Concomitant EML4-ALK rearrangement and EGFR mutation in non-small cell lung cancer patients: a literature review of 100 cases

The discovery of EGFR mutations and EML4-ALK gene rearrangements has radically changed the therapeutic scenario for patients with advanced non-small cell lung cancer. ALK and EGFR tyrosine-kinase inhibitors showed better activity and efficacy than standard chemotherapy in the first and second line treatment settings, leading to a clear advantage in overall survival of advanced non-small cell lung cancer patients harboring these genetic alterations. Historically the coexistence of EGFR mutations and EML4-ALK rearrangements in the same tumor has been described as virtually impossible. Nevertheless many recent observations seem to show that it is not true in all cases. In this review we will discuss the available literature data regarding this rare group of patients in order to give some suggestions useful for their clinical management. Furthermore we report here two cases of concomitant presence of both alterations that will help us in the development of discussion.


INTRODUCTION
Lung tumors are very common, both in men and women. They represent the leading cause of cancer death in males and the third most common cause in women. Although the 5-years survival rate has moderately increased in the last years (from 8% to 13%) lung cancer remains associated with a very poor prognosis. [1] The disease is frequently metastatic at diagnosis and therefore treatments are rarely curative. However recently, few but significant steps forward have been made in the treatment of advanced stage non-small cell lung cancer (NSCLC). [2,3] In recent years research has focused on molecular characteristics of lung cancer, highlighting the role of specific genes involved in tumor growth. These genes have been shown to be important therapeutic targets, especially in advanced adenocarcinomas. [4] In particular activating EGFR mutations were identified in 10-15% of Caucasian and in 40% of Asian patients. Their presence is the most important predictor of good response to EGFRtyrosine-kinase inhibitors (TKIs) therapy. In patients with advanced EGFR mutated NSCLC many trials compared chemotherapy and EGFR-TKIs in first line setting and showed the clear superiority of the latter. [5] Furthermore after the progression to first and second generation EGFR-TKIs (gefitinib, erlotinib and afatinib), the third-generation drugs, have showed very high activity, especially in patients harboring EGFR T790M mutation, that is the main cause of resistance to first line TKI therapy. [6,7] The rearrangement of ALK with EML4 oncogene on the chromosome 2 short arm is another important therapeutic target. Such aberration activates a specific tyrosine kinase, involved in the processes of survival and cell proliferation and it is found in about 3-7% of lung www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 35), pp: 59889-59900 Review adenocarcinomas in Caucasian patients. The determination of EML4-ALK gene rearrangement is necessary to select patients to be treated with specific ALK-TKIs. These drugs have shown to be more active than chemotherapy in all the lines of treatment in which they have been testes, even in patients progressing after one or two different ALK-TKIs. [8] These outstanding results underline the need of a thorough molecular characterization of advanced NSCLC, especially in never smokers with non squamous histology. Historically, the coexistence of EGFR mutations and EML4-ALK rearrangements have been described as extremely rare, perhaps because of its low prevalence and the sensitivity of diagnostic tools. In contrast recently published case reports and case series suggest that this is not true for all patients [9][10][11]. In this paper we performed an accurate literature review and we reported two new cases of patients diagnosed with lung adenocarcinoma, exhibiting both EGFR mutation and EML4-ALK rearrangement. Finally we did a qualitative synthesis of results of the review.

Case 1
Patient One was a 52-year-old never smoker Caucasian male, diagnosed with stage IV lung adenocarcinoma in May 2012. A CT scan showed lung and bone metastasis at diagnosis. A biopsy of the lung mass revealed the presence of EGFR exon 19 deletion and therapy with gefitinib was initiated. In August 2012, after 2 months of gefitinib (250 mg daily), the drug was stopped for liver and skin toxicity. The patient then received two cycles of chemotherapy with cisplatin and gemcitabine and reported progressive disease (PD) as best response. Considering the previous evidence of EGFR exon 19 deletion and the discontinuation of gefitinib due to toxicity, we decided to start treatment with erlotinib (150mg daily). Nevertheless lung and bone PD was evident after 4 months. So the patient was treated with pemetrexed and obtained partial response (PR) after 4 cycles and PD on mediastinal nodes after 8 cycles. In December 2013 palliative radiotherapy (RT) (total 30 Gy) on mediastinum and pulmonary hilum was performed. After a new pleural PD the assessment of EML4-ALK rearrangement on the original biopsy demonstrated the presence of EML4-ALK fusion gene. Crizotinib (500mg daily) was administered from February 2014 until October 2015 when a brain MRI showed the presence of brain metastases. So the patient underwent whole brain radiotherapy (WBRT) and started chemotherapy with vinorelbine. After 3 cycles, complicated by gastrointestinal toxicity, a CT scan showed lung PD and the drug was stopped. In April 2016 the ECOG performance status (PS) of the patient was 2 and he started treatment with ceritinib (750 mg daily) in the context of a clinical trial. Unfortunately, the disease progressed rapidly and the patient died in July 2016.

Case 2
Patient Two, a 43-year-old never smoker Caucasian woman, underwent a right lower lobectomy with systematic nodal dissection in April 2012. The initial pathologic diagnosis was lung adenocarcinoma stage IIIa (pT3 pN1 Mx), KRAS wild-type, harboring EGFR exon 19 deletion and EML4-ALK rearrangement. We decided to start adjuvant chemotherapy but in June 2012, before the start of treatment, a brain MRI showed multiple brain metastases and a CT scan demonstrated the presence of lung and nodal disease. So the patient underwent WBRT and started therapy with gefitinib (250 mg daily) which continued until March 2014. She obtained a PR as best response. In March 2014 a new brain MRI demonstrated a slight cerebral PD and the patient was switched to crizotinib (500 mg daily), that was stopped after 4 days because of a G4 gastrointestinal toxicity. Due to the patient's refusal of chemotherapy, she started erlotinib (150 mg daily). She continued therapy until January 2016, obtaining a stabilization of thoracic disease and a slow asymptomatic progression of cerebral disease. In January 2016 the brain MRI showed a significant cerebral PD and a CT scan evidenced lung and pleural PD. The patient was still in good clinical condition (ECOG PS = 0) thus she was enrolled in a clinical trial and received ceritinib (750 mg daily). After 2 cycles a chest/abdomen CT scan and brain MRI showed PR in both lung and brain. At present the patient keeps a good PS and continues ceritinib. The dose was reduced to 450 mg daily due to G2 gastrointestinal toxicity. The disease is stable.

Descriptive analysis and qualitative synthesis
We identified 98 cases of concomitant EGFR mutation and ALK rearrangement in NSCLC patients from a literature search. In this paper we will describe a total of 100 cases, including our 2 patients ( Table 1). The main patient characteristics are consistent with previous findings about patients with the sole EGFR mutation or EML4-ALK translocation [52]. There is a prevalence of women (51 female, 31 male, 18 not reported), Asian (53 Asian, 30 Caucasian, 17 not reported) and never smoker patients (58 never smokers, 15 smokers, 27 not reported). We are here considering all patients treated with at least one TKI (EGFR or ALK-TKIs) and since some patients had more than one treatment with the same class of TKI we are analyzing together all treatments administered for each class (lines of therapy). Fifty one patients received at least one EGFR-TKI, with a total of 56 EGFR-TKI lines administered. Considering only the 53 EGFR-TKI lines evaluable for response, independently from the setting, disease control (CR or PR or SD) was achieved in 37 cases (69.8%) with 23 objective responses reported (43.4%). Thirty-seven patients were treated with at least one ALK-TKI (39 evaluable lines of treatment). In these patients disease control was achieved in 31 cases (79.5%)

DISCUSSION
In recent years the study of molecular characteristics of lung cancer has underlined the pathogenic and therapeutic importance of specific genes involved in tumor growth. The most significant discovery was undoubtedly the role as therapeutic target of EGFR and ALK in the non squamous histology, but many other genes involved in the cell proliferation process have EML4-ALK = echinoderm microtubule associated protein like 4 -anaplastic lymphoma kinase, EGFR = epidermal growth factor receptor, pts = patients, TKI = tyrosine-kinase inhibitor, F = female, Y = years, mts = months, NS = never smokers, NR = not reported, BR = best response, N = number, SD = stable disease, PFS = progression free survival, PR = partial response, PD = progressive disease, M = male, CS = current smokers, NT = not treated, HS = high smokers, CR = complete response, LH = light smokers, FS = former smokers, NE = not evaluable. www.impactjournals.com/oncotarget been investigated. The oncogene KRAS is mutated in 20-30% of lung adenocarcinomas, particularly in heavy smokers and in the Caucasian population. However its role is still controversial both as a prognostic and predictive factor. Other molecular alterations have already demonstrated a promising therapeutic impact, especially in lung adenocarcinomas. They are represented by the ROS1 and RET gene rearrangements (about 1-2% of lung adenocarcinomas) and by the activating mutations of BRAF (V600E and others) and HER2. Also the FGFR1 and PDGFR amplification and the mutations of PI3KCA, PTEN and DDR2 could have future therapeutic implications, especially in squamous cell lung cancer. [4,12]. Anyhow at present only the discovery of EGFR and ALK molecular alterations has radically changed the clinical history of two specific subgroup of patients (EGFR mutated and EML4-ALK rearranged patients). [13,14] In the first years after its discovery the presence of ALK rearrangement has been described as mutually exclusive with the presence of EGFR mutations [15][16][17][18][19][20]. Some authors reported that ALK rearrangements were more common in patients with poorly differentiated adenocarcinoma whilst EGFR mutations were typical of well-differentiated cancers [15]. In the same way the coexistence of KRAS mutations with the other two alterations was described as virtually impossible. Thus the initial suggested algorithms for bio-molecular characterization of non squamous NSCLC proposed testing the histological samples for KRAS mutations followed by EGFR test only if KRAS was wild type. ALK testing was to be performed only if no alterations were found in previous tests [16][17][18][19][20].
However a higher number of subsequent observations demonstrated that the concomitant presence of EML4-ALK rearrangements and EGFR mutations is rare but not irrelevant [9][10][11] (Table 1), and also the coexistence with KRAS and other mutations is possible [46][47][48][49][50][51][52]. So at present this diagnostic algorithm cannot be considered the gold standard.
In regards to patients with both EML4-ALK rearrangements and EGFR mutations, literature data are conflicting. Yang et al [34] and Ulivi et al [49] reported a percentage of concurrent alterations of 1.3% and 1.6% respectively, whilst Lee et al described only 6 cases of concomitant alterations on 6637 NSCLC cases [48]. In some cases the presence of EGFR mutations is described as a resistance mechanism occurring in patients with EML4-ALK translocations treated with ALK-TKIs [46,51]. At the same time the development of EML4-ALK rearrangements could be an acquired resistance mechanism after EGFR-TKIs treatment, although the second event is considered more rare [23,41]. Other authors reported that both genetic alterations could be present from the beginning of tumor proliferation [23,34,35,39]. The different cellular clones represent the expression of the heterogeneity of tumors [39]. Moreover these two bio-molecular alterations may coexist in the same tumor cell [35]. In our opinion all mechanisms are possible and can probably occur in the same patient during the course of the disease.
Another hot topic is the variable response reported to both ALK-TKIs and EGFR-TKIs in the different published cases. Some authors hypothesized that the coexistence of two alterations could be predictive of poor response in patients treated with TKIs [24,43,46]. These data however, were not confirmed by other authors, that reported a good sensitivity to these treatments [22,25,33]. Furthermore some papers showed a better outcome only in patients treated with first line ALK-TKIs [34,40,48] whilst an unsatisfactory response was reported with first line EGFR-TKIs [30,31,35]. Again these results are not confirmed by other experiences [22,25,33,52].
Despite the increasing number of reported cases of coexistent alterations, these are still not frequent enough to allow the authors to define specific guidelines and at present it is not possible to establish the best sequence of treatments. The two cases reported here are an example of this ambiguous context. Both patients had concomitant EGFR mutation and ALK rearrangement ab initio. They were both treated upfront with an EGFR-TKI (gefitinib) followed by another drug of the same family (erlotinib). The first patient did not respond to either treatments, while in the second case we obtained a PR with gefitinib and a long SD with erlotinib. Both patients were also treated with two ALK-TKIs (crizotinib and ceritinib). The first patient achieved a long SD after crizotinib but did not respond to ceritinib. In the second case the response to crizotinib was not evaluable because it was withdrawn due to an important toxicity, but the patient reported an outstanding response with ceritinib. Thus also in our experience, data are very dishomogeneous and do not allow us to draw any conclusions. Nevertheless we can give some suggestions on the basis of the qualitative data synthesis.
Considering literature data and the great efficacy of both ALK and EGFR-TKIs, it is strongly advised that EGFR and ALK tests are performed ab initio in all advanced non squamous NSCLC. Although there is no literature consensus, and our analysis is only a descriptive non statistical data synthesis, ALK-TKIs seem to be slightly more effective than EGFR-TKIs in patients with both alterations (Table  2). Disease control or disease response are reported as best response in 69.8% and 43.4 % versus 79.5% and 51.3% of reviewed cases treated with EGFR and ALK-TKIs respectively. Twenty three patients received both EGFR and ALK-TKIs. In the great majority of cases (22 patients) EGFR-TKIs were administered before ALK-TKIs. Despite this fact, in this subgroup of patients, disease control was achieved in 61.5% versus 73.1% and disease response was seen in 23.1% versus 42.3% of cases treated with EGFR versus ALK-TKIs respectively. Furthermore also the role of chemotherapy in these patients is still debated [52]. Obviously, the very limited number of evaluable patients does not allow us to draw definitive conclusions but in our opinion, ALK-TKIs treatment could be the preferred first line approach, if there aren't any other data that could guide the therapeutic choice.
EGFR mutations, ALK translocations and many other bio-molecular alterations [53] may appear in the course of the disease as acquired resistance mechanisms after EGFR and ALK-TKIs treatment. Thus, if technically feasible, it is crucial to re-biopsy all patients with PD in order to detect all possible new alterations and design therapy on the basis of the new bio-molecular status.
Patients with EGFR mutations and EML4-ALK translocations have a higher incidence of brain metastases [54]. These are also present in our two cases. To date WBRT and Stereotactic Radiosurgery (SRS) represent the most used treatments in patients with intracranial involvement [55]. Nevertheless EGFR and ALK-TKIs, especially the new-generation drugs, have shown relevant intracranial activity [56][57][58]. Therefore the TKIs role and their correct place within the therapeutic strategy are still debated, especially for patients with brain metastasis and both alterations.
Since the two alterations may coexist ab initio, another very interesting future perspective could be to test the safety and the potential efficacy of a double inhibition of both ALK and EGFR. Designing clinical trials specific for these patients could be very fascinating and could answer many open questions. However, due the rarity of this condition, it would be difficult to recruit the right number of patients.
In conclusion, an accurate bio-molecular characterization is crucial to drive the therapeutic strategy of NSCLC with multiple driver alterations. Perhaps the allelic alteration's fraction in NGS or the liquid biopsies could be useful in defining the best therapeutic choice [52]. All available clinical data should be shared and analyzed in order to increase our experiences and correctly manage these patients.

MATERIALS AND METHODS
We performed a literature review on reported cases and case series of concomitant EML4-ALK rearrangement and EGFR mutation in patients with NSCLC. We collected all available clinical data and investigated treatment strategies used in the management of this rare group of tumors. We searched the digital databases using these keywords: "Non-Small Cell Lung Cancer", "EML4-ALK rearrangements", "EGFR mutations", variously associated with "concomitant", "coexistence", "coexistent", "concurrent", "both", "same tumor", "same cancer", "same neoplasm". We did not restrict the search to English language but we included in the analysis only the publications for which an English abstract was available. We focused our attention mainly on the drugs used and treatment efficacy. We read the full text of selected papers (abstracts were analyzed when the English version of the article was not present). The reference list of selected papers was also used in the selection process.
We examined 923 papers, 865 works were excluded based on title and abstract content while, the remaining 58 papers were included in our review. Thirty three studies, were selected for qualitative synthesis (Table 1), (Figure 1).

Author contributions
GLR, MI, GC, FDB, MCG: study concept and design; acquisition of data; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript; study supervision. CP, DS, MV, MG, LB, NZ: acquisition of data; analysis and interpretation of data; critical revision of the manuscript; study supervision.