Evolutionary Precision Medicine: A Role for Repeat Epidermal Growth Factor Receptor Analysis in ALK-Rearranged Lung Adenocarcinoma?

Abstract

Genetic alterations in the anaplastic lymphoma kinase gene (ALK) that result in fusion proteins are detected in approximately 1% to 2% of unselected non–small-cell lung cancer (NSCLC) tumors. In NSCLC, the most common fusion partner is EML4, but several other ALK fusion partners that result in a constitutively active kinase domain with potent transforming capability in vitro have been described. NSCLC tumors positive for the ALK fusion protein are more commonly of adenocarcinoma histology, and the patients tend to be younger with little or no smoking history. ALK rearrangements have previously been mutually exclusive with activating epidermal growth factor receptor gene (EGFR) and KRAS mutations, and ALK-rearranged NSCLC is highly sensitive to targeted therapy with crizotinib; response rates are approximately 65%. Unfortunately, most patients experience disease progression within 1 year, and crizotinib resistance often is mediated through secondary acquired mutations in ALK itself. In 2014, ceritinib, a second-generation ALK inhibitor, received US Food and Drug Administration approval. Ceritinib overcomes most resistance mutations in ALK and shows an overall response rate of 56% among patients previously treated with crizotinib (55%, partial response; 1%, complete response). However, in one study of acquired resistance to ALK inhibition in 11 patients with ALK-positive disease who experienced relapse after crizotinib treatment, three patients had outgrowths of clones without the ALK rearrangement, which suggests that ALK-negative subclones were present in the pretreatment primary tumor, the ALK rearrangement was lost from the cancer genome during tumor progression, or that the patient was suffering from two tumors of independent clonal origins. One of these patients who experienced relapse had an activating EGFR mutation, which indicated that both the ALK fusion and the EGFR mutation may have existed in distinct cells or subclones within the tumor as a result of intratumor heterogeneity (ITH). ITH has in recent years been shown to be an inherent property of cancer across many types, including lung cancer. By serving as a considerable reservoir of diversity, it is likely a contributing factor to acquired therapy resistance and subsequent disease relapse. Indeed, we and others have previously shown how mutations in driver genes may display an illusion of clonality, which presents as clonal in one region of a tumor but may be entirely absent or subclonal in another region of the same tumor. However, despite its widespread prevalence, the true clinical relevance of ITH remains to be tested in prospective clinical studies, such as TRACERx. In the article that accompanies this editorial, Cai et al report on the prevalence of ALK and EGFR mutations in a study of 629 patients with lung adenocarcinoma who underwent surgical resection at Shanghai Pulmonary Hospital between 2004 and 2010. Among these, 30 tested positive for ALK fusions, and 364, for EGFR mutations. The authors acquired formalin-fixed paraffin-embedded sections from 20 patients with ALK-rearranged tumors who underwent surgical resection in 2010 and an additional 20 patients randomly selected among the population with somatic EGFR mutations. Laser capture microdissection of spatially separated tumor cell populations was performed to assess the ITH of ALK fusions and EGFR mutations. Selected areas were tested for both aberrations with multiplex reverse-transcriptase polymerase chain reaction. The authors found considerable heterogeneity in lung adenocarcinoma among the microdissected tumors; importantly, among the 20 patients with ALK rearrangements assessed for ITH, two patients harbored concomitant ALK fusions and EGFR mutations. Coexistence of ALK fusions with EGFR mutations has been previously reported in a cohort of 46 patients with lung adenocarcinoma, in which one co-occurrence was detected even though the authors specifically selected a cohort enriched for EGFR wild-type status. Others have detected rare concurrent ALK translocations and BRAF mutations. Cai et al have taken these findings further and analyzed four separate tumor areas from the two patients with both genes altered for ALK fusions and EGFR mutations. Here, they found that in patient 1, one area was negative for both aberrations, two were positive for both, and one was positive for EGFR but negative for ALK. In patient 2, three areas were positive for EGFR, whereas one area was positive for both somatic alterations, highlighting the importance of considering regional heterogeneity of driver events. DNA from three separate areas from a new formalin-fixed paraffin-embedded section from patient 1 was subsequently submitted for targeted next-generation sequencing, and the type of EGFR mutation and the relative abundance of each type of aberration were estimated in each area. The authors found that all three areas had EGFR mutations but only two also had ALK fusions. In the overlapping areas, the relative abundance of ALK fusions and EGFR mutations did not match, which suggests that at least some JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 33 NUMBER 32 NOVEMBER 1