Clinical spectrum of lung cancer across hospitals under the National Cancer Registry Programme in India: Challenges and opportunities

Abstract

Background and Objective:

Lung cancer is one among the top five cancers in India, both in incidence and mortality and is thus, a significant public health challenge. The economic disparities among nations significantly contribute to differences observed in the management of lung cancer.

Methods:

This study analysed the clinical spectrum of lung cancer from several hospitals using data from the National Cancer Registry Programme concerning demographic characteristics of patients, histological variants, and diagnostic and management practices between 2012 and 2019. For this descriptive study, data was drawn from 96 Hospital-Based Cancer Registries. Altogether, all cases classified under ICD-10, C34.90 were included in this study.

Results:

The study findings revealed that most lung cancer cases occurred in males aged 50–74 years and females aged 45–69 years. Adenocarcinoma were the most common (39.7%) variety, almost twice higher than squamous cell carcinoma subtypes. The majority (50.7%) of the patients with lung cancer were detected with distant metastases. Low rates of surgery were observed among the patients who had localised/locoregional spread, while one third of the patients diagnosed at another facility reported to the treating hospital between 7 to 30 days after diagnosis.

Conclusion:

This study highlights that delay in referral and subsequent treatment initiation are critical challenges in lung cancer care, including delayed diagnosis, limited treatment options, and a lack of streamlined referral processes. The study findings will be crucial for identifying the gaps in care and guiding strategies to improve early diagnosis and treatment of lung cancer.

INTRODUCTION

Lung cancer is the second most common cancer and the most common cause of cancer-related deaths worldwide.[1] In India, lung cancer has been recognised as one of the top five cancers both in terms of incidence and death, being the most common cause of cancer-related mortality in males.[2,3] The disability-adjusted life years (DALYs) of lung cancer among males in India is 183.3 per 100,000, contributing 13.5% to total cancer DALYs in 2016.[4] Economic disparities between countries primarily drive inequalities in lung cancer management. Being a diverse country in terms of the health-seeking behaviour of patients as well as the distribution of the facilities to diagnose and treat this lethal disease, the pattern of lung cancer treatment also varies, reflecting the availability and accessibility of the care facilities.

Although few studies from India have tried to describe the diagnostic and therapeutic aspects according to their institutional experiences,[5,6,7,8] there is a need for large-scale data representing a broad nationwide scenario related to the methods of diagnosis and clinical management of lung cancer in India. Registry data is a vital tool for improving the understanding, management, and outcomes of lung cancer treatment. It supports a data-driven approach to healthcare that benefits patients, healthcare providers, and policymakers alike.

The National Centre for Disease Informatics and Research (NCDIR), under the aegis of the Indian Council of Medical Research (ICMR), has been implementing the National Cancer Registry Program (NCRP) over more than four decades, collecting data sourced through various cancer registries distributed all over the country. This national-level data regularly provides cancer burden estimates and projections. In the past two decades, lung cancer demographics have progressively shifted, with a rising number of cases among younger patients, women, and cases of lung adenocarcinomas. However, a timely diagnosis remains crucial for improving disease outcomes, and minimising delays in specialist referrals is vital for early detection. The current analysis aims to describe the clinical profile of lung cancer in terms of patient demographics, histological variants, and diagnostic and management practices across different hospitals in India and accordingly highlight the challenges in lung cancer care. This information would be critical for identifying opportunities for developing targeted strategies that address the rising incidence of lung cancer in India.

MATERIALS AND METHODS

In India, a network of Population Based Cancer Registries (PBCRs) and Hospital Based Cancer Registries (HBCRs) have been compiling and providing cancer-related data since 1981. This descriptive study’s primary cancer diagnosis and management data source was collated from 96 HBCRs nationwide for 2012-19.

Data collection at a cancer registry is conducted by using case abstraction from patient records and investigation reports by trained investigators under the guidance of a cancer-treating physician. The abstracted data is recorded on a standardised core form and electronically transmitted to the central coordinating unit at ICMR-NCDIR. All the reported cases with the International Classification of Diseases, version 10 of C34.90, were considered eligible for inclusion in this analysis. The collated data was verified to ensure its completeness, and duplicates were removed using an in-house computer application for data capture.

This study essentially involved descriptive analysis of lung cancer parameters related to age and gender distribution, methods of diagnosis, broad histological types, the clinical extent of disease at presentation, the intention to treatment and treatment modality and the time between first diagnosis and commencement of treatment at the reporting institutions (RIs), i.e., HBCRs. Histologies were classified into broad groups based on ICD for oncology, 3^rd edition. The clinical extent was classified as follows: localised (cancer restricted to the primary site), locoregional (direct extension/lymph nodes) and distant metastasis.

Statistical analysis

Descriptive analysis using frequencies and percentages was used to demonstrate the study results.

Ethics clearance

This study was approved by the Institutional Ethics Committee, ICMR-NCDIR; No. NCDIR/IEC/2017/5, dated March 1, 2017.

RESULTS

Demographics and methods of diagnosis of lung cancer

A total of 45,228 patients (34,395 males and 10,833 females, male to female ratio 3:1) with lung cancer were included in this study. Nearly three-fourths of the lung cancer cases among males were diagnosed between 50–74 years of age, with incidence peak at 60–64 years. Most cases in females were observed to present 5 years earlier than in males, with the majority (69.4%) occurring between 45 and 69 years [Figure 1].

Figure 1.

Distribution of lung cancer patients by age and gender from 96 HBCRs in 2012–19

Microscopic tissue examination (97.8%) was the most frequently employed diagnostic method in both sexes. Other diagnosis methods, such as imaging studies or clinical-based diagnosis, contributed to the remaining 2.2% of techniques for final diagnosis. Among those detected microscopically, confirmation of cancer by tissue diagnosis from the primary tumour was possible in 69.9% of patients, which was the most common method. Using primary tumour cytology as a primary cancer detection tool contributed to up to one-sixth of the cases (16.2% in males and 17.8% in females), while tumour characterisation from metastases using either histology or cytology was used in 7%.

Major histological types and disease extent

Histological analysis of lung cancers suggested epithelial tumours as the prime histologic type (85.2%), followed by carcinoma not otherwise specified [Table 1].

Table 1.

Number (n) and proportion (%) of broad histological classification by gender for Lung cancer from 96 HBCRs in 2012–19

Broad histological classification		Males				Females				Total
		n		%		n		%		n		%
Epithelial tumours
Adenocarcinomas		11801		35.1		5745		54.2		17546		39.7
Squamous cell carcinoma		7844		23.3		1192		11.3		9036		20.4
Non-small cell carcinoma, NOS		5280		15.7		1229		11.6		6509		14.7
Small cell carcinoma		3342		9.9		618		5.8		3960		9.0
Other neuroendocrine tumours		434		1.3		171		1.6		605		1.4
Carcinoma, NOS		1848		5.5		530		5.0		2378		5.4
Mesenchymal tumours		91		0.3		58		0.6		149		0.3
Tumours of ectopic origin
Germ cell tumours		5		<0.1		1		<0.1		6		<0.1
Others*		2979		8.9		1050		9.9		4029		9.1
Total		33624		100		10594		100		44218		100

HBCRs=Hospital Based Cancer Registries, NOS=Not otherwise specified. *Include malignant tumours: non-specific histology, small cell type, clear cell type, fusiform cell type, PNET

Within the epithelial subtype, adenocarcinomas (ACC) were the most common (39.7%) variety, almost twice higher than squamous cell carcinoma (SQCC) subtypes. Adenocarcinoma was the predominant histology (54.2%) among females, whereas in males, adenocarcinoma and SQCC were present in 35.1% and 23.3%, respectively. Other subtypes were similar in distribution among both sexes.

Small cell and other neuroendocrine varieties were observed in less than 10% of cases among epithelial tumours, while mesenchymal and germ cell tumours were observed in less than 0.5%.

The majority of the patients with lung cancer were detected with distant metastases followed by locoregional and localised extent [Figure 2]. This pattern was nearly similar for both males and females [Figure 2] and across different education level groups [Supplementary Table 1]. Based on the available tumour node metastasis (TNM) staging data, nearly a quarter (23%) of the patients were diagnosed with Stage 4 lung cancer. The most frequently identified sites of metastases were the brain, bones, articular cartilage, and liver.

Figure 2.

Distribution of clinical extent of disease (%) of Lung cancers among males and females in 96 HBCRs in 2012–1

Supplementary Table 1.

Distribution of Clinical Extent of Disease Before Treatment (CEDBT) according to education

Education		Localised				Locoregional				Distant metastasis				Unknown				Total
		n		%		n		%		n		%		n		%		n		%
Illiterate		1375		14.0		3830		39.0		4602		46.9		9		0.1		9816		100.0
Literate		888		20.3		1732		39.7		1747		40.0		1		0.0		4368		100.0
Primary/ Middle/ Secondary		1613		9.7		5909		35.5		9120		54.7		19		0.1		16661		100.0
Technical - after matric/ College and above		496		15.4		894		27.8		1815		56.4		14		0.4		3219		100.0
Unknown		2126		19.6		3222		29.7		5511		50.7		3		0.0		10862		100.0
Total		6498		14.5		15587		34.7		22795		50.7		46		0.1		44926		100.0

Treatment paradigms in lung cancer

Information about the intention of treatment was known in almost 90% of patients. As illustrated in Figure 3, most patients with localised and locoregional disease were intended to receive radical treatment (77.9% and 61.4%, respectively). At the same time, a fifth of them were planned to be managed with palliative modality. About one-fifth of the patients with distant metastasis were intended to receive palliative treatment.

Figure 3.

Distribution of Intention to treat by clinical extent of disease (%) of patients with Lung cancers among both sexes in 96 HBCRs in 2012–19

Actual treatment details provided for lung cancer were available, with missing data for <1% of cases. Chemotherapy was the primary modality of treatment for most patients, followed by a combination of chemotherapy and radiotherapy across all clinical extents of the disease in both sexes [Figure 4a and b].

Figure 4.

(a) Distribution of type of treatment according to the clinical extent of disease (%) for Lung cancer in 96 HBCRs for the year 2012–19—Males. (b) Distribution of type of treatment according to the clinical extent of disease (%) for Lung cancer in 96 HBCRs for the year 2012–19—Females

Time between diagnosis, first attendance and treatment commencement

The diagnosis of lung cancer occurred in two possible ways: i) Diagnosis took place at another facility, but referral/administration of treatment was provided at the RI where there was an HBCR (Group 1), ii) Both diagnoses and treatment took place at RI (Group 2).

A total of 24,618 patients were in the first category. For this group of patients, the time taken between diagnosis of lung cancer at a health facility other than the RI and attendance at the RI for cancer-directed treatment was calculated. Most patients (over 80%) reached the RI within 30 days of diagnosis (Median time of 9 days, IQR 4–20) from the first institute, irrespective of the stage of disease, while an average of patients 11% reached the RI within two months after diagnosis [Figure 5a, Supplementary Table 2]. Figure 5b shows the time between diagnosis and treatment commencement for patients belonging to Group I. Less than a quarter of the patients received cancer-directed treatment within a week of diagnosis, whereas almost 40% of the patients had their treatment initiated between 1 and 4 weeks from diagnosis (median time 24 days, IQR: 11–45 days), shown in Supplementary Table 2. This pattern remained similar across all the clinical extents of lung cancer.

Figure 5.

(a) Time between diagnosis and first attendance at the reporting institution for group 1. (b) Time between first diagnosis and commencement of cancer-directed treatment at the Reporting Institution for group 1. (c) Time between first diagnosis and commencement of cancer-directed treatment at Reporting Institution for group 2. Note: Group 1: Diagnosis took place in another facility but referred/availed the treatment at the Reporting Institution, Group 2: Both diagnosis and treatment took place at RI

Supplementary Table 2.

Median time with Interquartile range for Group 1 patients according to Clinical Extent of Disease Before Treatment (CEDBT)

CEDBT		No. of Days - Median (IQR)
Time between diagnosis and first attendance at reporting institution
Localised		8 (4, 20)
Locoregional		10 (4, 21)
Distant metastasis		9 (4, 19)
Total		9 (4, 20)
Time between first diagnosis and commencement of cancer directed treatment
Localised		21 (9, 42)
Locoregional		25 (12, 48)
Distant metastasis		23 (11, 43)
Total		24 (11, 45)

Figure 5c presents the time between diagnosis and treatment commencement for group 2. This group comprised 20280 lung cancer patients. Compared to Group I patients, a higher proportion of Group II patients across all disease extents were initiated on treatment within a week after diagnosis (median time: 15 days, IQR 4–33 days), shown in Supplementary Table 3. The median waiting times did not display much variation across patients with differing levels of education [Supplementary Tables 4 and 5]. The median waiting time appeared to be higher for surgery as compared to other treatment modalities [Supplementary Tables 6 and 7].

Supplementary Table 3.

Median time with Interquartile range for Group 2 patients according to Clinical Extent of Disease Before Treatment (CEDBT)

CEDBT		No. of Days - Median (IQR)
Time between first diagnosis and commencement of cancer directed treatment
Localised		10 (0, 32)
Locoregional		17 (4, 37)
Distant metastasis		16 (6, 32)
Unknown		24 (12, 42)
Total		15 (4, 33)

Supplementary Table 4.

Median time with Interquartile range for Group 1 patient according to education

Education Grouped		Median (IQR)
Time between diagnosis and first attendance at reporting institution
Illiterate		9 (4, 22)
Literate		9 (4, 22)
Primary/ Middle/ Secondary		9 (4, 20)
Technical - after matric/ College and above		9 (4, 18)
Unknown		9 (5, 18)
Total		9 (4, 20)
Time between first attendance and commencement of cancer directed treatment
Illiterate		7 (0, 22)
Literate		5 (0, 17)
Primary/ Middle/ Secondary		12 (2, 29)
Technical - after matric/ College and above		11 (2, 25)
Unknown		4 (0, 12)
Total		8 (1, 23)
Time between first diagnosis and commencement of cancer directed treatment
Illiterate		24 (11, 46)
Literate		20 (8, 42)
Primary/ Middle/ Secondary		28 (14, 50)
Technical - after matric/ College and above		26 (13, 43)
Unknown		18 (10, 33)
Total		24 (11, 45)

Supplementary Table 5.

Median time with Interquartile range for Group 2 patients according to education

Both Diagnosed and Treated at RI
Time between first diagnosis and commencement of cancer directed treatment
Education Grouped		Median (IQR)
Illiterate		14 (1, 33)
Literate		4 (0, 20)
Primary/ Middle/ Secondary		22 (11, 40)
Technical - after matric/ College and above		20 (10, 35)
Unknown		10 (2, 26)
Total		15 (4, 33)

Supplementary Table 6.

Median time between diagnosis and treatment initiation, with Interquartile range for Group 1 patients according to different treatment modalities

Treatment		No. of Days - Median (IQR)
Any Surgery		30 (16, 54)
Any Radiotherapy		25 (12, 48)
Any Chemotherapy		23 (11, 43)
Any targeted therapy		25 (10, 65)
Total		24 (11, 45)

Supplementary Table 7.

Median time between diagnosis and treatment initiation, with Interquartile range for Group 2 patients according to different treatment modalities

Treatment		No. of Days - Median (IQR)
Any Surgery		22 (4, 47)
Any Radiotherapy		16 (5, 34)
Any Chemotherapy		16 (5, 33)
Any targeted therapy		10 (1, 26)
Total		16 (5, 34)

DISCUSSION

This study comprehensively describes the clinical profile of 45,228 lung cancer patients who received care at 96 nationwide hospitals between 2012 and 2019. We observed that the majority of lung cancer cases occurred at the age of 50–74 years in males and 45–69 years in females, which is almost a decade earlier than in the Western world.[8] In India, the younger age at diagnosis of lung cancer may be attributed to a combination of the country’s population demographics and distinct regional risk factors, such as air pollution and genetic mutations, which increase the likelihood of developing lung cancer unrelated to smoking.[9,10] Most cases in females were observed to present 5 years earlier than in males, similar to findings from previous studies, which warrants further understanding.[11,12] The use of smoked tobacco, including both cigarettes and beedis, is the primary risk factor for lung cancer in Indian men.[13] The relative risk of developing lung cancer is observed to be 2.64 times higher for those who smoke Beedis and 2.23 times more for those who smoke cigarettes, with an overall smoking-related risk of 2.45.[14] The National Noncommunicable Disease Monitoring Survey (NNMS) 2017-2018 reported that the current smoked tobacco use rate was 23% among males.[15] The average age of initiation of tobacco use was 21 years, which is a contributory factor to the increased burden of cancers associated with tobacco use in India.[15]

Adenocarcinoma (ACC) and SQCC were the two most common types of lung cancers in our analysis, with the proportion of adenocarcinomas being higher. Globally, ACC is the most common variant, followed by SQCC, and is more prevalent among non-smokers.[16,17] Several Indian studies have reported a similar distribution recently.[7,18] Although ACC was more frequent than SQCC among males, the proportion was higher among females, the reasons for which are not precisely known. The prevalence of smoking in females in India (1.3%) is much lower than in males, and other factors like exposure to second hand tobacco smoke over the past 30 days in more than a third of the females (37.5%) may have a role to play.[15] Numerous studies conducted in developing nations have substantiated a heightened risk of lung cancer among women who are exposed to biomass fuel and engaged in cooking activities.[19,20] In this context, the combustion by-products include potential carcinogens such as 1,3-butadiene, cyclopentane/cd/pyrene and dibenz/a, h/anthracene.[21]

Histology of the primary tumour or metastases was the foremost way to derive a diagnosis in three-fourths of the patients in our registry, followed by cytologic evaluation of the primary site. This parallels the data from major cancer centres in India, where microscopic diagnosis was used in nearly all the patients.[18] Although a chest radiograph is typically the first diagnostic test performed in suspected cases of lung cancer, the high prevalence of tuberculosis (TB) and the overlap in clinical and radiological features often result in misdiagnosis.[22,23,24] This underscores the critical need for timely and definitive pathological confirmation, while also highlighting the potential value of novel biomarkers in supporting early and accurate diagnosis of lung cancer.

Most lung cancer patients had distant metastasis at the time of diagnosis in our study, which is similar to some other hospital-based studies.[18,25] In India, late detection of lung cancer is common, which could be attributed to a tendency for an empirical diagnosis of tuberculosis infection at primary and secondary level centers (as many of the symptoms and X-ray appearances are overlapping), with higher rates observed in low- and middle-income countries (LMICs).[26] The study conducted in Mysore found that lung cancer patients who received ATT had higher median diagnostic delay of 57.09 days compared to 25.01 days in those who did not receive ATT.[27] Similarly, a study from Delhi reported that patients with a history of tuberculosis treatment took significantly longer to initiate chemotherapy than those without such a history 187 [134–261.5] days vs. 113 [75–180] days.[28] While dedicated tuberculosis centres in district hospitals have diagnostic facilities focused on the diagnosing tuberculosis, differentiating and establishing a lung cancer diagnosis is not often possible at these centres. Establishing bronchoscopy and image-guided lung biopsy services in all district-level hospitals may boost the diagnostic output. There also exists a need to sensitize and raise awareness among primary and secondary care physicians through mass campaigning about lung cancer causes and symptoms. Using electronic media may go a long way toward early detection by affecting the health-seeking behaviour of the public.

Concerning treatment, one noteworthy finding is low surgery rates among patients with localised/locoregional spread. Surgery remains the most effective treatment for early-stage lung cancer.[29] Previous studies have shown that individuals diagnosed at a thoracic surgical centre may be up to 51% more likely to undergo thoracic surgery for lung cancer.[30] Disparities in demographic and socioeconomic factors such as availability of healthcare, timely access to treatment, income and insurance status also impact the receipt of surgery as a definitive treatment.[31] Chemotherapy was the mainstay of treatment for all patients, irrespective of the clinical extent of the disease. Chemotherapy for Non-Small Cell Lung Carcinoma (NSCLC) has advanced significantly in recent years, becoming increasingly targeted and personalised for each patient by identifying driver genetic mutations.[32] In LMICs like India, the scarcity of biomarker testing and limited access to newer therapies have hindered the widespread adoption of targeted therapy and immunotherapy.[33] For instance, in a case series of 169 patients with advanced NSCLC who underwent NGS testing at a tertiary care centre in New Delhi, rearranged during transfection (RET) rearrangements were identified in only 2.9% (n = 5) of cases, yet most were treated with conventional chemotherapy due to limited availability of targeted agents.[34] This reflects the broader challenges in integrating precision oncology approaches into routine clinical care in resource-constrained settings. The use of radiotherapy is increasing among patients who refuse/are unfit for surgery. Stereotactic ablative radiotherapy (SABR) has shown survival benefits in the early stages of lung cancer when compared to conventional radiotherapy.[30] About a third of the patients diagnosed at another facility reported to the treating hospital between 7 to 30 days after diagnosis. The initiation of treatment among these patients was also slightly delayed compared to patients diagnosed and treated at the same facility, which points towards delays in referral and timely care seeking. These challenges underline the scope for improvement in the availability and uptake of appropriate treatment modalities, streamlining patient referrals and smooth transition of patient care across different health facilities.

The strength of this study lies in its details on the diagnosis and treatment of lung cancer using multicentric national-level data through a large dataset from 96 hospitals, wherein the data was collected using a standardized core form and subjected to rigorous quality checks. Limitation of this study include the chance of selection bias, as the data were collected from many hospital-based cancer registries distributed almost all over the country, and hospitals were purposively invited to set up cancer registries in their facilities. Additionally, since cancer registry data collection relies on data abstraction from the patient’s medical records, data on factors including co-morbidities and the patient’s performance status that could have influenced treatment decisions could not be analysed.

In conclusion, managing lung cancer in India presents significant challenges and promising opportunities. Key challenges, which are mainly delayed diagnosis, limited treatment options and lack of streamlined referral, also create opportunities for improvement. By leveraging these opportunities, India can make strides toward more effective lung cancer management and improved survival rates.

Data availability statement

The data generated in this study are available upon request from the corresponding author.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

The study was funded by an intramural funding from the Indian Council of Medical Research, New Delhi, and the Ministry of Health and Family Welfare, New Delhi.