Article Text

Download PDFPDF

Detection of EGFR and KRAS mutations on trans-thoracic needle aspiration of lung nodules by high resolution melting analysis
  1. A Fassina1,
  2. A Gazziero1,
  3. D Zardo1,
  4. M Corradin1,
  5. E Aldighieri2,
  6. G P Rossi2
  1. 1
    Department of Diagnostic Medical Sciences and Special Therapies, Pathology Section, School of Medicine, University of Padova, Padova, Italy
  2. 2
    Department of Clinical and Experimental Medicine, School of Medicine, University of Padova, Padova, Italy
  1. Correspondence to Professor A Fassina, Department of Diagnostic Medical Sciences and Special Therapies, Pathology Section, University of Padova, via Gabelli 61, 35100 Padova, Italy; ambrogio.fassina{at}


Background: EGFR and KRAS are the target genes for tumour response to epidermal growth factor receptor (EGFR) inhibitors.

Aims: To investigate EGFR and KRAS mutational status with high resolution melting (HRM) analysis applied to cytological material obtained from trans-thoracic needle aspiration (TTNA) in order to better select patients for targeted therapy.

Methods: DNA was extracted from fixed material of 108 TTNAs under CT guidance, from 108 consecutive patients. In 77 TTNAs (71.3.%) that were positive for non-small cell lung cancer, the variant in exon 21 (the missense mutation at codon 858, L858R) and the deletion in exon 19 (in frame deletion at codons 747–749) of the EGFR gene, and the point mutation in exon 2 of KRAS were investigated with HRM assay using sequencing as the reference “gold standard”.

Results: Nine (11.7%) samples were positive for KRAS exon 2 mutations, and two (2.6%) samples were positive for the EGFR exon 21 missense mutation by HRM assay. No deletion at exon 19 for EGFR was detected by HRM analysis. All HRM results were confirmed by direct DNA sequencing.

Conclusions: HRM analysis of cytological material was accurate for the detection of two major EGFR mutations and KRAS mutations in exon 2. HRM analysis was fast, easy to apply, cheap, highly reproducible, and could be used with small amounts of material, such as is obtainable with needle lavage. Therefore, it may be useful as an adjunct to the cytological report that yields valuable molecular information.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Lung cancer (LC) is the leading cause of cancer-related death in many developed countries, with more than 1.3 million patient deaths per year worldwide.1 Microscopically, LC is classified into small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), which comprises 80% of all LC cases. These tumours can be further divided into squamous cell carcinoma, adenocarcinoma, and large cell carcinoma.2 3

Despite current treatments, NSCLC has a poor prognosis, with an overall 5-year survival rate of about 15%.4 Low-stage NSCLCs are treated with surgery,5 but the majority of patients present with advanced disease and die due to recurrences and metastases.6 Current clinicopathological features do not allow for the satisfactory prediction of outcome; hence, novel biomarkers that predict response to treatment must be identified in order to improve the selection of more effective and tailored therapies.

The epidermal growth factor receptor (EGFR) belongs to the ErbB receptor kinase family and is over-expressed in NSCLC, as well as in many other epithelial malignancies,7 8 resulting in activation of several cellular signalling pathways and leading to DNA synthesis and cell proliferation.9

Tyrosine kinase inhibitors (TKIs) have been developed as EGFR-targeted molecules. They include gefitinib (Iressa, Astra Zeneca) and erlotinib (Tarceva, Genentec/Roche) and are designed to compete with ATP for the ATP-binding site in the tyrosine kinase domain of EGFR, with the effect of blocking signal transduction, inhibiting cell growth and inducing apoptosis. Up to one third of all NSCLC patients have been reported to show a rapid and dramatic response to TKIs; however, unfortunately, most of the patients do not benefit from this treatment. Of note, an association between EGFR mutations and tumour response to TKIs has been suggested.9 10 11 12 13 14 15 16 Aberrant functions of other components in the ErbB signalling pathway have also been described in NSCLC, and attention has been focused on KRAS, a downstream GTPase that belongs to the Ras superfamily.17 The activating form of KRAS plays a key role in Ras/MAPK signalling involved in multiple pathways, including actin cytoskeletal integrity, proliferation, differentiation, cell adhesion, apoptosis and cell migration.18 KRAS gene mutations are commonly found at codons 12 and 13 in exon 2 and rarely at codons 59 and 61 in exon 3, and are present in 33% of NSCLC.18 19 They can alter protein conformation to impair GTP binding with a resulting constitutive protein activation.

Studies suggest that EGFR and KRAS mutations are mutually exclusive in NSCLC,20 21 22 reflecting different mechanisms of carcinogenesis. Of note, NSCLC tumours carrying KRAS mutations are held to be refractory to TKIs, suggesting that KRAS activation confers TKI resistance by activating pathways downstream to EGFR.12 In a retrospective study, Eberhard et al reported worse time to progression and survival among KRAS mutant patients treated with chemotherapy with erlotinib, compared with chemotherapy alone.23 Moreover, in the randomised phase III trial of erlotinib versus placebo in previously treated patients with advanced NSCLC, Zhu et al showed a reduced response to erlotinib among KRAS mutant patients when compared to wild-type patients.24 Thus, identification of EGFR and KRAS mutation status can be crucial for selecting the appropriate pharmacotherapy for NSCLC patients. EGFR and KRAS mutations are commonly detected by direct sequencing, which requires a large amount of good quality tissue, such as that available from surgical specimens.25 However, this is relatively expensive, time consuming, and unfeasible for non-resectable tumours, and therefore unsuitable for routine pre-therapeutic screening. Many researchers have tried to validate new methods for detecting EGFR and KRAS mutations, as summarised in table 1.

Table 1

Advantages and disadvantages of the major techniques applied on cytological material

For example, in a recent paper, Molina-Vila et al eliminated the need for the DNA purification of microdissecting tumour cells in PCR buffer directly. Detection for EGFR mutations was performed by length analysis of fluorescently labelled PCR products for EGFR deletion in exon 19 and by TaqMan assay for EGFR mutation in exons 20 and 21, respectively.36 The advantage of this method consists of the detection of EGFR mutations in samples containing less than 150 cancer cells. In addition, this method is laborious and expensive.

Lim et al have explored successfully whole genome amplification (WGA) of small amounts of DNA from low-volume diagnostic lung biopsies.37 The technique permits one to expand starting DNA from small biopsies for molecular analyses; however, it requires an efficient, appropriate and expensive WGA kit, and it is time consuming.

Recently, the development of high resolution melting (HRM) analysis has provided a novel, rapid, easy, and inexpensive tool for detecting mutations.26 27 28 29 Furthermore, HRM has been applied to formalin-fixed paraffin embedded biopsies,28 and fine needle aspiration slides,29 31 with excellent results. In this study, we investigated 77 NSCLC cytological samples by HRM analysis to detect EGFR and KRAS activating mutations.

Materials and methods

Sample collection

A total of 108 consecutive patients who required TTNA under CT guidance at the radiology Unit of Padova University Hospital for the diagnosis of a peripheral lung lesion were recruited from January 2005 to March 2008. The study protocol was approved by our Institutional Board, and written informed consent was obtained in all patients. Aspirates were smeared on glass slides and immediately fixed using a spray preparation with polyethylene glycol base for Papanicolau staining, or air-dried for Giemsa. Immediate examination of the material was used to assess the adequacy and quantity of tumour cells and to provide a preliminary diagnosis. If the sample was judged to be appropriate, the remaining material and the needle lavage were fixed for preservation of the cells in FineFix (Milestone Srl, Bergamo, Italy). This technique has been shown to maintain good morphological results and to satisfactorily preserve the quality and integrity of nucleic acids.32 33

Of 108 cases, 77 were classified as NSCLC and were considered for the clinical and pathological findings, while cases of SCLC (n = 5), metastases (n = 8), and non-malignancy (n = 18) were included as negative controls.

DNA extraction

The samples collected in FineFIX were centrifuged at the maximum speed for one minute to remove the excess fixative, followed by DNA extraction with the High Pure PCR Template Preparation Kit (ROCHE Diagnostic Spa, Indianapolis, Indiana, USA), according to the manufacturer’s instructions. DNA quality was checked on gel electrophoresis, and its quantity was measured by biophotometer (Eppendorf, Hamburg, Germany).

High resolution melting analysis

The DNA of lung TTNAs was amplified using specific primers for human EGFR and KRAS genes. Table 2 lists the sequences of primers for the EGFR gene of DEL region in exon 19 and the region of the L858R mutation in exon 21, as well as for the KRAS gene of a region in exon 2.

Table 2

EGFR and KRAS primer sequences used for HRM analysis

DNA was amplified via real-time PCR in the presence of a proprietary saturating DNA dye contained in the LightCycler 480 High Resolution Melting Master. A melting curve was produced using high data acquisition rates, and data were analysed with the LightCycler 480 Gene Scanning Software Module for deletion and mutation identification. The reactions were performed on a gene scanning platform (LightCycler 480, ROCHE Diagnostic Spa).

DNA sequencing

After HRM assay, the reaction mixture was purified by GenEluteTM PCR Clean-Up kit (Sigma-Aldrich, USA), according to the manufacturer’s instructions, and the amplified products were sequenced. DNA sequencing was performed by BMR Genomics (Padova, Italy) on all the samples analysed with the HRM assay. Sequence analysis was performed with Chromas 2.31 software.

Statistical analysis

The correlation between the presence of EGFR or KRAS mutational status and the clinicopathological variables were assessed by Fisher’s exact test.


Cytological analyses

Cytological evaluation was based on cyto-morphology of TTNA smears stained by Papanicolau and Giemsa, according to WHO classification and international cytology principles for cytological diagnosis.3 34 Of the 108 cases, 82 (75.9%) were primary LC, with 5 (6.1%) being SCLC and 77 (93.9%) being NSCLC. Eight TTNAs (7.4%) were metastatic tumours, while the remaining 18 were non-malignant (1.7%) (table 3).

Table 3

Cytological diagnoses by fine needle aspiration

EGFR and KRAS mutations detection by HRMA

In the evaluation of EGFR exon 19, we obtained identical melting curves for all 77 primary LC samples; the direct sequencing confirmed the absence of deletion in exon 19 (fig 1). HRM analysis for EGFR mutation in exon 21 reported three different types of melting curves (fig 2). At direct sequencing, these curves were found to correspond to the substitution in 2573T>G (amino acid substitution L858R) in two cases, and the single nucleotide polymorphism 2508C>T in one case (table 4); the remaining 74 samples were confirmed to be wild type.

Figure 1

The amplification plot of EGFR exon 19 shows no aberrant shifting of melting curves from the wild type (blue line). No deletion was found as confirmed by the sequencing chromatograms.

Figure 2

The difference plot of EGFR exon 21 shows three different melting profiles corresponding to the substitution in 2573T>G (amino acid substitution L858R) in red, the single nucleotide polymorphism 2508C>T in green and the wild type samples in blue (WT) as confirmed by the sequencing chromatograms.

Table 4

Genotyping analyses data for the patients with non-small cell lung cancer

HRM analysis for KRAS mutation in exon 2 reported 3 different types of melting curves (fig 3). At direct sequencing, they corresponded to mutations at codon 12 in 8 cases, and to the mutation at codon 13 in 1 case (table 4); the remaining 68 samples were negative for KRAS mutation. EGFR and KRAS mutations were mutually exclusive at both HRM analysis and direct sequencing.

Figure 3

The difference plot of KRAS exon 2 shows three different melting profiles corresponding to nucleotide substitutions: in codon 12 in red, in codon 13 (37G>T) in green, and the wild type samples in blue (WT). Sequencing of patients showed four different amino acid changes at codon 12 (34 G>T, 35 G>C, 35 G>A, 35 G>T).

Relationship between EGFR and KRAS mutations and clinicopathological factors

Table 5 summarises the correlation between clinicopathological characteristics of patients and EGFR and KRAS mutations.

Table 5

Clinicopathological characteristics of patients

EGFR and KRAS mutational status did not show any significant association with male gender, age, histotype and differentiation. However, KRAS mutations were more frequent in smokers (current and former, 87%) than in those who had never smoked (13%), and were also associated with ⩾ II tumour stages (Fisher’s exact test, p = 0.008).


The development of targeted molecular therapies has promoted the investigation of tailored treatments, with the final aim of replacing the current approach of using generalised chemotherapy and/or radiotherapy. Currently, NSCLC can be treated with EGFR TKIs in the presence of mutations in the targeted gene, since activating mutations in the tyrosine kinase domain of the EGFR gene were reported to be associated with a dramatic improvement with TKIs.11 12 13 14 Investigation of KRAS mutations in association with EGFR gene variations also could be relevant for tailored treatment, since KRAS mutations are held to imply lack of responsiveness to TKIs.20 21 22 New methods for mutation screening are desirable because the aforementioned DNA sequencing is impractical in the routine clinical use. Ideally, they should be easy to apply, fast, cheap, and applicable to small amount of bioptic material, such as TTNA specimens. Several methods have been proposed; some have been reported to give good results for detection of EGFR mutations, but have not been adopted in clinical practice because they require intensive labour or sophisticated instruments (table 1).25 35 36 37 38 39 40 41 42 43 44 HRM analysis has been introduced as a screening method that can provide a sensitivity of 100% for the detection of heterozygous germ line mutations.45 46 For somatic mutations in tumours, its accuracy could be compromised by a low proportion of tumour cells in biopsies; however, it has been shown that for EGFR and KRAS mutations, as low as 5–10% mutant alleles could be accurately ascertained by HRM assay.28 47 Currently, molecular analysis is performed on formalin-fixed, paraffin-embedded specimens; however, primary LC is often diagnosed on cytological preparation and additional biopsies or specimen are often unavailable for molecular analysis.31

We investigated EGFR and KRAS mutational status by the evaluation of cytology specimens from TTNAs of lung nodules.

In HRM analysis for EGFR exon 21, it was possible to discriminate three different melting curves referring to the wild type, the mutation 2573T>G, and the single nucleotide polymorphism 2508C>T. Our study has shown only two (2.6%) mutations in exon 21 of the EGFR gene, but no deletion at exon 19 of the EGFR gene has been detected using HRM assay, which was confirmed by direct sequencing. The reported EGFR mutation rate varies from 2% to 7% and can reach 24% in series of pre-selected patients.11 16 48 Several factors influence the rate of mutations: Asian ethnicity, female gender, and lack of a smoking history, as discussed by Fukui et al.16 Our prospective study was based on Caucasian patients of local Italian origin, who were mainly men (68.8%) and smokers (87% current or former). This could reasonably explain the low mutational rate observed.

HRM analysis for the KRAS gene detected nine mutations; the technique prevents the discrimination of single nucleotide mutations in codon 12, while the mutation in codon 13 has a different melting curve. In our analysis, the rate of KRAS mutations was 11.7%, in accordance with the literature.12 22 30

Limitations of the study

DNA obtained from TTNAs is usually scarce. Thus, it is important to consider the characteristics of the fixative to use, opting for one able to preserve nucleic acids. We used FineFix as a formalin-free fixative because it allows for the preservation of high quality DNA, as shown by Stanta et al and Gazziero et al.32 33 Moreover, in this study, each sample was assessed before HRM testing by cytological smear, and only those with at least 50% in tumour cells were analysed. This was held to be necessary to avoid false negatives due to an increased percentage of the normal cellular component. Hence, whether these results can be generalised to specimens obtained without this preliminary evaluation remains to be determined.


Since TTNA is often used in non-operable patients, it is useful to obtain molecular information from the cytological specimens to better select patients for the targeted pharmacotherapy. The findings of altered HRM assay can therefore be used to select patients to be further characterised with sequencing to confirm the presence of somatic mutations in EGFR and KRAS.

Our results show the feasibility of complementing the evaluation of cytology specimens from TTNAs of lung nodules with investigation of EGFR and KRAS mutational status by HRM assay.

Take-home messages

  • High resolution melting (HRM) analysis can be applied to cytological material obtained from trans-thoracic needle aspirates of lung nodules to detect EGFR and KRAS mutations and provides oncologists with molecular information for targeted pharmacotherapy.

  • The HRM technique is simple, rapid, reliable and less expensive than direct sequencing.



  • Competing interests None.

  • Ethics approval Ethics approval was obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.