- Email: [email protected]
Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer
G Model CANEP 880 No. of Pages 6
Cancer Epidemiology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Cancer Epidemiology The International Journal of Cancer Epidemiology, Detection, and Prevention journal homepage: www.cancerepidemiology.net
Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer Ning Lia,b,1, Miao Xub,1, Mu-Yan Caic,1, Feng Zhoud, Chao-Feng Lie, Bao-Xiao Wanga , Wei Oua , Si-Yu Wanga,* a Department of Thoracic Surgery, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China b Department of Experimental Research, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China c Department of Pathology, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China d Department of Medical Affairs, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China e Department of Information Technology, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 5 March 2015 Received in revised form 11 June 2015 Accepted 28 June 2015 Available online xxx
Background: Bilirubin levels have been associated with risk of several malignancies. The association between pretreatment serum bilirubin levels and survival of curatively resected non-small-cell lung cancer (NSCLC) is unclear. Methods: This analysis was performed retrospectively in a cohort of 1617 consecutive patients with bilirubin levels within the range considered normal, who received curative resection for NSCLC. The receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cut-off points. The significance of pretreatment serum total bilirubin (TBIL), direct bilirubin (DBIL), and indirect bilirubin (IBIL) levels in the prognosis of patients with curatively resected NSCLC was investigated. Results: The cutoff points of serum TBIL, DBIL and IBIL were 9.50 mmol/L, 3.45 mmol/L and 6.95 mmol/L, respectively. High TBIL was observed in 65.2% of entire patient population, high DBIL 50%, and high IBIL 56.8%. The high-TBIL group had significantly lengthened overall survival (OS; hazard ratio [HR], 0.73; 95% confidence interval [CI] 0.63–0.84; P < 0.001), disease-free survival (DFS; HR, 0.72; 95% CI 0.64–0.82; P < 0.001) and distant metastasis-free survival (DMFS; HR, 0.74; 95% CI 0.60–0.91; P = 0.004). Similarly, high-DBIL and high-IBIL levels were associated with longer OS, DFS, and DMFS with significant differences. In multivariable analysis, IBIL level was identified as an independent significant prognostic factor. Conclusions: Moderately elevated pretreatment bilirubin levels are associated with longer OS, DFS, and DMFS for patients with curatively resected NSCLC. IBIL is an independent prognostic factor in curative resected NSCLC. ã 2015 Elsevier Ltd. All rights reserved.
Keywords: Bilirubin Lung cancer Prognosis Survival
1. Introduction Annually, among more than 1 million new lung cancer cases worldwide [1], non-small-cell lung cancer (NSCLC) constitutes 80– 85% [2]. Despite improvements in chemotherapy and target
* Corresponding author at: Department of Thoracic Surgery, Sun Yat-sen University Cancer Center, 651 Dongfeng Road East, Guangzhou 510060, China. Fax: +86 20 873434393. E-mail address: [email protected] (S.-Y. Wang). 1 These authors contributed equally to this article.
therapy, the 5-year survival of all stages of lung cancer remains under 15%, with less than 7% sufferers living more than 10 years [3]. Diagnosis and treatment of early-staged NSCLC are valued since improving the prognosis of locally advanced and metastatic NSCLC is a tough impediment. Identifying novel prognostic factors of this usually fatal disease is important for early diagnosis, prognosis assessment and more appropriate treatment. Bilirubin, the major end product of heme degradation, is cleared from the liver, where it is conjugated to form water-soluble direct bilirubin that is secreted into bile [4]. Unconjugated bilirubin, namely indirect bilirubin, accounts for more than 80% of total
http://dx.doi.org/10.1016/j.canep.2015.06.007 1877-7821/ ã 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
2
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
bilirubin. Owning to its potential properties of antioxidation, antiinflammation and anticancer [5], bilirubin has no longer been regarded only as a marker of hepatobiliary and hematological disorders. In recent years, the relationship between serum bilirubin and cancer risk has drawn much interest. The inverse association between bilirubin and cancer risk has been suggested in lung cancer [6,7], breast cancer [8] and colorectal cancer [9]. Elevated bilirubin is also associated with longer survival in nonmetastatic breast cancer [10]. In vitro studies indicated that bilirubin induces apoptosis in colon cancer cells [11], exhibits an antiproliferative effect on human adenocarcinoma cells [12], and inhibits cell growth in various tumor cells [13]. However, the relationship between bilirubin levels and survival of lung cancer has not been studied previously. Bilirubin levels may be a useful marker for cancer prognosis. We hypothesized that lung cancer patients with moderately elevated pretreatment serum bilirubin levels would have a longer survival, given the cytoprotective effects of bilirubin. Thus, we conducted this study to investigate whether serum bilirubin levels at the time of diagnosis are associated with the survival of NSCLC patients who received curative resection as the primary treatment. 2. Patients and methods This analysis was performed retrospectively in a cohort of 1617 consecutive patients that received curative resection for stage IIIIA NSCLC at Sun Yat-sen University Cancer Center (SYSUCC) from January 1, 2005 and December 31, 2009. All patients were restaged by the seventh international classification system for lung cancer [14]. The medical Ethics Committee and Clinical Trial Review Committee of SYSUCC approved this study. Information was collected from electronic and papery individual patient records, and information about survival was obtained from the follow-up registry of SYSUCC. The information collected included age, sex, smoking history, histology, pathological stage, time of diagnosis and relapse/metastasis, and pretreatment serum total bilirubin (TBIL), direct bilirubin (DBIL) and indirect bilirubin (IBIL) levels. All biochemical analyses were performed in the Department of Clinical Laboratory, SYSUCC. Potentially eligible patients had to have curatively resected pathologically confirmed stage I-IIIA NSCLC without previous therapy other than surgical resection and neoadjuvant chemotherapy. Exclusion criteria included evidence of hepatobiliary or hemolytic disease, previous malignancies, perioperative death, and insufficient data of survival or pretreatment hematology test data. We also excluded patients with total bilirubin levels lower than 3 mmol/L for both sexes, and higher than 40 mmol/L for men and 30 mmol/L for women. To avoid the potential impact of surgery or chemotherapy on bilirubin level, we ensured that each eligible patient had a pretreatment biochemical test.
the time of locoregional relapse and distant metastasis, respectively. Kaplan–Meier curves were drawn for these endpoints, and differences were compared by the log-rank test. The Cox proportional hazards model was used to perform multivariate analyses. Variables reaching a significant level of 0.1 in univariate analyses suggested a trend, and were included in multivariate analysis. Two-tailed P values of less than 0.05 were considered as statistically significant. 4. Results 4.1. Patient population A total of 1617 patients with curative resection for primary stage I-IIIA NSCLC were included in this study (Fig. 1). Table 1 shows the baseline characteristics of the study patient population. The TBIL cut-off points for OS, DFS, LRFS, and DMFS were 9.50, 9.55, 14.95, and 9.95 mmol/L, respectively (Supplementary Fig. 1). The TBIL cutoff point of 9.50 mmol/L for OS was selected as the uniform point in the survival analyses. Similarly, DBIL level of 3.45 mmol/L and IBIL level of 6.95 mmol/L were selected as the cut-off points for survival analyses (Supplementary Figs. 2 and 3). High TBIL was observed in 65.2% of entire patient population, high DBIL 50%, and high IBIL 56.8%. All of the high-TBIL group, the high-DBIL group and the high-IBIL group had more older patients. There were more never smokers, more adenocarcinomas and more advanced tumors in the high-TBIL group and the high-IBIL group. Male patients were more prevalent in the high-DBIL group than the low-DBIL group (Table 1). The distribution of serum TBIL was right-skewed (Fig. 2). Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. 5. Survival After a median follow-up of 80.9 (95% CI, 78.8–83.0) months, 360 patients developed locoregional relapse, 371 developed distant metastases, and 815 died for the entire cohort. The highTBIL group had significantly lengthened OS compared with the low-TBIL group, with a median OS of 78.0 months versus 56.9 months (HR, 0.73; 95% CI 0.63–0.84; P < 0.001; Supplementary Fig. 4A). The 5-year OS rates were 58.7% and 48.2%, respectively, among the patients with high TBIL and those with low TBIL. Also, DFS (HR, 0.72; 95% CI 0.64–0.82; P < 0.001; Supplementary Fig. 4B) and DMFS (HR, 0.74; 95% CI 0.60–0.91; P = 0.004; Supplementary Fig. 4D) were significantly longer among those in the high-TBIL group. The 5-year DFS and DMFS were, respectively, 47.3% and 76.5% for patients in the high-TBIL group, and 36.3% and 69.9% for patients in the low-TBIL group. No significant difference was
3. Statistical analysis The receiver operating characteristic (ROC) curve analysis was performed to evaluate the ability of bilirubin (TBIL, DBIL and IBIL) to predict for long-term outcomes and to determine the optimal cut-off points. The cut-off points with highest sum of sensitivity and specificity dichotomized the entire cohort. Chi-square test was used to compare categorical variables. The following end points were evaluated: overall survival (OS), defined as the interval from the time of being diagnosed to the time of death from any cause; disease-free survival (DFS), defined as the interval from the time of being diagnosed to the time of disease recurrence/metastasis or death from any cause; locoregional relapse-free survival (LRFS) and distant metastasis-free survival (DMFS), defined as the interval from the time of being diagnosed to
Fig. 1. Flow diagram.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
3
Table 1 Basic characteristics by TBIL, DBIL and IBIL blood groupa . TBIL group
DBIL group
Characteristics
High TBIL
Low TBIL
Total Age <55 year 55–64 year >64 year Gender Male Female Smoking status Never Ever Pathology Adenocarcinoma Squamous cell carcinoma Adenosquamous carcinoma Large cell carcinoma Other pTNM stage Ia Ib IIa IIb IIIa pT stage 1 2 3 4 pN stage 0 1 2 Surgery type Lobectomy Pneumonectomy Other Chemotherapyc Neoadjuvant Adjuvant No
1055
562
b
P
IBIL group
High DBIL
Low DBIL
809
808
<0.01 335 (31.8) 368 (34.9) 352 (33.4)
200 (35.6) 228 (40.6) 134 (23.8)
764 (72.4) 291 (27.6)
405 (72.1) 157 (27.9)
463 (43.9) 592 (56.1)
189 (33.6) 373 (66.4)
628 (59.5) 341 (32.3) 59 (5.6) 15 (1.4) 12 (1.1)
270 (48.0) 238 (42.3) 35 (6.2) 13 (2.3) 6 (1.1)
154 (14.6) 376 (35.6) 27 (2.6) 156 (14.8) 342 (32.4)
50 (8.9) 199 (35.4) 13 (2.3) 80 (14.2) 220 (39.1)
203 (19.2) 663 (62.8) 113 (10.7) 76 (7.2)
76 (13.5) 373 (66.4) 66 (11.7) 47 (8.4)
616 (58.4) 142 (13.5) 297 (28.2)
299 (53.2) 74 (13.2) 189 (33.6)
855 (81.0) 157 (14.9) 43 (4.1)
452 (80.4) 89 (15.8) 21 (3.7)
63 (5.9) 693 (65.7) 329 (31.2)
28 (5.2) 375 (66.7) 172 (30.6)
b
P
High IBIL
Low IBIL
918
699
296 (32.2) 320 (34.9) 302 (32.9)
239 (34.2) 276 (39.5) 184 (26.3)
656 (71.5) 262 (28.5)
513 (73.4) 186 (26.6)
421 (45.9) 497 (54.1)
231 (33.0) 468 (67.0)
569 (62.0) 275 (30.0) 51 (5.6) 14 (1.5) 9 (1.0)
329 (47.1) 304 (43.5) 43 (6.2) 14 (2.0) 9 (1.3)
136 (14.8) 322 (35.1) 25 (2.7) 131 (14.3) 304 (33.1)
68 (9.7) 253 (36.2) 15 (2.1) 105 (15.0) 258 (36.9)
180 (19.6) 576 (62.7) 94 (10.2) 68 (7.4)
99 (14.2) 460 (65.8) 85 (12.2) 55 (7.9)
531 (57.8) 123 (13.4) 264 (28.8)
384 (54.9) 93 (13.3) 222 (31.8)
733 (79.8) 147 (16.0) 38 (4.1)
574 (82.1) 99 (14.2) 26 (3.7)
52 (3.1) 615 (64.0) 278 (30.3)
39 (2.8) 453 (62.3) 223 (31.9)
<0.01 251 (31.0) 276 (34.1) 282 (34.9)
284 (35.1) 320 (39.6) 204 (25.2)
640 (79.1) 169 (20.9)
529 (65.5) 279 (34.5)
321 (39.7) 488 (60.3)
331 (41.0) 477 (59.0)
460 (56.9) 282 (34.9) 41 (5.1) 16 (2.0) 10 (1.2)
438 (54.2) 297(36.8) 53 (6.6) 12 (1.5) 8 (1.0)
113 (14.0) 285 (35.2) 19 (2.3) 121 (15.0) 271 (33.5)
91 (11.3) 290 (35.9) 21 (2.6) 115 (14.2) 291 (36.0)
150 (18.5) 499 (61.7) 94 (11.6) 66 (8.2)
129 (16.0) 537 (66.5) 85 (10.5) 57 (7.1)
475 (58.7) 109 (13.5) 225 (27.8)
440 (54.5) 107 (13.2) 261 (32.3)
652 (80.7) 127 (15.3) 30 (4.0)
655 (80.9) 119 (15.8) 34 (3.3)
45 (2.2) 536 (62.0) 248 (30.7)
47 (3.1) 531 (64.0) 253 (31.3)
0.04
<0.01
0.88
<0.01
0.39
<0.01
0.60
<0.01
<0.01
0.52
<0.01
0.19
0.01
0.02
0.92
0.03
0.01
0.05
0.84
0.20
0.77
0.70
Pb
0.52
0.94
0.74
Abbreviations: TBIL: total bilirubin; DBIL: direct bilirubin; IBIL: indirect bilirubin; pTNM stage: pathological tumour-node-metastasis stage. a Some percentages may not sum to 100 due to rounding. b x2 test. c Some patients received both neoadjuvant and adjuvant chemotherapy.
Fig. 2. The distribution of serum total bilirubin (TBIL) level.
observed for LRFS between the high-TBIL group and the low-TBIL group (HR, 0.88; 95% CI 0.71–1.09; P = 0.236; Supplementary Fig. 4C). The rates of 5-year LRFS were 75.2% in the high-TBIL group and 73.4% in the low-TBIL group.
Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. Similarly, the high-DBIL group had significantly longer OS (HR, 0.85; 95% CI 0.74–0.98; P = 0.023; Supplementary Fig. 5A), DFS (HR, 0.82; 95% CI 0.72–0.92; P = 0.001; Supplementary Fig. 5B), and DMFS (HR, 0.76; 95% CI 0.62–0.93; P = 0.007; Supplementary Fig. 5D) than the low-DBIL group. The difference in LRFS between the high-DBIL group and the low-DBIL group was not significant (HR, 0.94; 95% CI 0.77–1.16; P = 0.561; Supplementary Fig. 5C). The 5-year OS, DFS, LRFS and DMFS were, respectively, 57.5%, 47.3%, 75.2% and 77.5% for patients in the high-DBIL group, and 52.7%, 39.5%, 73.9% and 71.1% for patients in the low-DBIL group. Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. The high-IBIL group was associated with longer OS (HR, 0.75; 95% CI 0.65-0.86; P < 0.001; Fig. 3A), DFS (HR, 0.77; 95% CI 0.68– 0.87; P < 0.001; Fig. 3B), and DMFS (HR, 0.73; 95% CI 0.60–0.90; P = 0.002; Fig. 3D) with significant difference. LRFS did not differ significantly between high-IBIL group and low-IBIL group (HR, 0.96; 95% CI 0.78–1.19; P = 0.716; Fig. 3C). The 5-year OS, DFS, LRFS and DMFS were, respectively, 59.2%, 47.4%, 74.9% and 77.0% for patients in the high-IBIL group, and 49.7%, 38.2%, 74.1% and 70.5% for patients in the low-IBIL group.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
4
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
Fig. 3. Kaplan–Meier curves for (a) overall survival, (b) disease-free survival, (c) locoregional relapse-free survival, and (d) distant metastasis-free survival stratified by indirect bilirubin (IBIL) level.
To validate the survival results, we also performed Kaplan– Meier analysis of the entire cohort stratified by the median serum TBIL value of 11.0 mmol/L. The results demonstrated that patients with a level of TBIL 11.0 mmol/L had lengthened OS, DFS, and DMFS (Supplementary Fig. 6). Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. 6. Univariate analysis and multivariate analysis OS was further analyzed in univariate and multivariate analyses. In the univariate analysis, gender, smoking status, pTNM stage, pT stage, pN stage, surgery type, chemotherapy, TBIL level, DBIL level, and IBIL level were identified as significant prognostic factors. When these variables were further analyzed in the multivariate analysis, we found that smoking status (P = 0.019), pTNM stage (P = 0.003), pT stage (P < 0.001), pN stage (P = 0.002), chemotherapy (P < 0.001), and IBIL level (P = 0.004) have significant HRs, indicating that they are independent, significant predictors for OS (Table 2). High serum IBIL level was a positive prognostic factor. 7. Discussion In this study we investigated the prognostic value of serum bilirubin levels in NSCLC, and to our knowledge, report for the first time that moderately elevated bilirubin levels at diagnosis represent positive prognostic factors for survival in completely resected NSCLC. Patients with relatively high TBIL, DBIL or IBIL level had significantly longer OS, DFS and DMFS than patients with relatively low TBIL, DBIL or IBIL levels. IBIL level was identified as a significant prognostic predictor for OS, independent of any other clinicopathological features of NSCLC, including pTNM stage.
Apart from the association between bilirubin levels and cancer risk and mortality, earlier studies did not demonstrate clear evidence regarding the relationship of bilirubin levels and survival of NSCLC. Consistent with previous studies, our study suggested that high bilirubin levels within the relatively normal range were protective factors against cancer, owing to bilirubin’s power of antioxidation, anti-inflammation and antiproliferation, given the critical role of oxidative stress as a contributing factor to carcinogenesis [15]. The prospective study in Belgium by Temme et al. suggested that high total bilirubin was significantly associated with decreased cancer mortality for men, and the association was in the same direction for women [16]. The cohort study by Horfall et al. investigated the relation between bilirubin levels and risk of respiratory disease (including lung cancer) and death among 504,206 adults from the Health Improvement Network database in UK. Their results showed that relatively elevated bilirubin within 40 mmol/L for men and 30 mmol/L for women was correlative with decreased risk of lung cancer and allcause mortality [6]. The results of our study combined with those suggest bilirubin can not only protect against the development of malignancies, but also prolong the survival of cancer patients. Several studies have shown the inverse association between smoking and bilirubin levels [6,7,17,18]. In our studies, there were more never smokers in the high-TBIL and high-IBIL groups than the low-TBIL and low-IBIL groups, respectively. Our results are in consistence with prior results. Smoking is a well-established risk factor for lung cancer [19–21] and evidences suggested that it is a negative prognostic factor for the survival of NSCLC [22,23]. This can partly explain our survival results. Of note, the survival benefit of bilirubin was specific to IBIL in our study, because in multivariate analysis, among TBIL, DBIL and IBIL levels, only IBIL level was identified as an independent, significant predictor for survival. This observation is consistent with the report by Keshavan et al. which
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
5
Table 2 Univariate and multivariate analysis for overall survival. Univariate Hazard ratio (95% CI) Age <55 year 55–64 year >64 year Gender Male Female Smoking status Never Ever Pathology Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Large cell carcinoma Other pTNM stage I II III pT stage 1 2 3 4 pN stage 0 1 2 Surgery type Lobectomy and other Pneumonectomy Chemotherapy Yes No TBIL level High Low DBIL level High Low IBIL level High Low
Multivariate P
Hazard ratio (95% CI)
P –
0.240 – – –
Reference 1.02 (0.86–1.20) 1.15 (0.97–1.36) 0.048 Reference 1.17 (1.00–1.37)
0.486 Reference 1.09 (0.87–1.37)
0.017 Reference 1.19 (1.03–1.37)
0.019 Reference 1.19 (1.03–1.37) –
0.332 – – – – –
Reference 0.98 (0.85–1.14) 1.31 (0.98–1.75) 1.05 (0.60–1.84) 0.74 (0.35–1.57) <0.001 Reference 2.43 (1.99–2.96) 3.65 (3.12-4.28)
0.003 Reference 1.79 (1.28–2.50) 1.80 (1.15–2.81)
<0.001 Reference 2.05 (1.63–2.57) 2.95 (2.22–3.92) 5.08 (3.78–6.81)
<0.001 Reference 1.74 (1.38–2.20) 1.83 (1.32–2.55) 3.06 (2.15–4.37)
<0.001 Reference 2.31 (1.89–2.83) 3.13 (2.69–3.64)
0.002 Reference 1.56 (1.12–2.18) 2.07 (1.37–3.11)
<0.001 Reference 1.79 (1.40–2.28)
0.374 Reference 1.13 (0.88–1.46)
<0.001 Reference 1.37 (1.20–1.58)
<0.001 Reference 1.41 (1.20–1.65)
<0.001 Reference 1.37 (1.19–1.58)
0.220 Reference 1.18 (0.90–1.54)
0.023 Reference 1.17 (1.02–1.35)
0.680 Reference 1.00 (0.83–1.20)
<0.001 Reference 1.34 (1.17–1.54)
0.004 Reference 1.22 (1.07–1.41)
Abbreviations: CI: confidence interval; pTNM stage: pathological tumor-node-metastasis stage; TBIL: total bilirubin; DBIL: direct bilirubin; IBIL: indirect bilirubin.
revealed that bilirubin induces apoptosis in colon adenocarcinoma cell lines, and this effect appears to be specific to the unconjugated bilirubin, as ditaurobilirubin and biliverdin did not show this effect [11]. Reasons for this may largely relate to the unbound nature of IBIL, which can freely diffuse across cell membranes. The study by Zucker et al. suggested that unconjugated bilirubin can readily cross cellular membranes [24]. The study by Weisiger et al. indicated that if the serum unconjugated bilirubin level is 10 mmol/L, the level of unbound (free) bilirubin will be 7.4 nmol/L [25]. We speculate that unbound bilirubin may be the main effective portion of total bilirubin to act as a factor of antioxidation, anti-inflammation and antiproliferation. This study was retrospective in nature and therefore had some limitations. First, like other retrospective studies, we cannot completely rule out the impact of selection bias. The patients were included over a relatively long period (2005–2009), and many patients were initially staged according the sixth international classification system for lung cancer, whereas they were restaged according to the seventh classification system during the analysis. Second, Cantonese constitute most of our study population. Our results may not apply to other ethnic populations. Third, we were unable to collect other sufficient information that appears to be
significant prognostic factors for NSCLC, such as epidermal growth factor receptor (EGFR) mutational status, EGFR copy number and anaplastic lymphoma kinase (ALK) rearrangement status. 8. Conclusion We have demonstrated in a retrospective analysis that NSCLC patients with moderately elevated pretreatment serum bilirubin levels who are treated with curative resection have longer OS, DFS, and DMFS; IBIL is an independent, prognostic predictor for NSCLC patients with curative resection. Further well-designed studies involving diverse ethnic populations are warranted to confirm the prognostic value of serum bilirubin levels in NSCLC, as well as taking into account other potential prognostic factors for NSCLC. Conflicts of interest None. Funding None.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
Author contributions LN, XM and WSY were involved in the design of the study. WSY, CMY and XM supervised the study. LN, XM, CMY, ZF, LCF, WBX, OW were involved in data acquisition. LN, XM, CMY and WSY were involved in data analysis and interpretation. All authors were involved in writing the article and approved the final version. References [1] A. Jemal, F. Bray, M.M. Center, J. Ferlay, E. Ward, D. Forman, Global cancer statistics, CA Cancer J. Clin. 61 (2011) 69–90. [2] G. D’Addario, M. Fruh, M. Reck, P. Baumann, W. Klepetko, E. Felip, Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 21 (Suppl. 5) (2010) v116–9. [3] L. Crino, W. Weder, J. van Meerbeeck, E. Felip, Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 21 (Suppl. 5) (2010) v103–15. [4] J.D. Ostrow, J.H. Jandl, R. Schmid, The formation of bilirubin from hemoglobin in vivo, J. Clin. Invest. 41 (1962) 1628–1637. [5] R. Stocker, Y. Yamamoto, A.F. McDonagh, A.N. Glazer, B.N. Ames, Bilirubin is an antioxidant of possible physiological importance, Science 235 (1987) 1043– 1046. [6] L.J. Horsfall, G. Rait, K. Walters, D.M. Swallow, S.P. Pereira, I. Nazareth, et al., Serum bilirubin and risk of respiratory disease and death, JAMA 305 (2011) 691–697. [7] C.P. Wen, F. Zhang, D. Liang, C. Wen, J. Gu, H. Skinner, et al., The ability of bilirubin in identifying smokers with higher risk of lung cancer: a large cohort study in conjunction with global metabolomic profiling, Clin. Cancer Res. (2014) . [8] S. Ching, D. Ingram, R. Hahnel, J. Beilby, E. Rossi, Serum levels of micronutrients, antioxidants and total antioxidant status predict risk of breast cancer in a case control study, J. Nutr. 132 (2002) 303–306. [9] S.D. Zucker, P.S. Horn, K.E. Sherman, Serum bilirubin levels in the U.S. population: gender effect and inverse correlation with colorectal cancer, Hepatology 40 (2004) 827–835. [10] X. Liu, Q.H. Meng, Y. Ye, M. Hildebrandt, J. Gu, X. Wu, Prognostic significance of pretreatment serum levels of albumin, LDH and total bilirubin in patients with non-metastatic breast cancer, Carcinogenesis (2014) . [11] P. Keshavan, S.J. Schwemberger, D.L. Smith, G.F. Babcock, S.D. Zucker, Unconjugated bilirubin induces apoptosis in colon cancer cells by triggering mitochondrial depolarization, Int. J. Cancer 112 (2004) 433–445.
[12] P. Rao, R. Suzuki, S. Mizobuchi, T. Yamaguchi, S. Sasaguri, Bilirubin exhibits a novel anti-cancer effect on human adenocarcinoma, Biochem. Biophys. Res. Commun. 342 (2006) 1279–1283. [13] R. Ollinger, P. Kogler, J. Troppmair, M. Hermann, M. Wurm, A. Drasche, et al., Bilirubin inhibits tumor cell growth via activation of ERK, Cell Cycle 6 (2007) 3078–3085. [14] P. Goldstraw, J. Crowley, K. Chansky, D.J. Giroux, P.A. Groome, R. Rami-Porta, et al., The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours, J. Thorac. Oncol. 2 (2007) 706–714. [15] L.J. Marnett, Oxyradicals and DNA damage, Carcinogenesis 21 (2000) 361–370. [16] E.H. Temme, J. Zhang, E.G. Schouten, H. Kesteloot, Serum bilirubin and 10-year mortality risk in a Belgian population, Cancer Causes Control 12 (2001) 887– 894. [17] K. Frost-Pineda, Q. Liang, J. Liu, L. Rimmer, Y. Jin, S. Feng, et al., Biomarkers of potential harm among adult smokers and nonsmokers in the total exposure study, Nicotine Tob. Res. 13 (2011) 182–193. [18] S.S. O'Malley, R. Wu, S.T. Mayne, P.I. Jatlow, Smoking cessation is followed by increases in serum bilirubin, an endogenous antioxidant associated with lower risk of lung cancer and cardiovascular disease, Nicotine Tob. Res. 16 (2014) 1145–1149. [19] R. Doll, R. Peto, Mortality in relation to smoking: 20 years’ observations on male British doctors, Br. Med. J. 2 (1976) 1525–1536. [20] R. Doll, R. Peto, K. Wheatley, R. Gray, I. Sutherland, Mortality in relation to smoking: 40 years’ observations on male British doctors, BMJ 309 (1994) 901– 911. [21] D.S. Ettinger, D.E. Wood, W. Akerley, L.A. Bazhenova, H. Borghaei, D.R. Camidge, et al., Non-small cell lung cancer, version 1. 2015, J. Natl. Compr. Canc. Netw. 12 (2014) 1738–1761. [22] Y.Y. Janjigian, K. McDonnell, M.G. Kris, R. Shen, C.S. Sima, P.B. Bach, et al., Packyears of cigarette smoking as a prognostic factor in patients with stage IIIB/IV nonsmall cell lung cancer, Cancer 116 (2010) 670–675. [23] T. Kawaguchi, M. Takada, A. Kubo, A. Matsumura, S. Fukai, A. Tamura, et al., Performance status and smoking status are independent favorable prognostic factors for survival in non-small cell lung cancer: a comprehensive analysis of 26,957 patients with NSCLC, J. Thorac. Oncol. 5 (2010) 620–630. [24] S.D. Zucker, W. Goessling, A.G. Hoppin, Unconjugated bilirubin exhibits spontaneous diffusion through model lipid bilayers and native hepatocyte membranes, J. Biol. Chem. 274 (1999) 10852–10862. [25] R.A. Weisiger, J.D. Ostrow, R.K. Koehler, C.C. Webster, P. Mukerjee, L. Pascolo, et al., Affinity of human serum albumin for bilirubin varies with albumin concentration and buffer composition: results of a novel ultrafiltration method, J. Biol. Chem. 276 (2001) 29953–29960.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
Cancer Epidemiology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Cancer Epidemiology The International Journal of Cancer Epidemiology, Detection, and Prevention journal homepage: www.cancerepidemiology.net
Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer Ning Lia,b,1, Miao Xub,1, Mu-Yan Caic,1, Feng Zhoud, Chao-Feng Lie, Bao-Xiao Wanga , Wei Oua , Si-Yu Wanga,* a Department of Thoracic Surgery, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China b Department of Experimental Research, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China c Department of Pathology, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China d Department of Medical Affairs, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China e Department of Information Technology, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, Guangdong 510060, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 5 March 2015 Received in revised form 11 June 2015 Accepted 28 June 2015 Available online xxx
Background: Bilirubin levels have been associated with risk of several malignancies. The association between pretreatment serum bilirubin levels and survival of curatively resected non-small-cell lung cancer (NSCLC) is unclear. Methods: This analysis was performed retrospectively in a cohort of 1617 consecutive patients with bilirubin levels within the range considered normal, who received curative resection for NSCLC. The receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cut-off points. The significance of pretreatment serum total bilirubin (TBIL), direct bilirubin (DBIL), and indirect bilirubin (IBIL) levels in the prognosis of patients with curatively resected NSCLC was investigated. Results: The cutoff points of serum TBIL, DBIL and IBIL were 9.50 mmol/L, 3.45 mmol/L and 6.95 mmol/L, respectively. High TBIL was observed in 65.2% of entire patient population, high DBIL 50%, and high IBIL 56.8%. The high-TBIL group had significantly lengthened overall survival (OS; hazard ratio [HR], 0.73; 95% confidence interval [CI] 0.63–0.84; P < 0.001), disease-free survival (DFS; HR, 0.72; 95% CI 0.64–0.82; P < 0.001) and distant metastasis-free survival (DMFS; HR, 0.74; 95% CI 0.60–0.91; P = 0.004). Similarly, high-DBIL and high-IBIL levels were associated with longer OS, DFS, and DMFS with significant differences. In multivariable analysis, IBIL level was identified as an independent significant prognostic factor. Conclusions: Moderately elevated pretreatment bilirubin levels are associated with longer OS, DFS, and DMFS for patients with curatively resected NSCLC. IBIL is an independent prognostic factor in curative resected NSCLC. ã 2015 Elsevier Ltd. All rights reserved.
Keywords: Bilirubin Lung cancer Prognosis Survival
1. Introduction Annually, among more than 1 million new lung cancer cases worldwide [1], non-small-cell lung cancer (NSCLC) constitutes 80– 85% [2]. Despite improvements in chemotherapy and target
* Corresponding author at: Department of Thoracic Surgery, Sun Yat-sen University Cancer Center, 651 Dongfeng Road East, Guangzhou 510060, China. Fax: +86 20 873434393. E-mail address: [email protected] (S.-Y. Wang). 1 These authors contributed equally to this article.
therapy, the 5-year survival of all stages of lung cancer remains under 15%, with less than 7% sufferers living more than 10 years [3]. Diagnosis and treatment of early-staged NSCLC are valued since improving the prognosis of locally advanced and metastatic NSCLC is a tough impediment. Identifying novel prognostic factors of this usually fatal disease is important for early diagnosis, prognosis assessment and more appropriate treatment. Bilirubin, the major end product of heme degradation, is cleared from the liver, where it is conjugated to form water-soluble direct bilirubin that is secreted into bile [4]. Unconjugated bilirubin, namely indirect bilirubin, accounts for more than 80% of total
http://dx.doi.org/10.1016/j.canep.2015.06.007 1877-7821/ ã 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
2
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
bilirubin. Owning to its potential properties of antioxidation, antiinflammation and anticancer [5], bilirubin has no longer been regarded only as a marker of hepatobiliary and hematological disorders. In recent years, the relationship between serum bilirubin and cancer risk has drawn much interest. The inverse association between bilirubin and cancer risk has been suggested in lung cancer [6,7], breast cancer [8] and colorectal cancer [9]. Elevated bilirubin is also associated with longer survival in nonmetastatic breast cancer [10]. In vitro studies indicated that bilirubin induces apoptosis in colon cancer cells [11], exhibits an antiproliferative effect on human adenocarcinoma cells [12], and inhibits cell growth in various tumor cells [13]. However, the relationship between bilirubin levels and survival of lung cancer has not been studied previously. Bilirubin levels may be a useful marker for cancer prognosis. We hypothesized that lung cancer patients with moderately elevated pretreatment serum bilirubin levels would have a longer survival, given the cytoprotective effects of bilirubin. Thus, we conducted this study to investigate whether serum bilirubin levels at the time of diagnosis are associated with the survival of NSCLC patients who received curative resection as the primary treatment. 2. Patients and methods This analysis was performed retrospectively in a cohort of 1617 consecutive patients that received curative resection for stage IIIIA NSCLC at Sun Yat-sen University Cancer Center (SYSUCC) from January 1, 2005 and December 31, 2009. All patients were restaged by the seventh international classification system for lung cancer [14]. The medical Ethics Committee and Clinical Trial Review Committee of SYSUCC approved this study. Information was collected from electronic and papery individual patient records, and information about survival was obtained from the follow-up registry of SYSUCC. The information collected included age, sex, smoking history, histology, pathological stage, time of diagnosis and relapse/metastasis, and pretreatment serum total bilirubin (TBIL), direct bilirubin (DBIL) and indirect bilirubin (IBIL) levels. All biochemical analyses were performed in the Department of Clinical Laboratory, SYSUCC. Potentially eligible patients had to have curatively resected pathologically confirmed stage I-IIIA NSCLC without previous therapy other than surgical resection and neoadjuvant chemotherapy. Exclusion criteria included evidence of hepatobiliary or hemolytic disease, previous malignancies, perioperative death, and insufficient data of survival or pretreatment hematology test data. We also excluded patients with total bilirubin levels lower than 3 mmol/L for both sexes, and higher than 40 mmol/L for men and 30 mmol/L for women. To avoid the potential impact of surgery or chemotherapy on bilirubin level, we ensured that each eligible patient had a pretreatment biochemical test.
the time of locoregional relapse and distant metastasis, respectively. Kaplan–Meier curves were drawn for these endpoints, and differences were compared by the log-rank test. The Cox proportional hazards model was used to perform multivariate analyses. Variables reaching a significant level of 0.1 in univariate analyses suggested a trend, and were included in multivariate analysis. Two-tailed P values of less than 0.05 were considered as statistically significant. 4. Results 4.1. Patient population A total of 1617 patients with curative resection for primary stage I-IIIA NSCLC were included in this study (Fig. 1). Table 1 shows the baseline characteristics of the study patient population. The TBIL cut-off points for OS, DFS, LRFS, and DMFS were 9.50, 9.55, 14.95, and 9.95 mmol/L, respectively (Supplementary Fig. 1). The TBIL cutoff point of 9.50 mmol/L for OS was selected as the uniform point in the survival analyses. Similarly, DBIL level of 3.45 mmol/L and IBIL level of 6.95 mmol/L were selected as the cut-off points for survival analyses (Supplementary Figs. 2 and 3). High TBIL was observed in 65.2% of entire patient population, high DBIL 50%, and high IBIL 56.8%. All of the high-TBIL group, the high-DBIL group and the high-IBIL group had more older patients. There were more never smokers, more adenocarcinomas and more advanced tumors in the high-TBIL group and the high-IBIL group. Male patients were more prevalent in the high-DBIL group than the low-DBIL group (Table 1). The distribution of serum TBIL was right-skewed (Fig. 2). Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. 5. Survival After a median follow-up of 80.9 (95% CI, 78.8–83.0) months, 360 patients developed locoregional relapse, 371 developed distant metastases, and 815 died for the entire cohort. The highTBIL group had significantly lengthened OS compared with the low-TBIL group, with a median OS of 78.0 months versus 56.9 months (HR, 0.73; 95% CI 0.63–0.84; P < 0.001; Supplementary Fig. 4A). The 5-year OS rates were 58.7% and 48.2%, respectively, among the patients with high TBIL and those with low TBIL. Also, DFS (HR, 0.72; 95% CI 0.64–0.82; P < 0.001; Supplementary Fig. 4B) and DMFS (HR, 0.74; 95% CI 0.60–0.91; P = 0.004; Supplementary Fig. 4D) were significantly longer among those in the high-TBIL group. The 5-year DFS and DMFS were, respectively, 47.3% and 76.5% for patients in the high-TBIL group, and 36.3% and 69.9% for patients in the low-TBIL group. No significant difference was
3. Statistical analysis The receiver operating characteristic (ROC) curve analysis was performed to evaluate the ability of bilirubin (TBIL, DBIL and IBIL) to predict for long-term outcomes and to determine the optimal cut-off points. The cut-off points with highest sum of sensitivity and specificity dichotomized the entire cohort. Chi-square test was used to compare categorical variables. The following end points were evaluated: overall survival (OS), defined as the interval from the time of being diagnosed to the time of death from any cause; disease-free survival (DFS), defined as the interval from the time of being diagnosed to the time of disease recurrence/metastasis or death from any cause; locoregional relapse-free survival (LRFS) and distant metastasis-free survival (DMFS), defined as the interval from the time of being diagnosed to
Fig. 1. Flow diagram.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
3
Table 1 Basic characteristics by TBIL, DBIL and IBIL blood groupa . TBIL group
DBIL group
Characteristics
High TBIL
Low TBIL
Total Age <55 year 55–64 year >64 year Gender Male Female Smoking status Never Ever Pathology Adenocarcinoma Squamous cell carcinoma Adenosquamous carcinoma Large cell carcinoma Other pTNM stage Ia Ib IIa IIb IIIa pT stage 1 2 3 4 pN stage 0 1 2 Surgery type Lobectomy Pneumonectomy Other Chemotherapyc Neoadjuvant Adjuvant No
1055
562
b
P
IBIL group
High DBIL
Low DBIL
809
808
<0.01 335 (31.8) 368 (34.9) 352 (33.4)
200 (35.6) 228 (40.6) 134 (23.8)
764 (72.4) 291 (27.6)
405 (72.1) 157 (27.9)
463 (43.9) 592 (56.1)
189 (33.6) 373 (66.4)
628 (59.5) 341 (32.3) 59 (5.6) 15 (1.4) 12 (1.1)
270 (48.0) 238 (42.3) 35 (6.2) 13 (2.3) 6 (1.1)
154 (14.6) 376 (35.6) 27 (2.6) 156 (14.8) 342 (32.4)
50 (8.9) 199 (35.4) 13 (2.3) 80 (14.2) 220 (39.1)
203 (19.2) 663 (62.8) 113 (10.7) 76 (7.2)
76 (13.5) 373 (66.4) 66 (11.7) 47 (8.4)
616 (58.4) 142 (13.5) 297 (28.2)
299 (53.2) 74 (13.2) 189 (33.6)
855 (81.0) 157 (14.9) 43 (4.1)
452 (80.4) 89 (15.8) 21 (3.7)
63 (5.9) 693 (65.7) 329 (31.2)
28 (5.2) 375 (66.7) 172 (30.6)
b
P
High IBIL
Low IBIL
918
699
296 (32.2) 320 (34.9) 302 (32.9)
239 (34.2) 276 (39.5) 184 (26.3)
656 (71.5) 262 (28.5)
513 (73.4) 186 (26.6)
421 (45.9) 497 (54.1)
231 (33.0) 468 (67.0)
569 (62.0) 275 (30.0) 51 (5.6) 14 (1.5) 9 (1.0)
329 (47.1) 304 (43.5) 43 (6.2) 14 (2.0) 9 (1.3)
136 (14.8) 322 (35.1) 25 (2.7) 131 (14.3) 304 (33.1)
68 (9.7) 253 (36.2) 15 (2.1) 105 (15.0) 258 (36.9)
180 (19.6) 576 (62.7) 94 (10.2) 68 (7.4)
99 (14.2) 460 (65.8) 85 (12.2) 55 (7.9)
531 (57.8) 123 (13.4) 264 (28.8)
384 (54.9) 93 (13.3) 222 (31.8)
733 (79.8) 147 (16.0) 38 (4.1)
574 (82.1) 99 (14.2) 26 (3.7)
52 (3.1) 615 (64.0) 278 (30.3)
39 (2.8) 453 (62.3) 223 (31.9)
<0.01 251 (31.0) 276 (34.1) 282 (34.9)
284 (35.1) 320 (39.6) 204 (25.2)
640 (79.1) 169 (20.9)
529 (65.5) 279 (34.5)
321 (39.7) 488 (60.3)
331 (41.0) 477 (59.0)
460 (56.9) 282 (34.9) 41 (5.1) 16 (2.0) 10 (1.2)
438 (54.2) 297(36.8) 53 (6.6) 12 (1.5) 8 (1.0)
113 (14.0) 285 (35.2) 19 (2.3) 121 (15.0) 271 (33.5)
91 (11.3) 290 (35.9) 21 (2.6) 115 (14.2) 291 (36.0)
150 (18.5) 499 (61.7) 94 (11.6) 66 (8.2)
129 (16.0) 537 (66.5) 85 (10.5) 57 (7.1)
475 (58.7) 109 (13.5) 225 (27.8)
440 (54.5) 107 (13.2) 261 (32.3)
652 (80.7) 127 (15.3) 30 (4.0)
655 (80.9) 119 (15.8) 34 (3.3)
45 (2.2) 536 (62.0) 248 (30.7)
47 (3.1) 531 (64.0) 253 (31.3)
0.04
<0.01
0.88
<0.01
0.39
<0.01
0.60
<0.01
<0.01
0.52
<0.01
0.19
0.01
0.02
0.92
0.03
0.01
0.05
0.84
0.20
0.77
0.70
Pb
0.52
0.94
0.74
Abbreviations: TBIL: total bilirubin; DBIL: direct bilirubin; IBIL: indirect bilirubin; pTNM stage: pathological tumour-node-metastasis stage. a Some percentages may not sum to 100 due to rounding. b x2 test. c Some patients received both neoadjuvant and adjuvant chemotherapy.
Fig. 2. The distribution of serum total bilirubin (TBIL) level.
observed for LRFS between the high-TBIL group and the low-TBIL group (HR, 0.88; 95% CI 0.71–1.09; P = 0.236; Supplementary Fig. 4C). The rates of 5-year LRFS were 75.2% in the high-TBIL group and 73.4% in the low-TBIL group.
Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. Similarly, the high-DBIL group had significantly longer OS (HR, 0.85; 95% CI 0.74–0.98; P = 0.023; Supplementary Fig. 5A), DFS (HR, 0.82; 95% CI 0.72–0.92; P = 0.001; Supplementary Fig. 5B), and DMFS (HR, 0.76; 95% CI 0.62–0.93; P = 0.007; Supplementary Fig. 5D) than the low-DBIL group. The difference in LRFS between the high-DBIL group and the low-DBIL group was not significant (HR, 0.94; 95% CI 0.77–1.16; P = 0.561; Supplementary Fig. 5C). The 5-year OS, DFS, LRFS and DMFS were, respectively, 57.5%, 47.3%, 75.2% and 77.5% for patients in the high-DBIL group, and 52.7%, 39.5%, 73.9% and 71.1% for patients in the low-DBIL group. Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. The high-IBIL group was associated with longer OS (HR, 0.75; 95% CI 0.65-0.86; P < 0.001; Fig. 3A), DFS (HR, 0.77; 95% CI 0.68– 0.87; P < 0.001; Fig. 3B), and DMFS (HR, 0.73; 95% CI 0.60–0.90; P = 0.002; Fig. 3D) with significant difference. LRFS did not differ significantly between high-IBIL group and low-IBIL group (HR, 0.96; 95% CI 0.78–1.19; P = 0.716; Fig. 3C). The 5-year OS, DFS, LRFS and DMFS were, respectively, 59.2%, 47.4%, 74.9% and 77.0% for patients in the high-IBIL group, and 49.7%, 38.2%, 74.1% and 70.5% for patients in the low-IBIL group.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
4
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
Fig. 3. Kaplan–Meier curves for (a) overall survival, (b) disease-free survival, (c) locoregional relapse-free survival, and (d) distant metastasis-free survival stratified by indirect bilirubin (IBIL) level.
To validate the survival results, we also performed Kaplan– Meier analysis of the entire cohort stratified by the median serum TBIL value of 11.0 mmol/L. The results demonstrated that patients with a level of TBIL 11.0 mmol/L had lengthened OS, DFS, and DMFS (Supplementary Fig. 6). Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.canep.2015.06.007. 6. Univariate analysis and multivariate analysis OS was further analyzed in univariate and multivariate analyses. In the univariate analysis, gender, smoking status, pTNM stage, pT stage, pN stage, surgery type, chemotherapy, TBIL level, DBIL level, and IBIL level were identified as significant prognostic factors. When these variables were further analyzed in the multivariate analysis, we found that smoking status (P = 0.019), pTNM stage (P = 0.003), pT stage (P < 0.001), pN stage (P = 0.002), chemotherapy (P < 0.001), and IBIL level (P = 0.004) have significant HRs, indicating that they are independent, significant predictors for OS (Table 2). High serum IBIL level was a positive prognostic factor. 7. Discussion In this study we investigated the prognostic value of serum bilirubin levels in NSCLC, and to our knowledge, report for the first time that moderately elevated bilirubin levels at diagnosis represent positive prognostic factors for survival in completely resected NSCLC. Patients with relatively high TBIL, DBIL or IBIL level had significantly longer OS, DFS and DMFS than patients with relatively low TBIL, DBIL or IBIL levels. IBIL level was identified as a significant prognostic predictor for OS, independent of any other clinicopathological features of NSCLC, including pTNM stage.
Apart from the association between bilirubin levels and cancer risk and mortality, earlier studies did not demonstrate clear evidence regarding the relationship of bilirubin levels and survival of NSCLC. Consistent with previous studies, our study suggested that high bilirubin levels within the relatively normal range were protective factors against cancer, owing to bilirubin’s power of antioxidation, anti-inflammation and antiproliferation, given the critical role of oxidative stress as a contributing factor to carcinogenesis [15]. The prospective study in Belgium by Temme et al. suggested that high total bilirubin was significantly associated with decreased cancer mortality for men, and the association was in the same direction for women [16]. The cohort study by Horfall et al. investigated the relation between bilirubin levels and risk of respiratory disease (including lung cancer) and death among 504,206 adults from the Health Improvement Network database in UK. Their results showed that relatively elevated bilirubin within 40 mmol/L for men and 30 mmol/L for women was correlative with decreased risk of lung cancer and allcause mortality [6]. The results of our study combined with those suggest bilirubin can not only protect against the development of malignancies, but also prolong the survival of cancer patients. Several studies have shown the inverse association between smoking and bilirubin levels [6,7,17,18]. In our studies, there were more never smokers in the high-TBIL and high-IBIL groups than the low-TBIL and low-IBIL groups, respectively. Our results are in consistence with prior results. Smoking is a well-established risk factor for lung cancer [19–21] and evidences suggested that it is a negative prognostic factor for the survival of NSCLC [22,23]. This can partly explain our survival results. Of note, the survival benefit of bilirubin was specific to IBIL in our study, because in multivariate analysis, among TBIL, DBIL and IBIL levels, only IBIL level was identified as an independent, significant predictor for survival. This observation is consistent with the report by Keshavan et al. which
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
5
Table 2 Univariate and multivariate analysis for overall survival. Univariate Hazard ratio (95% CI) Age <55 year 55–64 year >64 year Gender Male Female Smoking status Never Ever Pathology Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Large cell carcinoma Other pTNM stage I II III pT stage 1 2 3 4 pN stage 0 1 2 Surgery type Lobectomy and other Pneumonectomy Chemotherapy Yes No TBIL level High Low DBIL level High Low IBIL level High Low
Multivariate P
Hazard ratio (95% CI)
P –
0.240 – – –
Reference 1.02 (0.86–1.20) 1.15 (0.97–1.36) 0.048 Reference 1.17 (1.00–1.37)
0.486 Reference 1.09 (0.87–1.37)
0.017 Reference 1.19 (1.03–1.37)
0.019 Reference 1.19 (1.03–1.37) –
0.332 – – – – –
Reference 0.98 (0.85–1.14) 1.31 (0.98–1.75) 1.05 (0.60–1.84) 0.74 (0.35–1.57) <0.001 Reference 2.43 (1.99–2.96) 3.65 (3.12-4.28)
0.003 Reference 1.79 (1.28–2.50) 1.80 (1.15–2.81)
<0.001 Reference 2.05 (1.63–2.57) 2.95 (2.22–3.92) 5.08 (3.78–6.81)
<0.001 Reference 1.74 (1.38–2.20) 1.83 (1.32–2.55) 3.06 (2.15–4.37)
<0.001 Reference 2.31 (1.89–2.83) 3.13 (2.69–3.64)
0.002 Reference 1.56 (1.12–2.18) 2.07 (1.37–3.11)
<0.001 Reference 1.79 (1.40–2.28)
0.374 Reference 1.13 (0.88–1.46)
<0.001 Reference 1.37 (1.20–1.58)
<0.001 Reference 1.41 (1.20–1.65)
<0.001 Reference 1.37 (1.19–1.58)
0.220 Reference 1.18 (0.90–1.54)
0.023 Reference 1.17 (1.02–1.35)
0.680 Reference 1.00 (0.83–1.20)
<0.001 Reference 1.34 (1.17–1.54)
0.004 Reference 1.22 (1.07–1.41)
Abbreviations: CI: confidence interval; pTNM stage: pathological tumor-node-metastasis stage; TBIL: total bilirubin; DBIL: direct bilirubin; IBIL: indirect bilirubin.
revealed that bilirubin induces apoptosis in colon adenocarcinoma cell lines, and this effect appears to be specific to the unconjugated bilirubin, as ditaurobilirubin and biliverdin did not show this effect [11]. Reasons for this may largely relate to the unbound nature of IBIL, which can freely diffuse across cell membranes. The study by Zucker et al. suggested that unconjugated bilirubin can readily cross cellular membranes [24]. The study by Weisiger et al. indicated that if the serum unconjugated bilirubin level is 10 mmol/L, the level of unbound (free) bilirubin will be 7.4 nmol/L [25]. We speculate that unbound bilirubin may be the main effective portion of total bilirubin to act as a factor of antioxidation, anti-inflammation and antiproliferation. This study was retrospective in nature and therefore had some limitations. First, like other retrospective studies, we cannot completely rule out the impact of selection bias. The patients were included over a relatively long period (2005–2009), and many patients were initially staged according the sixth international classification system for lung cancer, whereas they were restaged according to the seventh classification system during the analysis. Second, Cantonese constitute most of our study population. Our results may not apply to other ethnic populations. Third, we were unable to collect other sufficient information that appears to be
significant prognostic factors for NSCLC, such as epidermal growth factor receptor (EGFR) mutational status, EGFR copy number and anaplastic lymphoma kinase (ALK) rearrangement status. 8. Conclusion We have demonstrated in a retrospective analysis that NSCLC patients with moderately elevated pretreatment serum bilirubin levels who are treated with curative resection have longer OS, DFS, and DMFS; IBIL is an independent, prognostic predictor for NSCLC patients with curative resection. Further well-designed studies involving diverse ethnic populations are warranted to confirm the prognostic value of serum bilirubin levels in NSCLC, as well as taking into account other potential prognostic factors for NSCLC. Conflicts of interest None. Funding None.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007
G Model CANEP 880 No. of Pages 6
6
N. Li et al. / Cancer Epidemiology xxx (2015) xxx–xxx
Author contributions LN, XM and WSY were involved in the design of the study. WSY, CMY and XM supervised the study. LN, XM, CMY, ZF, LCF, WBX, OW were involved in data acquisition. LN, XM, CMY and WSY were involved in data analysis and interpretation. All authors were involved in writing the article and approved the final version. References [1] A. Jemal, F. Bray, M.M. Center, J. Ferlay, E. Ward, D. Forman, Global cancer statistics, CA Cancer J. Clin. 61 (2011) 69–90. [2] G. D’Addario, M. Fruh, M. Reck, P. Baumann, W. Klepetko, E. Felip, Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 21 (Suppl. 5) (2010) v116–9. [3] L. Crino, W. Weder, J. van Meerbeeck, E. Felip, Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 21 (Suppl. 5) (2010) v103–15. [4] J.D. Ostrow, J.H. Jandl, R. Schmid, The formation of bilirubin from hemoglobin in vivo, J. Clin. Invest. 41 (1962) 1628–1637. [5] R. Stocker, Y. Yamamoto, A.F. McDonagh, A.N. Glazer, B.N. Ames, Bilirubin is an antioxidant of possible physiological importance, Science 235 (1987) 1043– 1046. [6] L.J. Horsfall, G. Rait, K. Walters, D.M. Swallow, S.P. Pereira, I. Nazareth, et al., Serum bilirubin and risk of respiratory disease and death, JAMA 305 (2011) 691–697. [7] C.P. Wen, F. Zhang, D. Liang, C. Wen, J. Gu, H. Skinner, et al., The ability of bilirubin in identifying smokers with higher risk of lung cancer: a large cohort study in conjunction with global metabolomic profiling, Clin. Cancer Res. (2014) . [8] S. Ching, D. Ingram, R. Hahnel, J. Beilby, E. Rossi, Serum levels of micronutrients, antioxidants and total antioxidant status predict risk of breast cancer in a case control study, J. Nutr. 132 (2002) 303–306. [9] S.D. Zucker, P.S. Horn, K.E. Sherman, Serum bilirubin levels in the U.S. population: gender effect and inverse correlation with colorectal cancer, Hepatology 40 (2004) 827–835. [10] X. Liu, Q.H. Meng, Y. Ye, M. Hildebrandt, J. Gu, X. Wu, Prognostic significance of pretreatment serum levels of albumin, LDH and total bilirubin in patients with non-metastatic breast cancer, Carcinogenesis (2014) . [11] P. Keshavan, S.J. Schwemberger, D.L. Smith, G.F. Babcock, S.D. Zucker, Unconjugated bilirubin induces apoptosis in colon cancer cells by triggering mitochondrial depolarization, Int. J. Cancer 112 (2004) 433–445.
[12] P. Rao, R. Suzuki, S. Mizobuchi, T. Yamaguchi, S. Sasaguri, Bilirubin exhibits a novel anti-cancer effect on human adenocarcinoma, Biochem. Biophys. Res. Commun. 342 (2006) 1279–1283. [13] R. Ollinger, P. Kogler, J. Troppmair, M. Hermann, M. Wurm, A. Drasche, et al., Bilirubin inhibits tumor cell growth via activation of ERK, Cell Cycle 6 (2007) 3078–3085. [14] P. Goldstraw, J. Crowley, K. Chansky, D.J. Giroux, P.A. Groome, R. Rami-Porta, et al., The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours, J. Thorac. Oncol. 2 (2007) 706–714. [15] L.J. Marnett, Oxyradicals and DNA damage, Carcinogenesis 21 (2000) 361–370. [16] E.H. Temme, J. Zhang, E.G. Schouten, H. Kesteloot, Serum bilirubin and 10-year mortality risk in a Belgian population, Cancer Causes Control 12 (2001) 887– 894. [17] K. Frost-Pineda, Q. Liang, J. Liu, L. Rimmer, Y. Jin, S. Feng, et al., Biomarkers of potential harm among adult smokers and nonsmokers in the total exposure study, Nicotine Tob. Res. 13 (2011) 182–193. [18] S.S. O'Malley, R. Wu, S.T. Mayne, P.I. Jatlow, Smoking cessation is followed by increases in serum bilirubin, an endogenous antioxidant associated with lower risk of lung cancer and cardiovascular disease, Nicotine Tob. Res. 16 (2014) 1145–1149. [19] R. Doll, R. Peto, Mortality in relation to smoking: 20 years’ observations on male British doctors, Br. Med. J. 2 (1976) 1525–1536. [20] R. Doll, R. Peto, K. Wheatley, R. Gray, I. Sutherland, Mortality in relation to smoking: 40 years’ observations on male British doctors, BMJ 309 (1994) 901– 911. [21] D.S. Ettinger, D.E. Wood, W. Akerley, L.A. Bazhenova, H. Borghaei, D.R. Camidge, et al., Non-small cell lung cancer, version 1. 2015, J. Natl. Compr. Canc. Netw. 12 (2014) 1738–1761. [22] Y.Y. Janjigian, K. McDonnell, M.G. Kris, R. Shen, C.S. Sima, P.B. Bach, et al., Packyears of cigarette smoking as a prognostic factor in patients with stage IIIB/IV nonsmall cell lung cancer, Cancer 116 (2010) 670–675. [23] T. Kawaguchi, M. Takada, A. Kubo, A. Matsumura, S. Fukai, A. Tamura, et al., Performance status and smoking status are independent favorable prognostic factors for survival in non-small cell lung cancer: a comprehensive analysis of 26,957 patients with NSCLC, J. Thorac. Oncol. 5 (2010) 620–630. [24] S.D. Zucker, W. Goessling, A.G. Hoppin, Unconjugated bilirubin exhibits spontaneous diffusion through model lipid bilayers and native hepatocyte membranes, J. Biol. Chem. 274 (1999) 10852–10862. [25] R.A. Weisiger, J.D. Ostrow, R.K. Koehler, C.C. Webster, P. Mukerjee, L. Pascolo, et al., Affinity of human serum albumin for bilirubin varies with albumin concentration and buffer composition: results of a novel ultrafiltration method, J. Biol. Chem. 276 (2001) 29953–29960.
Please cite this article in press as: N. Li, et al., Elevated serum bilirubin levels are associated with improved survival in patients with curatively resected non-small-cell lung cancer, Cancer Epidemiology (2015), http://dx.doi.org/10.1016/j.canep.2015.06.007