This article is mentioned in:

Introduction

Lung cancer is the most common malignancy in men and the second most common among women (1). It remains one of the leading causes of cancer mortality worldwide (2) and is classified broadly into small-cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), the latter being significantly more common and further subdivided into various histological types (2). Numerous advances in lung cancer treatment have been made in recent years, perhaps the most important being molecular testing for NSCLC to establish targeted therapy towards driver mutations (3,4). However, the frequency of these mutations appears to vary significantly in different geographic regions and ethnicities (4). For example, EGFR, PTEN, ALK, ROS1 and RET mutations are predominant in East Asia and are also more commonly identified in females and non-smokers (5,6), while KRAS, TP53, BRAF non-V600E, STK11 and JAK2/3 mutations are more common in smokers (7). Programmed death-ligand 1 (PD-L1), a transmembrane protein involved in immunosuppression, is also expressed in several malignancies, including NSCLC. It binds to programmed cell death 1 (PD-1), inhibiting effector T cells and protects malignant cells from the immune system, hence blocking this phenomenon. In addition, it assists in the death of malignant cells and has become an essential entity of lung cancer treatment. Therefore, mutation testing of PD-L1 expression, has shown promising results in targeted therapy (8). As with other molecular targets, PD-L1 expression has a significant geographic and epidemiologic variation with higher prevalence in East Asia (9) and in females (9).

Data on the prevalence of molecular drivers of NSCLC in the Middle East is scarce, with only one cross-sectional study reporting the EGFR mutation pattern in the region (10). Therefore, in the present study, the histological patterns and molecular drivers of NSCLC in the United Arab Emirates (UAE), were reviewed.

Materials and methods

The electronic health records (EHR) of all patients diagnosed with lung cancer between April 2015 and September 2022 were retrospectively analysed at the Respiratory Institute, Cleveland Clinic Abu Dhabi (Abu Dhabi, UAE) tertiary care hospital. Approval (REC approval number A-2018-006) for the study was obtained from The Cleveland Clinic Abu Dhabi Research Ethics Committee (Abu Dhabi, UAE) and patient consent was waived due to the retrospective nature of the present study.

Study population

EHRs were searched to locate all patients diagnosed with lung cancer. Patients met inclusion criteria if they were adults and had a new diagnosis of lung cancer. While the patients with SCLC and those without pathology results were excluded from the final analysis. The data were collected between May 2019 and December 2022.

Study variables

Recorded data points included demographics (age, sex, ethnicity, smoking status, height and weight) and tumour characteristics (focality, type, histological grade, visceral/lymphatic invasion, clear margins, TNM classification, staging, tumour mutations and PD-L1 expression).

Data analysis

Quantitative variables are expressed as the mean and standard deviation for normally distributed data and the median and interquartile range for all other data. Categorical variables are expressed as numbers and percentages. Statistical comparisons between continuous characteristics were carried out using an unpaired t-test while comparison for categorical variables was performed thorough chi-squared test, and P<0.05 was considered to indicate a statistically significant difference. The data analysis was conducted using MS Excel 2019 (Microsoft Corp.). Logistical regression model was created to study the relationship between EGFR and other categorical variables using R-studio version 23.12.0 (Posit Software).

Results

A total of 224 patients were noted to have a diagnosis of lung cancer on a search of EHR; 32 were excluded (Fig. 1) due to incomplete records or diagnosis other than NSCLC, and hence, 192 patients (138 males and 54 females) were included in the final analysis. The mean age of patients was 66.3 years (std deviation, ±12.52), and the mean BMI was 26.5 (std deviation, ±5.56). A total of 155 patients (81%) were either current or ex-smokers, while 19% had never smoked. The baseline characteristics and results are shown in Table I. Adenocarcinoma was the most common type of lung cancer and accounted for 134 patients (70%), followed by squamous cell cancer in 47 (24%) patients, while large cell lung cancer was noted in only 5 (3%) patients.

Figure 1 Lung cancer cohort flow diagram. After the exclusion criteria, 192 were included in the final cohort analysis. SCLC, small-cell lung carcinoma.

Table I Baseline characteristics and results for lung cancer cohort.

Table I

Baseline characteristics and results for lung cancer cohort.

Variables	n	%	Mean	Std dev
Baseline characteristics
Age	192		66.3	12.52
BMI	192		26.5	5.56
Male	138	72
Female	54	28
Current or ex-smoker	155	81
Tumour type
Adenocarcinoma	134	70
Squamous cell carcinoma	47	24
Large cell carcinoma	5	3
Other	6	3
Tumour characteristics
Visceral pleural invasion	12	6.3
Lymphovascular invasion	13	6.8
Positive margins on surgical biopsy	8	6.3
Cavitation on CT	13	6.9
Ground glass area present on CT	32	16.8
Endobronchial involvement	40	21.3
Molecular markers
EGFR	29	15
BRAF	3	1.6
ALK1	8	4.2
ROS1	2	1
MET	1	0.5
KRAS	16	8.3
RET	0	0
PD-L1	56	24.7

[i] Std dev, standard deviation; BMI, body mass index; CT, computed tomography; ALK1, activin A receptor like type 1; ROS1, ROS proto-oncogene 1, receptor tyrosine kinase; MET, MET proto-oncogene, receptor tyrosine kinase; RET, Ret proto-oncogene; PD-L1, programmed death-ligand 1.

In terms of tumour focality, 138 patients (71.8%) had a single tumour, and 24 patients (12.5%) had separate tumour nodules of the same histopathologic type. The tumour focality breakdown is shown in Fig. 2. Diagnosis in most of the patients was performed using bronchoscopy (108 patients); 45 were diagnosed by CT-guided biopsy and 39 patients by surgery. A total of 106 (55%) patients had stage IV cancer (stage IVA, 53% and stage IVB, 47%), followed by 43 (22%) with stage III and 33 (18%) with stage I, and 10 (5%) patients had stage II cancer. There was a slight reduction in the number of patients in stage I, with a slight increase in stage II and III on pathological staging, while the number of patients with stage IV remained the same (Fig. 3).

Figure 2 Cohort breakdown according to tumor foci and histopathological types, with 70% diagnosed with a single tumour, and 13% had separate tumour nodules of the same histopathological type.

Figure 3 Lung cancer stages in patients by initial clinical and final pathological classification. There was a slight reduction in the number of patients in stage I, with a slight increase in stage II and III as regards pathological staging, while the number of patients with stage IV remained the same.

PD-L1 expression was present in 56 (24.7%) patients and was expressed in >50% of the tumour cells in 27 (12%) patients. Of the seven mutations tested, EGFR mutations were the most common and were detected in 29 (15%) patients, followed by KRAS in 16 (8.3%) patients, activin A receptor like type 1 (ALK1) in 8 (4.2%) patients, while BRAF, ROS proto-oncogene 1, receptor tyrosine kinase (ROS1) and MET proto-oncogene, receptor tyrosine kinase (MET) were present in 3 (1.6%), 2 (1%) and 1 (0.5%), respectively (Table I). Ret proto-oncogene (RET) mutation was not detected in any patients. Distribution of molecular markers is shown in Fig. 4. A total of 22 out of 29 patients with EGFR mutation were non-smokers, which was statistically significant (χ2=22.2, dof 1; P<0.005) (data not shown). Conversely, only 27 out of 56 PD-L1-positive patients were non-smokers (χ2=0.21, dof 1; P=0.64). No difference was observed with regard to the age of PD-L1-positive and -negative patients (mean age, 66.7 and 64.6, respectively; P=0.479) (data not shown).

Figure 4 Distribution of molecular mutations in a lung cancer cohort. PD-L1 expression was the most common mutation, followed by EGFR mutations and a smaller number of other mutations. PD-L1, programmed death-ligand 1; ALK1, activin A receptor like type 1; ROS1, ROS proto-oncogene 1, receptor tyrosine kinase; MET, MET proto-oncogene, receptor tyrosine kinase; RET, Ret proto-oncogene.

There was no significant difference between EGFR-positive or negative patients in terms of age (mean age, 67.2 and 64.6, respectively; P=0.168) (data not shown). Logistical regression also showed a significantly reduced odds ratio for male sex and smoking but and increased odds ratio for adenocarcinomas. The results obtained regarding the relationship between, age and tumour stage were not significant. The results of logistical regression are detailed in Table II.

Table II Results of logistical regression, detailing the relationship between EGFR positivity and dependant variables.

Table II

Results of logistical regression, detailing the relationship between EGFR positivity and dependant variables.

Characteristic	OR	95% CI	P-value
Age	0.98	0.95, 1.02	0.4
Sex M	0.25	0.1, 0.64	<0.005
Smoker	0.24	0.09, 0.61	<0.005
Stage 3,4	1.07	0.05,7.45	0.4
Adenocarcinoma	0.08	0, 0.41	0.016

[i] OR, odds ratio; CI, confidence interval.

Discussion

The present study investigated a cohort of 192 cases of NSCLC at a tertiary care hospital in the UAE. The population of the present study comprised 138 males and 54 females, with an average age of 66.3 years. Notably, the mean age of the population of the present study was older than previously published studies in the Gulf and Asian region (10-12). A significant majority of the patients, 81%, had a history of smoking, while 19% had never smoked. The prevalence of adenocarcinoma was notably high, accounting for 70% of the cases, followed by squamous cell carcinoma (24%) and large cell lung cancer (3%). The prevalence of smoking and the type of NSCLC was similar to another reported study in the region (10). Tumour focality, stage at diagnosis and modality of diagnosis in the present study was similar to another study from the region (10). No correlation between the tumour focality and the cancer stage was established, which appears counterintuitive. However, the data is heavily skewed towards stage III and IV cancer with very few patients having early-stage disease. Similarly, most of the patients had only a single tumour or satellite nodules in the same lobe. This discrepancy is one of the limitations of the present study and larger datasets would be more appropriate to show such a relationship.

Molecular characteristics

EGFR mutations were detected in 15% of the population of the present study. Significant heterogenicity exists in the frequency of EGFR mutations in different ethnicities, with East Asian populations showing greater mutations in comparison to European populations (13). The EGFR gene is located on the short arm of chromosome 7. It is 200-kb long and contains 28 exons encoding for the EGFR protein that contains 1,210 amino acids (14). Mutation in this gene is usually associated with adenocarcinoma, Asian ethnicity and light-smoking but not with sex. There are several variant subtypes of EGFR mutations most of which are located on exons 18-21 with exon 19 mutations being the most common. Exon 19 deletions are more commonly associated with the male sex while exon 21 deletions appear to be prevalent in females (15).

In the population of the present study, the mutation frequency was lower than in the East Asian study but higher than the reported frequencies in the European study (14,15). The results obtained in the present study were similar to a recent systematic review, which revealed an overall prevalence of 17% for EGFR mutations in the Middle East and North African region; however, they reported significant variations within countries ranging from 11-30% (16). In the present study, EGFR mutations were significantly more frequent in non-smokers, with 22 out of 29 patients with EGFR positive mutations falling into this category. These findings are well reported in a number of studies (16-18); in particular, the meta-analysis by Ren et al (18) which indicated that the odds ratio for EGFR mutation in non-smokers was 4.8 compared with smokers.

PD-L1

The present study revealed that PD-L1 expression was present in 24.7% of the patients. The prevalence of PD-L1 expression in the present study was similar to data published in the only other study examining PD-L1 expression from our region (19); however, both studies are corroborated by Dietel et al (20) in their large multicentre study investigating PD-L1 expression in 18 different countries. PD-L1 positivity did not show a significant association with smoking status, as 27 out of 69 PD-L1-positive patients were non-smokers.

Other mutations

KRAS mutations were present in 8.3% of patients, followed by ALK1 in 4.2%. Other mutations, such as BRAF, ROS1 and MET, were observed in smaller proportions, while no RET mutations were detected. Τhe findings of the present study offer important insights into the epidemiological, clinical, and molecular characteristics of NSCLC in the UAE. The data also revealed presentation with advanced stage, which underlines the need for early detection and intervention strategies. The molecular analysis provided valuable information about the prevalence of genetic mutations in patients with NSCLC. EGFR mutations were notably frequent and were strongly associated with non-smoking status. PD-L1 expression, on the other hand, did not show a significant association with smoking. Major limitations of the present study included the retrospective design. Furthermore, the molecular testing was at the discretion of the multidisciplinary tumour board. In conclusion, the present study contributes to the understanding of NSCLC by providing a comprehensive overview of patient demographics, tumour characteristics, and molecular profiles in the UAE. Further research is warranted to explore the clinical implications of these findings and can serve as a guide for future research, clinical decision-making, and treatment approaches for patients with NSCLC in the UAE.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

IS conceptualized and designed the manuscript. SI, NK, MU, ZZ, SS and AW collected the data. IS, NK, MU analyzed the data. NK, MU, ZZ, SS and AW performed the critical review of the manuscript. IS and SI drafted the manuscript. SI and AW confirm the authenticity of all the raw data. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

The present study was approved (REC approval number A-2018-006) by The Cleveland Clinic Abu Dhabi Research Ethics Committee (Abu Dhabi, UAE) and patient consent was waived due to the retrospective nature of the present study.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References