An overview of mortality & predictors of small-cell and non-small cell lung cancer among Saudi patients
- DOI
- 10.1016/j.jegh.2017.09.004How to use a DOI?
- Abstract
Lung cancer ranks as the top cancer worldwide in terms of incidence and constitutes a major health problem. About 90% of lung cancer cases are diagnosed at advance stage where treatment is not available. Despite evidence that lung cancer screening improves survival, guidelines for lung cancer screening are still a subject for debate. In Saudi Arabia, only 14% of lung cancers are diagnosed at early stage and researches on survival and its predictors are lacking. This overview analysis was conducted on predictors of lung cancer mortality according to the two major cancer types, small-cell lung cancers (SCLCs) and non-small cell lung cancers (NSCLCs) in Saudi Arabia. A secondary data analysis was performed on small-cell lung cancers (SCLCs) and Non-small cell lung cancers (NSCLCs) registered in the Saudi Cancer Registry (SCR) for the period 2009–2013 to estimate predictors of mortality for both lung cancer types. A total of 404 cases (197 SCLC and 207 NSCLC) were included in the analysis, all Saudi nationals. A total of 213 (52.75%) deaths occurred among lung cancer patients, 108 (54.82%) among SCLCs and 105 (50.72%) among NCSLCs. Three quarter of patients are diagnosis with advance stage for both SCLC & NSCLC. Univariate analysis revealed higher mean age at diagnosis in dead patients compared to alive patients for SCLCs (p = 0.04); but not NSCLCs, a lower mortality for NSCLCs diagnosed in 2013 (p = 0.025) and a significant difference in stage of tumor (p = 0.006) and (p = 0.035) for both SCLC and NSCLC respectively. In multiple logistic regression, stage of tumor was a strong predictor of mortality, where distant metastasis increased morality by 6-fold (OR = 5.87, 95% CI: 2.01 – 17.19) in SCLC and by 3-fold (OR = 3.29, 95% CI: 1.22 – 8.85) in NSCLC, compared to localized tumors. Those with NSCLC who were diagnosed in 2013 were less likely to die by 64% compared to NSCLC diagnosed in 2009 (OR = 0.36, 95% CI: 0.14 – 0.93). Age, sex, topography and laterality were not associated with mortality for both types of lung cancer. We observed that the stage of the tumor is the strongest predictor of mortality for both SCLCs and NSCLs. This confirms the impact of diagnostic stage on survival. However, establishing Saudi-specific lung cancer screening guidelines will require further research on the benefits and harms of screening modalities in the Saudi population.
- Copyright
- © 2017 Ministry of Health, Saudi Arabia. Published by Elsevier Ltd.
- Open Access
- This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Lung cancer is the most commonly diagnosed cancer in the world. According to GLOBACAN, 1.8 million cases of lung cancer were registered during 2012, making up 13% of all cancer deaths [1].
Lung cancer incidence shows a growing trend globally given that in 2002 the number of registered cases was 1.35 million cases which increased to 1.6 million cases by 2008 [2]. This change in incidence contributed to changes in exposure to risk factors, especially smoking, as well as increased lifespan and population size [2,3].
Mortality rate of lung cancer is relatively close to incidence rate, that, global mortality-to-incidence ratio being 0.88 for males and 0.84 for females [3,4]. Lung cancer was responsible for 2.3% of total deaths worldwide during 2004 and ranked as the 8th leading cause of death [5,6].
Saudi Arabia has low incidence rate of lung cancer compared to global incidence. In 2013, age standardized ratio (ASR) was 5.5 per 100,000 for males and 1.8 per 100,000 for females [7]. In contrast, the average global ASR during 2008 was 33.8 per 100,000 for males and 13.5 for females [8]. But, a growing population in Saudi Arabia, along with an expected increase in elderly population size by sevenfold, accompanied by high smoking prevalence; which is gradually increasing by 1.5% for males and 2.0% for females, are all factors that can potentially result in more incidence of lung cancer in Saudi Arabia [9–12].
Lung cancers staging globally shows that only 15% of cases are diagnosed at an early stage [8]. Saudi Arabia falls within the global range, with only 14% of cases diagnosed early with localized tumors [7]. However, in developed countries this percentage is higher. In Canada for example, 20–30% of cases are diagnosed at an early stage [13].
Annual lung cancer screening has been recommended since 2012 by several health organizations and expert panels, which showed a significant decrease in lung cancer mortality by 20.0% (95% CI: 6.8–26.7) (P = 0.004) [14]. However, screening implications are still subject to discussion [15].
Currently, Saudi Arabia does not have national screening guidelines for lung cancer. Developing country-specific screening guidelines requires a strong research base on lung cancer mortality to offer more evidence for decision makers about the value of instituting and implementing national screening guidelines.
The existing gap in detailed knowledge on lung cancer mortality in Saudi Arabia needs to be offset to provide information-for-action for the development of Saudi-customized screening and management guidelines for this malignancy. Hence, main purpose of this paper is focused on the predictors of mortality among lung cancer patients, both small-cell and non-small cell, in Saudi Arabia for the period of 2009–2013.
2. Methods
2.1. Study design
A secondary data analysis was performed on lung cancer cases reported to the (SCR) during the period of 2009–2013 after their permission to use and disseminate the data. Determination of IRB was reviewed by Emory university.
2.2. Setting and data source
Data was requested and readily available from the SCR, as part of their objective is to support dissemination and utilization of registered data.
2.3. Study variables
Data obtained by the SCR includes information on patients demographic data (sex, age, nationality), date of diagnosis, tumor characteristics (topography, morphology, behavior, grade, extension [SEER stage], and laterality), date of last contact (via telephone) and mortality indicators (status and cause of death) [7,16].
2.4. Data management
The dataset included 452 registered cases of lung cancer during 2009–2013, all of which were of Saudi nationality. Diagnoses were restricted to NSCLC and SCLC types of lung cancers. We exclude 48 cases; observations without a valid vital status (N = 1) were excluded as well as unknown cause of death (N = 1) or non-reported data on cancer extension (N = 47; 23 dead and 24 alive). Therefore, the final dataset included 404 observations with a total exclusion of 48 observations.
2.5. Statistical analysis
The distribution of variables was examined by conducting a univariate analysis stratified by SCLC and NSCLC to find differences at each variable within vital status.
To estimate predictors of mortality among lung cancer patients, logistic regression was performed with stratification by histological type. A dichotomous mortality indicator variable was regressed on age, sex, topography, extension, laterality and year of diagnosis. Odd ratios (OR) and 95% confidence intervals (95% CI) were calculated for each predictor. Tumor grade was not included in this study’s model due to the large number of missing data (N = 272) and due to the lack of a widely-accepted grading system for lung cancer that is used consistently by all health facilities [17,18]. Tumor behavior variable also was not included in the model due to collinearity, where all observations are recorded as malignant behavior. The level of statistical significance was set at 0.05. Stata (version SE64, Stata Corporation, College Station, TX) was used for analysis.
3. Results
3.1. Demographic and tumor characteristics for overall cases, SCLC and NSCLC
A total of 404 lung cancer cases were diagnosed during the period 2009–2013 and included in the final analysis. SCLC composed 48.8% (N = 197) of the cases, out of which 108 (54.8%) did not survive; and NSCLC composed 51.2% of observations (N = 207), out of which 105 (50.7%) did not survive. Out of a total of 404 cases, there were 337 (83.4) cases in males and 67 (16.6%) in females. The overall median age of lung cancer diagnosis was 63.3 year (SD 12.46 years). The median age of diagnosis was 63.7 years (SD 12.4 years) in males and 61.02 (SD12.37) in females (p = 0.10). majority of patient are diagnosed at advance stage, 74.1% of SCLC and 76.8% of NSCLC has diagnosed with metastasis stage [Table 1].
| SCLC N = 197 | NSCLCN = 207 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Alive 89 (45.18%) | Dead 108 (54.82%) | Total | P-value | Alive 102 (49.28%) | Dead 105 (50.72%) | Total | P-value | ||
| Age at diagnosis, mean (SD) | 62.39 (SD 13.26) | 66.04 (SD 11.30) | 64.39(SD 12.33) | 0.04 * | 61.42 (SD 12.09) | 62.94(SD 12.94) | 62.19(SD 12.52) | 0.3833* | |
| Sex | Male | 73 (42.69%) | 98 (57.31%) | 171 (86.80%) | 0.07** | 79 (77.45%) | 87 (82.86%) | 166 (80.19%) | 0.329** |
| Female | 16 (61.54%) | 10 (38.46%) | 26 (13.20%) | 23 (22.55%) | 18 (17.14%) | 41 (19.81%) | |||
| Year of diagnosis | 2009 | 16 (17.98%) | 21 (19.44%) | 37 (18.78%) | 0.433 ** | 29 (28.43%) | 30 (28.57%) | 59 (28.5%) | 0.025** |
| 2010 | 19 (21.35%) | 28 (25.93%) | 47 (23.86%) | 22 (21.57%) | 22 (20.95%) | 44 (21.26%) | |||
| 2011 | 17 (19.10%) | 19 (17.59%) | 36 (18.27%) | 9 (8.82%) | 22 (20.95%) | 31 (14.98%) | |||
| 2012 | 14 (15.73%) | 23 (21.30%) | 37 (18.78%) | 17 (16.67%) | 20 (19.05%) | 37 (17.87%) | |||
| 2013 | 23 (25.84%) | 17 (15.74%) | 47 (20.30%) | 25 (24.51%) | 11 (10.48%) | 36 (17.39%) | |||
| Topography | Main Bronchus | 7 (7.87%) | 12 (11.11%) | 19 (9.64%) | 0.597*** | 6 (5.88%) | 4 (3.81%) | 10 (4.83%) | 0.851*** |
| Upper lobe | 24 (26.97%) | 27 (25%) | 51 (25.89%) | 33 (32.35%) | 34 (32.38%) | 67 (32.37%) | |||
| Middle lobe | 4 (4.49%) | 5 (4.63%) | 9 (4.57%) | 3 (2.94%) | 6 (5.71%) | 9 (4.35%) | |||
| Lower lobe | 19 (21.35%) | 14 (12.96%) | 33 (16.75%) | 21 (20.59%) | 20 (19.05%) | 41 (19.81%) | |||
| Overlapped lesion | 3 (3.37%) | 7 (6.48%) | 10 (5.08%) | 5 (4.90%) | 3 (2.86%) | 8 (3.86%) | |||
| Not otherwise specified | 32 (35.96%) | 43 (39.81%) | 75 (38.07%) | 34 (33.33) | 38 (36.19%) | 72 (34.78%) | |||
| Carcinoma Of other ill-defined sites | 0 | 0 | 0 | 0 | 0 | 0 | |||
| Stage | Localized | 16 (17.98%) | 6 (5.56%) | 22 (11.17) | 16 (15.69%) | 7 (6.67%) | 23 (11.11%) | ||
| Regional by direct extension | 7 (7.87%) | 2 (1.85%) | 9 (4.54%) | 7 (6.86%) | 3 (2.86%) | 10 (4.83%) | |||
| Regional by lymph node | 6 (6.74%) | 4 (3.7%) | 10 (5.08%) | 7 (6.86%) | 2 (1.9%) | 9 (4.35%) | |||
| Regional by both direct extension and lymph node | 4 (4.49%) | 5 (4.63%) | 9 (4.57%) | 2 (1.96%) | 2 (1.9%) | 4 (1.93%) | |||
| Regional-NOS | 0 | 1 (0.93%) | 1 (0.51%) | 1 (0.98%) | 1 (0.95%) | 2 (0.97%) | |||
| Distance metastasis/Systematic disease | 56 (62.92) | 90 (83.33% | 146 (74.11) | 69 (67.65%) | 90 (85.71%) | 159 (76.81%) | |||
| Laterality | Not paired | 1 (1.12%) | 1 (0.93%) | 2 (1.02%) | 0.876 *** | 0 | 0 | 0 | 0.453*** |
| Right origin | 44 (49.44%) | 51 (47.22%) | 95 (48.22%) | 57 (55.88%) | 64 (60.95%) | 121 (58.45%) | |||
| Left origin | 32 (35.96%) | 35 (32.41%) | 67 (34.01%) | 34 (33.33%) | 32 (30.48%) | 66 (31.88%) | |||
| Only one side involved (right or left) unspecified | 0 | 0 | 0 | 6 (5.88%) | 2 (1.90%) | 8 (3.86%) | |||
| Bilateral (side of origin unknown or single primary) | 2 (2.25%) | 4(3.70%) | 6 (3.05%) | 5 (4.90%) | 7 (6.67%) | 12 (5.80%) | |||
| Midline origin | 0 | 0 | 0 | 0 | 0 | 0 | |||
| paired no information concerning laterality | 10 (11.24%) | 17 (15.74%) | 27 (13.71%) | 0 | 0 | 0 | |||
T-test
Chi-square test
Fisher’s exact test.
Characteristics of lung cancer patients stratified by histological type.
In univariate analysis, SCLC cases showed a statistically significant difference in mean age between those who are alive and those dead (p = 0.0386), however, in NSCLC, age showed no significant difference between alive and dead (p = 0.3833). For year of diagnoses there was no statistically significant difference between alive and dead at SCLC (p = 0.433), but it showed a statistically significant difference for NSCLCs (p = 0.025). Stage of tumor showed a significant difference for both, SCLC (p = 0.006) and NSCLC (p = 0.035) [Table 1].
3.2. Lung cancer mortality, for SCLC
Using multiple logistic regression analysis for SCLC, extension was found to be the strongest predictor of mortality. Having regional extension by both direct extension and lymph node increase the odd of mortality by 6-fold compared to having localized disease (OR = 6.08, 95% CI: 1.05–35.18), and having distance metastasis increased the odds of mortality by 5-fold compared to having localized disease (OR = 5.87, 95% CI: 2.01–17.19).
None of the other variables were found to be statistically significantly associated with mortality in SCLC, including age (OR = 1.02, 95% CI 1.00–1.03) gender (OR = 2.30, 95% CI: 0.90–5.89) and year of diagnosis [Table 2 ]
| Mortality | Odd Ratio | 95% Confidence interval | P-Value |
|---|---|---|---|
| Age | 1.02 | 1.00–1.05 | 0.10 |
| Sex | |||
| Female (Ref.) | 1 | ||
| Male | 2.30 | 0.90–5.89 | 0.08 |
| Extention | |||
| Localized (ref.) | 1 | ||
| Regional by direct extension | 0.84 | 0.12–5.82 | 0.86 |
| Regional by lymph node | 1.62 | 0.29–9.13 | 0.58 |
| Regional by both direct extension and lymph node | 6.08 | 1.05–35.18 | 0.04 |
| Regional nos | – | – | - * |
| Distance Metastasis/systematic disease | 5.87 | 2.01–17.19 | 0.00 |
| Topographgy | |||
| Main bronchus (ref.) | 1 | ||
| Upper lobe | 0.41 | 0.11–1.61 | 0.20 |
| Middle lobe | 0.67 | 0.09–5.32 | 0.71 |
| Lower lobe | 0.29 | 0.07–1.23 | 0.09 |
| Overlapped lesion | 0.56 | 0.08–3.76 | 0.55 |
| Not otherwise specified | 0.41 | 0.11–1.52 | 0.18 |
| Laterality | |||
| Left (ref.) | |||
| Not paired | 0.18 | 0.01–4.12 | 0.28 |
| Right origin | 0.86 | 0.42–1.75 | 0.68 |
| Bilateral (side of origin unknown or single primary) | 2.20 | 0.32–15.24 | 0.43 |
| Paired no information concerning laterality | 1.48 | 0.51–4.31 | 0.47 |
| Year of diagnosis | |||
| 2009 (ref.) | 1.00 | ||
| 2010 | 1.29 | 0.48–3.44 | 0.61 |
| 2011 | 0.83 | 0.30–2.30 | 0.71 |
| 2012 | 1.59 | 0.57–4.50 | 0.38 |
| 2013 | 0.54 | 0.20–1.48 | 0.23 |
Estimates for this category could not be obtained because count = 1 for SCLC.
Mortality Predictors among SCLC.
3.3. Lung cancer mortality, for NSCLC
Multiple regression model for NSCLC showed a statistically significant increase in mortality among cases with distance metastasis compared to those with local disease (OR = 3.29, 95% CI: 1.22–8.85). However, in contrast to SCLC, regional extension by both direct and lymph node extension did not significantly increase the odds of mortality. Besides that, year of diagnosis in cases showed a decrease in the odds of mortality for those diagnosed in 2013 compared to those diagnosed in 2009 (OR = 0.36, 95% CI: 0.14–0.93).
None of the other variables were found to be statistically significantly associated with mortality in NSCLC, including age (OR = 1.01, 95% CI 0.99–1.04) gender (OR = 1.26, 95% CI: 0.59–2.69) [Table 3 ].
| Mortality | Odd ratio | 95% confidence interval | P-value |
|---|---|---|---|
| Age | 1.01 | 0.99–1.04 | 0.31 |
| Sex | |||
| Female (ref.) | |||
| Male | 1.26 | 0.59–2.69 | 0.56 |
| Extension | |||
| Localized (ref.) | 1 | ||
| Regional by direct extension | 0.77 | 0.14–4.16 | 0.76 |
| Regional by lymph node | 0.81 | 0.12–5.22 | 0.82 |
| Regional by both direct extension and lymph node | 2.15 | 0.22–20.70 | 0.51 |
| Regional nos | 1.26 | 0.06–27.56 | 0.88 |
| Distance metastasis/systematic disease | 3.29 | 1.22–8.85 | 0.02 |
| Topography | |||
| Main bronchus (ref.) | 1 | ||
| Upper lobe | 1.67 | 0.38–7.29 | 0.50 |
| Middle lobe | 2.67 | 0.34–21.08 | 0.35 |
| Lower lobe | 1.42 | 0.31–6.58 | 0.65 |
| Overlapped lesion | 1.45 | 0.17–12.25 | 0.73 |
| Not otherwise specified | 2.23 | 0.50–10.01 | 0.30 |
| Laterality | |||
| Left (ref.) | 1.00 | ||
| Not paired | 1 (empty) | ||
| Right origin | 1.04 | 0.54–2.03 | 0.90 |
| Bilateral (side of origin unknown or single primary) | 0.25 | 0.04–1.51 | 0.13 |
| Paired no information concerning laterality | 0.84 | 0.20–3.55 | 0.81 |
| Year of diagnosis | |||
| 2009 (ref.) | 1.00 | ||
| 2010 | 0.96 | 0.41–2.23 | 0.92 |
| 2011 | 2.14 | 0.77–5.95 | 0.14 |
| 2012 | 1.10 | 0.45–2.67 | 0.83 |
| 2013 | 0.36 | 0.14–0.93 | 0.04 |
Mortality Predictors among NSCLC.
4. Discussion
In this secondary data analysis, the overall mortality rate of lung cancer among Saudi national patients in the period 2009–2013 for both SCLC and NSCLC was 52.75% (54.82% in SCLC and 50.72% in NSCLC). Mortality was strongly predicted by tumor extension. In SCLC, regional extension by both direct and lymph node extension increase mortality by 6-folds compared to localized tumor and distance metastasis/systematic increase mortality by 5-fold compared to localized tumor. In NSCLC, distance metastasis/systematic disease increased mortality by 3-fold compared to localized tumor.
The lung cancer survival rate shows a difference by histological type, with NSCLC having a better prognosis compared to SCLC [19]. However, the staging of a tumor is the strongest determinant factor of lung cancer survival. Chansky et al. conducted a retrospective study on 9137 patients and observed a strong correlation between Tumor, Node, Metastasis (TNM) stage and survival, where the median survival for patients at stage IIIA was 19 months, compared to 95 months for patients at stage IA [20]. Our study, confirms the strong effect of disease stage on mortality for both NSCLC and SCLC.
Several efforts have tried to develop an accurate prediction model for lung cancer prognosis by adding further factors to tumor staging, although, tumor staging is still the major mortality predictor [25]. Other factors that showed a prognostic effect independent of disease stage include performance status (PS), age and gender [26,27]. Additionally, several genetic biomarkers associated with lung cancer were found to have a prognostic effect [28–31]. Other factors like obesity [32] and smoking history [33] were also found to impact survival in lung cancer. Hence, to achieve higher prediction accuracy, a more complex approach that integrates individual, pathological markers and genetics factors is needed [34]. Our study did not allow for a full investigation of important prognostic factors because of the limited nature of data collected within cancer registries.
Analysis of mortality by year of diagnosis showed significantly lower odds for those diagnosed in 2013 compared to 2009 for NSCLC but not for SCLC. It is unclear whether any significant improvements were introduced in clinical management of lung cancer during 2013 [64]. However, the observed drop in mortality in 2013 and the reason why it applied to NSCLCs but not SCLCs requires further investigation.
The overall mean age of diagnosis for both SCLC and NSCLC was 63.27 years (SD = 12.46), which is 6.73 years younger than the average age at diagnosis in the United States. Additionally, the age range in our study overlapped with the age range of lung cancer screening (55–74 years) used in the NLST. The NLST age range was determined by the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial’s 2012 prediction model (PLCOm2012) which determined that the 55–74 year age range as an inclusion criterion in the trial [21]. It is worth to note that the screening recommendations of the USPTF recommended an age range for screening of 55–80 years. For Saudi Arabia, it is not clearly evident whether screening beyond the range of 55–74 years could be an advantage or not, especially when viewing this issue from the angle of premature mortality impact of lung cancer, and by taking into consideration the expected age difference between the US and Saudi Arabia.
Various studies revealed that the effect of cigarette smoking on lung cancer varies by histological types, where cigarette smoking and early initiation of smoking show a stronger association with SCLC compared to NSCLC [22,23]. In the US, the proportion of SCLC was tracked over from 1973 to 2002 and showed a gradual decrease from 17.26% in 1986 to 12.95% in 2002; this decrease in SCLC proportion is potentially a result of strengthened in tobacco control policies during the 1980s [24]. However, the histological distribution of lung cancer in Saudi Arabia supports this association. Our study showed that SCLC constituted 50% of lung cancer cases among males and 37% among females. This could be explained, in part, by differences in smoking prevalence rates between men and women in Saudi Arabia, where the age-standardized smoking prevalence rate among males during 2013 was 26.8%, compared to 3% among women [12].
A main limitation of this study is the unavailability of time-to-event data which prevented the use of survival analysis methods and the ascertainment of an accurate estimate of survival rates for each cancer type. Another limitation is the restricted data on predictors. For a better characterization of lung cancer mortality, a prediction model that integrates pathological variables, biological markers, genetics and patient physical status is needed [68]. The lack of such factors is assumed to have a residual confounding role in the results. Lastly, the study population is restricted to Saudi nationals and it include only the registered cases in SCR as the completeness and accuracy of SCR’s data are questionable, and the results may not be generalizable to other populations.
A major strength of our study is the use of data from the SCR, which is a population-based registry with nation-wide coverage of diagnosed cancers that uses standardized methods for data collection.
As a conclusion, this study showed a strong effect of disease stage on mortality, especially in SCLC. It also showed that, majority of lung cancer patients in Saudi Arabia are diagnosed at an advanced stage. For establishment of a Saudi-specific lung cancer screening guidelines, further research on economic analysis, cost-effectiveness, safety, and complication will be required to make a final recommendation on lung cancer screening.
Acknowledgements
We acknowledge with gratitude the efforts of the Saudi Cancer Registry who provided us with the data to conduct this research. We also thank Ms. Yasmin Zaki who provided editing support for the manuscript.
References
Cite this article
TY - JOUR AU - Hatim I. Alghamdi AU - Ali F. Alshehri AU - Ghada N. Farhat PY - 2017 DA - 2017/09/28 TI - An overview of mortality & predictors of small-cell and non-small cell lung cancer among Saudi patients JO - Journal of Epidemiology and Global Health SP - S1 EP - S6 VL - 7 IS - Supplement 1 SN - 2210-6014 UR - https://doi.org/10.1016/j.jegh.2017.09.004 DO - 10.1016/j.jegh.2017.09.004 ID - Alghamdi2017 ER -