TP53 mutations are predictive and prognostic when co-occurring with ALK rearrangements in lung cancer

The entity now known as anaplastic lymphoma kinase (ALK) gene rearrangement positive non-small-cell lung cancer (NSCLC) was first identified in 2007 [1] and the enrollment of patients with these NSCLCs into the original clinical trial of the multitargeted tyrosine kinase inhibitor (TKI) crizotinib dates back to 2008 [2–5]. Fast forward a decade and the preclinical to clinical advances seen in ALK rearranged NSCLC place it at the forefront of the precision oncology revolution that has transformed the palliation of advanced NSCLCs [6]. Four different ALK TKIs—crizotinib (since 2011), ceritinib (since 2014), alectinib (since 2015) and brigatinib (since 2017)—are available worldwide with additional inhibitors, such as lorlatinib, in the last stages of regulatory approval [7–10]. The pace of clinical trial development has been so brisk that evidence-based standards have shifted multiple times and the ‘oldest’ ALK TKI crizotinib has been displaced in the initial line of treatment by the ‘next-generation’ ALK TKI alectinib, which is associated with exceedingly high initial overall response rate and median progression-free survival (PFS) times that can almost reach 3 years [11]. Although biological resistance, mainly through the development of ALK kinase resistant mutations (one example ALK-G1202R), is invariable with ALK TKI monotherapy and the central nervous system a common site of progression; significantly active ALK inhibitors such as lorlatinib can transiently control these resistant clones or brain sanctuary sites, respectively [12, 13]. The majority of patients with advanced ALK rearranged NSCLC can expect to receive sequential oral TKI monotherapy for prolonged periods of time with reported median overall survival (OS) that exceeds 4 years and nearly half of initially treated patients will be 5-year survivors following diagnosis [14, 15], a true shit in the natural history of this subtype of lung cancer.

However, there is significant heterogeneity in how an individual patient with ALK rearranged tumor will benefit from ALK TKIs and other therapies. A minority of patients can have exceedingly high (>5 years) PFS times with crizotinib [14] while others can rapidly progress in the central nervous system or systemically with below 1 year in OS times [16]. This heterogeneity remains unexplained but prevailing hypotheses harken to differences in patient characteristics (smoking status, age, co-morbidities, metabolism of ALK TKIs, immune system status) and more importantly in co-occurring genomic aberrations that can modulate the benefit of ALK TKIs (i.e. can be putative predictive biomarkers) or survival (i.e. can be putative prognostic biomarkers). The latter are of extreme importance not only for ALK rearranged NSCLC—where initial efforts have indicated that even the ALK fusion partner may alter clinical outcomes [17]—but also for other oncogene-driven NSCLCs. As an example, epidermal growth factor receptor (EGFR) mutated NSCLCs are known to harbor a multitude of co-occurring genomic events when analyzed by comprehensive genomic profiling, including mutations in: tumor protein P53 (TP53), phosphatidylinositol-4, 5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), phosphatase and tensin homolog (PTEN), Erb-B2 receptor tyrosine kinase 2 (ERBB2), MET proto-oncogene receptor tyrosine kinase (MET) among others [18, 19]. Some of these co-occurring genomic events can completely abrogate the ability to induce a response to EGFR TKI monotherapy, as in the case of high-level ERBB2 or MET amplification [20], while others have less impact on initial EGFR TKI response, as is the case of PIK3CA mutations and heterozygous PTEN mutations [18, 19]. TP53 mutations in these EGFR mutated cohorts are both predictive of shorter duration of tumor responses to EGFR TKIs and also prognostic of shorter lifespans in these patients [18, 19]. TP53 mutations are the most common mutations found in NSCLCs and co-occur frequently with driver oncogenes. Indeed, the Lung Cancer Mutation Consortium (LCMC) confirms that TP53 mutations are the most common co-occurring event with EGFR mutations or ALK rearrangements or ROS proto-oncogene 1 receptor tyrosine kinase (ROS1) or other driver alterations; and the presence of TP53 mutations is one of the strongest prognostic markers for shorter survival times in advanced lung cancers [21]. The underlying biological basis for the prognostic impact of TP53 mutations—a surrogate for P53 protein loss—is still an active area of investigation but it is already known that this tumor suppressor has critical antiproliferative/antiapoptotic functions [22], its loss can accelerate the transforming potential of oncogenes in lung cancers [23] and its loss can hamper tumor response to TKIs [24].

With this background, Kron reports in this issue of Annals of Oncology a detailed analysis of co-occurring mutations in ALK rearranged NSCLC [25]. A total of 216 tumors are analyzed and pathogenic TP53 mutations are identified in close to a quarter (23.8%) of these cancers. Other genomic aberrations using the author’s limited gene panel have frequencies below 1%–5%. As seen in the aforementioned cases from the LCMC of co-occurring driver oncogene mutations (in EGFR, ALK, ROS1 and others) with TP53 mutations, the presence of an ALK rearrangement with a TP53 mutation is both predictive of shorter durations of PFS to ALK TKIs and is prognostic for shorter OS times. The authors carry out a detailed statistical analysis that takes into account patient characteristics (age, sex, smoking history, current smoker status, performance status and number of brain metastases), treatment choices (number of treatment lines before crizotinib and number of treatment lines before ceritinib) and the genomic TP53 mutation status. Only current smoker status and TP53 mutations are significant negative prognostic factors for OS in univariate analysis, and merely TP53 mutations remain a negative prognostic factor in multivariate Cox regression analysis.

The results highlight the powerful prognostic role of TP53 mutations but it is unclear how they will alter the current treatment paradigms in ALK rearranged NSCLC. Unfortunately, therapeutic efforts to re-establish P53 protein function in cancer have been disappointing and no candidate is available clinically [26]. Therefore, the prevailing question remains if a provider will omit the use of an ALK TKI in a tumor with an ALK rearrangement co-occurring with a TP53 mutation? The answer is likely no, as ALK TKIs are significantly superior to other forms of approved NSCLC therapy—be them cytotoxic chemotherapy or immune checkpoint inhibitors—irrespective of TP53 mutational status [27, 28]. Indeed, worldwide only a fraction of ALK rearranged NSCLCs are diagnosed by comprehensive genomic profiling assays that incorporate analysis of TP53 mutation status [6, 29]. This scenario compiled by the lack of recommendation for TP53 mutation analysis in diagnostic specimens with NSCLC [30, 31] limit the real-world applicability of considering TP53 mutations status as a predictive and prognostic marker in the offices of most oncologists that treat patients with these tumors. Kron concludes their work affirming that ‘future clinical trials stratification of this patient subgroup should be considered’ and that ‘new treatment strategies should be investigated to improve the outcome of ALK/TP53 co-mutated patients’ [25]. I agree that as ALK TKI monotherapies improve the duration of control for ALK rearranged NSCLC and novel combination therapies are tested in clinical trials, it may be important to understand prognostic markers such as TP53 mutation status in future trial designs. And I wholeheartedly recommend that the improvement and/or development of therapies for tumors driven by an oncogene with P53 function loss should be prioritized by academic plus pharmaceuticals consortiums evaluating advanced NSCLCs.

Funding

This work was funded in part through an American Cancer Society grant RSG 11-186 (to DBC) and National Institutes of Health (NIH)/National Cancer Institute (NCI) grant CA218707 (to DBC).

Disclosure

DBC has received consulting fees and honoraria from Pfizer, Astra-Zeneca and ARIAD/Takeda.