0.8

0.8

0.7 0.6

"--..............

I

"

0.2,

N 0 Treatment 150 56 Retinol 157 75 Control --.----

0.1

Logrank P=0.054

0.3

ca

r 0.4

1

2

3

0.2

N O Treatment 150 18 Retinol 157 29 Control ------

0.1

Logrank P= 0.09

0.3

nn o

...... .

0,6 0.7

0.5 a 0.4 a_

..

4

5

6

TIME (years)

0.0 0

1

2

3

4

5

6

TIME (years) Fig 1. Kaplan-Meier estimate of disease-free interval (time to pulmonary relapse and/or new primary tumor at any site) by treatment. N, Number of patients in arm; O, number of observed end points in curve.

Fig 2. Time to new primary cancer at any site by treatment. N, Number of patients in arm; O, number of observed end points in curve.

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

1220

PASTORINO ET AL

tumors. Except for histologic type, which was used as a stratification parameter, the two arms were perfectly balanced with respect to other prognostic factors, such as age, resection volume, and pathologic tumor extent (TI NO 37% v 38%). The smoking habits of patients were carefully monitored during the intervention phase to rule out the potential confounding factor of prior or current tobacco consumption. In our study, only 16 (11%) patients in the treatment arm and 18 (11%) patients in the control arm continued or resumed smoking habits during the followup.

1.0 0.9 .........

L-

0.8 0.7 S0.6

0.5 X 0.4

a-

N

0.3

0

Treatment

0.2

150 13 Retinol 157 25 Control -------

0.1

Logrank P= 0.045

At

0.on 0

-

1

2

3

4

5

6

years

from

6

TIME (years) Fig 3. Time to primary tumor in the chemo head and neck, or bladder) by treatment. N, Numberofpatientsld (luin arm; 0, number of observed end points in cur ve.

The overall estimated survival rate a t5 years was 62% versus 54% (P = .44, Fig 4). The resualts of multivariate analysis are summarized in Table 5. T here was a trend in favor of treatment for all end points cconsidered, but the treatment difference was statistically significant only regarding time to new primaries in the fiefl of prevention (P = .045, log-rank test). The treatment difference in terms ofdisease-free interval was close to sta tistical significance (P = .054, log-rank test) and just signifi when adjusted ,ant for primary tumor classification (P=.0 model). DISCUSSION Chemoprevention of lung cancer is one of the most innovative fields of clinical research in oncology. Vitamin A and retinoids seem to be potentially useful tools for this purpose, and in patients cured of head and neck or of lung cancer, their adjuvant and chemopreventive properties can be concurrently tested. However, the optimal substance, dosage, and duration for large-scale intervention trials still have to be defined. In the present phase, careful randomized clinical studies are required to assess the effectiveness of different chemoprevention plans and to determine their feasibility with respect to compliance and toxicity. This randomized trial was designed to test the activity of high-dose vitamin A as adjuvant therapy for non-smallcell stage I lung cancer. A population of limited-stage, curatively resected cancers was considered adequate to investigate the ability of vitamin A to reduce primary tumor recurrence, as well as the incidence of second primary

the

beginning

of

the

study,

vitamin

A

tdviai fte h bgnig 6 er fo At~~~-~-~~ mor within the target field. The overall probability of disease-free survival was close to statistical significance (P = .055), and became significant (P = .038) when adjusted for primary tumor classification with the Cox regression model. These data seem consistent with the results of the randomized study from Hong et a112 in which isotretinoin administration prevented second primary tumors in head and neck patients. However, the power of the comparisons is lower in our trial in view of the limited number of events observed thus far. One should also consider that in lung cancer patients the frequency of new primary tumors is lower than that observed in head and neck cancer patients, and primary relapses in the lungs may be difficult to differentiate from second tumors at this site. In the present analysis, all cases of isolated relapse in the lung have been reevaluated blindly, and a few cases of local or distant recurrence have been reclassified as new primary lung cancers. In fact, administration

improved

the

tim

u-

1.0 0.9 0.8 0.7 0.6 S0.5

2 0.4 0.2

N 0 Treatment 150 55 Retinol 157 64 Control ------

0.1

Logrank P=0.44

0.3

0.0 0

1

2

3

4

5

6

TIME (years) Fig 4. Overall survival by treatment. N, Number of patients in arm; O, number of observed end points in curve.

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

LUNG CANCER CHEMOPREVENTION

1221 Table 5. Results of Multivariate Analysis

End Point

Prognostic Factors in Cox Regression

Test for Treatment (P)

Relative Risk

Confidence Interval

Survival Disease-free interval Time to relapse Time to new primary Time to new in field

pT, smoking habits Histology Treatment

.44 .054* .36 .09 .045

1.16 1.4 1.22 1.65 1.96

0.80-1.66 0.99-1.98 0.80-1.88 0.91-3.00 1.00-3.84

*P = .038 adjusting for pT.

both the overall failure rate and the incidence of second primary lung cancer observed in the control group are close to the expected range for surgically treated stage I lung cancer.5,7,29,30 On the basis of experimental evidence, even a reduction of distant relapses is compatible with the mechanism of action of high-dose vitamin A, which may be more active against micrometastatic deposits than against synchronous occult primary tumors. Given that second primary tumor incidence remains constant over time and that stage I lung cancers have a high expected survival, a longer follow-up will hopefully clarify the real impact of vitamin A administration on primary cancer failure versus second primary tumor incidence. The lack of benefit for overall survival, despite a marginal difference in the total number of deaths, may be partially attributable to specific factors related to lung cancer surgery. One is comorbidity, that is, mortality due to cardiovascular diseases or other nonneoplastic events, which affected I I cases in each arm. The other is salvage surgery, which in this series was applied to 44% of new primary tumors in the retinol arm and 69% of new primary tumors in the control arm. This study demonstrates that high-dose vitamin A is a relatively safe treatment and may be well tolerated for a period of 1 to 2 years.31 The high estimated compliance rate is compatible with that found in other investigations that used monodose daily schedule and reinforcement programs. 32 ,33 The preparation of retinol palmitate in aqueous emulsion fulfilled the expectations of greater bioavailability (sixfold that of oily solutions) and lower toxicity than retinol."1 In fact, the majority of patients were able to tolerate over a long period of time the daily dose of 300,000 IU that was recommended by the available

dermatologic experience.34-36 As for clinical symptoms, our results are in agreement with those of two large reviews on acute and chronic side effects of vitamin A supplementation. 37, 38 The concern for retinol-induced liver damage, usually related to overconsumption of vitamin A supplements without medical prescription or control, 38 was not con-

firmed in our randomized experimental setting. Our results confirmed that high-dose vitamin A causes a significant increase in serum triglycerides,3 9' 40 but not in serum cholesterol. This phenomenon was not as serious as the one observed in patients with preexisting risk factors or hypertriglyceridemia 4t and subsided soon after discontinuation of treatment. Our long-term followup did not show a higher risk of cardiovascular diseases in cancer patients treated over a long period of time, and the number of deaths not related to cancer was equal in the two arms. Overall, the hypothesis that emulsified retinol palmitate is indeed less toxic than other forms of retinol may well explain the discrepancy between our data and the experience in the United States, where oily or water-soluble retinol have been for years the most used forms of vitamin A. We conclude from this analysis that high-dose vitamin A is effective in reducing the number of new primary malignancies related to tobacco consumption and may improve the disease-free interval in patients curatively resected for stage I lung cancer. The impact of such a treatment on survival needs to be further explored. It is our present belief that the search for optimal chemoprevention regimens should be focused on the combination of various agents with different preventive properties. Based on the evidence accumulated during the initial phase of this trial, a larger European cooperative study, named Euroscan, was designed in 1988.42 Two thousand patients with prior cancer of the larynx, oral cavity, or lung were planned to be entered onto the new study and randomized to receive high-dose vitamin A and/ or n-acetyl-cysteine over a period of 2 years or to standard follow-up. The accrual of patients in the present study ended October 1989, as soon as the pilot phase of Euroscan trial was completed. The experience and results achieved with this study should reinforce the goals of the larger European trial, the size of which will provide adequate power to detect even moderate survival benefits in the near future.

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

1222

PASTORINO ET AL

REFERENCES 1. Hansen HH (ed): Lung Cancer. European Commission Series for General Practitioners. Berlin, Germany, Springer-Verlag, 1990 2. Fontana RS, Sanderson DR, Woolner LB, et al: Lung cancer screening: The Mayo program. J Occup Med 28:746-750, 1986 3. Holmes EC: Surgical adjuvant therapy of non-small-cell lung cancer. J Surg Oncol 1:26-33, 1989 (suppl) 4. Shields TW, Robinette CD: Long-term survivors after resection of bronchial carcinoma. Surg Gynecol Obstet 136:759-768, 1973 5. Fontana RS: Early diagnosis of lung cancer. Am Rev Resp Dis 116:399-402, 1977 6. Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 96:440-447, 1988 7. Thomas P, Rubinstein L, the Lung Cancer Study Group: Cancer recurrence after resection: TI NO non-small cell lung cancer. Ann Thorac Surg 49:242-247, 1990 8. Lotan R: Effects of vitamin A and its analogs (retinoids) on normal and neoplastic cells. Biochim Biophys Acta 605:33-91, 1980 9. Sporn MB, Newton NL: Chemoprevention of cancer with retinoids. Fed Proc 38:2528-2534, 1979 10. Lippman SM, Kessler JF, Meyskens FL: Retinoids as preventive and therapeutic anticancer agents (part II). Cancer Treat Rep 71:493-515, 1987 11. Hong WK, Endicott J, Itri L, et al: 13-cis-retinoic acid in the treatment of oral leukoplakia. New Engl J Med 315:1501-1505, 1986 12. Hong WK, Lippman JM, Itri L, et al: Prevention of second primary tumors with isotretinoin in squamous cell carcinoma of the head and neck. N Engl J Med 323:795-801, 1990 13. Greenberg ER, Baron JA, Stukel TA, et al: A clinical trial of beta carotene to prevent basal-cell and squamous-cell cancers of the skin. N Engl J Med 323:789-796, 1990 14. Bollag W: From vitamin A to retinoids: Chemical and pharmacological aspects, in Orfanos CE, et al (eds): Retinoids-Advances in Basic Research and Therapy. Berlin, Germany, Springer-Verlag, 1981, pp 5-11 15. Elias PM, Williams ML: Retinoids, cancer, and the skin. Arch Dermatol 117:160-180, 1981 16. Kamm JJ, Ashenfelter KO, Ehmann CW: Preclinical and clinical toxicology of selected retinoids, in Sporn MB, et al (eds): The Retinoids, vol 2. Orlando, FL, Academic Press, 1984, pp 287-326 17. Moon RC, McCormick DL, Mehta RG: Inhibition of carcinogenesis by retinoids. Cancer Res 43:2469-2475, 1983 18. Goodman DS: Vitamin A and retinoids in health and disease. N Engl J Med 310:1023-1031, 1984 19. Korner WF, Vollm J: New aspects of the tolerance of retinol in humans. Int J Vit Nutr Res 45:363-372, 1975 20. Kessler JF, Meyskens FL: Clinical results with retinoids as preventive and therapeutic anticancer agents, in Prasad K (ed): Vitamins, Nutrition and Cancer. Basel, Switzerland, Karger, 1986 21. Windhorst DB, Nigra T: General clinical toxicology of oral retinoids. J Am Acad Dermatol 6:675-682, 1982 22. Bertram JS, Kolonel LN, Meyskens FL: Rationale and strategies for chemoprevention of cancer in humans. Cancer Res 47:30123031, 1987

23. Pastorino U, Soresi E, Clerici M, et al: Lung cancer chemoprevention with retinol palmitate. Acta Oncologica 27:773-782, 1988 24. Mountain CF: A new international staging system for lung cancer. Chest 89:225-233, 1986 (suppl) 25. Infante M, Pastorino U, Chiesa G, et al: Laboratory evaluation during high-dose vitamin A administration: A randomised study on lung cancer patients after surgical resection. J Cancer Res Clin Oncol 117:156-162, 1991 26. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958 27. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: Analysis and examples. Br J Cancer 35:1-39, 1976 28. Cox DR: Regression models and life-tables. J R Stat Soc 34: 187-220, 1972 29. Pairolero PC, Williams DE, Bergstrahl EJ, et al: Postsurgical stage I bronchogenic carcinoma: morbid implications of recurrent disease. Ann Thorac Surg 38:331-338, 1984 30. Pastorino U, Valente M, Bedini V, et al: Limited resection for stage I lung cancer. Eur J Surg Oncol 17:42-46, 1991 31. Pastorino U, Chiesa G, Infante M, et al: Safety of high-dose vitamin A. Randomized trial on lung cancer chemoprevention. Oncology 48:131-137, 1991 32. Cramer JA, Mattson RH, Prevey ML: How often is medication taken as prescribed? JAMA 261:3273-3277, 1989 33. Norell SA: Improving medication compliance: A randomised clinical trial. Br Med J 2:1031-1033, 1979 34. Frey JR, Schoch MA: Therapeutische versuche bei psoriasis mit vitamin A, zugleich ein beitrag zur A-hypervitaminose. Dermatologica 104:80-86, 1952 35. Rapaport HG, Herman H, Lehman E: Treatment of ichtyosis with vitamin A. J Pediatr 21:733-746, 1942 36. Silverman S, Renstrup G, Pindborg J: Studies in oral leukoplakias: III. Effects of vitamin A comparing clinical, histopathological, cytologic, and hemathologic responses. Acta Odont Scand 21:271292, 1963 37. Bauernfeind JC: The Safe Use of Vitamin A: A Report of the International Vitamin A Consultative Group (IVAG). Washington, DC, The Nutrition Foundation, 1980 38. Bendich A, Langseth L: Safety of vitamin A Am J Clin Nutr 49:358-371, 1989 39. Goodman GE, Alberts DS, Peng YM, et al: Plasma kinetics of oral retinol in cancer patients. Cancer Treat Rep 68:1125-1133, 1984 40. Kligman AM, Leyden JJ, Mills O: Oral vitamin A (retinol) in acne vulgaris, in Orfanos CE, et al (eds): Retinoids-Advances in Basic Research and Therapy. Berlin, Germany, Springer-Verlag, 1981, pp 245-253 41. Nigra TP, Katz RA, Jorgensen H: Elevation of serum triglyceride levels from oral 13-cis-retinoic acid, in Orfanos CE, et al (eds): Retinoids-Advances in Basic Research and Therapy. Berlin, Germany, Springer-Verlag, 1981, pp 363-369 42. De Vries N, Van Zandwijk N, Pastorino U: The Euroscan study. Br J Cancer 64:985-989, 1991

Downloaded from ascopubs.org by 37.44.207.46 on January 26, 2017 from 037.044.207.046 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.