CME
Physicians: Maximum of 0.75 AMA PRA Category 1 Credit™
Released: October 29, 2020
Expiration: October 28, 2021
In this module, Marwan Fakih, MD, discusses therapeutic options in development for the treatment of solid tumors with KRAS p.G12C mutations, including novel agents such as sotorasib (previously AMG 510) and MTRX849 as well as methods to possibly address resistance in this patient population.
The key points discussed in this module are illustrated with thumbnails from the accompanying downloadable PowerPoint slideset that can be found here or downloaded by clicking any of the slide thumbnails in the module alongside the expert commentary.
Clinical Care Options plans to measure the educational impact of this activity. Some questions will be asked twice: once at the beginning of the activity, and then once again after the discussion that informs the best choice. Your responses will be aggregated for analysis, and your specific responses will not be shared.
Before continuing with this educational activity, please take a moment to answer the following questions.
KRAS, together with other RAS isoforms, have long been described as one of the most prevalent oncogenes in cancer. A study by Khan and colleagues2 demonstrates the incidence of KRAS mutations (shown in blue) and amplifications (shown in orange) across several solid tumor types. KRAS mutations have demonstrated variable prognostic value depending on the type and stage of each malignancy.
In CRC, KRAS mutations have been associated with a poor prognosis, particularly in the metastatic disease.2 There are mixed reports regarding the impact of specific KRAS mutations on overall outcome in patients with stage III CRC, but KRAS mutations are associated with decreased survival after relapse in patients with stage III disease.3,4 There is no strong evidence, however, that KRAS mutations are associated with an increased risk of recurrence in this patient population.
In lung cancer, KRAS mutations demonstrate a worse disease-free survival and OS and an increased risk of recurrence in resected NSCLC.5,6 Therefore, KRAS mutations are known as a poor prognostic indicator, through decreased OS, in patients with NSCLC compared with patients with wild-type NSCLC.7
KRAS p.G12C is an activating oncogenic mutation, which results in KRAS remaining predominantly in the GTP-bound active form, thereby enhancing cellular proliferation and survival signaling, driving tumor growth across various cancers.8-10 It is one of the most common KRAS mutations in cancer and occurs in approximately 13% of NSCLC, and 1% to 3% in CRC and other solid tumors.1 KRAS p.G12C mutations have been associated with a clear negative prognostic indication on OS in patients with CRC in the setting of metastatic disease.11
Presently, KRAS p.G12C is the only KRAS mutation that is targetable with small molecule inhibitors such as sotorasib and adagrasib. These inhibitors bind to GDP-bound KRASG12C at the mutant cysteine 12 moiety and prevent activation of KRASG12C into the GTP-bound state, blocking downstream signaling from KRAS.10 Both sotorasib and adagrasib have been associated with clinical benefit in patients with the KRAS p.G12C mutation.1,12
When examining the clinical and genomic characteristics associated with KRAS p.G12C vs other KRAS mutations and wild-type KRAS in patients with cancer, a higher tumor mutational burden, greater association with tobacco signatures and PD-L1 expression, and slightly higher incidence in women are associated with the G12C mutation.13
To investigate the utility of targeting KRASG12C with small molecules, my colleagues and I designed the CodeBreaK100 trial to evaluate the safety and efficacy of sotorasib in patients with advanced solid tumors.1 CodeBreaK100 was a multicenter, open-label phase I clinical trial that enrolled patients with advanced solid tumors and a KRAS p.G12C mutation. Patients were required to have an Eastern Cooperative Oncology Group performance status ≤ 2 and disease progression on standard systemic therapy or had no appropriate standard therapy available. Patients with active brain metastases were excluded.
The dose escalation in this trial ranged from 180 mg to 960 mg once daily. The highest dose—960 mg—was identified as the dose for the expansion cohort. The primary endpoint of the study was safety and tolerability and the secondary endpoints included ORR, duration of response, PFS, and pharmacokinetics.1
Overall, sotorasib was well tolerated, but 56.6% of patients experienced a treatment-related adverse event (AE), with 11.6% with grade ¾ AE and 1.6% with a serious AE.1 The most common AEs reported in this trial were diarrhea (29.5%), fatigue (23.3%), and nausea (20.9%). Of note, 2 patients (1.6%) experienced a grade 3/4 elevation in alanine aminotransferase/aspartate aminotransferase. Although 1 patient with a grade 3 AE chose to discontinue therapy, the patient with the grade 4 treatment-related AE resolved to baseline levels following dose reduction and tapering of glucocorticoids.1 Otherwise, no dose-limiting AEs were observed, and no treatment-related AE led to death.
Following the identification of the recommended phase II dose of sotorasib 960 mg/day, which was associated with good pharmacokinetics and full inhibition of the target KRASG12C, the study enriched enrollment of 3 subgroups of patients with previously treated advanced cancers harboring the KRAS p.G12C mutation: NSCLC, CRC, and all other solid tumors.
In total, 59 patients with advanced NSCLC were enrolled on the trial, with 34 patients receiving sotorasib at the 960-mg dose.1 These 34 patients were included in the dose-escalation and dose-expansion cohort.
The waterfall plot for evaluable patients with advanced NSCLC shows that most patients had stable disease or an objective response. The ORR for all patients was 32.2% and 35.3% for patients who received the 960-mg dose. The disease control rate (DCR) was 88.1% across all dose levels and 91.2% at the 960-mg dose level.
Of note, one patient with progressive disease in the escalation cohort is not represented in this waterfall plot as the patient discontinued study treatment prior to the first tumor assessment. One additional patient was unable to be evaluated.1
The median time to response with sotorasib was 1.4 months (range: 1.1-9.5). The median duration of response was 10.9 months (range: 1.1+ to 13.6+ months), with responses still ongoing at data cutoff.
Among the 34 patients with advanced NSCLC who received sotorasib at the recommended phase II dose of 960 mg/day, 9 experienced durable clinical benefit and remained on treatment at the time of the data cutoff. The median PFS was 6.3 months in this patient population.1
In total, 42 patients with previously treated advanced KRAS p.G12C–positive CRC were also enrolled on CodeBreaK100 with 25 of these patients treated at the recommended phase II dose of 960 mg/day.1 At a median follow-up of 12.8 months, 39 out of these 42 patients had discontinued treatment; 37 discontinued due to disease progression.
Out of 42 patients with CRC, the ORR was 7.1%, with a DCR of 73.8% and a median PFS of 4 months. Patients treated with sotorasib 960 mg similarly achieved an ORR of 12% and a DCR of 80%.
Confirmed PRs were identified in 3 of the 42 patients for all dose levels and in 3 of the 25 patients in the 960-mg treatment group. These 3 patients remained on treatment at the time of data collection, and 1 patient has experienced an objective response. Of note, 68% of patients treated at 960 mg/day achieved stable disease and approximately 16% of patients had either a nonevaluable status or progressive disease. The median duration of stable disease in the 960 mg/day group was 4.2 months.1
The median PFS was 4 months and data from preliminary reports reflect a median OS of 10.1 months across all dose levels of sotorasib in the cohort of patients with advanced CRC.1,12 Furthermore, the median OS was not evaluable at the time of data cutoff for the 960 mg dose level with OS estimates of 96% at 3 months and 82.9% at 6 months.12
The third expansion cohort consisted of patients with solid tumors excluding CRC and NSCLC (n = 28) and included 12 patients with pancreatic cancer, 4 with appendiceal cancer, 2 with endometrial cancer, 2 with an unknown primary cancer, and 1 patient each with 8 other primary tumors. This cohort of patients was heavily pretreated with 60% of patients having received at least 3 previous lines of therapy.1
Four patients in this cohort achieved a PR: 1 patient with melanoma, 1 patient with endometrial cancer, 1 patient with appendiceal cancer, and 1 patient with pancreatic cancer. Another 17 patients achieved stable disease including 8 out of the 10 patients with pancreatic cancer. Stable disease was noted in multiple other tumor types including appendiceal, ampullary, bile duct, endometrial, gastric, sinonasal, small bowel, and 2 unknown primary cancers
In summary, the results of the CodeBreaK100 trial demonstrate that sotorasib has activity across various solid tumors with the KRAS p.G12C mutation. At the recommended phase II dose of 960 mg/day, the highest activity was seen in patients with KRAS p.G12C mutation–positive advanced NSCLC (ORR: 35.3%; DCR: 91.2%). Less robust activity at this dose level was noted in patients with KRAS p.G12C mutation–positive advanced CRC (ORR: 12%; DCR: 80%). Activity was also noted in other solid tumors, including KRAS p.G12C–mutated pancreatic cancer, endometrial cancer, appendiceal cancer, melanoma, and biliary cancer.
Another small molecule inhibitor of KRASG12C that binds at the cysteine residue, locking the protein in an inactive form and inhibiting downstream signaling, is adagrasib, previously known as MRTX849.
Adagrasib was evaluated in the KRYSTAL-1 study, a first-in-human phase I/II clinical trial that enrolled patients with KRAS p.G12C mutation–positive solid tumors who had progressed on standard of care or who had no available standard of care. The planned dose levels for the escalation cohorts were 150 mg, 300 mg, 600 mg, and 1200 mg daily given orally.
The dose expansion cohort was 600 mg twice daily given orally. The primary endpoints of this study were safety, pharmacokinetics, and ORR by Response Evaluation Criteria in Solid Tumors. Secondary endpoints included to define the maximum tolerated dose, tolerability in combination with other therapeutics, safety, and pharmacokinetics.14
Treatment-related adverse events were evaluated for all 110 patients in the phase I/Ib and II trial, regardless of tumor type. Adverse events of any grade included nausea (54%), diarrhea (51%), vomiting (35%), fatigue (32%), and increased levels of an enzyme that indicates minor liver irritation (20%).
The maximum tolerated dose was not identified during escalation, but the capsule burden became intolerable at the 1200 mg/day dose level, which may have contributed to the selection of the oral 600 mg twice daily dose level on expansion.14
Among the 51 evaluable patients in the NSCLC cohort, 45% of patients had an objective response, and the disease control rate was 96%. Overall, the median time to response was 1.5 months and the median follow-up reported was 3.6 months.
Of note, for the 14 patients included in the phase I/Ib trial who experienced a longer follow-up period, the confirmed ORR was 43%, with 67% of responders on treatment for at least 11 months. The median time on treatment was 8.2 months.
At the time of data cutoff, objective responses and stable disease appeared to be durable with the majority of patients (65%) still receiving treatment.15
In the CRC cohort, confirmed objective responses were observed in 17% of patients, with a broad disease control rate of 94%.Twelve of the 18 patients remain on treatment at the time of data cutoff, with up to 55% of patients on treatment for 4 months or longer.16
Adagrasib also demonstrated activity in other solid tumors, including pancreatic, ovarian, and endometrial cancers and cholangiocarcinoma. Of the 6 patients evaluated in the other solid tumor cohort, confirmed PRs were observed in a patient with endometrial cancer and a patient with pancreatic cancer, and unconfirmed PRs were observed in a patient with ovarian cancer and a patient with cholangiocarcinoma. All 6 patients remain on treatment.16
Both the phase I CodeBreaK100 trial with sotorasib and the phase I/II KRYSTAL-1 trial with adagrasib have shown promising efficacy and safety. Based on data from these early trials of 2 different small molecule inhibitors in patients with solid tumors and KRAS p.G12C mutations, where do we go from here?
The first step will be to understand the various testing modalities for the KRAS p.G12C mutation in solid tumors and continue clinical trials to explore how these agents can fit into current clinical practice. Finally, we will need to address the issue of acquired resistance in certain tumor types with KRAS p.G12C mutations.
It is important to realize that some patients receiving these new targeted agents are developing resistance, particularly patients with advanced CRC. Thus, it will be important to assess mechanisms of resistance and evaluate novel combination strategies in different tumor types to identify methods to overcome this resistance.
First, we must be able to identify which patients should be tested for the KRAS p.G12C mutation and understand the best testing modality for these patients who may benefit from testing.
Owing to the evidence that supports KRAS mutations as a prognostic biomarker, all patients with CRC and NSCLC should be tested for KRAS mutations. Furthermore, with the understanding that EGFR, KRAS, ROS1, BRAF, METex14-skipping mutations, RET rearrangements, and ALK genetic variants do not usually overlap, testing for KRAS mutations may help identify those who would not require further molecular testing.17
In addition, how we test patients is largely dependent on access to test tumor tissue with NGS via Clinical Laboratory Improvement Amendments (CLIA)–certified assays using either central lab assays or institutional assays. The advantage of NGS assays is largely that they provide information regarding numerous driver mutations and genetic alterations in a single test.
Another testing option is to perform hotspot mutation analysis for the KRAS oncogene in tumor biopsy tissue. A limitation of a hotspot mutation assay is that it only provides information related to KRAS so other driver mutations in the patient’s tumor may be missed.
Liquid biopsies or circulating tumor DNA assays can also be used to identify KRAS p.G12C mutations. These tests can be done using several CLIA-certified NGS assays that are able to identify multiple genetic alterations including KRAS p.G12C in your patients.
The concordance between liquid biopsies and tissue biopsies for KRAS mutations is very high. Recent studies comparing CLIA-certified ctDNA assays with tissue genomic assays show a concordance > 90%. However, when tissue genomic assays are feasible, they should still be considered the gold standard for testing. In addition, it should be recognized that low burden of disease may be associated with limited ctDNA circulation, which may affect the reliability of the ctDNA assay. Liquid biopsy cases where the ctDNA is unable to identify any genomic alteration should be considered as uninformative. In a similar fashion, ctDNA assays associated with very low mutation allele frequencies should be considered less reliable in detecting tumorigenic alterations, especially subclonal alterations.
In summary, the presence of a clonal KRAS p.G12C alteration on ctDNA is considered diagnostic for the presence of this alteration in the corresponding tissue. Similarly, the lack of KRAS p.G12C alteration on ctDNA in the presence of other driver alteration that are present at high mutation allele frequencies would rule out, with high certainty, the presence of a KRAS p.G12C mutation in the tumor.
There are several other research assays (ie, not certified) that have demonstrated accuracy and high sensitivity when examining hotspot KRAS mutations that may offer another way of testing specifically for KRAS p.G12C in patients with solid tumors. Again, these assay would be limited to a KRAS analysis only and therefore other additional genetic alterations would be missed.18,19
Preclinical evidence suggest resistance to KRASG12C inhibitors is dependent on receptor tyrosine kinase dependency and rebound EGFR activation, as well as activation of other alternative downstream signaling events that bypass RAS inhibition.20 Specifically in patients with colorectal cancer, a compensatory EGFR phosphorylation mechanism has been identified in preclinical models.21 Other mechanisms of limited antitumor activity have been identified, including KRAS nucleotide cycling as well as pathways that induce feedback reactivation and/or bypass KRAS dependence.
Efforts in the preclinical and clinical settings are underway to mitigate resistance to KRASG12C inhibitors by targeting the pathways that may contribute to resistance. For example, combination regimens of KRASG12C inhibitors with other agents that target RTKs, mTOR, or cell cycle pathways demonstrate enhanced response and marked tumor regression in cell models, even those that are refractory to KRASG12C targeted agents.22 In addition, Amodio and colleagues21 reported that concomitant inhibition of EGFR and KRASG12C resistance can be overcome in colorectal tumors. Finally, inhibition of SHP2, a protein required for RAS/ERK pathway activation, prevented adaptive resistance to MEK inhibitors in multiple cancer cell lines and the combination of SHP2 inhibition with a MEK inhibitor was effective in preclinical cell models expressing mutant or wild-type KRAS.23
Our team is currently exploring this in the clinical setting, investigating the proposed synergistic effect among KRASG12C inhibition and various classes of agents including TKIs, PD-1/PD-L1 inhibitors, mTOR inhibitors, and SHP2 inhibitors.12
In the next sections, I discuss specific methods to overcome resistance patterns, both in the preclinical, and clinical setting.
Amodio and colleagues21 confirmed that EGFR phosphorylation and escape MAP kinase pathway activation can be seen in CRC models. The figure on the left with multiple KRAS p.G12C–mutated CRC cell lines shows that after 1 hour of inhibition with sotorasib, there is a profound dose-dependent inhibition of ERK phosphorylation with sotorasib. However, at 24 hours, the inhibition of ERK phosphorylation is attenuated in most of the cell lines, supporting a compensatory mechanism that results in reactivation of ERK phosphorylation and, therefore, an escape mechanism against the antitumor activity of sotorasib.
The 2 patient-derived xenograft models of CRC with KRAS p.G12C mutation, show synergism with the combination of sotorasib with cetuximab, with enhanced antitumor activity in comparison to sotorasib alone or cetuximab alone. This synergy or mechanism of resistance appears to be selective for CRC in comparison to NSCLC and is a potential explanation for the discordance in response rates that were seen on the CodeBreaK100 clinical trial between the CRC and NSCLC patient populations. These data suggest that one way to address mechanisms of resistance in patients with advanced CRC would be to combine a KRASG12C inhibitor with an anti-EGFR agent.
Adagrasib has also demonstrated benefit in preclinical mouse models in combination with anti–PD-1 agents. As shown by Briere and colleagues,24 the combination of adagrasib and anti–PD-1 agents lead to durable CRs and a survival advantage relative to either single-agent therapy that was observed in most mice in the study. Of note, an antitumor adaptive response was also seen as evidenced by the lack of tumor development when treated mice were rechallenged with CT26 KRASG12C cells compared with treatment-naive mice that did develop tumors.
Similarly, given the preclinical studies suggesting synergy with KRASG12C inhibitors and mTOR inhibitors as well as MEK inhibitors and SHP2 inhibitors, there are ongoing combinations of small molecule inhibitors of KRASG12C with these agents. Specifically, my team is completing dose-exploration and dose-expansion trials on sotorasib in combination with MEK inhibitors in the previously studied patient populations (NSCLC, CRC, and other advanced solid tumors), TKIs in patients with NSCLC with or without previous KRAS inhibitor therapy, and PD-1/ PD-L1 agents in a variety patient populations and using various dosing schemas.25
Additional strategies that do not involve KRASG12C inhibitors include the inhibition of SHP2 with the small molecule RMC-4630, which blocks the cycling of KRAS. RMC-4630 is being evaluated as monotherapy as well as in combination with a MEK inhibitor, given the synergistic activity of the combination.26
Other strategies targeting KRAS mutation include T-cell therapy or cellular therapeutics based on intriguing data showing a profound response to T-cell infiltrating lymphocytes in a patient with KRAS mutation treated with KRAS targeting T-cell infiltrating lymphocytes.27
Finally, there are studies evaluating vaccines targeting KRAS, with one example including the combination of a KRAS RNA vaccine plus a PD-1 inhibitor, pembrolizumab, in patients with solid tumors with KRAS mutations.
Additional directions in targeting KRAS include the study of multiple additional small molecule inhibitors of KRASG12C. There are other alternative ways of inhibiting KRAS, including a highly selective PLK1 inhibitor, onvansertib, which is being evaluated in combination with systemic chemotherapy in patients with CRC and KRAS mutations, as well as inhibitors of the switch I/II pocket through BI 2852, and an SOS1 inhibitor with BAY 293 small molecule.
Knowledge of the mechanism of action, prognostic value, and incidence of the KRAS p.G12C mutation across different tumor types can help guide testing and treatment decisions including referral to clinical trials. To date, the 2 targeted agents that are furthest along in clinical development are sotorasib and adagrasib. Of note, alternative agents and novel combination regimens are in very early stages of development with goals of optimize treatment efficacy in patients with cancers harboring the KRASG12C mutation and mitigating resistance to treatment.
With great hope on the horizon for this patient population with difficult-to-treat disease, the opportunity is in the hands of the clinician to be aware of the incidence of KRAS p.G12C mutations among patients with various cancers, to identify these patients through testing, and to optimize treatment through patient counseling and referral to ongoing clinical trials.
Back
