Long-term survivors of small-cell lung cancer (SCLC): A French multicenter study

Annals of Oncology 8: 1009-1014, 1997. © 1997 Kluwer Academic Publishers. Printed in the Netherlands.

Original article Long-term survivors of small-cell lung cancer (SCLC): A French multicenter study P. Jacoulet, A. Depierre, D. Moro, A. Riviere, B. Milleron, E. Quoix, E. Ranfaing, D. Anthoine, J. J. Lafitte, B. Lebeau, J. P. Kleisbauer, F. Massin, P. Fournel, M. Zaegel, J. P. Leclerc, G. Garnier, E. Brambilla & F. Capron on behalf of the Groupe d'Oncologie de Langue Frangaise * See page 1013 for list of institutions

Summary Background: The aim of this study was to analyze SCLC patients beyond 30 months, particularly their outcome, their way of life, and factors which could influence relapses, secondprimary cancers and death. Patients and methods: Between January 1986 and May 1995, 263 SCLC patients who survived longer than 30 months were included from 52 French institutions. The analysis was performed on the 155 cases confirmed by a pathologic review. Results: Physical, mental and psychological states were considered as normal at 30 months in respectively 70.3%, 87.7% and 67.7% of patients, not influenced by prophylactic cranial irradiation, number of chemotherapy cycles, CCNU or cisplatin. Therapeutic sequelae were neurological impairment (13%), pulmonary fibrosis (18%) and cardiac disorders (11%) at 30 months. Return to work was possible for 40% of patients in the first two years following diagnosis. Among 43 relapsing patients, 33 benefited from a second-line treatment. Their median survival was 12 months since retreatment, and seven patients have survived again longer than 30 months. Age > 60 at the time of diagnosis was found as an independent factor

Introduction Small-cell lung cancer (SCLC) represents about 20% of all cases of bronchogenic carcinoma recognized in France. The Mary Matthews study [1] was the first devoted to long-term survival in patients with SCLC. The initial chemotherapy trials in which were reported longterm survivors were published in 1975 [2,3]. In 1977, an actuarial survival rate of 10% at 36 months was observed with a multiagent chemotherapy [4]. Since 1986, the American Cooperative Groups have extensively studied the prognostic factors of SCLC in patients included in various therapeutic randomized trials: the Cancer And Leukemia Group B (CALGB) [5], the South West Oncology Group (SWOG) [6,7]. Two Scandinavian institutions [8,9] performed as well as Peto [10] an analysis of predictors of long-term survival in SCLC. These studies were based essentially on pre-therapeutic and sometimes therapeutic determinants of the SCLC outcome. We

increasing the risk of relapse beyond 30 months (OR = 2.46, IC 95% (1.16-5.26), P = 0.01). The risk of relapse became less than 10% beyond five years. Twenty patients (13%) developed a second primary cancer in a mean time of 58.6 months. The risk of second primary cancer was increased by a number of chemotherapy cycles > 6 (OR - 3.25, IC 95% (1.08-9.8) P = 0.02) and by an age > 60 (OR = 2.92, IC 95% (1.07-7.97), P = 0.03). Five- and 10-year survival rates were respectively 68% and 44%. In these patients having reached a 30-month survival, three independent factors were predictive of a survival longer than five years: age s£ 60 at the time of diagnosis (OR = 2.85, IC 95% (1.23-6.6), P = 0.01), chest radiotherapy (OR = 3.1, IC 95% (1.28-7.69), P = 0.006) and absence of relapse (OR = 4.5, IC 95% (1.75-12.5), P = 0.002). This study suggests that: 1) therapeutic sequelae are rather mild, allowing return to work in 40% of patients; 2) relapsing 30-month survivors can benefit from second-line treatment; 3) SCLC cure can be achieved with a 10-year follow-up.

Key words: long-term survival, prognostic factors, small-cell lung cancer, treatment

initiated in 1986 a multicenter study in order to collect data on survival, quality of life, morbidity and mortality of a SCLC population having reached a 30-month survival. The question was: as soon as a patient reached a 30-month survival, are there factors which could influence his outcome beyond 30 months ?

Patients and methods In January 1986 we initiated on behalf of the GOLF (Thoracic Oncology Group of the Societe de Pneumologie de Langue Francaise) a national data base of long-term survivors in patients with SCLC. Registration was made on a volunteering basis from 52 institutions involved in lung cancer management (see Appendix). To be included, patients had to be in complete response 30 months after the initiation of treatment. Complete response according to the WHO criteria [11] was defined as complete disappearance of all measurable lesions and all other objective signs and symptoms of disease, including paraneoplastic syndromes. Each medical doctor was sent a pre-established

1010 written questionnaire made of three sheets containing 30 items. In the first sheet were retrospectively collected the pre-therapeutic characteristics of the 30-month survivors: age, sex, occupation, stage of disease, metastatic sites. The second sheet of the questionnaire was to describe the treatment modalities. At 30 months and then every year until death, the following data were collected in the third sheet, late toxicity of treatment, return to work, occurrence of a second primary (a histologlcal proof was mandatory), relapse of SCLC, physical activity (normal/ restricted/limited to self care), mental state (normal/memory loss/ confusion/language disorders/lack of interest in life), psychological state (normal/anxious/sleep disorders/depression/asthenia/abnormal behavior). To fulfill the third sheet of the questionnaire, the physician had to ask specific questions to each patient and to perform at least a complete clinical examination and a chest X-ray. Copies of the original questionnaire are available on request. A panel of three pathologists had to review all the slides and worked blinded to the clinical data; the WHO criteria [12] were used. Two hundreds and ten out of 263 (79.8%) slides were available, and the diagnosis of SCLC was confirmed for those which presented all the histological WHO criteria [12]. This analysis, depending on the quality of samples, allowed the pathologists to confirm SCLC in 101 cases (48%). Despite crushing artefacts, 54 (25.7%) more slides were added to the present study, including a total of 155 SCLC patients. An atypical carcinold was found in 9.5% of cases, a non SCLC in 7.6% of cases. In 9% of cases, the sample was inadequate to conclude.

Table 1. Characteristics of the population (total 155 cases). Sex Age

Mean (range) $60 Extent of disease Limited Diffuse Site of metastasis Liver Bone Brain Bone marrow Peripheral node Pancreas Adrenal glands Treatment - Chemotherapy More than six cycles - Radiotherapy Chest radiotherapy Cranial radiotherapy - Surgery

131 men 57 (27-88) 104

24 women

135(87%) 20(13%) 8 6 3 1 1 1 2 154(99.3%) 77 97 (62.6%) 79(51%) 27(17.4%)

smokers, 31 who stopped smoking several years before the diagnosis. Thirteen percent of patients had initially Statistical methods an extensive disease but the majority of them had a All the data of the 155 long-term survivors were computerized and single metastatic site; liver was involved in eight cases. updated to 1 June 1997. Initial description and examination of the data The initial treatment reflects the habits of each instituwere conducted using a standard data base. Univanate analyses were tion in the past decade. No patient received only a based on 2 x 2 tables with the covariates expressed as dichotomous 2 supportive care. A chemotherapy was performed in all variables and tested by x tests with continuity correction. When the cases but one who was operated. The most frequently P-value was less than 0.05, the differences were regarded as significant. The logistic regression model was used to analyze inside the long-term administered drugs were: adriamycin (138 of 154), cyclosurvivors population, relationships between pretherapeutic/therapeuphosphamide (137 of 154) and VP16 (126 of 154). The tic factors and probability of second primary cancer/relapse/five-year association of these three drugs (plus one or several survival. Covariates for which the /"-value was < 0.30 were included in more) was done in 110 of 154 (71,4%). Cisplatin was the multivariate model using BMDP software [13]. Estimates of the used in 67 of 154 cases (43%). The number of chemosurvival function for cases were made using the Kaplan-Meier method [14], given from initial treatment. therapy cycles varied from two to 18. Seventy-seven patients (50%) received more than six chemotherapy cycles. A chest and prophylactic brain irradiation was delivered in respectively 62,6% and 51% of cases. The mean dose Results for brain irradiation was 28,6 Gy (range 18-36). A In order to estimate the global rate of 30-month survi- surgical treatment was associated to other therapies in vors, nine of 52 institutions were assigned to give the 27 cases. total number of SCLC patients seen for the period Therapeutic sequelae were evaluated 30 months after during which long-term survivors were collected. These the beginning of treatment in all patients. The questionnine institutions, chosen because they had an exhaustive naire and a standard neurological examination revealed computerised data base of their SCLC patients, had neurological disorders in 20 of 155 patients (12.9%). registered a total of 1627 unselected patients with SCLC. They consisted in headaches, memory loss, hearing loss, Among them, 75 patients had been included in the na- fatigue, mental impairment, peripheral neuropathy. tional data base, representing 35% of all the pathologi- Pulmonary fibrosis was mainly attributed to radiothercally confirmed patients. Their characteristics were com- apy and occured in 28 of 155 patients (18%). Nine parable to those of the entire population of long-term patients developed a chronic respiratory failure requirsurvivors, so that a 4.6% 30-month survival rate could ing oxygen therapy. Cardiomyopathy and pericardial be finally estimated. tamponade were attributed to anthracyclins and /or chest irradiation, and were encountered in 17 of 155 The analysis was made on the 155 pathologically confirmed SCLC. Characteristics of the population at patients (11%). the time of diagnosis and treatment modalities are deCriteria for integration in a normal lifetime, i.e., scribed on Table 1. There were 131 men, 24 women, mean- physical, mental and psychological states, were evaluated aged 57 (27-82). One hundred and seven patients (71.8%) as previously described, at 30 months. Physical, mental were smokers at the time of diagnosis and among them, and psychological states were considered as normal in 15 did not stop smoking afterwards. There were 11 non respectively 70.3%, 87.7% and 67.7% of patients. For

1011 100

100

0

0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 46 48 51 54 57 60

12

24

36

48

60

72

84

96 108 120 132 144 156

Months

months

Figure 1. Survival of relapsing patients.

patients who suffered from abnormalities, there was no correlation with prophylactic cranial irradiation. Return to work was possible in 31 of the 77 patients (40%) who had an occupation at the time of diagnosis and aged less than 60 at 30 months. Return to work was not influenced by therapeutic sequelae, relapse, SPT, type of occupation (x 2 tests). Forty-three patients (27%) relapsed, 33 died of relapse. Relapses were loco-regional in 22 patients, metastatic in 19 patients, unprecised in two patients who were lost of follow-up. The last registered relapse occurred at the 104th month. Fifty percent of patients relapsed before the 39th month. Treatment of relapses consisted in chemotherapy (20 cases), radiotherapy (nine cases), chemoradiotherapy (three cases), palliative care (eight cases), unknown (three cases). Survival data of relapsing patients are given in Figure 1. The overall median survival was 12 months since the relapse (range 2-56+ months). Among the 23 patients retreated with chemotherapy for relapse, 13 received exactly the same chemotherapy as initially, and 10 received completely or partially differents schemes. Those 13 who were retreated with the same drugs had a median survival of 12 months, the other 10 had a median survival of 13 months. Among the 32 relapsing patients who received a second-line treatment, seven (21.8%) have survived again longer than 30 months since retreatment. Several parameters were studied through a univariate analysis with respect to the risk of relapse beyond 30 months in this population: age and stage of the disease at the time of diagnosis, sex, tobacco abuse after the diagnosis, cisplatin based regimen, number of cycles of chemotherapy > 6, surgery and chest irradiation for limited SCLC. The first seven parameters were entered in the multivariate model. The only independent predictor of relapse beyond 30 months was age > 60 at the time of diagnosis: OR = 2.46, IC 95% (1.15-5.27), P = 0.02. Twenty-three second primary tumors (SPT) occurred in 20 of 155 patients (12.9%). These tumors were non SCLC (nine cases), urinary tract tumors (seven cases), gastro intestinal cancers (four cases) head and neck cancers (two cases) and leukemia (one case). The mean

Figure 2. Survival of the 30-month survivors.

time of occurrence of these second primary tumors was 58.6 months from the initial treatment for SCLC (range 7-131 months). All the 20 patients with second primary tumors had been tobacco smokers. Among these 20 patients, 13 patients died; seven from their SPT, and six from other causes. Risk factors of SPT were examined first by a univariate method: age > 60 at the time of diagnosis and more than six chemotherapy cycles were of significant value and remained independent factors in the multivariate regression analysis with OR = 2.92, IC (1.07-7.9), P, = 0.03 and OR = 3.25, IC (1.08-9.8), P = 0.02 respectively. Sex, tobacco abuse after the diagnosis, stage of the disease at the time of diagnosis, therapeutic sequelae, cisplatin, CCNU and chest irradiation were not significant predictors of SPT. Non-cancer related deaths occurred in 30 patients (16.1%). They were due to strokes (seven), coronary or heart disease (eight), infectious diseases (eight), miscellaneous (seven). Five- and 10-year actuarial survival rates were respectively 68% and 44% (Figure 2). The survival curves show an excessively high mortality during the first five years. After five years, the mortality rate decreases. Disease-free survival rates at five and 10 years were respectively 75% and 65% (Figure 3). The risk of recurrence is less than 30% beyond three years and less than 10% beyondfiveyears. 100

12

24

36

48

60

72

84

96

108 120 132 144 156 months

Figure 3. Disease-free survival.

1012 thanfiveyears, had a normal lifestyle, free of recurrence or late toxic effects. It must be emphasized that social reintegration was possible since 40% of patients could return to work within the first two years following the diagnosis of SCLC. We did not find any predictor for that. In a previous study [24] it was reported that return to work depended on the kind - heavy or light - of occupation but not in ours. There are three main issues in the outcome of longterm survivors: the relapse of SCLC, the occurrence of a second primary malignancy and probably an excessive non-cancer related mortality. Relapses are more frequent in the first three years after diagnosis, but can still occur in the late follow-up. The median survival from relapse was 12 months and long-term survivals were observed after relapse, suggesting that there is a place for second-line therapy in patients who live longer than 30 months. A second-line treatment with either a same or a different chemotherapy as initially seems to give similar results in term of survival (12 and 13 month Discussion median survival respectively). We found that age > 60 Long-term survival was defined as a survival for at least at the time of diagnosis was an independent predictor of 30 months after the initiation of treatment. The 30- relapse, and as a consequence, that age ^ 60 and abmonth time corresponds to a significant change in the sence of relapse were independent predictors of five-year natural history of SCLC [15]. The most frequent cause of survival. Spiegelman and Albain found as well a unfavmortality within the two to three first years is relapse of orable effect of age > 60 on the survival of limited SCLC, while the mortality thereafter is primarily due to SCLC [5,6]. The influence of age was recently studied in second malignancies and intercurrent diseases [16]. The limited SCLC with a cut-off point of 70. In this study, pathological review confirmed 73% SCLC, and this is in age was not a significant prognostic variable on survival agreement with previous results. Davis [17] reported [25]. The favorable role of chest radiotherapy on five78 long-term survivors, with 66% of confirmed SCLC. year survival is in agreement with the current tendency Tison [18] collected 91 cases of long-term survival SCLC [26]. It must be stated that in our study, chest radiotherapy was delivered by different ways (early, late, alterof which 46% were confirmed as SCLC. natively or concomitantly, with various doses) and preTherapeutic sequelae have been studied in previous dominantly but not only in limited stage. reports. Central nervous system (CNS) toxicity is the most important problem. Armstrong [15], in a recent The risk of SPT has been often estimated. It exceeds review, noted it in 13%-78% of cases. Most of the clinical that of recurrent SCLC after five years [27-29]. The signs are memory loss, gait disturbances, seizures, de- actuarial risk of SPT may reach 50% at eight years. The mentia, psychomotor retardation. Some patients develop relative risk is 3.7 for all histologies of SPT, and 6.8 for CT scan abnormalities such as cerebral atrophy and non SCLC [27]. The discrepancy in the incidence of SPT ventricular dilation. There is a controversy about the between several reports can be explained by the difficulimpact of therapies on these CNS disorders. It has been ties of the diagnosis between a SPT and a relapse of the suggested that chemotherapy and particularly CCNU original cancer. Martini [29] defines SPT as histologimay contribute to CNS toxicity [19, 20]. The role of cally different tumors from the initial cancer, occuring in prophylactic cranial irradiation (PCI) is unclear [21]. a different site, at least two years after, in patients conHowever, a recent study presented by Komaki [22], per- sidered to be in complete remission. But, in many studies formed with a rigorous evaluation scale, failed to dem- on survival in SCLC (as in ours) an histological diagonstrate any direct CNS toxicity of PCI. Patients who nosis of relapses was not always established. In addition, presented cognitive impairment symptoms after PCI, it has been demonstrated that 30% to 38% of SCLC are were before treatment suffering from the same abnor- in fact combined tumors in which non SCLC clones of malities. We were unable to find a statistical difference in cells cannot easily be disclosed by small biopsies [30]. the physical, mental and psychological states between These clones could develop belately after SCLC treatpatients who received PCI and those who did not. ment. However, most of SPT are smoking related in Chemotherapy with more than six cycles, cisplatin or previous reports. Richardson [31] showed that smoking CCNU did not either affect these states. The majority of cessation significantly reduces the risk of SPT in longpatients had normal physical, mental and psychological term survivors. In our study, a SPT occurred in 12.9% of states at 30 months. Albain [23] showed as well, that patients, and 12 SPT out of 23 were smoking-related. 50% of patients with limited SCLC who survived more One case of acute leukemia was observed in a patient Furthermore, in this population of patients having reached a 30-month survival, we examined which factors could be predictive of a survival duration longer than five years. Among age at the time of diagnosis, sex, smoking cessation, stage of disease, therapeutic sequelae, chest radiotherapy, prophylactic cranial irradiation, number of chemotherapy cycles, cisplatin, surgery, relapse, second primary tumor, four parameters had a /•-value less than 0.30 in the univariate analysis: age s£60 at the time of diagnosis (P = 0.001), absence of relapse beyond 30 months (P < 10~4), chest radiotherapy (P - 0.07), no therapeutic sequelae at 30 months (P - 0.07). In the multiple regression model, three factors remained significant predictors of a >five-yearsurvival: age < 60 (OR = 2.85, IC 95% (1.23-6.6), P = 0.016), chest radiotherapy (OR = 3.1, IC 95% (1.28-7.69), P = 0.006) and absence of relapse (OR = 4.5, IC 95% (1.7512.5), P = 0.002).

1013 who presented a total of three SPT. Hematological malignancies are well described in long-term SCLC survivors [32]. The decrease of late haematological malignancies has been pointed out in some studies [27, 32] and might be due to the limitation of both the number of cycles of chemotherapy, and the use of CCNU and procarbazine. This is in agreement with our study: a ~nurnber~of"chemotherapy cycles greater than sfxTwas^m independent factor of SPT. An excessive non-cancer related mortality was noted in previous reports [33], which exceeds age-adjusted mortality. Brown [33] compared the rates of non cancer deaths in cancer patients and in US population. Cancer patients' data were obtained from the National Cancer Institute's Surveillance Epidemiology and End Results (SEER). The overall non-cancer relative hazard for lung cancer was 2.73 and it decreased dramatically with the number of years after diagnosis until the fourth year.

Conclusion This work could be carried out owing to the collaboration of 52 institutions. It gives a rather optimistic view of the long-term SCLC survivors, outlining three important points: 1. Therapeutic sequelae at 30 months are moderate and do not prevent further normal lifestyle and return to work which was possible in 40% of patients. 2. Second-line treatment should be considered in patients who relapse beyond 30 months. Retreatment, even with the same drugs as initially, can be highly effective, allowing again long-term survival. 3. SCLC cure can be achieved even in case of diffuse stage, with a 10-year follow-up. This selected population of long-term survivors probably have other prognostic factors which will have to be searched in the future. Acknowledgements The authors wish to thank Dr. A. Dubiez for the statistical analysis, B. Depierre for carrying out the data collection. 'Appendix List of the French institutions participating to the study: Amiens (Jounieaux); Angers (Maffre); Annecy (Datchary); Besancon (Depierre); Bordeaux (Taytard); Brest (Clavier, Gouva, Robinet); Caen (Riviere); Chartres (Zaegel); Chateaudun (Monnot); Chevilly-Larue (Homasson); Dijon (Massin, Coudert); Dreux (Fichet); Grenoble (Nagy - Mignotte, Moro); La Rochelle (Vincent); Le Coudray (Benoit); Le Havre (Duhamel, Destombes); Le Mans (Piron, Malbos); Lille (Tonnel, Lafitte); Lisieux (Abouz); Lyon Berard (Mornex, Rebattu); Lyon Louis Pradel (Trillet); Lyon Croix Rousse (Guerin); Marseille Conception (Boutin); Marseille Sainte Marguerite (KJeisbauer); Meaux (Blanchon); Montfermeil (Fabre); Montlucon (Geraads); Mulhouse

(Miech); Nancy (Anthoine, Menard, Feintrenie); Nantes (Dabouis); Nice (Lagrange); Percy (Allard); Pau (Malet); Poitiers (Patte, Boita); Reims (Marechal, Coninx); Rennes (Delaval); Pans La Salpetriere (Dautzenberg, Traveriat); Paris Hotel Dieu (Brechot); Paris Laennec (Mann); Paris Saint Antoine (Lebeau); Paris Tenon (Milleron, Francois); Saverne (Brolly, Malen); Saint Brieuc (Coetmeur); Saint Etienne (Fournel); Saint Germain en Laye (Leclerc); Strasbourg (Quoix, Roeslin); Tarbes (Silvani, Prud'homme); Tours (Lemarie); Troyes ^Tisserant);-Verdun -(David)r ViHejuif (Ruffie)rVillers-Sauit Denis (Renard).

References 1. Matthews MJ, Rozencweig M, Staquet MJ et al. Long-term survivors with small cell carcinoma of the lung. Eur J Cancer 1980; 16: 527-31. 2. Nixon DW, Carey RW, Suit HD. Combination chemotherapy in oat cell carcinoma of the lung. Cancer 1975; 36: 867-72. 3. Laing AH, Berry RJ, Newman CR. Treatment of small cell carcioma of bronchus. Lancet 1975; 1 (7899)- 129-32. 4. Israel L, Depierre A, Choffel A et al. Immunochemotherapy in 34 cases of oat cell carcinoma of the lung with 19 complete remissions. Cancer Treat Rep 1977; 61: 343-7. 5. Spiegelman D, Maurer LH, Ware JH et al. Prognostic factors in small cell carcinoma of the lung: An analysis of 1521 patients. J Clin Oncol 1989; 7 (3): 344-54. 6 Albain K.S, Crowley JJ, Leblanc M, Livingston RB. Determinants of improved outcome in small cell lung cancer: An analysis of the 2580 patients southwest oncology group data base. J Clin Oncol 1990; 8 (9): 1563-74. 7. Sagman U, Naki E, Evans NK. Small cell carcinoma of the lung: derivation of a prognostic staging system. J Clin Oncol 1991; 9. 1639-49. 8. Osterlind K, Hansen HH, Hansen M et al. Long-term diseasefree survival in small-cell carcinoma of the lung: A study of clinical determinants. J Clin Oncol 1986; 4: 1307-13. 9. Lassen U, Osterlind K, Hansen M et al. Long-term survival in small-cell lung cancer: Posttreatment characteristics in patients surviving 5 to 18+ years-an analysis of 1,714 consecutive patients. J Clin Oncol 1995, 13:1215-20. 10. Rawson NSB, Peto J. Overview of prognostic factors in small cell lung cancer. Br J Cancer 1990; 61: 597-604. 11. Miller AB, Hoogstraten B, Staquet M. Reporting results of cancer treatment. Cancer 1981; 47: 207-14. 12. Kreyberg L. Histological Typing of Lung Tumors: International Histological Classification of Tumors. Geneva: World Health Organization 1967. 13. BMDP Statistical Software. Berkeley, CA: University of California Press 1981. 14. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457-81. 15. Armstrong JG. Long-term outcome of small cell lung cancer. Cancer Treat Rev 1990; 17: 1-13. 16. Stephens RJ, Bailey AJ, Machin D. Lung Cancer Working Party. Long-term survival in small cell lung cancer: The case for a standart definition. Lung Cancer 1996; 15: 297-309. 17. Davis S, Wright PW, Schulman SF et al. Long-term survival in small cell carcinoma of the lung: A population experience. J Clin Oncol 1985; 3(1): 80-91. 18. Tison V, Rosti G, Caruso G et al Pathologic review of SCLC long-term survival. Third International Conference on SCLC, Ravenna, Italy, 1993 (Abstr). 19. Frytak S, Shaw J, O'Neill B et al. Leucoencephalopathy in smallcell lung cancer patients receiving prophylactic cranial irradiation. Am J Clin Oncol 1989; 12: 27-33. 20. Looper J, Einhorn L, Garcia S et al. Severe neurologic problems following successful therapy for small cell lung cancer (SCLC). Proc Am Soc Clin Oncol 1984; 3: A 231.

1014 21. Rosen S, Makuch R, Lichter A et al. Role of prophylactic cranial irradiation in prevention of central nervous system metastases in small cell lung cancer. Potential benefit restricted to patients with complete reponse. Am J Med 1983; 74: 615-24. 22. Komaki R, Meyers CA, Shin DM et al. Evaluation of cognitive function in patients with limited small-cell lung cancer prior to shortly following prophylactic cranial irradiation. Int J Radiat Oncol Biol Phys 1995; 33 (1): 179-82. 23. Albain KS, Crowley JJ, Livingston RB. Long-term survival and toxicity in small-cell lung cancer: Expanded South West Oncology Group experience. Chest 1991; 99: 1425-32. 24. Bergman B, Sorenson S. Return to work among patients with small cell lung cancer. Eur J Respir Dis 1987; 70. 49-53. 25. Siu LL, Shepherd FA, Murray N et al. Influence of age on the treatment of limited-stage small cell lung cancer. J Clin Oncol 1996; 4 (3): 821-8. 26. Pignon JP, Arriagada R, Ihde D. A meta-analysis of thoracic radiotherapy for small all lung cancer. N Engl J Med 1992; 327: 1618-24. 27. Heyne KH, Lippman SM, Lee JJ et al. The incidence of second primary tumors in long-term survivors of small-cell lung cancer. J Clin Oncol 1992; 10: 1519-24. 28. Ihde DC,Tucker MA. Second primary malignancies in small-cell lung cancer — a major consequence of modest success J Clin Oncol 1992; 10: 1511-3.

29. Martini N, Melamed M. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1975, 70: 606-12. 30. Sehested M, Hirsch FR, Osterlind K Morphologic variations of small cell lung cancer: A histopathologic study of pretreatment and posttreatment specimens in 104 patients. Cancer 1986; 57: 804-7. 31. Richardson GE, Tucker MA, Venzon DJ et al. Smoking cessation after successfull treatment of small cell lung cancer is associated with fewer smoking-related second primary cancers. Ann Intern Med 1993; 119 (5): 383-90. 32. Souhami RL, Law K. Longevity in small cell lung cancer a report to the lung cancer subcommittee of the United Kingdom Coordinating Committee for cancer research. Br J Cancer 1990; 61: 584-9. 33. Brown B, Brauner C, Minnotte M. Non cancer deaths in white adults cancer patients. J Natl Cancer Inst 1993; 85: 979-87. Received 11 July 1997; accepted 4 September 1997. Correspondence to: A. Depierre, MD Service de Pneumologie Villers Saint Denis CHU 25000 Besancon France