EGFR, KRAS, BRAF, ALK, and cMET genetic alterations in 1440 Sardinian patients with lung adenocarcinoma
Lung cancer is one of the most incident neoplastic diseases, and a leading cause of death for cancer worldwide. Knowledge of the incidence of druggable genetic alterations, their correlation with clinical and pathological features of the disease, and their interplay in cases of co-occurrence is crucial for selecting the best therapeutic strategies of patients with non-small cell lung cancer. In this real-life study, we describe the molecular epidemiology of genetic alterations in five driver genes and their correlations with the demographic and clinical characteristics of Sardinian patients with lung adenocarcinoma.
Data from 1440 consecutive Sardinian patients with a histologically proven diagnosis of lung adenocarcinoma from January 2011 through July 2016 were prospectively investigated. EGFR mutation analysis was performed for all of them, while KRAS and BRAF mutations were searched in 1047 cases; ALK alterations were determined with fluorescence in situ hybridization in 899 cases, and cMET amplifications in 788 cases.
KRAS mutations were the most common genetic alterations involving 22.1% of the cases and being mutually exclusive with the EGFR mutations, which were found in 12.6% of them. BRAF mutations, ALK rearrangements, and cMET amplifications were detected in 3.2, 5.3, and 2.1% of the cases, respectively. Concomitant mutations were detected only in a few cases.
Almost all the genetic alterations studied showed a similar incidence in comparison with other Caucasian populations. Concomitant mutations were rare, and they probably have a scarce impact on the clinical management of Sardinians with lung adenocarcinoma. The low incidence of concomitant cMET amplifications at diagnosis suggests that these alterations are acquired in subsequent phases of the disease, often during treatment with TKIs.
KeywordsLung cancer Adenocarcinoma Targeted therapies EGFR KRAS BRAF ALK cMET
Anaplastic lymphoma kinase
v-raf murine sarcoma viral oncogene homolog B
Epidermal growth factor receptor
Fluorescence in situ hybridization
Kirsten rat sarcoma viral oncogene homolog
Non-small cell lung cancer
ROS proto-oncogene 1
Small cell lung cancer
Tyrosine kinase inhibitors
Lung cancer is one of the most incident neoplastic diseases and a leading cause of death for cancer worldwide [1, 2]. Its incidence has been increasing in developing countries and in women in the last decade, while it began to decline in males in most developed countries . Mortality rates remain high, despite recent advances in the prevention, screening, surgical and medical management of patients with lung cancer. Surgery is an effective treatment in the early stages of the non-small cell lung cancer (NSCLC) subtypes; unfortunately, approximately 80% of the sufferers are at an advanced stage at the time of diagnosis, and approximately 20% of them are affected by small cell lung cancer (SCLC), which has no substantial benefits from surgery . Chemotherapy has been the main treatment available for advanced stage patients for years. Last-generation chemotherapy drugs combined with a platinum regimen showed a 5-year survival improvement of 11%, but with a median survival time of only 8–10 months [4, 5]. In addition, chemotherapy drugs cannot differentiate tumour cells and normal cells, leading to dramatically strong adverse reactions that compromise the effectiveness and completeness of therapies.
Efforts to improve the results of the oncological treatments for NSCLC, together with the technological advances in DNA sequencing, led to the development of new therapeutic strategies based on the knowledge and classification of specific molecular features of the disease. Subsets of patients with adenocarcinoma and activating mutations within the kinase domain of the epidermal growth factor receptor (EGFR) gene have been successfully treated with selective tyrosine kinase inhibitors (TKIs), such as erlotinib, gefitinib, afatinib, and osimertinib, which are also characterized by reduced adverse events in comparison with traditional chemotherapy [6, 7]. In addition, anaplastic lymphoma kinase (ALK) and ROS proto-oncogene 1 (ROS1) fusions have been demonstrated to be effectively druggable with targeted inhibitors such as crizotinib, alectinib, and ceritinib and are currently recommended for the treatment of advanced stage adenocarcinoma harbouring that kind of genetic alteration [8, 9, 10]. Furthermore, active research is ongoing for the evaluation of the clinical impact of additional draggable genetic alterations, such as Kirsten rat sarcoma viral oncogene homologue (KRAS) and v-raf murine sarcoma viral oncogene homologue B (BRAF) mutations or proto-oncogene cMET amplifications involved in the pathogenesis of lung cancer, and have proven effective in treating other malignancies [11, 12]. In addition, the coexistence of driver mutations in the same tumours has been demonstrated to consistently impact the therapeutic outcomes and survival rates of patients undergoing chemotherapy or targeted therapies for NSCLC, as they can alter the responses to target therapies . For these reasons, recent guidelines suggest that 9 genes related to targeted therapies should be detected, including EGFR, KRAS, HER2, ALK, ROS1, cMET, BRAF, RET, and NTRK .
Knowledge of the incidence of such genetic alterations, their correlation with clinical and pathological features of the disease, and their interplay in cases of co-occurrence is crucial for selecting the best therapeutic strategies of patients with NSCLC. In the present study, we describe the molecular epidemiology of EGFR, KRAS, BRAF, ALK and MET genetic alterations and their correlations with the demographic and clinical characteristics of 1440 Sardinian patients with lung adenocarcinoma.
Patients and samples
A total of 1440 consecutive Sardinian patients with a histologically proven diagnosis of adenocarcinoma of the lung from January 2011 through July 2016 were prospectively enrolled and investigated. For all the patients enrolled, medical records and pathology reports were used to retrieve the demographic and clinical data at the time of diagnosis; sex, age, smoking habits, type of sample (primary tumour or metastasis) and the origin of the sample (biopsy or surgery) were assessed. To avoid any bias, the patients were consecutively enrolled regardless of age at diagnosis and disease characteristics of the primary tumour. Sardinian origin was ascertained through verification of the place of birth for all patients. All patients were informed about the aims of this study and, before the tissue sample was collected, provided written informed consent. The study was performed in accordance with the principles of the Declaration of Helsinki and was approved by the Committee for the Ethics of the Research and Bioethics of the National Research Council (CNR).
Genomic DNA was isolated from tissue sections using a standard protocol, and DNA quality was assessed for each specimen, as previously reported . Briefly, paraffin was removed from formalin-fixed paraffin-embedded (FFPE) samples by treatment with Bio-Clear (Bio-Optica, Milan, Italy), and DNA was purified using the QIAamp DNA FFPE Tissue Kit (Qiagen Inc., Valencia, CA, USA) following the manufacturer’s instructions. Yields of purified DNA were assessed by the Qubit dsDNA High-Sensitivity Assay Kit on the Qubit 2.0 Fluorometer (Life Thermofisher, Waltham, MA USA).
Mutation analysis was conducted in the coding sequence of the following genes: EGFR (exons 18, 19, and 21, where all mutations predicting the response of treatment with EGFR tyrosine kinase inhibitors are located), KRAS (the entire coding portion: exons 2, 3, and 4), and BRAF (exon 15, where nearly all of the oncogenic mutations are located). Quantitative measurements of mutations were based on pyrosequencing methodology, which is a real-time sequencing-by-synthesis approach that allows for the quantification of mutated alleles with a detection limit of 5–7% . Pyrosequencing represents a good compromise between specificity and sensitivity among commonly used mutational analysis methods (Sanger-based sequencing: specificity 100%, sensitivity 15–20%; pyrosequencing: specificity 90%, sensitivity 5–7%; real-time PCR assay: specificity - for each single variant only − 100%, sensitivity 2–3%) . Pyrosequencing assays were performed on a PyroMark Q24 system (Qiagen Inc., USA) following the manufacturer’s instructions.
Fluorescence in situ hybridization (FISH) analysis was carried out in interphase tumour cells using the following: for cMET, the specific CTB.13 N12 BAC probe (at the 7q31.2 locus) and the control centromere, labelled with Spectrum-Orange and Spectrum-Green (Vysis, Downer’s Grove, IL, USA), respectively; for ALK, the ALK Break Apart FISH Probe Kit (Vysis, USA). Protocols for FISH analysis were as previously described by our group .
Descriptive analysis for qualitative and quantitative variables was conducted using proportions and the mean ± standard deviation (SD), respectively. Variable distribution was assessed by the Shapiro-Wilcoxon test. Statistical differences between groups were compared using unpaired Student’s t-test, Mann-Whitney rank sum test, chi-square test or Fisher’s exact test as appropriate. P ≤ 0.05 was considered statistically significant. Data were analysed using STATA 13® statistical software (StataCorp LP, College Station, TX, USA).
Nine hundred sixty-three (67%) of the 1440 patients enrolled were males, and the mean age was 67 (range 30–88). Most of the cases (1064, 74%) involved individuals with more than 61 years of age. Only 13% (186 cases) were never smokers, 475 (33%) were active smokers, and 538 (37%) were former smokers; data regarding smoking habits were not available in 241 (17%) cases. The samples were obtained from the primary tumour in 1243 (86%) of the cases and from metastatic lesions in the remaining cases. Finally, in only 242 (17%) cases, the specimen was obtained by surgery, reflecting the advanced stage of the disease at diagnosis in most cases in which exclusively a biopsy was performed.
Distribution of EGFR mutations according to patients’ characteristics
Number of patients (%)
EGFR mutated cases
Total cases analyzed
Age at diagnosis
≤ 50 years
> 70 years
Type of sample
Distribution of KRAS and BRAF mutations according to patients’ characteristics
Number of patients (%)
KRAS mutated cases
BRAF mutated cases
Total cases analyzed
Age at diagnosis
≤ 50 years
> 70 years
Type of sample
Distribution of ALK rearrangements according to patients’ characteristics
Number of patients (%)
ALK mutated cases
Total cases analyzed
Age at diagnosis
≤ 50 years
> 70 years
Type of sample
Distribution of cMET rearrangements according to patients’ characteristics
Number of patients (%)
cMET mutated cases
Age at diagnosis
≤ 50 years
> 70 years
Type of sample
ALK rearrangements were detected in 48 (5.3%) out of the 899 cases examined; they were significantly more common in females and individuals younger than 50 years of age (Table 3). Furthermore, examining the 778 patients in which both ALK rearrangement and cMET amplification analysis were carried out, genetic alterations were found in 8% of the cases, the most common being ALK rearrangements (43 cases, 5.9%), while cMET amplifications occurred in 16 (2.1%) cases (Fig. 3). The only concomitant genetic alterations found in these patients involved two cases (0.3%) with an ALK rearrangement and an amplification of cMET, and two cases harbouring an EGFR mutation and an amplification of cMET. cMET amplification showed no predilection for any of the clinical parameters evaluated (Table 4). In summary, in our series, EGFR mutations were significantly more incident in females and never smokers, KRAS mutations in males and in smokers, and ALK rearrangements in females and individuals with less than 50 years of age.
The study of the genetic alterations in patients with NSCLC has profoundly changed the therapeutic landscape of the disease. Considering the 1047 patients in whom EGFR, KRAS, and BRAF mutation analysis was simultaneously carried out in our study, approximately 37% were found to have a genetic alteration in one of the examined genes. This percentage is slightly lower than those of previous studies, which reported approximately half of the patients with lung adenocarcinoma harbouring an EGFR or KRAS mutation . This may depend on the genetic characteristics of the target population in our study, characterized by high levels of genetic homogeneity due to geographical reasons. In any case, the concept that mutations in driver genes occur in a consistent percentage of lung adenocarcinomas remains, but its impact seems to be different in different populations; indeed, the incidence of EGFR mutations is significantly higher in Asian populations (even higher than 50%) compared to western countries .
In our series, EGFR mutations were searched in 1440 patients and were found in 12.6% of them, which is consistent with the partial results published in a previous report involving patients from the same population . This figure is slightly lower than that described in recent prospective studies performed in other Caucasian populations [19, 20]. In addition, EGFR mutations significantly more frequent in females (24.1%) and never smokers (58.6%), a finding that has been extensively reported in previous studies and from different geographical areas [6, 19, 21]. The incidence of EGFR mutations has been reported as low as 28% in American never smokers and as high as 68% in Asian never smokers ; the rate found in our series is closer to those reported in Asian populations. As mere speculation, it is interesting that Sardinians, who have long been recognized as forming a distinct outlier within contemporary European genetic diversity, experienced an immigration of individuals belonging to the initial wave of migration from the Asian areas (mainly the Middle East) into southeastern Europe during the early Neolithic transition, leading to the observed genetic affinity of the ancients descending from these migrants to present-day Sardinians [23, 24].
A recent meta-analysis evaluated the EGFR, ALK-EML4 and KRAS mutational patterns in smokers and non-smokers of various ethnicities . The authors confirmed that there was a significantly increased risk of presenting EGFR mutations and ALK-EML4 fusions in never smokers compared to ever smokers with adenocarcinoma. In addition, as the smoking history increased, there was a decreased risk for exhibiting the EGFR mutation, particularly for cases > 30 pack-years. Compared to ever smokers, never smokers had a decreased risk of KRAS mutations in all the populations examined . Both the meta-analysis results on EGFR and KRAS mutations were confirmed in our study.
Regarding the types of the specific EGFR mutations, L858R in exon 21 and deletion in exon 19 were the most frequent, accounting for 38 and 29% of all the observed EGFR alterations, respectively. Again, these mutations are also the most frequent in studies in Asian populations, but with lower percentages . Concomitant L858R mutation and deletion in exon 19 were found only in one case in our series; such a concomitance seems to be more frequent in studies in Asian populations . Examining the subset of patients without EGFR mutations, we found that half of them had no additional genetic alteration. As expected, most of the remaining EGFRwild-type patients harboured KRAS mutations (approximately one-third of the total EGFRwild-type cases), followed by ALK rearrangements and to a lesser extent by BRAF mutations and cMET amplification.
KRAS mutations were detected in 22.1% of the cases examined, while the only BRAF mutation described was V600E in exon 15 detected in 3.2% of the cases examined. In a study performed at the Memorial Sloan-Kettering Cancer Center, testing of 2529 cases for KRAS mutations (codons 12 and 13) detected 670 (26%) mutations, including G12C (39%), G12 V (21%), G12D (17%), G12A (11%), and other G12 and G13 mutations (12%) . Additionally, in our series, the most common KRAS alterations were missense mutations in exon 2, namely, G12C (39.8%), G12 V (16.5%), and G12D (13.9%); missense mutations in exon 3 accounted globally for 13.9% of the total. KRAS mutations in our series were significantly associated with male sex and smoking history of the patients, as previously mentioned .
Additionally, BRAF mutations were more frequent in males in our cohort. Involved in the RAS-MEK-ERK signalling pathway, BRAF is a serine/threonine kinase that lies downstream of RAS and has gained the most attention in malignant melanomas, where a V600E mutation is a common driver that is the therapeutic target of the selective BRAF inhibitors (vemurafenib, dabrafenib, encorafenib) and MEK inhibitors (cobimetinib, trametinib, binimetinib) . Up to 8% of lung adenocarcinomas harboured BRAF mutations in recent studies (including Italian cohorts), most of them being the V600E mutation, which was the only BRAF alteration detected in our cohort in 3.2% of the cases examined [28, 29, 30, 31, 32]. Nevertheless, in a recent study performed using a next-generation sequencing approach on 36 lung adenocarcinomas, BRAFV600E mutations occurred in 28% of the cases, mostly in smokers (90%), and in concomitance with AKT or PIK3CA mutations, non-V600E mutations occurred in 72% of the cases and in concomitance with KRAS mutations in four cases . These findings suggest that the epidemiological landscape of BRAF and other genetic alterations in NSCLC will be further cleared as new technologies for genetic testing become available for routine diagnostic purposes.
The ALK rearrangements are druggable targets in NSCLC patients with specific inhibitors. Considering the 778 patients examined for both ALK rearrangements and cMET amplifications, we found that 8% of them harboured ALK or cMET genetic alterations. The rates of ALK rearrangements (5.3%) and cMET amplifications (2.1%) found in our cohort were similar to those reported in the scientific literature [34, 35]. ALK translocations are common in young patients with non-smoking history and with no apparent ethnic differences ; in our study, they were more frequent in young females, without any association with smoking status. cMET gene amplification causes 1st generation EGFR-TKI resistance by activating EGFR-independent phosphorylation of ERBB3 and downstream activation of the PI3K/AKT pathway, providing a bypass mechanism. This redundant activation of ERBB3 permits cells to transmit the same downstream signalling in the presence of EGFR-TKIs. This mechanism involves 5–22% of resistant adenocarcinomas and is not related to that dependent on the EGFRT790M mutation on exon 20 (not searched in this study), which represents approximately 60% of resistance cases [37, 38]. Considering that the incidence of cMET amplifications in our cohort was 2.1%, most of them seem to occur in subsequent phases of the disease and during treatment with TKIs. This dictates the need for a double inhibition of both EGFR and cMET to overcome the development of drug resistance.
cMET was amplified in all four cases in which two concomitant driver genetic alterations were found. Two of them harboured an EGFR mutation and a cMET amplification, while the remaining two cases presented an ALK rearrangement with a simultaneous cMET amplification. Indication for a starting therapy combining inhibitors of both altered pathways may be necessary in those cases. No coexistence of EGFR, KRAS, or BRAF mutations was detected in our cohort, confirming the widely described mutually exclusive mutational pattern. The concomitant EGFR-KRAS mutations are described mainly in case reports; in a large cohort Chinese study on 5125 patients, 153 cases harbouring concomitant aberrations were found, and among them, 30 carried concomitant EGFR-KRAS mutations . Nevertheless, recent large cohort studies report a higher grade of the concomitance of ALK mutations in NSCLC patients, especially those harbouring EGFR mutations [40, 41, 42, 43]. ALK mutations are reported to occur in concomitance with EGFR mutations in 0–6% of cases [40, 41, 42, 43]; in our cohort, no such cases were found. Lee et al. analysed the clinical features of six patients harbouring EGFR-KRAS mutations and six patients with EGFR-ALK mutations, evidencing different morphological features of the tumours and behaviour to treatments . Most EGFR-KRAS mutation patients showed papillary and acinar histologic patterns with hobnail cells, while all EGFR-ALK mutation patients showed solid or cribriform patterns, and three had signet ring cells. Responses to treatment in patients with genetic co-alterations were recently evaluated in a large cohort Chinese study including 3774 cases . The authors reported 63 (1.7%) samples with more than one driver gene mutation; among these, 43 were co-alterations with an EGFR mutation, and 20 had an ALK rearrangement. In this study, 1st-line EGFR-TKI treatment did not significantly improve the progression-free survival (PFS) of patients harbouring concomitant EGFR mutations compared to patients harbouring a single EGFR mutation. However, for concomitant EGFR mutation patients, TKI therapy was more effective than chemotherapy (median PFS of 10.8 vs 5.2 months, P = 0.023) . In any case, the interaction of concomitant genetic alterations in terms of synergism versus the possible dominance of one rather than the other oncogene and the subsequent impact on targeted therapies are currently not completely clarified.
Our study has some limitations, mainly the non-homogeneous distribution of the genetic analyses performed; this simply depended on the gradual introduction of such analyses in clinical practice and the availability of sample tissues for testing. Furthermore, analyses did not include the T790 M mutation on exon 18 or the histological subtypes of the tumours examined. Nevertheless, the consistent number of the global cases analysed taken from real-life clinical practice, the genetic homogeneity of the population examined, and the quality of the methods employed for the tests represent the strengths of our work.
Our data showed that KRAS mutations are the most common genetic alterations in Sardinian patients with lung adenocarcinoma, involving 22.1% of the cases examined and being mutually exclusive with the EGFR mutations, which were found in 12.6% of the cases studied. BRAF mutations, ALK rearrangements, and cMET amplifications were detected in 3.2, 5.3, and 2.1% of them, respectively; these figures are relatively low in comparison with most studies in other Caucasian populations. Concomitant mutations were detected only in a few cases, suggesting that they rarely may represent a factor of drug resistance in Sardinians with lung adenocarcinoma, as opposed to other populations in which such concomitance is more common. The low incidence of concomitant cMET amplifications at diagnosis suggests that these alterations are acquired in subsequent phases of the disease, often during treatment with TKIs.
The Sardinian Lung Cancer (SLC) Study Group includes the following members who participated as investigators in this study and should be considered as co-authors: Antonio Pazzola, Giovanni Maria Fadda, Pietro Pirina, Alessandro Fois, Carlo Putzu, Giorgio Ginesu, Alberto Porcu (Azienda Ospedaliera Universitaria, Sassari, Italy); Giorgio Astara, Mario Scartozzi (Azienda Ospedaliera Universitaria, Cagliari, Italy); Anna Maria Carta, Efisio Defraia, Daniela Guerzoni, Giuseppe Porcu (Azienda Ospedaliera Brotzu, Cagliari, Italy); Gianfranco Bardino, Claudio Sini (Ospedale Civile, Olbia, Italy); Francesca Capelli, Maria Giuseppina Sarobba (Ospedale Zonchello, Nuoro, Italy).
MC, PP, AC and GP made substantial contributions in the conception and design of the study, as well as in data analysis and interpretation. MC, PP, DAS and GP contributed in drafting the manuscript. All the members of SLC Study Group substantially contributed in clinical data collection and interpretation, and performed critical revisions of the manuscript. MC, MCS, MiC, GrP, AM, MP, VD contributed in performing and interpreting molecular analyses. AC and VD interpreted pathological data. PP and GP performed the final revision of the manuscript. All authors read and approved the final manuscript.
The mutational analyses performed were partially funded by the Sardinian Regional Government (Regione Autonoma della Sardegna). The funding source had no role in study design, data collection, decision to publish, or preparation of the manuscript.
Ethics approval and consent to participate
All study participants provided written consent. The study was approved by the Committee for the Ethics of the Research and Bioethics of the National Research Council (CNR).
Consent for publication
All the patients provided written informed consent for the anonymous use of their clinical data for the purposes of the study.
Giuseppe Palmieri has/had an advisory role for Bristol Myers Squibb, Incyte, Merck Sharp & Dohme, Novartis, Pierre Fabre, and Roche-Genetech. All the remaining authors declare no conflict of interest.
- 5.Li HD, Liu SL. Molecular targeted therapy for non-small cell lung cancer: the reality in China and coping strategy. Prac J Med Pharm. 2018;35:373–9.Google Scholar
- 6.Paliogiannis P, Attene F, Cossu A, Defraia E, Porcu G, Carta A, et al. Impact of tissue type and content of neoplastic cells of samples on the quality of epidermal growth factor receptor mutation analysis among patients with lung adenocarcinoma. Mol Med Rep. 2015;12:187–91.PubMedPubMedCentralCrossRefGoogle Scholar
- 10.Lindeman NI, Cagle PT, Aisner DL, Arcila ME, Beasley MB, Bernicker EH, et al. Updated molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: guideline from the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. J Mol Diagn. 2018;20:129–59.PubMedCrossRefGoogle Scholar
- 14.NCCN. Clinical practice guidelines in oncology. Non-small cell lung cancer. https://www.nccn.org/professionals/physician_gls/default.aspx. Accessed 10 May 2019.
- 15.Ihle MA, Fassunke J, König K, Grünewald I, Schlaak M, Kreuzberg N, et al. Comparison of high resolution melting analysis, pyrosequencing, next generation sequencing and immunohistochemistry to conventional Sanger sequencing for the detection of p.V600E and non-p.V600E BRAF mutations. BMC Cancer. 2014;14:13.PubMedPubMedCentralCrossRefGoogle Scholar
- 18.Ulivi P, Chiadini E, Dazzi C, Dubini A, Costantini M, Medri L, et al. Nonsquamous, non-small-cell lung cancer patients who carry a double mutation of EGFR, EML4-ALK or KRAS: frequency, clinical-pathological characteristics, and response to therapy. Clin Lung Cancer. 2016;17:384–90.PubMedCrossRefPubMedCentralGoogle Scholar
- 21.Pao W, Miller V, Zakowski M, Doherty J, Politi K, Sarkaria I, et al. EGF receptor gene mutations are common in lung cancers from “never smokers” and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A. 2004;101:13306–11.PubMedPubMedCentralCrossRefGoogle Scholar
- 23.Sikora M, Carpenter ML, Moreno-Estrada A, Henn BM, Underhill PA, Sánchez-Quinto F, et al. Population genomic analysis of ancient and modern genomes yields new insights into the genetic ancestry of the Tyrolean Iceman and the genetic structure of Europe. PLoS Genet. 2014;10:1004353.CrossRefGoogle Scholar
- 45.Zhuang X, Zhao C, Li J, Su C, Chen X, Ren S, et al. Clinical features and therapeutic options in non-small cell lung cancer patients with concomitant mutations of EGFR, ALK, ROS1, KRAS or BRAF. Cancer Med. 2019. https://doi.org/10.1002/cam4.2183.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.