CodeBreaK 200: Sotorasib (AMG510) Has Broken the KRAS G12C+ NSCLC Enigma Code

Targeting KRAS G12C Mutation NSCLC

From the largest pan-cancer KRAS mutation survey, KRAS mutations occurred in 35% of non-squamous cell carcinomas (SqCCs) and non-small-cell lung cancers (NSCLCs).¹ Among the KRAS mutations, KRAS G12C isotype mutation was the most common, occurring in 40% and 36% of non-SqCC and SqCC lung cancer, respectively.¹ Physiologically, KRAS catalyzes the rapid hydrolysis of guanosine-5’triphosphate (GTP) to guanosine diphosphate (GDP), hence it alternates between the active GTP-bound and inactive GDP-bound states. KRAS G12C mutations result from a cysteine substitution for glycine at codon 12, which shifts KRAS into a constitutively active state, resulting in hyperactivation of downstream signaling of cell proliferation, survival, and tumorigenesis pathways.² Until recently, targeted therapies against KRAS G12C mutations have been largely unsuccessful due to small binding pocket size, rapid conversion from active GTP to inactive GDP, and abundance of GTP in the cytoplasm.

Accelerated Approval Sotorasib (AMG-510) Based on CodeBreaK 100 (NCT03600883)

Sotorasib (AMG510) is a first-in-class covalent KRAS G12C inhibitor that targets the inactive form of KRAS through covalent binding to the exposure cysteine residue on the “switch pocket II” of the KRAS-GDP isoform.³ Sotorasib was granted accelerated approval by the US Federal Drug Administration (FDA) on May 21, 2021 for the treatment of patients with KRAS G12C-mutated locally advanced or metastatic non--small-cell lung cancer (NSCLC), following at least one prior systemic therapy based on overall response rate (ORR) and duration of response (DOR) from the expansion cohort of CodeBreaK 100 patients.^4,5 Prior to this accelerated approval, sotorasib was granted Orphan Drug Designation for the treatment of KRAS G12C-mutated NSCLC on May 1, 2019.⁵

All KRAS G12C mutations in the pivotal Phase II trial were confirmed by central laboratory using the Therascreen KRAS RGQ PCR Kit.⁵ Overall, treatment with 960 mg once daily of sotorasib demonstrated an ORR of 36% (95% CI: 28–45) among 124 patients with KRAS G12C mutation-positive NSCLC, who experienced disease progression after receiving immunotherapy and/or chemotherapy (but not all had been treated with chemotherapy and immunotherapy). The median DOR was 10.0 months (95% CI: 6.9–NR).^4,5 The most common adverse reactions (≥20%) were diarrhea, musculoskeletal pain, nausea, fatigue, hepatotoxicity, and cough. Adverse reactions resulting in permanent discontinuation of sotorasib occurred in 9% of patients. For the 100 patients who had received both chemotherapy and immunotherapy (IO) (entry criteria for CodeBreaK 200), the ORR was 31% (95% CI: 22–41) and DOR of 10.0 months (95% CI: 6.9–NR).⁵ See Table 1.

Table 1 Comparison of Efficacy and Safety of Sotorasib in Phase II and Phase III Trials

As with all FDA accelerated approvals, at least one randomized confirmatory Phase III trial had to be initiated, and patients enrolled at the time of accelerated approval. Continued approval is contingent upon verification and description of clinical benefits in a confirmed trial, which, in this instance is CodeBreaK 200. The results of this trial were presented at the annual meeting of the European Society of Medical Oncology (ESMO) 2022 and subsequently published in The Lancet.⁶

CodeBreaK 200 (NCT04303780) (Confirmatory Randomized Phase III Trial of Sotorasib versus Docetaxel)

CodeBreaK 200 is a global open-label study randomizing KRAS G12C+ NSCLC patients who had received both prior chemotherapy and immunotherapy treatment in a 1:1 ratio to either sotorasib 960 mg once daily⁷ or docetaxel at 75 mg/m² IV every three weeks. Stratification factors included number of prior lines of therapy (1 vs 2 vs >2), race (Asian vs non-Asian), and history of brain metastasis (yes/no). Of note, patients with active brain metastases were excluded. The primary endpoint was Blind Independent Review Committee (BIRC) progression-free survival (PFS), with overall survival (OS), ORR, safety, and quality of life as secondary endpoints. Per US FDA guidance, the protocol was amended to reduce planned enrollment from 650 to ~330 patients, and crossover from docetaxel to sotorasib was permitted, thus an OS was unlikely to be achieved.⁶

Overall, 171 patients were randomized to sotorasib and 174 patients to docetaxel. The median BIRC-assessed PFS with sotorasib was 5.6 months (95% CI: 4.3–7.8) versus 4.5 months for docetaxel (95% CI: 3.0–5.7), with a stratified proportional HR of 0.66 (95% CI: 0.51–0.86; p = 0.002). Importantly, 12-month PFS rate was 24.8% for sotorasib versus 10.1% for docetaxel. Sotorasib achieved a significantly higher ORR of 28.1% (95% CI: 21.5–35.1) compared to ORR of 13.2% (95% CI: 8.6–19.2) for docetaxel (p <0.001) and DCR of 60.3% (95% CI: 52.7–67.7). The DOR was 8.6 months (95% CI: 7.1–18.0) with sotorasib compared to 6.8 months (95% CI: 4.3–8.3) achieved by docetaxel. There was no OS benefit with 10.6 months achieved by sotorasib (95% CI: 8.9–14.0) versus 11.3 months achieved by docetaxel (95% CI: 9.0–14.9)⁶ (Table 1).

CodeBreaK 200 Subgroup Analysis

Of the three stratification factors, both Asians and non-Asians benefited from sotorasib over docetaxel. In fact, Asian patients seemed to gain more benefit from sotorasib (HR = 0.33 95% CI: 0.14, 0.80), with a median PFS of 8.3 months for Asian patients treated with sotorasib; this finding must be viewed with caution, however, because the number of Asians enrolled was limited. Patients without brain metastasis (HR = 0.74; 95% CI: 0.53–1.03) and patients with only 1 prior line of therapy (HR = 0.70; 95% CI: 0.47–1.04) just missed the significance threshold for PFS benefit of sotorasib over docetaxel. Nevertheless, there was numerical improvement in PFS of sotorasib over docetaxel.

Docetaxel “Over-Performed” in CodeBreaK 200

Finally, the hazard ratio in a randomized Phase III trial is a comparison between the investigational arm versus the standard of care arm. The median PFS of 4.5 months achieved by patients in the docetaxel arm is the highest median PFS achieved by docetaxel in a randomized trial in the second line setting. In the REVEL trial that established a docetaxel and ramucirumab combination trial, the median PFS for docetaxel was only 3 months whereas for the combination arm it was 4.5 months.⁸ In the two docetaxel versus immunotherapy monotherapy randomized Phase III trials (Keynote-010 and OAK), the median PFS achieved by docetaxel was 4.0 months in patients with any PD-L1 expression.^9,10

Sotorasib in CodeBreaK 200 Achieved Its Pre-Specified Primary Endpoint

Another criticism of CodeBreaK 200 was that there was no OS benefit despite only 34% of the docetaxel-treated patients receiving a KRAS G12C inhibitor. However, the sample size of CodeBreaK 200 was reduced as per US FDA recommendation to achieve a faster read-out of the clinical efficacy data. The magnitude of the improvement in PFS of sotorasib over docetaxel may be less than what most oncologists had expected or anticipated based on the Phase II results of sotorasib in KRAS G12C+ NSCLC.⁴ However, for regulatory purposes, CodeBreaK 200 achieved its primary endpoint of significant improvement of PFS and, per regulatory purposes, should result in full regulatory approval for sotorasib. CodeBreaK 200 results are a proof of principle that KRAS G12C is a targetable mutation. The commercial marketplace will determine how sotorasib is widely prescribed pending the results from other competing covalent KRAS G12C inhibitors.

Pending Questions About Sotorasib Dosing

In the Phase 1 dose escalation portion of sotorasib, sotorasib was dosed from 180 mg, 360 mg, 720 mg and 960 mg once daily.¹¹ Per the US FDA approval summary after analyzing the data submitted by Amgen, systemic exposures (ie, AUC_0–24h and C_max) of sotorasib after daily administration of doses ranging from 180 mg to 960 mg were similar (Table 1). In addition, overall response rates (ORRs) ranging from 25–50% were observed in patients with NSCLC who were treated with lower doses of sotorasib in the dose escalation portion of the trial (N ranging from 3–16 per dose level). As such, a post-marketing requirement (PMR) by the US FDA required a randomized study of two doses of sotorasib (240 mg once daily vs 960 mg once daily) (CodeBreaK 201) (NCT04933695).

The First Breakthrough Established a New Field but is Never the Last Breakthrough

We have learned in targeted therapy in lung cancer that the first positive trial established the field; however, it is not the final advance to be made. For example, when first-generation (1G) EGFR TKIs achieved greater clinical efficacy than platinum-based doublet chemotherapy, they became the standard of care.¹¹ However, with time, the now third-generation (3G) EGFRTKIs have consistently demonstrated statistically significant PFS over 1G EGFR TKIs.¹² Similarly, crizotinib, the first ALK TKI approved, demonstrated in the PROFILE1014 trial statistically significantly improved PFS over platinum-based chemotherapy.¹³ Of note, the randomized trial of crizotinib in ALK+ NSCLC involved crizotinib demonstrating statistically significantly improved PFS over single agent chemotherapy very similar to the current CodeBreaK 200.¹⁴

Sotorasib is the First but Not the Only Covalent KRAS G2C Inhibitor to Be Approved

There are many covalent KRAS G12C inhibitors being developed and some of the Phase II expansion cohort results in KRAS G12C+ NSCLC are presented and summarized in Table 2 (1adagrasib,¹⁵ JDQ443,¹⁶ GDC-6036,¹⁷ B-1553¹⁸). Overall, the ORR among the current KRAS G12Cs are similar and we look forward to clinical efficacy and safety data that may differentiate these covalent KRAS G12C inhibitors.

Table 2 Reported Phase II Trial of Covalent KRAS G12C Inhibitors

Indeed, adagrasib was granted accelerated approval by the US FDA on December 12, 2022.¹⁹ Approval was based on KRYSTAL-1, a multicenter, single-arm, open-label clinical trial (NCT03785249), which included patients with locally advanced or metastatic NSCLC with KRAS G12C mutations. Efficacy was evaluated in 112 patients whose disease had progressed on or after platinum-based chemotherapy and an immune checkpoint inhibitor, given either concurrently or sequentially. The main efficacy outcome measures were confirmed ORR according to RECIST 1.1, as evaluated by blind independent central review, and DOR. The ORR was 43% (95% CI: 34%, 53%) and median DOR was 8.5 months (95% CI: 6.2, 13.8).¹⁵ The full approval of adagrasib will hinge on demonstrating clinical benefit in a pivotal Phase III trial (Table 3).

Table 3 On-Going KRASG12C Covalent versus Docetaxel Trial

Other Confirmatory Randomized Trials of KRAS G12C Inhibitors versus Docetaxel (KRYSTAL-12, B-FAST, KontRASt-02)

Several of the KRAS G12C inhibitors have advanced past Phase II dosing and Phase III trials are on-going including adagrasib (KRYSTAL-12, NCT04685135), GDC (B-Fast, NCT03178552), JDQ433 (KontRASt-02, NCT05132075) (Table 3). The design of these pivotal trials is similar, with cross-over allowed and PFS as the primary endpoint.

Unmet Needs of Current Covalent KRAS G12C Inhibitors and Future Directions

As we look forward to more positive second-line randomized trials for the other KRAS G12C inhibitors, the results of CodeBreaK 200 and other KRAS G12C inhibitors have raised several therapeutic challenges and unknowns.

Efficacy of Covalent KRAS G12C Inhibitors in Patients with Brain Metastasis

There is a high propensity for KRAS G12C+ NSCLC patients to develop CNS metastasis.²⁰ The efficacy of these covalent KRAS G12C inhibitors has not been consistently demonstrated. In the CodeBreaK 100 and 200, patients with untreated brain metastases were not eligible for entry, thus limiting assessment of CNS efficacy of sotorasib.

Studies of patients with RECIST-defined measurable brain metastases, allowing asymptomatic KRAS G12C+ NSCLC patients with brain metastases to engage in clinical trials, are on-going.²⁰

Incorporating Immunotherapy into KRAS G12C Inhibitors and Deciphering the Role of STK11, KEAP1, and NFE2L2 Genomic Co-Alterations in Response Modulation

KRAS G12C+ NSCLC has heterogeneous genomic alterations but three different mutations (STK11, KEAP1, NFE2L2) are of therapeutic interest. All three have been shown to negatively modulate response to immunotherapy.^21,22 Indeed, in the largest survey of KRAS mutations in solid and liquid tumors, STK11 and KEAP1 mutations were positively associated with KRAS G12C mutation non-squamous NSCLC.¹ KRAS G12C+ NSCLC is more commonly found in current/former smokers whose tumors are high expressers of PD-L1.¹ Importantly, the percentage of KRAS G12C+ NSCLC which had PD-L1 expression of >1% were 72% (non-SqCC NSCLC) and 81% (SqCC NSCLC), respectively. Thus, combining KRAS G12C inhibitors with immunotherapy is the next holy grail of targeted KRAS G12C+ NSCLC. To date, single-agent KRAS G12C inhibitors (sotorasib and adagrasib) had similar efficacy in terms of ORR among STK11, KEAP1, and NFE2L2 mutations.^4,15

Combination of sotorasib with anti-PD-1 or anti-PD-L1 immunotherapy (pembrolizumab have been presented.²³ The researchers enrolled 58 KRAS G12C inhibitor-naïve KRAS G12C+ NSCLC patients and treated them in 12-dose exploration cohorts at varying doses of sotorasib (120–960 mg daily) in combination with either atezolizumab 1200mg or pembrolizumab 200mg, administered concurrently every three weeks until intolerability or disease progression. Half of the cohorts were lead-in cohorts, where patients received sotorasib monotherapy for either 21 or 42 days prior to their first dose of immunotherapy, then received atezolizumab or pembrolizumab together with sotorasib. The dose-limiting toxicity window was 21 days following initiation of combination treatment.

The 58 patients were treated with a median follow-up of 12.8 months (range: 1.6, 29.9). Median prior lines of therapy were 1 (range 0–7); 67% patients received prior IO. The median doses of sotorasib were 83 (range: 22–791), and median doses of immunotherapy were three (range: 1–26]. The most common grade 3–4 treatment-related adverse events were increased ALT and AST. Of patients with grade 3–4 hepatotoxicity, first adverse events occurrence was outside the dose-limiting window in 22 of 25 (88%) patients; most were managed with corticosteroids, and 97% of events resolved. No fatal adverse events occurred. Grade 3–4 adverse events and adverse events leading to treatment discontinuation occurred less often in the lead-in versus concurrent cohorts. Across all 12 cohorts, confirmed response was observed in 17 of 58 patients (29%; range, 0–67). Among the 17 responders, median duration of response was 17.9 months (range: 1.5+, 23.4). For all 58 patients across cohorts, median overall survival was 15.7 months (95% CI: 9.8–17.8).²³

Adagrasib 400 mg twice daily (less than the approved 600 mg twice daily) in combination with pembrolizumab 200 mg intravenous (IV) every 3 weeks has also reported preliminary results for 7 patients from KRYSTAL-1 and 75 patients from KRYSTAL-7 (11 with PD-L1 <1% and 62 with PD-L1 >/=1%).²⁴ The majority (99%) were current or former smokers.

Among the 53 evaluable patients, the ORR was 49% (95% CI: 35–63) and DCR was 89% (95% CI: 77–96) across all PD-L1 expression levels. Responses were observed in 59% (13/22) of patients with PD-L1 TPS ≥50%, in 48% (10/21) with PD-L1 TPS 1%−49%, and in 30% (3/10) with PD-L1 TPS <1%. Importantly for safety, aspartate aminotransferase (AST) and alanine transaminase (ALT) elevation each occurred in 21% of the patients with grade 3 toxicities, in 9% (AST elevation) and 8% (ALT elevation), respectively. The median time to onset for ALT and AST increase was 26 days and 37 days, respectively; only 1 patient reported new onset treatment-associated ALT/AST increase after 3 months.²⁴

Moving to First-Line Treatment of KRAS G12C+ NSCLC

The standard of care for first-line treatment of NSCLC without activating EGFR or ALK fusion is chemotherapy and immunotherapy for PD-L1 <50% and either chemotherapy and immunotherapy or immunotherapy monotherapy. Regardless, in order to encompass all KRAS G12C+ NSCLC, based on the adagrasib and sotorasib immunotherapy combination, several clinical trials are necessary to determine the most efficacious and safe treatment approach, such as immunotherapy vs KRAS G12C inhibition and immunotherapy + KRAS G12C inhibitor for PD-L1 ≥50%. For PD-L1 <50%, a separate pivotal Phase III chemoimmunotherapy ± KRAS G12C inhibitor will have to be conducted. However, safety and tolerability of chemotherapy + immunotherapy + KRAS G12C inhibitor must be firmly established first.

Overcoming Multiple Simultaneous on-Target and off-Target Resistances

Lastly, resistance to sotorasib²⁵ and adagrasib²⁶ monotherapy has been reported. The resistance mechanisms are driven by both on-target acquired resistance mutations (mutation at the cysteine residues that abolishes the binding of covalent KRAS inhibitors) and off-target such as MEK bypass pathway and even receptor tyrosine kinase fusions. Thus, combination therapy with KRAS G12C inhibitors with other targeted therapy is another challenge, given the multiple resistance pathways that appeared at the same time. One approach is to inhibit the upstream action of KRAS with SH2 containing protein tyrosine phosphatase-2 (SHP2) or SOS Ras/Rac guanine nucleotide exchange factor 1 (SOS1) inhibitors. Trials with SHP2 inhibitors are on-going with both sotorasib (with RMC-4630 [NCT05054725] and BBP-398 [NCT05480865]) and adagrasib (with TNO-155 [KRYSTAL-2, NCT04330664]). Combination of adagrasib with an SOS1 inhibitor (BI-1701963) (KRYSTAL-14, NCT04975256]) has been completed. We eagerly await the results of these combination trials.

Sotorasib Activity in Advanced KRAS G12C+ Colorectal Adenocarcinoma and Pancreatic Adenocarcinoma

Finally, sotorasib has demonstrated clinical activity in advanced KRAS G12C+ colorectal and pancreatic adenocarcinoma. Results from 62 patients in the colorectal cancer arm of CodeBreaK 100 were recently published. In response to sotorasib monotherapy, ORR was 9.7% (95% CI: 3.6–19.9). Grade 3 or higher treatment-related adverse events were seen in only 10% of patients, most commonly diarrhea (3%), creatine phosphokinase increase (2%), and acute kidney injury (2%).²⁷ Additionally, 38 patients from the pancreatic cancer arm of CodeBreaK 100 demonstrated an ORR of 21% (95% CI: 10–37), with median PFS 4.0 months (95% CI: 2.8–5.6). The median OS was 6.9 months (95% CI: 5.0–9.1), with treatment-related adverse events in 42% of patients.²⁸ While the clinical activity of sotorasib was modest in KRAS G12C+ CRC, the ORR of 21% in KRAS G12C+ pancreatic adenocarcinoma gives patients with this deadly disease real hope and affirms the activity of sotorasib against KRAS G12C mutation across KRAS G12C+ solid tumors and further highlights the clinical benefit in NSCLC.

Executive Summary

1. CodeBreaK 200 achieved its primary endpoint of statistical improvement in median PFS of sotorasib over docetaxel.
2. Sotorasib also achieved higher ORR and 12-month PFS rates than docetaxel monotherapy.
3. Sotorasib has demonstrated clinical activity against other KRAS G12C+ solid tumors. Hence, while the absolute improvement in median PFS in CodeBreaK 200 was modest, sotorasib in CodeBreaK 200 represents the beginning of targeted therapy in KRAS mutated solid tumors.