Did Everolimus Break the Rules?

Along with advances in genomic sequencing, there is a growing number of alterations that are being identified and evaluated for matched targeted therapy. Some of the successful examples of this approach include, but are not limited to, EGFR inhibitors for EGFR-mutated non–small cell lung cancer, BRAF inhibitors for BRAF V600–mutated non–colorectal cancers (although patients with colorectal cancer with BRAF V600E can be managed with the addition of anti-EGFR therapy), and Trk inhibitors for NTRK fusions in multiple cancer types. In fact, the latter example was approved in the United States based on the presence of the alteration rather than in a specific histology. Emerging from the successes are a few generalizations which we have called the “Rules of Targeted Therapy”. These include the ideas that the alterations are relatively uncommon in incidence but must be present for the agent to be effective; therapy is associated with high response rates but resistance eventually emerges; and toxicity of these agents is unique to the class and education of providers is essential to broad implementation.

In this issue of Clinical Cancer Research, Adib and colleagues, successfully conducted a phase II study with everolimus, an oral mTORC1 inhibitor, to target inactivating mutations in TSC1/TSC2 or activating MTOR mutations among patients with advanced solid tumors (1). Prior reports, albeit small case series, have demonstrated remarkable responses with mTORC1 inhibitors among patients with TSC1/TSC2 or MTOR alterations. Therefore, the authors conducted a pan-cancer basket study to further evaluate the efficacy of everolimus in this patient cohort. Among 30 patients enrolled, most patients had inactivating alterations in TSC1 (13/30) or TSC2 (15/30). One patient had concurrent TSC1 and TSC2 alterations and another patient had an MTOR mutation. Despite all patients having the biomarker of interest, the efficacy was limited with a low response rate [7% (2/30)], with short median progression-free survival [PFS; 2.3 months; 95% confidence interval (CI), 1.8–3.7 months] and median overall survival (OS; 7.3 months; 95% CI, 4.5–12.7 months).

It appears that these results do not follow the Rules. However, several important challenges may have been present that counteracted the conventional paradigms of targeted therapy. In this study, patients were relatively heavily pretreated with a median of 2.5 (range 0–8) prior systemic therapies. Hence, it is plausible that underlying genomic alterations may have changed at the time of intervention. Moreover, during the therapy with everolimus, the cancer may have developed rapid resistance that bypassed the therapeutic pressure. This reminds us that cancer genomic alterations are dynamic. Moreover, all patients harbored at least one concurrent genomic alteration that could, in theory, lead to primary resistance to monotherapy with everolimus. Furthermore, the PI3K/Akt/mTOR signaling pathway is regulated by multiple proteins and protein kinases with complicated interactions. mTOR itself, consists of two multiprotein complexes, mTORC1 and mTORC2, and targeting mTORC1 alone with everolimus may be insufficient. Thus, understating the alteration status of TSC1/TSC2 or MTOR alone may not translate into successful targeted therapy.

Although the clinical outcome was somewhat disappointing, we should not give up on finding alternative or complimentary ways to manage cancers harboring TSC1/TSC2 or MTOR alterations. As Thomas Edison said, “Many of life's failures are people who did not realize how close they were to success when they gave up”. To overcome the limitations of the single agent targeted therapy approach, we will likely need to explorer a customized targeted therapy approach based on patients' underlying cancer profiling. For example, alpelisib (PI3K inhibitor) in combination with fulvestrant (antiestrogen) is now FDA approved for metastatic breast cancer with PIK3CA mutation (which is targetable with alpelisib) and positive hormone receptor status (which is targeted with fulvestrant) – both agents are necessary to achieve optimal efficacy. However, in the pan-cancer setting, the situation is more complicated. As shown by the authors, all patients had at least one coexpressed genomic alteration along with mutations in TSC1/TSC2 or MTOR and every patient appears to have a unique genomic profile pattern. Thus, every patient may require an “N-of-one” customized therapy approach. For example, if patients have coalterations in BRAF, FGFR, or cell-cycle pathway, they may require combination therapy with BRAF, FGFR, and cell-cycle inhibitors, respectively, along with everolimus (Fig. 1). We have previously shown that a customized combination approach (N-of-one strategy), that was supported by molecular tumor board recommendations, was feasible and associated with significant improvement in clinical outcomes including response rate, PFS, and OS among patients with multiple solid tumors (2, 3). Moreover, NCI-ComboMATCH will be assessing the multiple combination approach to overcome drug resistance to the single-agent approach. Further assessment of the “N-of-one” genomically matched combination approach is warranted to understand the optimal dosage for safety and tolerability as well as the efficacy.

Another potential approach is to actively understand the heterogeneity and especially dynamic changes in genomic alterations that can bypass the therapeutic perturbation by obtaining serial biopsies, ideally from both blood and tissue, during therapy that may redefine the targets (Fig. 1). Such an effort is ongoing and the preliminary report, not surprisingly, demonstrated remarkable heterogeneity among biopsies (Serial Measurements of Molecular and Architectural Responses to Therapy trial; ref. 4).

Therefore, did everolimus in this study not play by the Rules or was it the exception that proved the Rules? Given the limitations discussed above, the latter conclusion may be more fitting. Notwithstanding, the authors should be congratulated for summarizing and reporting a seemingly negative study that showed minimal clinical efficacy. Reporting negative results can be challenging, however, it is critical to advance science since we can learn a great deal even when our hypotheses are not proven. One good example is BRAF V600E mutations in colorectal cancer where the initial clinical trial with a BRAF inhibitor alone was ineffective. Further investigation revealed colorectal cancer expresses high levels of EGFR that can abrogate the effects of BRAF inhibition, and the addition of anti-EGFR therapy can overcome resistance; this is now an FDA approved combination approach (5, 6). Even with the lack of activity seen with everolimus in the present study, the feasibly of enrolling an TSC1/TSC2/MTOR altered basket study was demonstrated. Hence a negative study is still a step forward to success.

S. Kato serves as a consultant for Foundation Medicine, NeoGenomics, and CureMatch. In addition, he reports speaker's fees from Roche and advisory board for Pfizer, and research funding from ACT Genomics, Sysmex, Konica Minolta, and OmniSeq. E.E. Cohen reports other support from Bayer, BioNTech, Eisai, Gilead, Merck, MSD, and Regeneron outside the submitted work.