Similitude and evolution of treatment algorithms

Key words::

• ALK-positive NSCLC
• Ceritinib
• Crizotinib
• Non-small-cell lung cancer

The discovery in 2007 that 3–5% of cases of non-small cell lung cancer (NSCLC) are ALK positive has led to important therapeutic advances for patients with metastatic or locally advanced ALK-positive NSCLC1. This molecular descriptor arises from a small inversion at chromosome 2 p resulting in a constitutively active chimeric fusion gene. This is most frequently EML4 (Echinoderm microtubule associated protein 4) ALK; however, other fusion partners may arise. ALK-rearranged tumors tend to have a solid signet ring or mucinous cribriform histology. Chromosomal abnormalities involving ALK can be translocations, amplifications or mutations. The prognostic implications of ALK translocations, the most common abnormality, are confounded by the current availability of therapeutically efficacious medications that target ALK. One study of resected lung adenocarcinomas, however, found that ALK or ROS1 fusion-positive tumors have a significantly poorer disease free survival on multivariate analysis after adjusting for age, gender, T stage, N stage and adjuvant chemotherapy use (p = 0.022; hazard ratio, 2.11)2. Another study of patients with non-small-cell lung cancer at various clinical stages found that stage IIIA EML4-ALK positive patients has a worse disease free survival on multivariate analyses, further establishing the oncogene addicted nature of these tumors3.

The study of Guérin et al. is a real world perspective on the outcome and cost implications for ALK-positive NSCLC patients that discontinue crizotinib monotherapy4. Crizotinib inhibits ALK and the receptor tyrosine kinases MET and ROS1. There have been important changes in practice since data accrual for this study in 2013. This retrospective evaluation compiled information from three US administrative claims database. Required enrollment criteria included a diagnosis of locally advanced or metastatic ALK-positive NSCLC. There has been a temporal evolution since 2013 in the treatment of these patients with crizotinib found to be superior to first line combination treatment with cisplatin and pemetrexed in metastatic ALK-positive NSCLC5. Median progression free survival for the crizotinib treated group was 10.9 months, compared to 7 months for chemotherapy treatment (hazard ratio for progression or death 0.45, p < 0.001). Previously crizotinib was not administered in the first line setting. A second advance has been the approval of ceritinib, a new small molecule ATP-competitive tyrosine kinase inhibitor of ALK, by the US Food and Drug Administration in April 2014. Ceritinib, in addition to ALK inhibition, also inhibits the insulin-like growth factor 1 receptor at 50 times lower potency than ALK. A phase 1 trial with dose escalation and extension cohorts found that within a subgroup of 80 previous recipients of crizotinib there was a response rate of 56% to ceritinib at a dose of at least 400 mg per day6. The recommended dose of ceritinib is 750 mg orally once per day. Median progression free survival was 7 months (95% CI 5.6–9.5 months). The reinduction of remission effected by ceritinib has again changed the treatment algorithm. Responses were observed to ceritinib in patients with established resistance mutations in ALK as well as in patients without detectable mutations. Ceritinib is also efficacious in treating CNS metastatic disease. In the study of Guérin et al. patients with brain metastasis had an overall survival of 44 days post discontinuation of crizotinib. Other emergent medications of relevance include alectinib, a potent second-generation ALK inhibitor that inhibits ALK with or without the L1196M gatekeeper mutation.

Acquired resistance in oncogene-addicted tumors almost invariably emerges from treatment with molecular therapeutics. EML4-ALK mutations can confer resistance to ALK inhibitors7. In a described case of a 28-year-old patient with metastatic adenocarcinoma of lung with acquired resistance to crizotinib after 5 months, 4374G to A and 4493C to A substitutions were identified. These caused cysteine to tyrosine and leucine to methionine changes, at amino acids 1156 and 1196 respectively. Leucine 1196 of ALK corresponds to threonine at position 315 in ABL and position 790 in EGFR. Each of these is the locus for the most commonly identified acquired mutations that confer resistance to tyrosine kinase inhibitors. Other mechanisms of resistance to crizotinib include amplification of EML4-ALK, as well as activation of alternative signaling pathways, including secondary mutations in EGFR and KRAS with loss of ALK translocation8.

The described survival of 61 days when patients discontinue crizotinib monotherapy is going to be obviated by the ascent of crizotinib up the treatment algorithm. It will also be changed by the availability of other targeted medications that treat secondarily resistant tumors or that have pharmacokinetic profiles, which include distribution into tumor sanctuary sites such as the CNS. But the study poses an interesting question. Can molecular therapeutics be completely discontinued at reaching the end of the treatment algorithm in ‘oncogene addicted tumors’? In the phase 3 placebo controlled RIGHT trial, of imatinib in metastatic or unresectable gastrointestinal stromal tumors (GIST) after failure of imatinib and sunitinib, median progression free survival was 1.8 months with imatinib compared with 0.9 months with placebo (hazard ratio for progression or death 0.46, 95% CI 0.27–0.78; p = 0.005)9. Of the placebo group 93% crossed over on progression. The inference is that in GIST refractory to the then standard tyrosine kinase inhibitors, the disease continues to have tumor subclones, which remain sensitive to kinase inhibitors. The analogous scenario in ‘oncogene addicted’ ALK-positive NSCLC remains untested. Just like in GIST with the validity of the question being partially diminished by the availability of regorafenib, the importance of the question in ALK-positive NSCLC is diminished by ceritinib10. The phenomena of tumor heterogeneity and Darwinian selection of resistant clones are relevant and the answer may only lie in individualized treatment decisions. These perhaps will require imaging modalities such as positron emission tomography to augmenting treatment selection11.

ALK-positive and EGFR mutant NSCLC are usually mutually exclusive. The phenomenon of treatment resistant EGFR mutant NSCLC is of interest. In that disease, resistance to tyrosine kinase EGFR inhibition may be conferred by a T790M EGFR substitution or MET amplification. A cancer cell population described as ‘drug tolerant persisters’ is recognized that is refractory to initial treatment via an IGF1-R mediated epigenetic program. EGFR-inhibitor-induced STAT3 signaling is another resistance mechanism. More recently it has been found that EGFR inhibition induces an EGFR-TRAF2-RIP1-IKK complex, which stimulates an NF-κB-mediated transcriptional survival program12. This adaptive survival mechanism can be inhibited by PBS-1086 raising the prospect that dual EGFR and NF-κB inhibition may subvert the emergence of treatment resistant subclones. A future analogous finding, not necessarily involving NF-κB but targeting an evolving resistance mechanism, in ALK-positive NSCLC raises the prospect that combination treatments may lessen the adaptive Darwinian selection of treatment resistant subclones. This would lead to a trend in treatment from successive reactive monotherapy steps in tumor management to a proactive combinatorial approach. The exact combinations remain to be defined but the temporal and spatial diversity of genomic instability within the evolution of lung cancer is already being characterized11. Branched evolution of lung adenocarcinomas which has been characterized by sequencing, has identified mutations consistent with APOBEC-mediated mutagenesis. These were more pronounced on the branches compared to the trunk. APOBEC, is an enzyme, which mediates mutagenesis by cytidine deamidase activity. Branched driver genes including PIK3CA, EP300, TGFBR1, AKAP9 and PTPRD have mutations compatible with an APOBEC mediated context inferring a possible impact of APOBEC on subclonal expansion.

Driver genetic aberrations such as EML4-ALK translocations tend to occupy truncal locations. While the published branched evolution maps are descriptions of tumor molecular progression, therapeutic inferences about treatment resistant subclones can be drawn based on the altered tumor microenvironment of ALK therapeutics. Targeting enzymes that mediate mutagenesis (APOBEC causes deleterious mutations in viruses within healthy individuals) may be a first step in the combinatorial approach. Rather that waiting for secondarily resistant mutations to emerge subverting the mechanism by which resistant mutations arise in the first instance may be the tumor’s ‘Achilles heel’.

Transparency

Declaration of funding

This editorial was not funded.

Declaration of financial/other relationships

F.C.K. has disclosed that he has no significant relationships with or financial interests in any commercial companies related to this study or article.

The CMRO peer reviewer on this manuscript has no relevant financial or other relationships to disclose.