EGFR mutation incidence in non-small-cell lung cancer of adenocarcinoma histology: a systematic review and global map by ethnicity (mutMapII)

Abstract

Mutations in the epidermal growth factor receptor (EGFR) gene are commonly observed in non-small-cell lung cancer (NSCLC), particularly in tumors of adenocarcinoma (ADC) histology (NSCLC/ADC). Robust data exist regarding the prevalence of EGFR mutations in Western and Asian patients with NSCLC/ADC, yet there is a lack of data for patients of other ethnicities. This review collated available data with the aim of creating a complete, global picture of EGFR mutation frequency in patients with NSCLC/ADC by ethnicity. Worldwide literature reporting EGFR mutation frequency in patients with NSCLC/ADC was reviewed, to create a map of the world populated with EGFR mutation frequency by country (a ‘global EGFR mutMap’). A total of 151 worldwide studies (n=33162 patients with NSCLC/ADC, of which 9749 patients had EGFR mutation-positive NSCLC/ADC) were included. There was substantial variation in EGFR mutation frequency between studies, even when grouped by geographic region or individual country. As expected, the Asia-Pacific NSCLC/ADC subgroup had the highest EGFR mutation frequency (47% [5958/12819; 87 studies; range 20%-76%]) and the lowest EGFR mutation frequency occurred in the Oceania NSCLC/ADC subgroup (12% [69/570; 4 studies; range 7%-36%]); however, comparisons between regions were limited due to the varying sizes of the patient populations studied. In all regional (geographic) subgroups where data were available, EGFR mutation frequency in NSCLC/ADC was higher in women compared with men, and in never-compared with ever-smokers. This review provides the foundation for a global map of EGFR mutation frequency in patients with NSCLC/ADC. The substantial lack of data from several large geographic regions of the world, notably Africa, the Middle East, Central Asia, and Central and South America, highlights a potential lack of routine mutation testing and the need for further investigations in these regions.

Introduction

Lung cancer represents a significant clinical burden worldwide. Recent statistics (WHO 2012) have indicated that lung cancer is the most common cancer in men globally, with an age-standardised rate (ASR; per 100,000) of 34.2 and an incidence of 1.2 million, and the fourth most common cancer in women (ASR 13.6; incidence 0.6 million) after cancer of the breast, colorectum, and cervix [1]. Consequently, lung cancer is the leading cause of cancer-related mortality worldwide, accounting for 19.4% of all cancer-related deaths [1].

Non-small-cell lung cancer (NSCLC) accounts for ~85% of primary lung cancers, and the majority of patients present with advanced or metastatic disease at diagnosis [2]. Adenocarcinoma (ADC) is one of the most common histological subtypes of NSCLC [3,4]. Molecular profiling of tumor samples from patients with NSCLC has identified driver mutations that may contribute to early carcinogenesis in more than 80% of ADC cases, including epidermal growth factor receptor (EGFR) mutations [5], which are now considered to be commonly associated with NSCLC tumors [5-13].

Mutation status improves predictions of the behavior and characterization of tumors of ADC histology when compared with the use of prognostic factors, such as tumor node metastasis stage [14]. Mutation testing has enabled many patients with EGFR mutation-positive NSCLC of ADC histology (referred to as NSCLC/ADC throughout) to receive personalized treatment with EGFR tyrosine kinase inhibitors (TKIs), which specifically target EGFR oncogenes [9-12], based on the molecular characteristics of their tumor, with the result that treatment outcomes are optimized in these patients [15,16]. It is now widely accepted that response to EGFR TKIs is greater in patients with tumors harboring EGFR mutations compared with patients without EGFR mutations; results from several clinical trials have demonstrated the successful use of mutation testing to ascertain the mutation status of the tumors of patients with NSCLC [16-18], and superior response to EGFR TKIs in patients with EGFR mutation-positive NSCLC compared with patients with EGFR mutation-negative NSCLC [16,17,19].

Current clinical guidelines now advocate the role of tumor EGFR mutation testing for patients with NSCLC during initial diagnosis [20,21], as do several working groups [22-24]. More clinicians are requesting EGFR mutation tests for their patients with NSCLC/ADC, and the number of facilities that conduct the tests has increased [25,26]. However, adoption of testing may be influenced by differences both within and between geographic regions/countries. It is important to understand where EGFR mutation testing is less routinely carried out, and why, to facilitate routes for clinicians/patients to access tests and have the opportunity to receive personalized therapy.

Studies of patients with NSCLC/ADC have provided robust data demonstrating the frequency of EGFR mutations in Western and Asian populations only [5,27]. There is limited understanding of the availability of mutation-status data indicating the presence of testing, and the prevalence and distribution of EGFR mutations in tumors of patients with NSCLC/ADC globally; further to this, the potential environmental [28] and genetic [29] influences on population differences in EGFR mutation status are yet to be fully elucidated. The aim of this systematic review is to describe the EGFR mutation frequency in patients with NSCLC/ADC from multiple ethnic backgrounds and create a global map of EGFR mutation prevalence (a ‘global EGFR mutMap’).

Methods

Literature search

A literature search of the MEDLINE database was carried out on February 18, 2014, to identify journal articles reporting studies of EGFR mutation frequency/incidence in patients with NSCLC/ADC. The following search criteria were used: [EGFR] AND [mutation] AND [lung cancer] AND [adenocarcinoma] AND [publication date after January 1, 2004]. The following were excluded: [non-English language papers] AND [Review] AND [Comment] AND [Editorial] AND [Letter].

Publications yielded in this initial search were reviewed and shortlisted; studies were only included if they had a population of ≥ 100 patients, although this criterion was waived in countries where data were sparse.

Where the initial search yielded no data from certain countries, additional literature searches of PubMed, BIOSIS, and Embase were carried out on June 23, 2014, using the previous search criteria with the addition of specific country names, to identify any additional relevant publications.

Analysis of EGFR mutation frequency

EGFR mutation frequency data were analyzed by country, gender, and smoking status (never- or ever-smokers, according to the definitions of the original studies). A map of the world was populated with available EGFR mutation frequency data by country (a ‘global EGFR mutMap’).

Results

Included studies

The initial literature search yielded 720 publications, from which 139 were selected, based upon the criteria that they contained relevant EGFR mutation frequency data in patients with NSCLC/ADC. Gaps were identified in Brazil, Denmark, Finland, Latin America, Mexico, Norway, Russia, Saudi Arabia, South America, Sweden, Turkey, and the UK; therefore, an additional literature search of PubMed, BIOSIS, and Embase was carried out on June 23, 2014, which yielded an additional 12 publications. This resulted in a total of 151 studies included in this review (n=33162 patients with NSCLC, of which n=9749 had EGFR mutation-positive NSCLC/ADC) (Figure 1; Supplementary Table 1). This included 70 smaller studies of < 100 patients with NSCLC/ADC (19/70 studies, n ≥ 70 patients; 40/70, n ≥ 50 patients), in the interest of providing as comprehensive a review as possible (Supplementary Table 1) from the following countries: Australia, Bangladesh, Brazil, China, Germany, Greece, Italy, Japan, Lithuania, Taiwan, Turkey, the UK, the USA, the Philippines, Pakistan, Republic of Korea, and Singapore.

Figure 1.

Flow of studies identified and included in the meta-analyses.

As expected from previous reports [5] and consistent with NSCLC incidence figures [1], there were more data available for NSCLC/ADC patient populations in the Asia-Pacific and European geographic regions (87 studies, 12819 patients; 39 studies, 10464 patients, respectively). Data available for patient populations of the other geographic regions included were as follows (in decreasing order): North America (19 studies, 7396 patients; including one study in US African-Americans [121 patients]); Indian subcontinent (5 studies, 1090 patients); South America (5 studies, 686 patients); Oceania (4 studies, 570 patients); and Africa (1 study, 137 patients) (Table 1).

Table 1.

Frequency of EGFR mutations in patients with NSCLC of ADC histology by country

Country	No. studies	EGFR mutation frequency
		No. patients with an EGFR mutation/total no. patients	Overall EGFR mutation frequency (%)	EGFR mutation frequency range (%)
Asia-Pacific
Overall	87	5958/12819	47	20-76
China	18 [27,34-50]	1403/2949	48	27-66
Hong Kong	3 [27,51,52]	312/585	53	47-58
Japan	33 [53-84]	2069/4619	45	21-68
Malaysia	2 [85,86]	272/599	45	39-47
Philippines	1 [27]	34/65	52	N/A
Republic of Korea	17 [87-103]	1248/2884	43	20-56
Singapore	2 [104,105]	57/142	40	39-43
Taiwan	9 [27,75,106-112]	423/739	57	36-76
Thailand	1 [27]	63/117	54	N/A
Vietnam	1 [27]	77/120	64	N/A
Europe
Overall	39	1527/10464	15	6-41
Austria	1 [113]	7/96	7	N/A
Czech Republic	1 [114]	21/101	21	N/A
Finland	1 [115]	58/398	10	N/A
France	2 [116,117]	193/1289	15	15-17
Germany	7 [118-124]	175/1573	11	6-41
Greece	2 [125,126]	30/137	22	20-22
Italy	9 [127-135]	306/2223	14	10-33
Lithuania	1 [136]	8/65	12	N/A
The Netherlands	2 [13,137]	162/1110	15	11-20
Norway	1 [138]	16/141	11	N/A
Poland	3 [139-141]	88/678	13	11-14
Portugal	1 [142]	29/216	13	N/A
Russia	2 [143,144]	53/240	22	20-31
Slovakia	1 [145]	56/285	20	N/A
Spain	1 [146]	283/1634	17	N/A
Sweden	1 [147]	16/148	11	N/A
Turkey	2 [148,149]	15/41	37	22-41
UK	1 [150]	11/89	12	N/A
North America
Overall	19	1638/7396	22	3-42
Canada	2 [151,152]	84/612	14	14
USA	16 [7,75,153-165]	1531/6663	23	3-42
USA (African Americans)	1 [166]	23/121	19	N/A
Indian subcontinent
Overall	5	278/1090	26	22-27
Bangladesh	1 [167]	14/61	23	N/A
India	3 [27,168,169]	261/1018	26	22-26
Pakistan	1 [170]	3/11	27	N/A
South America
Overall	5	250/686	36	9-67
Argentina	1 [30]	47/244	19	N/A
Brazil	3 [171-173]	67/239	28	9-34
Peru	1 [30]	136/203	67	N/A
Oceania
Overall	4	69/570	12	7-36
Australia	4 [75,174-176]	69/570	12	7-36
Africa
Overall	1	29/137	21	N/A
Morocco	1 [177]	29/137	21	N/A

ADC adenocarcinoma, EGFR epidermal growth factor receptor, N/A not available, NSCLC non-small-cell lung cancer.

EGFR mutation frequency in patients with NSCLC/ADC by geographic region

EGFR mutation frequency in patients with NSCLC/ADC is summarized in Table 1 and is represented as a ‘global EGFR mutMap’ in Figure 2.

Figure 2.

Global EGFR mutMap: EGFR mutation frequency in patients with NSCLC of ADC histology by country (where data available). See Table 1 for study and patient numbers for each country. ADC adenocarcinoma, EGFR epidermal growth factor receptor, NSCLC non-small-cell lung cancer.

When analyzed by geographic region, the Asia-Pacific NSCLC/ADC subgroup had the highest EGFR mutation frequency at 47% (n=5958/12819; 87 studies; range 20%-76%). Within this region, Taiwan had the highest EGFR mutation frequency in patients with NSCLC/ADC (57% [n=423/739; 9 studies; range 36%-76%]), while Singapore had the lowest (40% [n=57/142; 2 studies; range 39%-43%]), although comparisons between these countries were limited due to the differing patient population sizes (Table 1).

Of all the geographic regions included, the Oceania NSCLC/ADC subgroup, which only included data from Australia, had the lowest EGFR mutation frequency at 12% [n=69/570; 4 studies; range 7%-36%]) (Table 1; Figure 2).

The South American NSCLC/ADC subgroup had the widest EGFR mutation frequency range and second highest EGFR mutation frequency (36% [n=250/686; 5 studies; range 9%-67%]); although this result was skewed by data from one study of patients from Peru (EGFR mutation frequency, 67% [n=136/203]). The narrowest EGFR mutation frequency range was observed in the Indian subcontinent NSCLC/ADC subgroup (EGFR mutation frequency, 26% [n=278/1090; 5 studies; range 22%-27%]) (Table 1; Figure 2).

EGFR mutation frequency in patients with NSCLC/ADC by geographic region and gender

In all regions where data were available, EGFR mutation frequency in patients with NSCLC/ADC was higher in women compared with men: Europe, 22% versus 9%; Asia-Pacific 60% versus 37%; Indian subcontinent, 31% versus 23%; Africa, 48% versus 8%; and North America 28% versus 19% (Table 2). All country-specific NSCLC/ADC subgroups followed this pattern, apart from Bangladesh, where the EGFR mutation frequency was higher in men than in women (26% versus 14%, respectively) (Table 2).

Table 2.

Frequency of EGFR mutations in patients with NSCLC of ADC histology by gender and country

Country	No. studies	EGFR mutation frequency
		Males			Females
		No. patients with an EGFR mutation/total no. patients	Overall EGFR mutation frequency (%)	EGFR mutation frequency range (%)	No. patients with an EGFR mutation/total no. patients	Overall EGFR mutation frequency (%)	EGFR mutation frequency range (%)
Asia-Pacific
Overall	20	473/1250	37	20-80	763/1263	60	0-83
China	4 [36,40,46,49]	41/86	48	25-80	43/62	69	29-83
Hong Kong	1 [51]	28/84	33	N/A	87/131	66	N/A
Japan	10 [60,61,65,68,69,74,76,77,80,82]	345/893	39	34-67	534/866	62	0-76
Republic of Korea	4 [87,89,94,95]	41/152	27	20-36	75/162	46	38-80
Taiwan	1 [111]	18/35	51	N/A	24/42	57	N/A
Europe
Overall	9	88/1035	9	4-18	133/616	22	3-35
Austria	1 [113]	3/58	5	N/A	4/38	11	N/A
Germany	2 [118,120]	15/211	7	4-12	27/163	17	14-23
Greece	1 [125]	3/17	18	N/A	1/3	33	N/A
Italy	3 [127,133,135]	26/323	8	6-13	49/215	23	13-35
Poland	1 [140]	25/310	8	N/A	30/121	25	N/A
Russia	1 [143]	16/116	14	N/A	22/76	29	N/A
North America
Overall	2	175/923	19	8-20	459/1656	28	21-28
USA	2 [7,157]	175/923	19	8-20	459/1656	28	21-28
Indian subcontinent
Overall	3	160/689	23	23-26	99/318	31	14-33
Bangladesh	1 [167]	12/47	26	N/A	2/14	14	N/A
India	2 [168,169]	148/642	23	23	97/304	32	32-33
Africa
Overall	1	7/91	8	N/A	22/46	48	N/A
Morocco	1 [177]	7/91	8	N/A	22/46	48	N/A

ADC adenocarcinoma, EGFR epidermal growth factor receptor, N/A not available, NSCLC non-small-cell lung cancer.

EGFR mutation frequency in patients with NSCLC/ADC by geographic region and smoking status

In all regions where data were available, EGFR mutation frequency in NSCLC/ADC was higher in never-smokers compared with ever-smokers: Europe, 35% versus 8%; Asia-Pacific, 64% versus 33%; Indian subcontinent, 32% versus 17%; Africa, 41% versus 6%; and North America, 47% versus 14% (Table 3). Similar to EGFR mutation prevalence by gender, this pattern was followed by all country-specific subgroups apart from Bangladesh and Taiwan, where EGFR mutation frequency was higher in ever-smokers compared with never-smokers (24% versus 22% for Bangladesh; 57% versus 53% for Taiwan, respectively) (Table 3).

Table 3.

Frequency of EGFR mutations in patients with NSCLC of ADC histology by smoking status and country

Country	No. studies	EGFR mutation frequency
		Never-smokers*			Ever-smokers*
		No. patients with an EGFR mutation/total no. patients	Overall EGFR mutation frequency (%)	EGFR mutation frequency range (%)	No. patients with an EGFR mutation/total no. patients	Overall EGFR mutation frequency (%)	EGFR mutation frequency range (%)
North America
Overall	6a	680/1433	47	38-56	433/3151	14	6-16
USA	6a [7,155,157,159,161,178]	680/1433	47	38-56	433/3151	14	6-16
Asia-Pacific
Overall	20	847/1312	64	40-83	426/1242	33	11-100
China	4 [36,40,46,49]	97/134	72	50-79	28/66	42	11-100
Hong Kong	1 [51]	94/129	73	N/A	21/86	24	N/A
Japan	9 [60,61,68,69,74,76,77,80,82]	549/829	66	57-83	326/919	35	26-54
Republic of Korea	4 [87,89,94,95]	81/171	47	40-62	35/143	24	21-29
Taiwan	1 [111]	26/49	53	N/A	16/28	57	N/A
Europe
Overall	6	61/176	35	23-50	31/373	8	3-17
Austria	1 [113]	5/22	23	N/A	2/74	3	N/A
Germany	1 [120]	12/24	50	N/A	8/90	9	N/A
Greece	1 [125]	1/2	50	N/A	3/18	17	N/A
Italy	2 [133,135]	13/34	38	37-50	10/93	11	5-13
Russia	1 [143]	30/94	32	N/A	8/98	8	N/A
Indian subcontinent
Overall	2	38/118	32	22-35	19/109	17	14-24
Bangladesh	1 [167]	5/23	22	N/A	9/38	24	N/A
India	1 [169]	33/95	35	N/A	10/71	14	N/A
Africa
Overall	1	24/58	41	N/A	5/79	6	N/A
Morocco	1 [177]	24/58	41	N/A	5/79	6	N/A

^*As defined by original study criteria.

^aEver-smokers category had 5 studies.

ADC adenocarcinoma, EGFR epidermal growth factor receptor, N/A not available, NSCLC non-small-cell lung cancer.

Discussion

This review investigated EGFR mutation frequency in patients with NSCLC/ADC to create a ‘global EGFR mutMap’ to visually represent EGFR mutation frequency in countries where data were available, and to identify where there is still a need for more studies to be carried out.

Results of this review reflected the robust data available for EGFR mutation frequency in Western/Asian populations, building upon and supporting the results of previous studies and meta-analyses in these ethnicities. In the first ‘mutMap’ study of the incidence and coincidence of genetic mutations associated with NSCLC, EGFR mutation frequency was 47.9% in Asian patients with NSCLC/ADC, compared with 19.2% in Western patients with NSCLC/ADC [5]. These data are similar to our findings in related patient populations: Asia-Pacific subgroup, 47%; and European subgroup, 15%. The slight decrease in EGFR mutation frequency in the European subgroup of this study compared with the first mutMap study may be due to differing studies/patient populations used in the analyses. In addition, EGFR mutations in patients with NSCLC/ADC were more prevalent in, but not exclusive to, Asian patients compared with patients of other ethnicities included in this study; and across most countries where data were available, EGFR mutations in patients with NSCLC/ADC were more prevalent in, but not exclusive to, females and never-smokers, with only data from patients in Bangladesh and Taiwan differing from this pattern. Results of the first mutMap study similarly indicated that although EGFR mutations occurred more frequently in females, Asians, and never-/light-smokers, mutations were also found outside of these three subgroups [5]. Likewise, results of the PIONEER molecular epidemiology study across seven Asian countries/regions also found that although EGFR mutations occurred more frequently in females, Asians, and never-smokers, EGFR mutations were not restricted to patients with these clinical characteristics. In fact, more than 50% of patients with EGFR mutations in the PIONEER study were not female non-smokers [27]. These findings, therefore, support EGFR mutation testing in all patients with NSCLC/ADC.

The review also highlighted that despite the burden of NSCLC reported to exist in these large geographic regions by global registries [1], a substantial lack of data was observed from Africa, the Middle East, Central Asia, and Central and South America; with no data available at all for several countries within these regions (Figure 2). The importance of determining the prevalence of EGFR mutations in these regions/countries is emphasized by the interesting result that EGFR mutation frequency in NSCLC/ADC was particularly high in patients from Peru (67% [n=136/203]) [30], with the population of this study taken almost exclusively from the indigenous (‘Mestizo’), not historically Caucasian, population (68% versus 32% of the study sample, respectively). While only limited conclusions can be drawn from this single study, it does suggest NSCLC patient populations around the world may yet yield important findings. The paucity of published data in these regions/countries may reflect a lack of routine mutation testing in the clinical management of patients with NSCLC, which may be due to various factors, including tumor sampling practices, patients’ ability/willingness to undergo sampling, and availability/use of mutation tests. More global research into the infrastructure around diagnosis and management of NSCLC, and EGFR mutation frequencies, in regions outside of Europe and Asia will further understanding of this, and facilitate development of local clinical guidelines.

Where data were available, there was substantial variation in EGFR mutation frequency between studies, even when grouped by geographic region and country, which may be explained by several factors - for example, pre-selection of patients, differing patient baseline characteristics, underlying genetic differences between populations, variations in mutation-testing practices, and environmental factors-which are yet to be fully understood [28]. Considering genetic differences between populations, as well as different EGFR mutation subtypes existing [31], EGFR amplification and regulation of protein expression may also differ between patients [29]; a number of studies have identified polymorphisms in the EGFR promoter region and demonstrated that these polymorphisms have a functional impact on EGFR protein transcription or expression in non-cancerous tissues [32,33]. Furthermore, a study by Nomura et al. [29] has demonstrated that differences in prevalence of three of these polymorphisms exist between Asian and Western patients. This review found that EGFR mutation prevalence in the Indian subcontinent and South American populations lies between the prevalence in Asian and Western populations. One hypothesis may be that this intermediate EGFR mutation expression pattern may be explained, in part, by differences in EGFR promoter polymorphisms. Therefore, a study is potentially warranted that examines reasons for differences in prevalence of EGFR mutations and the link with, for example, environmental factors or EGFR promoter polymorphisms, and potentially EGFR expression in non-cancerous tissues.

This review comprises the most comprehensive summary to date of the global EGFR mutation frequency in patients with NSCLC/ADC. However, it has to be considered that the results presented here reflect only the available and relevant published data and, therefore, comprise a small sample size relative to the global population. This is especially relevant for this study, where data from 70 small studies (in Australia, Bangladesh, Brazil, China, Germany, Greece, Italy, Japan, Lithuania, Taiwan, Turkey, the UK, the USA, the Philippines, Pakistan, Republic of Korea, and Singapore), each of < 100 patients with NSCLC, were included in the interest of providing the most comprehensive review possible. Variations in study design and diagnostic and mutation-testing procedures between studies should also be considered.

Further work in less-studied regions-in particular, Africa, the Middle East, Central Asia, and Central and South America-is required to widen the scope and validity of this ‘global EGFR mutMap’. In addition, global mutation frequency data for multiple oncogenes known to be associated with NSCLC could be used to further inform global registries of the prevalence of NSCLC [1], to facilitate a registry of NSCLC incidence by mutation status. Building on the foundations laid here, further efforts to accurately establish the prevalence of EGFR mutations in a greater number of NSCLC/ADC patient populations will help to ensure that the appropriate knowledge, clinical guidelines, and resources are in place to improve treatment outcomes for more patients worldwide.

Disclosure of conflict of interest

Anita Midha, Simon Dearden and Rose McCormack are employees of AstraZeneca and hold shares in AstraZeneca.