Indian Journal of Cancer Home 

[Download PDF]
Year : 2017  |  Volume : 54  |  Issue : 1  |  Page : 285--290

Evolving epidemiology of lung cancer in India: Reducing non-small cell lung cancer-not otherwise specified and quantifying tobacco smoke exposure are the key

H Kaur1, IS Sehgal1, A Bal2, N Gupta3, D Behera1, A Das2, N Singh1,  
1 Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh, India
2 Department of Histopathology, Postgraduate Institute of Medical Education and Research, Chandigarh, India
3 Department of Cytology and Gynecological Pathology, Postgraduate Institute of Medical Education and Research, Chandigarh, India

Correspondence Address:
Dr. N Singh
Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh


BACKGROUND: Adenocarcinoma is the most prevalent histological type of lung cancer (LC) in developed countries while squamous cell carcinoma (SqCC) has so far been the most common type at our center. Herein, we report our continued assessment of the epidemiological trend of LC aimed at determining any change in the histological distribution. METHODS: Retrospective analysis involving all consecutive newly diagnosed LC patients over a 4-year period (March 2011–February 2015). Demographic characteristics, histology, and staging data for current data set were compared with our previously published data (2008–2011). As before, smoking index (SI) was used to group patients as never (SI = 0), light (SI = 1–100), moderate (SI = 101–300), and heavy (SI ≥301) smokers. RESULTS: Majority of 1301 patients had advanced disease (Stages IIIB = 30.1%; IV = 53.3%), were males (82.3%) and current/ex-smokers (76.9%). Adenocarcinoma and SqCC (36.4% each) were equally prevalent. As compared to our previous study, adenocarcinoma increased (36.4% vs. 27.5%) and nonsmall cell lung cancer-not otherwise specified (NSCLC-NOS) decreased (5.1% vs. 10.9%) significantly (P < 0.001). The current study had more heavy smokers (68.3% vs. 61.1%; P = 0.013) and median SI was also higher (500 vs. 400; P = 0.001). Among SI-based groups, significant differences were observed for age, gender, body mass index, histology, TNM stage, and metastatic disease distribution. CONCLUSION: Reduction in NSCLC-NOS has led to adenocarcinoma and SqCC being equally prevalent at our center in North India despite an increase in heavy smokers. Accurate histological NSCLC subtyping is necessary for optimal epidemiological assessment.

How to cite this article:
Kaur H, Sehgal I S, Bal A, Gupta N, Behera D, Das A, Singh N. Evolving epidemiology of lung cancer in India: Reducing non-small cell lung cancer-not otherwise specified and quantifying tobacco smoke exposure are the key.Indian J Cancer 2017;54:285-290

How to cite this URL:
Kaur H, Sehgal I S, Bal A, Gupta N, Behera D, Das A, Singh N. Evolving epidemiology of lung cancer in India: Reducing non-small cell lung cancer-not otherwise specified and quantifying tobacco smoke exposure are the key. Indian J Cancer [serial online] 2017 [cited 2020 Jul 5 ];54:285-290
Available from:

Full Text


Lung cancer (LC) is the leading cause of cancer-related deaths in India and globally. Tobacco smoking is the most important risk factor for LC in both genders. Bidi smoking is a more prevalent form of smoked tobacco in India than cigarette smoking.[1] In the recent past, adenocarcinoma incidence has increased globally replacing squamous cell carcinoma (SqCC) as the most prevalent type in developed countries. However, in several parts of North/East India, SqCC is still the most common form of LC.[1],[2],[3] Different regions of India also vary substantially in relation to age, gender, histology, smoking profile, and disease stage distribution.[4] The present study was conducted to investigate if there has been any change in the histological distribution among newly diagnosed LC patients at our center, a tertiary care referral institute in North India. We have also compared the epidemiological profile of LC observed in the current study with previously published data from our center.



This was a retrospective analysis of prospectively collected data for all consecutive newly diagnosed adult (≥18 years) patients who were registered in the LC clinic of our institute over a 4-year period (March 2011–February 2015). As per the normal protocol of our LC clinic, all patients were registered only after LC diagnosis was confirmed by histopathological and/or cytological examination of representative tissue specimens at the authors' institute.[1],[5] If necessary, more than one type of diagnostic modalities were used, and relevant procedures performed to obtain tissue for immunochemistry and/or molecular studies.[6],[7],[8] The following information was extracted from the database: demographic profile (age, gender, height/weight, performance status, and body mass index [BMI]), quantified smoking status, and histological type and stage (including presence/extent of extrathoracic metastases, if any). Detailed methodology regarding the above has been published previously and is summarized below.[1],[2],[5],[9] Quantification of smoking was done using smoking index (SI) - defined as the number of bidis + cigarettes smoked/day multiplied by a number of years smoked. Accordingly, patients were categorized into four SI-based groups: never smokers (SI = 0), light smokers (SI = 1–100), moderate smokers (SI = 101–300), and heavy smokers (SI ≥301).[1],[10]

Histological typing and lung cancer staging

Histological classification was done on the basis of morphology and if needed, relevant immunochemistry using the IASLC/ATS/ERS classification of lung tumors.[11] Staging for both nonsmall cell lung cancer (NSCLC) and small cell lung cancer (SCLC) was done using 7th edition of TNM classification based on tumor size and extension (T), lymph nodal involvement (N), and presence of distant metastasis (M).[12] SCLC was also staged as either limited (disease restricted to one hemithorax, with/without regional lymph node metastases) or extensive.[13] A computed tomography (CT) scan of thorax/upper abdomen (including liver and adrenals) was performed in all patients for staging with positron emission tomography-CT being reserved for patients with potentially resectable disease being considered for surgical resection.[5] Patients with symptoms suggestive of metastatic disease underwent relevant investigations (brain magnetic resonance imaging, bone scan, etc.).

Based on Indian guidelines, patients were categorized as normal (BMI = 18.50–22.99), overweight (BMI ≥23.0), mild/moderately underweight (BMI = 16.00–18.49), or severely underweight (BMI <16.0).[9],[14]

Statistical analysis

Statistical analysis was performed using the statistical software SPSS version 22.0 (SPSS Inc., Chicago, IL, USA). Data were expressed as the mean ± standard deviation (SD) or median (interquartile range [IQR]) or as number (percentages). Continuous and categorical variables in two groups were compared using Mann–Whitney U-test and Chi-square/Fisher's exact test, respectively. Results of the current study were compared with a previous studies from our center.[2],[9] P < 0.05 was considered statistically significant.


A total of 1301 LC patients were diagnosed during the study period. Mean and median ages were 58.6 (SD = 10.8) and 60 (IQR = 51–65) years, respectively [Table 1]. Majority were males (n = 1071; 82.3%). SqCC and adenocarcinoma (36.4% each) were the most common histological types. Most patients (83.4%) presented with advanced disease (Stages IIIB/IV). Extrathoracic disease was present in 361 (27.7%). Among 250 SCLC patients, distribution of extensive and limited disease was equal.{Table 1}

Majority (76.9%) were smokers, especially heavy smokers (SI ≥ 301; 51.9%). Significant differences were observed in relation to age, gender, BMI, histology, TNM stage, metastatic disease, and extrathoracic metastasis among SI-based groups [Table 2]. Mean age was highest among heavy smokers. Never smokers were predominantly females (55.5%), and their percentage progressively decreased as SI increased (23.4% in light smokers, 7.8% in moderate smokers, and only 3.1% in heavy smokers). BMI had an inverse relationship with severity of smoking being highest in never smokers and lowest in heavy smokers. Overweight/obese individuals (BMI ≥23 kg/m 2) were most frequent in never smokers (43.2%) and this reduced progressively as SI increased (25.0% in light smokers, 18.9% in moderate smokers, and 15.7% in heavy smokers). Conversely, underweight individuals were highest in heavy smokers and lowest in never smokers (41.2% vs. 20.7%). Even for histological distribution, as SI increased, a progressive decrease in the prevalence of adenocarcinoma (66.8% vs. 37.4% vs. 33.0% vs. 23.9%) accompanied by progressive increase in SqCC (9.6% vs. 32.7% vs. 40.4% vs. 47.7%) was observed. Frequency of SCLC was the lowest in never smokers. There was an inverse relationship of smoking status with disease stage at presentation (Stage IV being highest in never smokers [77.1%] and lowest in heavy smokers [43.6%]). Conversely, Stage IIIB increased progressively as SI increased (17.6% vs. 27.1% vs. 31.2% vs. 35.9%). All groups had <5% patients in Stages I–II. Extrathoracic metastatic disease at presentation was also highest amongst nonsmokers and reduced progressively in light, moderate, and heavy smokers. On subgroup analysis based on histology, stage distribution was significantly different with percentages of Stages I–IIIA vs. IIIB–IV being 24.9% vs. 75.1% in SqCC, 11.6% vs. 88.4% in adenocarcinoma, 20.4% vs. 79.6% in NSCLC-not otherwise specified (NOS)/large cell NSCLC, and 8.8% versus 91.2% in SCLC, P < 0.0001. Histological distribution was consistent during the time period of the current study [Table 3]. The current cohort's (time period 2011–2015) demographic profile was compared with previous (time period 2008–2011) from our center [Table 4].[2],[9] Although there was no change in age/gender distribution and ratio of current/ex-smokers to never smokers (3.3:1 in both cohorts), percentage of heavy smokers among all current/ex-smokers increased significantly from 59.8% to 67.5% in between the two study periods. Median (IQR) SI was also higher (500 [290–800] vs. 400 [240–700]; P = 0.001). Histological distribution also changed significantly (adenocarcinoma increased from 27.5% to 36.4% [P< 0.001]). However, on analyzing SqCC versus all other types (nonsquamous NSCLC and SCLC), statistical significance was not achieved (P = 0.47). Stage distribution could not be compared as different TNM classification schemes were used during the two study periods (6th and 7th edition).{Table 2}{Table 3}{Table 4}


Our continuous evaluation of LC clinico-epidemiological profile at our center has shown several interesting observations. While our initial publication involving 250 patients 5 years back showed no change as compared to the profile observed three decades earlier, the next publication with greater number of patients (>650) showed that quantified tobacco smoke exposure (QTSE) is able to categorize patients into groups that differ significantly among each other in terms of age, gender, histology, stage, and BMI distribution.[1],[2] Going further with the current analysis and with a cohort size of 1300+ patients, we are now able to document two important observations.

First, the observations related to SI-based groups (never, light, moderate, and heavy smokers) in our previous publication have been replicated in the current study. QTSE thus is able to segregate patients with distinct clinico-epidemiological profiles with heavy smokers (SI ≥ 301) characterized by higher age, predominantly male gender and squamous histology, lower BMI, and lesser prevalence of Stage IV/extrathoracic metastatic disease at one end and never-smokers characterized by younger age, predominantly female gender and adenocarcinoma histology, higher BMI, and higher prevalence of Stage IV/extrathoracic metastatic disease at the other end. The strength of the current analysis as compared to previous one is not only a two-fold increase in number of patients but also that SI-based groups continued to show these differences even when the denominator was all histological types and not just NSCLC (latter being the denominator for our previous analysis).[2] This also lends support to SCLC being equally well staged as NSCLC using 7th TNM staging system, something which we are already carrying out in routine clinical practice at our center.[15],[16],[17] Reasons for observed differences among SI-based groups have already been eluded to in detail in our previous publication, wherein we had also performed subgroup analysis and logistic regression analysis to show an independent association of QTSE with observed differences.[1] It is worthwhile to comment here that majority of patients had locally advanced/metastatic disease which is similar to our previously reported observations.[1],[17],[18]

Second, results of the current analysis and its comparison with our previous study reveal a changing trend in histological distribution of LC at diagnosis. Both previous and current cohorts included consecutive patients registered in our LC clinic subsequent to histological/cytological confirmation of diagnosis at our center.[1] Increasing adenocarcinoma prevalence in the current cohort has occurred in conjunction with a reduction in NSCLC-NOS (undifferentiated NSCLC). Numerically, SqCC appeared to be lesser in the current cohort. However, statistical analysis showed no difference even if patients were categorized into three broad groups – SCLC, SqCC, and nonsquamous NSCLC.

While one of the earliest reports (almost 25 years ago) from our center comprising >1000 patients had frequencies of SqCC and adenocarcinoma of approximately 34% and 26%, respectively, it refers to an era wherein histological classification was primarily based on microscopy alone with minimal/no use of immunochemistry.[19] Histological distribution observed was almost similar in a subsequent assessment and comparison published by us 5 years ago.[2] We believed that lack of change in epidemiological profile was attributable to “bidi” manufacturing (a cottage industry) not having changed over time. Bidi smoking is known to provide a concentration of carcinogens similar to that of an unfiltered cigarette, and that is why one bidi is considered to be equivalent to one cigarette for calculating time intensity (smoking pack-years/SI).[20] This hypothesis is to some extent substantiated even in the current study wherein SqCC continues to be >35% in the backdrop of increasing percentage of heavy smokers and similar overall percentage of current/ex-smokers. However, what is clearly different is that unlike our previous papers, frequency of NSCLC-NOS (undifferentiated NSCLC) has reduced successively with time (20% in initial studies [19] to 10% in our last analysis [1] and now down to 5% in the current cohort). The most likely reason for change in histological distribution in the current study, therefore, seems to be an improvement in histological classification. The current WHO classification puts more emphasis on using molecular and immune-histo/cyto-chemical techniques for accurate diagnosis in small biopsy/fine-needle aspiration cytology samples.[11]

The major limitation of the current study is that the patient population herein and consequently analysis was akin to that of a hospital-based cancer registry. Being retrospective in nature and having associated inherent flaws of such a study design, we cannot necessarily extrapolate these observations to population-based data. However, even the available population-based cancer registries in India ( do not include QTSE, and hence it is not feasible to carry out such an analysis from published data therein. The other major limitation is that since this a cross-sectional analysis, we are unable to provide prognostic role of histological subtype and stage similar to what we had done in longitudinal (prospective and retrospective) studies published by us recently.[16],[17],[18],[21],[22]

What are the clinical implications of the current study? The results suggest the persistence of smoking as a risk factor for LC in India with a greater percentage of heavy smokers in comparison to previous studies from our center. Implementation of strict tobacco control measures seems to be the only measure that is likely to reverse the ratio of current/ex-smokers to never smokers and also to prevent and reduce LC incidence. The current study also reinforces the need for reporting pathologists and cytologists to use immunochemistry to accurately classify LC histological type as per guidelines and as published by us previously; frequency of NSCLC-NOS continues to be lowest among major cancer centers in India – a country wherein geographical diversity is manifest even in a disease such as LC.[4]


Time trends indicate a change in histological distribution among newly diagnosed LC patients presenting to our center with adenocarcinoma becoming as prevalent as SqCC and this seems related to reduced prevalence of NSCLC-NOS achieved by better diagnostic accuracy of histological classification. Increasing percentage of heavy smokers observed in the current study and significant differences among SI-based groups for several key demographic characteristics (age, gender, BMI, histology, and stage) indicate need to incorporate QTSE during epidemiological assessment of LC.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Singh N, Aggarwal AN, Gupta D, Behera D, Jindal SK. Quantified smoking status and non-small cell lung cancer stage at presentation: Analysis of a North Indian cohort and a systematic review of literature. J Thorac Dis 2012;4:474-84.
2Singh N, Aggarwal AN, Gupta D, Behera D, Jindal SK. Unchanging clinico-epidemiological profile of lung cancer in north India over three decades. Cancer Epidemiol 2010;34:101-4.
3Dey A, Biswas D, Saha SK, Kundu S, Kundu S, Sengupta A. Comparison study of clinicoradiological profile of primary lung cancer cases: An Eastern India experience. Indian J Cancer 2012;49:89-95.
4Singh N, Behera D. Lung cancer epidemiology and clinical profile in North India: Similarities and differences with other geographical regions of India. Indian J Cancer 2013;50:291.
5Singh N, Aggarwal AN, Behera D. Management of advanced lung cancer in resource-constrained settings: A perspective from India. Expert Rev Anticancer Ther 2012;12:1479-95.
6Singh N, Bal A, Aggarwal AN, Das A, Behera D. Clinical outcomes in non-small-cell lung cancer in relation to expression of predictive and prognostic biomarkers. Future Oncol 2010;6:741-67.
7Maturu VN, Singh N, Bal A, Gupta N, Das A, Behera D. Relationship of epidermal growth factor receptor activating mutations with histologic subtyping according to International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society 2011 adenocarcinoma classification and their impact on overall survival. Lung India 2016;33:257-66.
8Bal A, Singh N, Agarwal P, Das A, Behera D. ALK gene rearranged lung adenocarcinomas: Molecular genetics and morphology in cohort of patients from North India. APMIS 2016;124:832-8.
9Singh N, Aggarwal AN, Gupta D, Behera D. Prevalence of low body mass index among newly diagnosed lung cancer patients in North India and its association with smoking status. Thorac Cancer 2011;2:27-31.
10Jindal SK, Malik SK, Dhand R, Gujral JS, Malik AK, Datta BN. Bronchogenic carcinoma in Northern India. Thorax 1982;37:343-7.
11Travis WD, Brambilla E, Noguchi M, Nicholson AG, Geisinger KR, Yatabe Y, et al. International Association for the study of Lung Cancer/American Thoracic Society/European Respiratory Society International Multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 2011;6:244-85.
12Vallières E, Shepherd FA, Crowley J, Van Houtte P, Postmus PE, Carney D, et al. The IASLC Lung Cancer Staging Project: Proposals regarding the relevance of TNM in the pathologic staging of small cell lung cancer in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol 2009;4:1049-59.
13Stahel RA, Ginsberg R, Havemann K, Hirsh FR, Ihde DC, Jassem J, et al. Staging and prognostic factors in small cell lung cancer: A consensus report. Lung Cancer 1989;5:119-26.
14Misra A, Chowbey P, Makkar BM, Vikram NK, Wasir JS, Chadha D, et al. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian Indians and recommendations for physical activity, medical and surgical management. J Assoc Physicians India 2009;57:163-70.
15Goldstraw P. International-Association-for-the-Study-of-Lung-Cancer Staging Manual in Thoracic Oncology. 1st ed. Orange Park, FL, USA: Editorial Rx Press; 2009.
16Singh PS, Aggarwal AN, Behera D, Kapoor R, Singh N. Simplified graded baseline symptom assessment in lung cancer patients undergoing first-line chemotherapy: Correlations and prognostic role in a resource constrained setting. J Glob Oncol 2017;3:54-63.
17Singh N, Singh PS, Aggarwal AN, Behera D. Comorbidity assessment using charlson comorbidity index and simplified comorbidity score and its association with clinical outcomes during first-line chemotherapy for lung cancer. Clin Lung Cancer 2016;17:205-13.e1.
18Singh N, Mootha VK, Madan K, Aggarwal AN, Behera D. Tumor cavitation among lung cancer patients receiving first-line chemotherapy at a tertiary care centre in India: Association with histology and overall survival. Med Oncol 2013;30:602.
19Jindal SK, Behera D. Clinical spectrum of primary lung cancer: Review of Chandigarh experience of 10 years. Lung India 1990;8:94-8.
20Gupta D, Boffetta P, Gaborieau V, Jindal SK. Risk factors of lung cancer in Chandigarh, India. Indian J Med Res 2001;113:142-50.
21Singh N, Aggarwal AN, Behera D, Jindal SK. Intercycle delays during chemotherapy of non-small cell lung cancer in a health care resource-constrained setting and their effect on overall survival. J Thorac Oncol 2010;5:236-9.
22Singh N, Aggarwal AN, Kaur J, Behera D. Association of graded folic acid supplementation and total plasma homocysteine levels with hematological toxicity during first-line treatment of nonsquamous NSCLC patients with pemetrexed-based chemotherapy. Am J Clin Oncol 2017;40:75-82.