Non-invasive approaches, includingctDNA material which are usually based on serum or plasma samples, are showinggreat potential to monitor EGFR-TKI treatment in NSCLC. ctDNA has a high degreeof specificity to detect EGFR mutations.
Moreover, ctDNA is capable ofmonitoring disease progression during EGFR-TKI treatment since it reflects thetumor burden. Effectively, these liquid specimens can complement withtumor tissues and help to guide EGFR-TKI therapy in NSCLC especially in thosewhere tissue biopsy is hard to obtain regularly. Thus, liquid samples obtainedby non-invasive approaches possess great potential to be valuable materialsapplied for guiding individual treatment (Sun et al., 2015). Nowadays, EGFR genemutations are the standard predictive biomarkers for selecting the treatment ofNSCLC patients: to receive EGFR-TKI treatment.
Multiple treatment options (suchas small-molecule inhibitors directed at these molecular targets) have beendeveloped, and some of them are migrating from bench to bedside; however,despite these rapid progresses, tyrosine kinase inhibitors of epidermal growthfactor receptor (EGFR-TKIs) are still the most successful example of targeted therapyin NSCLC. Compared with conventional treatment options for cancer patientsespecially chemotherapy, EGFR-TKIs are able to achieve prolongedprogression-free survival (PFS) with reduced side effects in NSCLC patientsharboring activating EGFR mutation (Sun et al., 2015). Somatic mutations in theepidermal growth factor receptor (EGFR) gene have been identified in patients withradiographic responses to EGFR-tyrosine kinase inhibitors (TKIs).
Currently,the response to EGFR-TKIs depends on patients harboring EGFR-sensitivemutations, and the median progression-free survival (PFS) is approximately 12months. Despite an initial response, most of these patients harboring EGFR mutationsacquire resistance to EGFR-TKIs (Hata et al., 2013).
Effectively, theresistance to treatment will lead to tumor progresses. For this reason,identification of the molecular mechanisms of EGFR-TKI resistance is necessary.This acquired resistance has many identified mechanisms including the secondT790M mutation of EGFR, amplification of MET or HER2, and mutations of PIK3CAor BRAF; but the secondary EGFR mutation, a pointmutation in exon 20 (T790M) accounts for approximately half of acquiredresistances to EGFR-TKIAs and monitoring T790M mutation is useful forestimating EGFR-TKI resistance.
The point mutation in exon 21 (L858R) ordeletion in exon 19 predicts good response to EGFR-TKIs, while the pointmutation (T790M) in exon 20 implies resistance to EGFR-TKIs. Taniguchi et al.performed a study to quantitatively detect the T790M-resistant mutations inctDNA. In 43.
5 % (10/23) of patients who had progressive disease afterEGFR-TKI treatment, the T790M mutation in ctDNA was detected (Sun etal., 2015). The introduction offirst-generation EGFR inhibitors (erlotinib, gefitinib) and subsequentlyafatinib for non–small cell lung carcinoma (NSCLC) patients with activatingsomatic mutations in EGFR has led to improvedtolerability and efficacy compared with first-line chemotherapy, as mentionedabove. But due to the emergence of resistance during treatment, a novel, oral,irreversible tyrosine kinase inhibitor was created for the treatment ofpatients with mutant EGFR NSCLC: Rociletinib(CO-1686). It has demonstrated efficacy against the activating mutations (L858Rand del19) and the acquired primary resistance mutation (T790M), while sparingwild-type EGFR. An ongoing phase I/II trial of rociletinib has demonstrated apromised clinical activity with a 59% ORR in T790M mutation-positive patientpopulation (Karlovich et al.
, 2016).In a study done by Thress et al., ctDNA wasshown to be more precise and informative than tissue as the blood mirrors theentire tumor burden. In addition, the rate of clinical response to AZD9291 (osimertinib, which is an orally administered EGFR-TKIis highlypotent against EGFR-TKI-sensitizing mutations and the T790M resistancemutation, but with a margin of selectivity against wild-type EGFR activity) wasalmost identical in patients positive for the T790M mutation in plasma and intissue, indicating that plasma detection may be a suitable alternative totissue genotyping in patients with NSCLC (Thress et al.
, 2015).On the otherhand, in a retrospective analysis done by Oxnard et al., the sensitivity ofplasma genotyping for detection of T790M was 70%; patients positive for T790Min plasma have outcomes with osimertinib that are equivalent to patientspositive by a tissue-based assay. This confirms that obtaining plasma samplescould avoid a tumore biopsy for T790M testing (Oxnard et al., 2016).To confirm more the importance of T790Mdetection by liquid biopsy, Crowley et al. showed that third-generationinhibitors have shown activity in the presence of this mutation (such asWZ4002114) and novel drug combinations have shown promising preclinicalactivity.
In addition, they emphasized on the utility of plasma samples forT790M detection by reporting that this mutation was first observed in relapsedpatients and later confirmed through the non-invasive analysis of plasmasamples, proving that resistance to targeted therapies can be monitoredin the blood (Crowley et al., 2013). Some evidence indicates thatT790M doesn’t only appear after TKI therapy but it may be present at low levelsin EGFR mutated AdenoCAs prior to EGFR drug therapy. Wang et al. useddPCR and ARMS to analyze T790M cfDNA mutations in 135 TKI-treated lung cancer patientswho had progression-free survival on the TKI therapy for over six months.
T790Mwas identified in 31.5% using dPCR and 5.5% using ARMS of these patients priorto TKI therapy, demonstrating that dPCR was a more sensitive test.
Post TKItherapy, T790M was identified in 43.0% of patients by dPCR, and in 25.2% by ARM.Patients with high pre-TKI therapy T790M levels had a poorer progression-freeand overall survival, demonstrating the importance and prognostic value of T790M detection (Ansariet al., 2016).More studies were done to confirm theaccuracy of plasma for mutation detection. The study done by Reckamp et al., usingQiagen therascreen EGFR Rotor-Gene Q PCR Kit for mutation analysis, reportedthe concordance data for matched tissue/cytologic and plasma samples testedwere: concordance 95% 131 of 138, sensitivity 73% 16 of 22 and specificity99% 115 of 116.
A comparable sensitivity of EGFR mutation detection wasobserved in plasma: 93% (38 of 41 specimens) for T790M, 100% (17 of 17) for L858R,and 87% (34 of 39) for exon 19 deletions. Moreover,the study showed EGFR mutation frequency that was9% for evaluable plasma samples. In addition, the frequency of T790M was 3%,57% for exon 19 deletions, 32% for L858R, 0% for exon 19 deletions and T790Moccurring together and 2% for L858R and T790M together (Reckamp et al.,2016).Another study reported thefrequency of EGFR mutations: three subgroups of EGFR mutationpositive patients were identified, including 19Dels or L858R alone, 19Dels or L858Rplus T790M, and T790M alone. The patients with EGFR 19Dels orL858R plus T790M accounted for the majority of the EGFR mutationpositive patients. Furthermore, EGFR19Dels or L858R alone was foundin 6.
5% to 14.3% of the patients in plasma throughout the study. T790M alonewas also detected in 3 to 6.
5% of the patients. Effectively, the study showedthat the majority of T790M positive patients carried simultaneously 19Dels orL858R in plasma throughout the course of their disease (Zheng et al., 2016). The mutation status concordance observedbetween 1162 matched tissue/cytologic and plasma samples (89%) suggests thatctDNA is a feasible sample for real-world EGFR mutation analysis ifrobust/sensitive DNA extraction and mutation analysis methodologies are used (Reckampet al.
, 2016).The 2011 publication by Liu etal. reported the successful EGFR mutation analysis of 86matched plasma-derived ctDNA and formalin-fixed, paraffin-embedded samplesusing a Scorpion-amplification refractory mutation system (ARMS). 27 EGFR mutation-positivewere identified in plasma out of the 40 tumor samples, showing a sensitivity of67.
5% (Douillard et al., 2014).In another study done by Wang et al., EGFRmutations were detected in the plasma samples from 17 NSCLC patients (12.7%). Withconsistency with results from the paired tumor tissues of two NSCLC patients, 11plasma samples had an Exon 19 deletion, 4 samples had a L858R mutation, and 2samples had a L861Q mutation. Finally, sensitivity and specificity for EGFRmutation detected in plasma were 22.06% (15/68) and 96.
97% (64/66) comparedwith tissue sample as a reference (Wang et al., 2014). Finally, in a study done by Kuang et al., usingthe SARMS assay they detected 12 patients with EGFR del 19 (E746_A750), 7patients with L858R, and 8 patients with EGFR T790M mutations. Theoverall concordance of tumor EGFR mutation withplasma EGFR mutation was 74% (32 of43), and this concordance as a function of the specific type of mutation (exon19 deletion versus L858R). 85% (17 of 20) concordance with the plasma EGFR mutationwas detected in the patients with a known tumor exon 19 deletion mutation whereasin those with a tumor L858R mutation the concordance rate was only 29% (2 of 7)with the plasma EGFR mutation. Moreover, therelationship with prior clinical response to gefitinib or erlotinib wasevaluated in patients in whom EGFR T790M was detected inplasma DNA.
There was no evidence for the presence of EGFR T790M(0 of 8; 0%) in patients with progressive disease to gefitinib or erlotinib orin patients who had never been treated with these agents (0 of 4; 0%). Thus, the EGFR T790Mmutation detected in plasma DNA is associated strongly with a prior clinicalresponse to gefitinib or erlotinib. Moreover, using the SARMStechnique, only 39% of EGFR mutations were detectedin plasma (7 out of 18 known EGFR-mutant patients); EGFR T790Mwas identified in 71% (5 of 7) in plasma of patients whose tumors were positivefor EGFR T790Mmutation (Kuang et al., 2009).Another study performed bySorensen et al.
confirms the pointthat prior TKI treatment is a must for the development of T790M. EGFR mutationswere examined in plasma samples collected during erlotinib treatment from 23lung cancer patients, where EGFR mutations were identified in their bloodsample taken before the initiation of treatment. 9 patients had T790M mutationcombined with the sensitizing EGFR mutation, 6 patients had the sensitizingmutation without the T790M mutation, and 8 patients had neither the sensitizingnor T790M mutations, with the appearance of T790M always occurring togetherwith an increase in the amount of the sensitizing EGFR mutation.
These results demonstratedthat the T790M mutation was not present in the blood before treatment witherlotinib in any of the 23 patients with presence of the sensitizing EGFRmutations in the pretreatment blood sample. In the end, the study reportedthat the T790M mutation continued to increase in all 9 patients until diseaseprogression while there was no trend toward an increase in the originalsensitizing EGFR mutation noted in patients who did not develop T790Mmutations (Sorensen et al., 2014).