Outcomes in recurrent small-cell lung cancer after one to four chemotherapy lines: A retrospective study of 300 patients

Lung Cancer 78 (2012) 112–120

Contents lists available at SciVerse ScienceDirect

Lung Cancer journal homepage: www.elsevier.com/locate/lungcan

Outcomes in recurrent small-cell lung cancer after one to four chemotherapy lines: A retrospective study of 300 patients Hélène Nagy-Mignotte a,b,c,∗ , Pascale Guillem a,b , Lucile Vignoud b , Marie Coudurier c , Aurélien Vesin b , Vincent Bonneterre d , Anne-Claire Toffart b,c , Linda Sakhri c , Christian Brambilla b,c , Elisabeth Brambilla b,e , Jean-Franc¸ois Timsit b,f , Denis Moro-Sibilot a,b,c , For the Multidisciplinary Thoracic Oncology Group Grenoble University Hospital, France a

Oncology Coordination Centre, Grenoble University Hospital, Grenoble, France INSERM Unit 823, Institute Albert Bonniot, Grenoble, France c Pneumology Clinic, Multidisciplinary Thoracic Oncology Group, Grenoble University Hospital, France d Centre for Occupational and Environmental Diseases, Grenoble University Hospital, France e Department of Pathological Anatomy and Cytology, Grenoble University Hospital, France f Medical Intensive Care Unit, Grenoble University Hospital, France b

a r t i c l e

i n f o

Article history: Received 20 February 2012 Received in revised form 31 May 2012 Accepted 9 June 2012 Keywords: Small cell lung cancer Recurrence Chemotherapy lines Survival Sensitivity to treatment Prognosis

a b s t r a c t Standard treatment of small-cell lung cancer (SCLC) is a combination of etoposide and platinum for patients with extensive disease, associated with radiotherapy for patients with limited disease (LD). Therapeutic strategies for relapse, although well characterized, are disappointing. Between 1997 and 2009, 300 patients were treated for SCLC at Grenoble University Hospital. We analyzed patients’ characteristics and outcomes at different treatment steps, to determine prognostic factors and propose subsequent treatment strategies according to “sensitive”, “resistant” or “refractory” status established after first-line treatment (L1). The median patient age was 63.2 years, 46.3% had LD, and 23% were female. The objective response rate (ORR) to first-line chemotherapy was 73% [CI95% : 67.6–77.9] and median survival was 13 months. After L1, comparison between “refractory” and “sensitive” groups showed more extensive disease (76.6% vs. 34.3%, p = 0.003), poorer Performance Status (PS 0–1: 48.4% vs. 67.8%, p = 0.008), more endocrine paraneoplastic syndrome (18.7% vs. 8.4%, p = 0.03) and more composite histology (17.2% vs. 4.9%, p = 0.004) in “refractory” patients. After second line (L2), ORR was 55.8% [CI95% : 45.2–66.0] in “sensitive”, 18.2% [CI95% : 8.2–32.7] in “resistant”, and 14.7% [CI95% : 4.9–31.0] in “refractory” groups; with partial response only for the last two groups. After L3 and L4, ORR was 24.0% [CI95% : 14.9–35.2] in “sensitive”, 9.1% [CI95% : 11.2–29.2] in “resistant” with partial response only. No response was observed for “refractory”. After L1, the median survival was respectively 23, 10 and 6.4 months for “sensitive”, “resistant” and “refractory” groups (p < 0.001). Multivariate analysis showed that LD and classical SCLC histology were positive predictors of belonging to the “sensitive” group. Positive factors for survival were sensitivity to L1, PS 0–1, LD, Charlson score <4, no endocrine paraneoplastic syndrome and no occupational exposure. Limited disease is the major predictive factor for sensitivity to treatments and survival. Factors linked to the patients’ clinical presentation also impact on survival. With currently recommended drugs, the “sensitivity” of the patient determined by the response to L1 indicates that it is pointless to treat “sensitive” with L4, “resistant” with L3 and “refractory” with L2, except for a few selected patients after multidisciplinary group discussion. © 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

∗ Corresponding author at: Centre de Coordination en Cancérologie-3C, CHU de Grenoble, BP 217, 38043 Grenoble Cedex 9, France. Tel.: +33 4 76 76 54 36/+33 4 76 76 94 34; fax: +33 4 76 76 75 55. E-mail addresses: [email protected], [email protected] (H. Nagy-Mignotte). 0169-5002/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.lungcan.2012.06.006

Currently, small cell lung cancer (SCLC) accounts for 12–15% of all cases of lung cancer [1,2]. Chemotherapy combined with radiotherapy is the mainstay of treatment for limited disease (LD) whereas chemotherapy alone is the standard therapeutic strategy for patients with extensive disease (ED). For about 25 years, four to six cycles of etoposide–platinum has been the standard first-line chemotherapy strategy [1,3,4] and has

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

been used in our institution since 1990. The rate of recurrence is high. Among the different therapeutic options for patients with recurrent disease there is consensus that second line chemotherapy should be considered. Regimens include topotecan, cyclophosphamide/doxorubicin/vincristine (CAV) and rechallenge with earlier first-line agents. Combinations including vinorelbine, gemcitabine and paclitaxel/carboplatin have shown activity in phase II studies [1,5]. However, only topotecan has been registered for second-line therapy by the regulatory agencies [1,6,7]. In the USA, 5-year survival has improved modestly from 4.9% in 1973 to 10% in 1998 for patients with limited disease [8]. In our institution 2-year survival has improved rising from 14.8 to 24% from 1982 to 2007, indicating a better control of first-line treatment. However, the results remain disappointing after relapse. In this retrospective study over 12 years, we describe the profile of patients at different stages of the disease and analyze the factors influencing the response to successive lines of chemotherapy and survival. 2. Methods 2.1. Population Consecutive cases of SCLC diagnosed by histology or cytology between October 1997 and September 2009 at Grenoble University Hospital and receiving at least one line of chemotherapy were prospectively included. The following parameters were recorded: age, gender, weight loss, WHO Performance Status, Charlson comorbidity score [9], history of other cancer, tobacco consumption and pack-years. Patients with history of occupational exposure were seen by an occupational health specialist and the exposure classified according to national and European codes [10]. Medical characteristics at diagnosis including stage (UICC 1987, 1998, 2003, 2007), limited or extensive form of the disease, histology (classical or composite SCLC according to WHO Systematized Nomenclature of Medicine Classification) and paraneoplastic syndrome were noted. The response to each line of chemotherapy, types of regimens and combined treatments such as surgery and/or radiotherapy, dates of diagnosis and treatments, and vital status at database closure (July 15, 2011) were recorded. Tumor responses were graded as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD), using the RECIST criteria [11]. The objective response rate (ORR) included partial and complete responses. We classified patients according to their response to first-line chemotherapy as follows: - “Sensitive” patients with an objective response 3 months after completing chemotherapy. - “Resistant” patients with an objective response lasting less than 3 months. - “Refractory” patients with no objective response to first-line chemotherapy. as described in published studies [1,5,12]. Patients who had undergone surgery and then received chemotherapy as adjuvant treatment were classed as “sensitive” if no relapse was reported in the 3 months following the end of treatment. End of treatment was considered to be the last cycle of chemotherapy or the end of mediastinal radiotherapy according to the treatments used. Patients alive and in complete remission 36 months after diagnosis were considered as “Long-term survivors” (LS). The standard treatment was etoposide–platinum. We grouped other chemotherapy regimens into four classes

113

according to the main active agent: topotecan, taxanes (mainly paclitaxel–carboplatin), doxorubicin (mainly CAV), and other drugs (eTable 1). Mediastinal and prophylactic cerebral radiotherapy were included in the first line regimen for limited disease only and was discussed on a case by case basis for extensive forms. Palliative radiotherapy is included in the “other” treatment category. 2.2. Statistical analysis Characteristics of the different study groups were compared using Chi2 or Fisher two-sided tests and the Student test for comparison of means. Characteristics between first-line and fourth-line groups were compared using a Chi2 test using the first-line group as the reference population. The clinical variables at diagnosis suspected to be associated with sensitivity were tested in multivariate analysis using a logistic regression model. Survival curves were calculated from the first day of each line of chemotherapy. Clinical variables potentially linked to survival were tested by multivariate analysis using the Cox proportional hazard model (overall survival). Age was recoded as a categorical variable in 10 year intervals from 50 onwards. Variables were dichotomized and were submitted to a stepwise selection procedure for multivariate analysis. Variables with p < 0.05 in the multivariate analysis were retained in the model. All statistical analyses were performed using STATA 9.2 (StataCorp, USA). 3. Results 3.1. Patient baseline characteristics We identified 311 patients, 300 received at least one line of chemotherapy and were included in the study (Table 1). None were lost to follow-up. Most patients were heavy smokers with a mean of 40 pack years and never smokers were exceptional. Endocrine paraneoplastic syndromes (EPS: Cushing and/or SIADH) were present in 13.3% of cases and were associated with extensive disease: 70% vs. 48.9% ED in the groups with and without EPS, respectively (p = 0.01). Occupational exposure was reported for 102 patients (34%), with asbestos accounting for 45% of this. Other types of occupational exposure are detailed in eTable 2. At diagnosis women were usually younger than men, 59.6 years [CI95% : 56.9–62.3] vs. 64.2 years [CI95% : 62.9–65.5] (p = 0.001), had less occupational exposure (15.9% vs. 39.4%, p < 10−3 ), less frequent comorbidity (Charlson < 4: 65.2% vs. 43.3%, p = 0.01), better PS (0–1) (65.2% vs. 47.6, p = 0.01) and more classical histological forms (97.1% vs. 89.6%, p = 0.05). Smoking habits, median packet years 41 [CI95% : 35.4–46.6] and 40 [CI95% : 35.2–44.8] and disease extension at diagnosis were similar between women and men respectively. 3.2. First line treatment At first-line treatment (Fig. 1 and eTable 1): 288 patients received etoposide–platinum including 174 (60.4%) with cisplatin, 94 (32.6%) with carboplatin and 20 (6.9%) cisplatin then carboplatin, 8 had another combination that included platinum and 4 received no platinum. Ninety-seven (32.3%) patients received mediastinal radiotherapy (90 LD and 7 ED), 74 (24.7%) received prophylactic brain irradiation (61 LD and 13 ED) and 24 (8%) had surgery. Among patients who had surgery and then adjuvant chemotherapy, 16 had no pre-operative diagnosis of SCLC, 3 had a diagnosis of malignancy without further precision, and one patient had a histological diagnosis. Eight patients had surgery after neoadjuvant chemotherapy. At the end of first-line chemotherapy 279 patients (93%) were classified: as “sensitive” 143 patients (47.7%), “resistant” 72

114

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

Table 1 Characteristics of patients according to the different lines of chemotherapy. Patients with 1st line n = 300 (%)

Patients with 2nd line n = 173 (%)

Patients with 3rd line n = 78 (%)

Patients with 4th line n = 27 (%)

Long term survivors n = 21 (%)

Female

69 (23.0)

42 (24.3)

23 (29.5)

11 (40.7)

6 (28.5)

Mean age Extremes (CI95% )

63.2 29–84 (62.0–64.4)

61.0

58.2

(59.3–62.7)

(55.8–60.6)

57.2 29–77 (53.1–61.3)

59.9 46–75 (56.2–63.7)

Age ≥ 60 years Age < 60

188 (62.7) 112 (37.3)

94 (54.3) 79 (45.7)

37 (47.4) 41 (52.6)

13 (48.1) 14 (51.9)

10 (47.6) 11 (52.4)

147 (85.0) 6 (3.5) 20 (11.6) 40 (35–45)

69 (88.5) 1 (1.3) 8 (10.2) 40 (33–47)

22 (81.5) 1 (3.7) 4 (14.8) 40 (28–52)

21 (100.0) 0 0 40 (28.5–51.5)

28 (16.2) 36 (20.8)

13 (16.7) 21 (26.9)

4 (14.8) 7 (25.9)

2 (9.5) 5 (23.8)

9 (5.2) 8 (4.6)

1 (1.3) 4 (5.2)

1 (3.7) 3 (11.1)

1 (4.7) 2 (9.5)

131 (75.7) 22 (12.7) 20 (11.6)

62 (79.5) 8 (10.3) 8 (10.3)

23 (85.2) 1 (3.7) 3 (11.1)

19 (90.5) 1 (4.8) 1 (4.8)

113 (65.3) 59 (34.1) 1 (0.6)

50 (64.1) 27 (34.6) 1 (1.3)

17 (63.0) 10 (37.0) 0

19 (90.5) 2 (9.5) 0

109 (63.0) 63 (36.4) 1 (0.5) 164 (94.8)

60 (76.9) 18 (23.1) 0 74 (94.9)

21 (77.8) 6 (22.2) 0 26 (96.3)

17 (80.9) 4 (19.0) 0 5 (23.8)

157 (95.7) 6 (3.7) 1 (0.6)

71 (96.0) 2 (2.7) 1 (1.3)

26 (100) 0 (0.0) 0 (0.0)

0 2 (40.0) 3 (60)

97 (56.0) 76 (43.9)

36 (46.1) 42 (53.8)

10 (37.0) 17 (63.0)

1 (4.8) 20 (95.2)

23 (13.3) 2 (1.1)

10 (12.8) 0 (0.0)

3 (11.1) 0 (0.0)

2 (9.5) –

161 (93.1) 12 (6.9)

72 (92.3) 6 (7.7)

25 (92.6) 2 (7.4)

19 (90.5) 2 (9.5)

28 (16.2) 144 (83.2) 1 (0.5)

12 (15.4) 66 (84.6) 0

5 (18.5) 22 (81.5) 0

14 (66.7) 7 (33.3) 0

34 (19.6) 44 (25.4) 95 (54.9) –

7 (9.0) 17 (21.8) 54 (69.2) –

1 (3.7) 5 (18.5) 21 (77.8) –

0 0 21 (100.0) –

11 (6.3)

4 (5.1)

0 (0.0)

9 (42.8)

56 (32.4) 47 (27.2) 25 (14.5)

33 (42.3) 31 (39.7) 6 (7.6)

14 (51.8) 13 (48.1) 0 (0.0)

16 (76.2) 12 (57.1) –

Tobacco consumption Smokers 260 (86.7) Never smokers 10 (3.3) 30 (10.0) Missing data 40 (36–44) Pack – years, median (CI95% ) Occupational exposure Asbestos 46 (15.3) Other 56 (18.7) Cancer history 21 (7.0) Lung/head and neck Other cancers 23 (7.6) Weight loss 220 (73.3) <10% 48 (16.0) ≥10% 32 (10.7) Missing data WHO Performance Status (PS) at diagnosis 170 (56.7) 0–1 ≥2 128 (42.7) 2 (0.7) Missing data Charlson score at diagnosis 155 (51.6) Score < 4 144 (48.0) Score ≥ 4 Missing data 1 (0.3) Death 275 (92.0) Cause of death 236 (85.8) Died of cancer Iatrogenic complication 21 (7.6) 18 (6.5) Other Form of disease 161 (53.7) Extensive-form Limited-form 139 (46.3) Paraneoplastic syndrome 40 (13.3) Endocrine 4 (1.3) Other Histologya “8041 3” – “Classical” SCLC 274 (91.3) “8045 3” – Composite SCLC 26 (8.7) Stage 61 (20.3) I, II, IIIA 237 (79.0) IIIB, IV 2 (0.6) Missing data First line sensitivity 64 (21.3) Refractory 72 (24) Resistant 143 (47.7) Sensitive 21 (7.0) Non evaluable Associated treatment 24 (8.0) Curative surgery Radiotherapy 97 (32.3) Mediastinal 74 (24.7) Cerebral prophylactic 40 (13.3) Other a

Systematized Nomenclature of Medicine (SNOMED) classification.

patients (24%) or “refractory” 64 patients (21.3%). Sixteen patients died of complications before evaluation and for 5 patients the date of relapse was not available.

in L4. Specifically, the complete response rate dropped from 31.3% to 2.5% from L1 to L3. In L4 only partial responses were obtained.

3.3. Overall response to subsequent lines of chemotherapy

3.4. Responses to subsequent lines of chemotherapy according to the first-line sensitivity

The rate of objective responses to the different lines of chemotherapy are given in Fig. 2a and eTable 3 contains the confidence intervals for these results. Overall, the objective response rate decreased from 73% in L1 to 38.1% in L2, 19.2% in L3 and 18.5%

3.4.1. “Sensitive” patients Objective response fell from 55.8% at L2 to 23.8% at L4, for which there were only partial responses. Complete response was only

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

115

n=311 patients n=11 palliative care

n=288 standard n=12 other

Line 1, n=300 161 - extensive disease - 139 limited disease

n=75 standard n=72 taxane n=12 topotecan n=9 doxorubicine n=5 other

- 97 extensive disease - 76 limited disease

n=8 standard n=32 taxane n=18 topotecan n=7 doxorubicine n=13 other

Line 3, n=78 - 36 extensive disease - 42 limited disease

N=127 censored : n=80 died of cancer n=15 died of complications n=17 died of other causes n=14 alive with complete response n=1 alive with cancer

Line 2, n=173

N= 95 censored : n=86 died of cancer n=4 died of complications n=5 alive with complete response

N= 51 censored :

n=9 taxane n=5 topotecan n=5 doxorubicine n=8 other

Line 4, n=27

- 10 extensive disease - 17 limited disease

n=45 died of cancer n=2 died of complications n=1 died of other cause n=1 alive with complete response n=2 alive with cancer

N=21 censored : n=21 died of cancer

n=1 standard n=1 taxane n=1 doxorubicine n=3 other

Line 5, n=6

- 6 limited disease

n=5 died of cancer n=1 alive with cancer

Fig. 1. Flow chart of the population according to the chemotherapy lines.

observed for a few patients in this group, 15 at L2 and 2 at L3 (Fig. 2b and eTable 3). 3.4.2. “Resistant” patients For “Resistant” patients only a few PR were obtained with L2 (18.2%) and even less with L3 (2 out of 17 patients). No response was obtained with L4 (Fig. 2c, eTable 3). 3.4.3. “Refractory” patients For “Refractory” patients, while a few partial responses were obtained with L2 (14.2%) and no response to further lines (Fig. 2d and eTable 3). 3.5. Drugs used in L2 according to the sensitivity to L1 (eTable 4) For “Sensitive” patients, 71.6% were re-treated with etoposide–platinum leading to an objective response in 65%,

with 20.6% achieving CR. Taxanes were proposed in 20% of cases. Among these, 36.8% patients obtained an ORR, with CR for one patient. For “Resistant” and “Refractory” patients, taxane based chemotherapy was proposed for 63.6% and 73.5% of patients respectively. Whatever the protocol used the response rate was very low, from no response to 21.4% with partial response. In these two groups no combination had significant activity. In the three groups, no response was obtained with topotecan. 3.6. Drugs used in L3 and L4 according to the sensitivity to L1 are shown in eTable 5 Due to the small number of patients concerned the data for lines 3 and 4 have been combined. For “Sensitive” patients, the standard treatment at L3 was possible for 8 patients and led to 5 ORR including 2 CR. Taxanes were

116

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

Fig. 2. (a–d) Objective response to chemotherapy for all patients and according to the group of sensitivity to first line treatment.

used after failure of standard treatment in 40.0% of patients and allowed 30.0% of PR in this group. A few PR were obtained with topotecan (17.6%). For “Resistant” patients whatever the treatment used there was only a very low response rate (9.1%). For “Refractory” patients none responded to any drug.

3.7. Differences in baseline characteristics between sensitive and refractory patients “Refractory” compared with “Sensitive” patients, were more often in poor physical condition (PS 0–1: 48.4% vs. 67.8%, p = 0.008), had more extensive disease (76.6% vs. 34.3%, p < 0.003), more often had EPS (18.7% vs. 8.4%, p = 0.03) and had more composite histologies (17.2% vs. 4.9%, p = 0.004). We found no difference between the groups in terms of sex, median age, smoking habits and occupational exposure (Table 2).

3.8. Differences in baseline characteristics of patients responding to L1–L4 During the subsequent lines of chemotherapy (from L1 to L4) the remaining patients were those initially “sensitive” to L1 (47.7% vs. 77.8%, p = 0.046) with few comorbidities (Charlson < 4: 51.6% vs. 77.8%, p = 0.046). Other comparisons between gender, disease extension at diagnosis, PS, weight loss, concomitant therapies (surgery and/or radiotherapy) were not discriminative (Table 1).

3.10. Survival Overall survival is shown in Table 3 and eFigs. 1–7. The median follow-up for patients still alive at the end of the study (n = 24) was 69.9 months [extremes: 17.3–154]. All “Refractory”, 98.6% of “Resistant” and 84% of “Sensitive” patients died. The median survival for all patients was 13 months, 21 for LD and 8.6 for ED (p < 10−3 ). After L1, median survival decreases to 7.4, 5.1 and 3.6 months for lines 2, 3 and 4 respectively. Female and male survival rates were similar (median 15.6 months vs. 12.4 months, p = 0.06). Concomitant EPS had a pejorative impact on survival (median 7.7 months vs. 14.5 months, p < 10−3 ). Survival was linked to sensitivity to L1 (Fig. 3 and Table 3), with medians of 23, 10 and 6.4 months for “Sensitive”, “Resistant” and “Refractory” patients respectively (p < 10−3 ). The median survival was significantly different between “Resistant” and “Refractory” groups (p = 0.005). Eighteen percent of “Sensitive” patients were alive at 5 years. The median survival from L2 was 9.3, 5.2 and 3.8 months for “Sensitive”, “Resistant” and “Refractory” groups

1.00

0.75

0.50

0.25

3.9. Differences in baseline characteristics of patients with “treatment failure” and “Long-term survivors”

0.00 0

We compared “Long-term Survivors” with patients who had received four cycles of chemotherapy (L4) and were considered as “treatment failure”. The LS group was associated with sensitivity to first-line treatment (100% of LS patients vs. 77.8% in L4, p = 0.03), limited disease (LD 95.2% vs. 63.0%, p = 0.01), surgery (42.8% vs. 0%, p < 10−3 ) and better PS (PS = 0–1: 90.5% vs. 63%, p = 0.03) (Table 1).

50

100 analysis time (months) Refractory Sensitive

150

Resistant

p<10-3 all groups; p= 0.005 “Resistant” vs “Refractory”. Fig. 3. Kaplan–Meier overall survival estimate by sensitivity to line 1.

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

117

Table 2 Patient characteristics by sensitivity to first line treatment. Sensitive N = 143 (%)

Resistant N = 72 (%)

Refractory N = 64 (%)

Female Mean age (CI95% )

33 (23.1) 63.0 (61.3–64.7)

19 (26.4) 63.6 (61.2–66.0)

13 (20.3) 61.2 (58.4–63.9)

Age ≥ 60 years Age < 60 years

91 (63.6) 52 (36.4)

48 (66.7) 24 (33.3)

32 (50.0) 32 (50.0)

Tobacco consumption Smokers Never smokers Missing data Pack – years, median (CI95% ) Occupational exposure Asbestos Other Cancer history Lung/head and neck Other cancers Weight loss <10% ≥10% Missing data WHO Performance Status (PS) at diagnosis 0–1 ≥2 Missing data Charlson score at diagnosis Score < 4 Score ≥ 4 Missing data Death Cause of death Died of cancer Iatrogenic complication Other Form of disease Extensive-form Limited-form Paraneoplastic syndrome Endocrine Other Histology “8041 3” – “Classical” SCLC “8045 3” – Composite SCLC Stage I, II, IIIA IIIB, IV Missing data Associated treatment Curative surgery Radiotherapy Mediastinal Cerebral prophylactic Other

128 (89.5) 4 (2.8) 11 (7.7) 31 (27–34) 25 (17.5) 30 (21.0)

60 (83.3) 2 (2.8) 10 (13.9) 26.5 (21–31)

53 (82.8) 3 (4.7) 8 (12.5) 31 (25–36)

8 (11.1) 15 (20.8)

12 (16.6) 9 (14.1)

6 (4.1) 11 (7.6)

5 (6.9) 5 (6.9)

5 (7.8) 4 (6.2)

112 (78.3) 17 (11.9) 14 (9.8)

46 (63.9) 15 (20.8) 11 (15.3)

45 (70.3) 14 (21.9) 5 (7.8)

97 (67.8) 44 (30.8) 2 (1.4)

37 (51.4) 35 (48.6) 0

31 (48.4) 33 (51.6) 0

84 (58.7) 59 (41.3) 0 120 (83.9)

36 (50.0) 36 (50.0) 0 71 (98.6)

32 (50.0) 31 (48.4) 1 (1.6) 64 (100.0)

108 (90.0) 4 (3.3) 8 (6.6)

64 (90.1) 5 (7.0) 2 (2.8)

63 (98.5) 1 (1.5) –

43 (34.3) 94 (65.7)

49 (68.1) 23 (31.9)

49 (76.6) 15 (23.4)

12 (8.4) 1 (0.7)

12 (16.7) 2 (2.8)

12 (18.7) 0 (0.0)

136 (95.1) 7 (4.9)

66 (91.7) 6 (8.3)

53 (82.8) 11 (17.2)

47 (32.9) 96 (67.1) 0

9 (12.5) 63 (87.5) 0

3 (4.8) 59 (92.2) 2 (3.1)

20 (14.0)

3 (4.1)

1 (1.5)

77 (53.8) 63 (44.0) 14 (9.8)

13 (18.0) 8 (11.1) 15 (20.8)

respectively (eFig. 8, p < 10−3 ). From L2 there was no longer a significant difference between the “Resistant” and “Refractory” groups (p = 0.8).

3.11. Predictive factors analysis for response and survival In logistic regression analysis, the only positive predictors of belonging to the “sensitive” group were: limited disease (Odds Ratio 5.1, [CI95% : 3.0–8.5], p < 10−3 ) and classical histological SCLC (Odds Ratio 3.19, [CI95% : 1.19–8.51], p = 0.021). In the multivariate Cox model, the factors favorably influencing survival were sensitivity to L1 (hazard ratio (HR) 0.19, p < 10−3 , [CI95% : 0.14–0.27]), Performance Status 0–1 (HR 0.63, p = 0.002, [CI95% : 0.48–0.84]) and limited disease (HR 0.42, p < 10−3 , [CI95% : 0.31–0.58]).

6 (9.4) 0 9 (14.1)

Negative predictors of survival were Charlson score ≥ 4 (HR 1.88, p < 10−3 , [CI95% : 1.4–2.53]), the presence of endocrine paraneoplastic syndrome (HR 1.63, p = 0.017, [CI95% : 1.09–2.42]), and occupational exposure (all types of exposure) (HR 1.39, p = 0.023, [CI95% : 1.05–1.85]) (eTable 2). This was confirmed when only asbestos exposure was considered (HR 1.48, p = 0.024, [CI95% : 1.05–2.1]). Mediastinal radiotherapy was not included in the multivariate model as it concerns only patients with limited disease (92.7% of patients receiving this treatment had LD).

4. Discussion This study highlights the difficulty of salvage chemotherapy for SCLC beyond first line treatment and the outcomes of successive

118

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

Table 3 Summary of survival.

All patients Limited disease Extensive disease Male Female EPSa Yes EPSa No Treatment < 2004 Treatment ≥ 2004 Survival from: Line 1 Line 2 Line 3 Line 4 Survival after L1 Sensitive Resistant Refractory Survival after L2 Sensitive Resistant Refractory a

N

Median (months)

[CI95% ]

Alive at 1 year N (%)

Alive at 3 years N (%)

Alive at 5 years N (%)

300 139 161 231 69 40 260 158 142

13 21 8.6 12.4 15.6 7.7 14.5 12.8 13.6

11.7–14.9 17.2–25.6 7.9–10.5 10.6–14.5 12.3–21.3 6–11.1 12.5–16.1 11.2–14.9 10.4–16.3

165 (50.7) 106 (71.2) 59 (32.9) 119 (47.6) 16 (60.9) 13 (22.5) 152 (55.0) 85 (51.9) 80 (49.3)

34 (12.3) 32 (24.7) 3 (1.4) 24 (11.1) 13 (16.5) 3 (5.0) 32 (13.4) 22 (13.3) 15 (11.2)

21 (8.7) 20 (17.9) 2 (0.6) 14 (6.9) 7 (13.0) 2 (2.5) 19 (9.2) 16 (9.5) 6 (8.1)

300 173 78 27

13 7.4 5.1 3.6

11.7–14.9 6.0–8.4 3.8–6.0 2.6–4.2

165 (50.7) 37 (20.8) 9 (9.7) 3 (6.7)

34 (12.3) 5 (3.3) 2 (1.9) –

21 (8.7) 3 (3.3) – –

143 72 64

23.1 10.0 6.4

20.4–27.0 8.1–11.1 4.6–7.7

125 (83.2) 25 (27.8) 12 (15.6)

34 (25.7) 0 0

21 (18.1) 0 0

95 43 34

9.3 5.2 3.8

8.3–10.9 2.8–6.3 3.1–6.4

31 (32.5) 4 (7.0) 4 (9.0)

5 (6.0) 0 0

3 (5.9) – –

EPS: endocrine paraneoplastic syndrome.

lines of treatment in our group of 300 patients reveal two previously overlooked patient characteristics. Firstly, as opposed to appearances, women are at a disadvantage. Contrary to what has been found in some other studies [3,5] their survival was not significantly better than that of men even though their condition at diagnosis was superior (better PS and less comorbidities). With a similar degree of smoking and less occupational exposure than men, women appear to be diagnosed with SCLC at a younger age. This supports the previously proposed hypothesis that they are more susceptible to pulmonary carcinogens [13–15]. Secondly, a history of occupational exposure has a negative impact on survival. Asbestos is known for its carcinogenic risk to the pleura and the bronchi, which is aggravated by smoking [16–19]. It has also been shown that pulmonary fiber content is directly correlated to survival [20]. Damage to the respiratory system, such as chronic obstructive pulmonary disease or fibrosis increases the risk of cancer [17], linked or not to other exposures and regardless of smoking [19]. Most occupational exposures contribute to worsening the co-morbidities of the patient, in addition to their specific risks. The other types of occupational exposure in our cohort of patients are nearly all implicated, either in processes that damage the respiratory system, or in tumoral pathologies affecting the lungs or other organs [19–22]. They are associated with lower socio-economic categories. This study suffers from some limitations. It was a “real-life” observational study, the selection of chemotherapy was determined according to the patient’s condition, his comorbidities and tolerance to L1 and the availability of the different drugs used varied over the course of this long study. Our aim was not to compare the efficacy of the different treatments as patients were not randomized and we did not evaluate the toxicity as most of the treatment regimens are widely used and have already been evaluated. The increasingly low response to successive lines of chemotherapy is well documented in SCLC [1,5,12]. Massarelli et al. have shown that chemotherapy for NSCLC is not beneficial for patients beyond L3 [23]. In our study a complete response at L3 was the exception, with partial responses obtained for less than 20% of treated patients. Survival sharply decreased after L2. This last point raises the ethical issue of the prescription of L3 and L4. One approach to this problem is to take into account the response to

L1. Our study clearly shows that “Sensitive” patients may benefit from L2 and L3 (depending on their condition). The indication for L4 is not obvious and should concern only a few highly selected patients. For “Resistant” patients the question as to whether to continue treatment begins from L3 and at L2 for “Refractory” patients. Their respective median survival of 6.4 months after L1 and 3.8 months after L2 corroborates this observation and raises the problem of chemotherapy-induced toxicity in addition to the disease itself. The reintroduction of etoposide–platinum in L2 for “Sensitive” patients, allowed us to obtain nearly 65% of ORR in agreement with other published studies [1,5,6,12]. Taxanes are a possible alternative with 25% of ORR in our study, as described by others [1,5,24–26]. Topotecan gave disappointing results, with responses far less than the reported benefits of about 30% ORR, considered preferable to palliative care in some published studies [1,5,6]. Possibly, topotecan, prescribed intravenously by our team, was used under non-optimal conditions [6,8,27]. However, our results may bring into question the legitimacy of the use of this expensive drug in patients with SCLC relapse. An examination of various criteria of poor prognosis (ED, high PS, high Charlson score, worsened by EPS) [28,29] predicts a high risk of belonging to a “Refractory” group. For those patients, optimization of their condition prior to chemotherapy and the best supportive care before, during and after the first cycles could improve their survival. Temel et al. have shown that the use of optimal palliative care improves the quality of life and survival in patients with incurable diseases [30,31]. Subsequent lines of treatment for “Refractory” patients do not appear to improve their condition. Consequently the choice of palliative care is an acceptable alternative. Lastly, in most cases the treatment of SCLC remains a therapeutic failure. In our cohort of 300 patients only 17 (5.6%) achieved complete response after relapse of first line treatment and 5 years survival was only 8.7%. The major prognostic factor is limited disease, reflecting restricted tumor bulk, as illustrated by our “Long-term survivors” (95.2% LD). These patients had “ultra-limited” disease, with 42.8% undergoing surgery. Patients with limited disease can be treated with a combination of chemotherapy and radiotherapy. The latter has been shown to have a positive impact on survival in case series [32–34]. In contrast only 23.4% of “Refractory” patients had LD.

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

Immunohistological studies of these tumors revealed a loss of ability to apoptose, high invasive capacity, the ability to grow in conditions of reduced oxygenation and little neo-angiogenesis [35–39]. Chemotherapy is countered by these aptitudes, and chemo-resistance mechanisms develop more readily when the tumor bulk is large [40]. Future therapies may come from an immunohistochemical approach which could restore the activity of tumor suppressor genes and hence apoptosis pathways. In the meantime, the pejorative prognosis of this type of cancer, which is strongly linked to tobacco consumption, should prompt the reinforcement of preventive measures against smoking and early screening of high-risk groups (particularly for women). Likewise progress needs to be made in the prevention of occupational exposure. 5. Conclusion Despite the fact that this study was a single centre retrospective observational study, we confirm the positive prognostic importance of limited forms of SCLC and the impact on the sensitivity to treatments of classical histological forms. Factors influencing survival include PS, co-morbidities and occupational exposure. The “sensitivity” of the patient, determined by the response to L1, indicates that it is probably futile to treat a “Sensitive” patient after L3, a “Resistant” patient after L2; and no treatment other than investigational new drugs should be proposed to “Refractory” patients. Study approval This is a retrospective long-term observational study and many patients had died. The anonymized database of the Multidisciplinary Thoracic Oncology Group respects the requirements of the French agency for computerized information (CNIL). Source of funding This study received no specific funding. Author contributions HNM, LV, MC, AV, CB, JFT, and DMS contributed to project design. HNM, MC, ACT, LS, DMS, and CB contributed to recruitment, clinical assessment and follow-up of the patients. HNM contributed to data collection. VB contributed to data on occupational exposure. PG, LV, AV, HNM, and JFT contributed to statistical analyzes. EB contributed to histo-pathological diagnosis. HNM, PG, VB, MC, DMS, and EB contributed to writing the manuscript. The final version of the manuscript has been approved by all coauthors.

Conflict of interest statement None of the authors declare any financial or moral conflict of interest with respect to this study.

Acknowledgements We thank Dr Alison Foote (employed by the Grenoble Clinical Research Center) for the translation and substantial critical editing of the manuscript and Dr Dan Veale for his comments.

119

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.lungcan. 2012.06.006.

References [1] Tiseo M, Ardizzoni A. Current status of second-line treatment and novel therapies for small cell lung cancer. J Thorac Oncol 2007;2:764–72. [2] Lee P. Lung cancer: recent advances. Ann Acad Med Singapore 2010;39(11):819–21. [3] Lally BE, Geiger AM, Urbanic JJ, Butler JM, Wentworth S, Perry MC, et al. Trends in the outcomes for patients with limited stage small cell lung cancer: an analysis of the Surveillance, Epidemiology, and End Results database. Lung Cancer 2009;64(2):226–31. [4] El Maalouf G, Rodier JM, Faivre S, Raymond E. Could we expect to improve survival in small cell lung cancer? Lung Cancer 2007;57(Suppl. 2):S30–4. [5] Stupp R, Monnerat C, Turrisi 3rd AT, Perry MC, Leyvraz S. Small cell lung cancer: state of the art and future perspectives. Lung Cancer 2004;45(1):105–17. [6] Riemsma R, Simons JP, Bashir Z, Gooch CL, Kleijnen J. Systematic review of topotecan (Hycamtin) in relapsed small cell lung cancer. BMC Cancer 2010;10:436. [7] Pirker R, Berzinec P, Brincat S, Kasan P, Ostoros G, Pesek M, et al. Therapy of small cell lung cancer with emphasis on oral topotecan. Lung Cancer 2010;70:7–13. [8] Govindan R, Page N, Morgensztern D, Read W, Tierney R, Vlahiotis A, et al. Changing epidemiology of small-cell lung cancer in the United States over the last 30 years: analysis of the surveillance, epidemiologic, and end results database. J Clin Oncol 2006;24(28):4539–44. [9] Fried L, Bernardini J, Piraino B. Charlson comorbidity index as a predictor of outcomes in incident peritoneal dialysis patients. Am J Kidney Dis 2001;37(2):337–42. [10] French national insurance system for salaried workers (CNAM-TS), European Occupational Diseases Statistics Classification of the causal agents of the occupational diseases (EODS causal agent codes) http://circa.europa.eu/ Public/irc/dsis/hasaw/library?l=/occupational statistics/working paper 18/ EN 1.0 . [11] Duffaud F, Therasse P. Nouvelles recommandations pour l’évaluation de la réponse tumorale dans les tumeurs solides. Bull du cancer 2000;87(12):881–6. [12] Garassino MC, Torri V, Michetti G, Lo Dico M, La Verde N, Aglione S, et al. Outcomes of small-cell lung cancer patients treated with secondline chemotherapy: a multi-institutional retrospective analysis. Lung Cancer 2011;72(3):378–83. [13] Nagy-Mignotte H, Guillem P, Vesin A, Toffart AC, Colonna M, Bonneterre V, et al. Primary lung adenocarcinoma: characteristics by gender and smoking habit. Eur Respir J 2011;38:1412–9. [14] Quoix E, Mennecier B. Cancer du poumon et pollution. Rev Pneumol Clin 2003;59:187–95. [15] Quoix E. Particularités du cancer bronchique chez la femme. Revue de pneumologie clinique: le poumon et le cœur 2004; 60 N◦ 5 cahier 3:27–30. [16] Levin JL, O’Sullivan MF, Corn CJ, Williams MG, Dodson RF. Asbestosis and small cell lung cancer in a clutch refabricator. Occup Environ Med 1999;56(9):602–5. [17] Littman AJ, Thornquist MD, White E, Jackson LA, Goodman GE, Vaughan TL. Prior lung disease and risk of lung cancer in a large prospective study. Cancer Causes Control 2004;15(8):819–27. [18] Mollo F, Bellis D, Andreozzi A, Burlo P, Bo P, Cravello M. Lung cancers attributable to asbestos on a pathological basis. G Ital Med Lav Ergon 1997;19(1):36–8. [19] Paris C. Facteurs de risques professionnels et extraprofessionnels des cancers bronchiques. In: Thiberville L, editor. Oncologie thoracique: synthèse des connaissances et perspectives. Paris: Editions Margaux Orange; 2002. p. 35–43. [20] Kayser K, Becker C, Seeberg N, Gabius HJ. Quantitation of asbestos and asbestoslike fibers in human lung tissue by hot and wet ashing, and the significance of their presence for survival of lung carcinoma and mesothelioma patients. Lung Cancer 1999;24(2):89–98. [21] MacArthur AC, Le ND, Fang R, Band PR. Identification of occupational cancer risk in British Columbia: a population-based case–control study of 2,998 lung cancers by histopathological subtype. Am J Ind Med 2009;52(3): 221–32. [22] National Toxicology Program. Final report on carcinogens background document for formaldehyde. Rep Carcinog Backgr Doc 2010;(10-5981):i-512. [23] Massarelli E, Andre F, Liu DD, Lee JJ, Wolf M, Fandi A, et al. A retrospective analysis of the outcome of patients who have received two prior chemotherapy regimens including platinum and docetaxel for recurrent non-small-cell lung cancer. Lung Cancer 2003;39(1):55–61. [24] Oze I, Hotta K, Kiura K, Ochi N, Takigawa N, Fujiwara Y, et al. Twenty-seven years of phase III trials for patients with extensive disease small-cell lung cancer: disappointing results. PLoS One 2009;4(11):e7835. [25] Isobe K, Kobayashi K, Kosaihira S, Kurimoto F, Sakai H, Uchida Y, et al. Phase II study of nimustine hydrochloride (ACNU) plus paclitaxel for refractory small cell lung cancer. Lung Cancer 2009;66(December (3)):350–4.

120

H. Nagy-Mignotte et al. / Lung Cancer 78 (2012) 112–120

[26] Park S, Ahn MJ, Ahn JS, Lee J, Hong YS, Park BB, et al. Combination chemotherapy with paclitaxel and ifosfamide as the third-line regimen in patients with heavily pretreated small cell lung cancer. Lung Cancer 2007;58(October (1)):116–22. [27] O’Brien ME, Ciuleanu TE, Tsekov H, Shparyk Y, Cuceviá B, Juhasz G, et al. Phase III trial comparing supportive care alone with supportive care with oral topotecan in patients with relapsed small-cell lung cancer. J Clin Oncol 2006;24(34):5441–7. [28] Mennecier B, Moreau L, Goichot B, Pauli G, Quoix E. Paraneoplastic Cushing’s syndrome and small cell bronchial carcinoma. Rev Pneumol Clin 1999;55(2):77–80. [29] Ilias I, Torpy DJ, Pacak K, Mullen N, Wesley RA, Nieman LK. Cushing’s syndrome due to ectopic corticotropin secretion: twenty years’ experience at the National Institutes of Health. J Clin Endocrinol Metab 2005;90(8):4955–62. [30] Temel JS, Greer JA, Muzikansky A, Gallagher ER, Admane S, Jackson VA, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med 2010;363(8):733–42. [31] El-Jawahri A, Greer JA, Temel JS. Does palliative care improve outcomes for patients with incurable illness? A review of the evidence. J Support Oncol 2011;9(3):87–94. [32] Lebeau B. Traitement des formes localisées des cancers à petites cellules. In: Thiberville L, editor. Oncologie thoracique: synthèse des connaissances et perspectives. Paris: Editions Margaux Orange; 2002. p. 355–60. [33] Arriagada R, Pignon JP, Ihde DC, Johnson DH, Perry MC, Souhami RL, et al. Effect of thoracic radiotherapy on mortality in limited small cell lung cancer.

[34]

[35]

[36]

[37]

[38]

[39] [40]

A meta-analysis of 13 randomized trials among 2,140 patients. Anticancer Res 1994;14(1B):333–5. Pignon JP, Arriagada R, Ihde DC, Johnson DH, Perry MC, Souhami RL, et al. A meta-analysis of thoracic radiotherapy for small-cell lung cancer. N Engl J Med 1992;327(23):1618–24. Brambilla E, Negoescu A, Gazzeri S, Lantuejoul S, Moro D, Brambilla C, et al. Apoptosis-related factors p53, Bcl2, and Bax in neuroendocrine lung tumors. Am J Pathol 1996;149(6):1941–52. Bolon I, Gouyer V, Devouassoux M, Vandenbunder B, Wernert N, Moro D, et al. Expression of c-ets-1, collagenase 1, and urokinase-type plasminogen activator genes in lung carcinomas. Am J Pathol 1995;147(5): 1298–310. Gazzeri S, Della Valle V, Chaussade L, Brambilla C, Larsen CJ, Brambilla E. The human p19ARF protein encoded by the beta transcript of the p16INK4a gene is frequently lost in small cell lung cancer. Cancer Res 1998;58(17): 3926–31. Brambilla E, Constantin B, Drabkin H, Roche J. Semaphorin SEMA3F localization in malignant human lung and cell lines: a suggested role in cell adhesion and cell migration. Am J Pathol 2000;156(3):939–50. Brambilla E, Gazdar A. Pathogenesis of lung cancer signalling pathways: roadmap for therapies. Eur Respir J 2009;33(6):1485–9. Stewart DJ, Johnson C, Lopez A, Glisson B, Rhee JM, Bekele BN. Extensive disease small cell lung cancer dose–response relationships: implications for resistance mechanisms. J Thorac Oncol 2010;5(11):1826–34.