Adagrasib with or without Cetuximab in Colorectal Cancer with Mutated KRAS G12C

Abstract

BACKGROUND

Adagrasib, an oral small-molecule inhibitor of mutant KRAS G12C protein, has shown clinical activity in pretreated patients with several tumor types, including colorectal cancer. Preclinical studies suggest that combining a KRAS G12C inhibitor with an epidermal growth factor receptor antibody could be an effective clinical strategy.

METHODS

In this phase 1–2, open-label, nonrandomized clinical trial, we assigned heavily pretreated patients with metastatic colorectal cancer with mutant KRAS G12C to receive adagrasib monotherapy (600 mg orally twice daily) or adagrasib (at the same dose) in combination with intravenous cetuximab once a week (with an initial loading dose of 400 mg per square meter of body-surface area, followed by a dose of 250 mg per square meter) or every 2 weeks (with a dose of 500 mg per square meter). The primary end points were objective response (complete or partial response) and safety.

RESULTS

As of June 16, 2022, a total of 44 patients had received adagrasib, and 32 had received combination therapy with adagrasib and cetuximab, with a median follow-up of 20.1 months and 17.5 months, respectively. In the monotherapy group (43 evaluable patients), a response was reported in 19% of the patients (95% confidence interval [CI], 8 to 33). The median response duration was 4.3 months (95% CI, 2.3 to 8.3), and the median progression-free survival was 5.6 months (95% CI, 4.1 to 8.3). In the combination-therapy group (28 evaluable patients), the response was 46% (95% CI, 28 to 66). The median response duration was 7.6 months (95% CI, 5.7 to not estimable), and the median progression-free survival was 6.9 months (95% CI, 5.4 to 8.1). The percentage of grade 3 or 4 treatment-related adverse events was 34% in the monotherapy group and 16% in the combination-therapy group. No grade 5 adverse events were observed.

CONCLUSIONS

Adagrasib had antitumor activity in heavily pretreated patients with metastatic colorectal cancer with mutant KRAS G12C, both as oral monotherapy and in combination with cetuximab. The median response duration was more than 6 months in the combination-therapy group. Reversible adverse events were common in the two groups. (Funded by Mirati Therapeutics; KRYSTAL-1 ClinicalTrials.gov number, NCT03785249.)

Kirsten rat sarcoma virus oncogene homologue (KRAS) is the most frequently mutated oncogene in human cancer and is one that drives oncogenesis in up to 50% of patients with colorectal cancer.^1–6 The KRAS protein cycles between “on” states with guanosine triphosphate (GTP) binding and “off” states with guanosine diphosphate (GDP) binding, with a protein resynthesis half-life of 23 hours.^6–9 The KRAS glycine-to-cysteine mutation at codon 12 (KRAS G12C) occurs in 3 to 4% of patients with colorectal cancer and impairs GTP hydrolysis, which shifts KRAS to the active GTP-binding state to drive protumorigenic effector signaling.^8,10–12

For more than two decades, the cornerstone treatment for metastatic colorectal cancer has been fluoropyrimidine-based chemotherapy regimens with oxaliplatin, irinotecan, or both.^13,14 In KRAS-mutated colorectal cancer, biologic therapies such as vascular endothelial growth factor inhibitors have improved outcomes in first- and second-line treatment when added to a chemotherapy backbone.¹³ Third-line treatment options are limited to regorafenib and trifluridine–tipiracil, which have extended median overall survival by less than 2 months over placebo and have been associated with a low response rate (<2%) and median progression-free survival of approximately 2 months, outcomes that highlight the need for new approaches.^15,16

KRAS mutations have historically been considered to be “undruggable,” and retrospective series of clinical data suggest that the presence of the KRAS G12C allele is associated with worse survival outcomes than other KRAS mutations in patients with colorectal cancer.^12,17 However, recent progress has led to the development of small molecules that can occupy the KRAS switch II binding pocket and covalently bind to the mutant cysteine to prevent GTP binding.^9,18 Currently, no therapies that specifically target KRAS G12C protein have been approved for the treatment of colorectal cancer.^13,19

Adagrasib (MRTX849) is an orally available, small-molecule covalent inhibitor of KRAS G12C that irreversibly and selectively binds the mutant protein in its inactive, GDP-bound state.^8,20 Adagrasib has favorable pharmacokinetic properties, including a long half-life (23 hours), dose-dependent pharmacokinetics, and central nervous system (CNS) penetration.^8,21,22 Maintaining continuous adagrasib exposure above a target threshold allows for the inhibition of KRAS-dependent signaling for the duration of the dosing interval and is predicted to maximize antitumor activity.^8,20 Preliminary data indicate clinical activity in pretreated patients across nine tumor types with KRAS G12C mutations, including those with colorectal cancer, non–small-cell lung cancer, pancreatic and biliary-tract cancers, and other gastrointestinal and nongastrointestinal cancers.⁸

Although early data from a study of adagrasib monotherapy are promising in patients with colorectal cancer with KRAS G12C mutations,²² reactivation may occur in the RAS–MAPK signaling pathway because of adaptive feedback mediated by epidermal growth factor receptor (EGFR).^23,24 Preclinical studies have shown that combining an antibody against EGFR with a KRAS G12C inhibitor could be an effective clinical strategy to mitigate EGFR reactivation.²⁴ Cetuximab is an anti-EGFR monoclonal antibody that is indicated for the treatment of RAS wild-type metastatic colorectal cancer, either as monotherapy or in combination with chemotherapy.¹³ Combining cetuximab with adagrasib may enhance the inhibition of KRAS-dependent signaling or overcome adaptive feedback to delay resistance and improve outcomes.²⁵

In the ongoing KRYSTAL-1 phase 1–2 trial, we evaluated the use of adagrasib as monotherapy or in combination with cetuximab in patients with previously treated metastatic colorectal cancer with mutant KRAS G12C.

METHODS

TRIAL OVERSIGHT

This open-label, nonrandomized trial was designed by employees of the sponsor (Mirati Therapeutics) and the investigators. The data were collected by the investigators and analyzed by statisticians employed by the sponsor. All the authors contributed to the interpretation of the data and the preparation of the manuscript and vouch for completeness and accuracy of the data and for the fidelity of the trial to the protocol. A sponsor-funded medical writer prepared the first draft of the manuscript and provided editorial assistance with an earlier version of the manuscript. The trial was conducted in accordance with the principles of the Declaration of Helsinki and the International Council for Harmonisation guidelines for Good Clinical Practice. The protocol (available with the full text of the article at NEJM.org) was approved by all the necessary ethics committees and institutional review boards at each trial site. Trial oversight was provided by the sponsor, investigators, local institutional review boards, a central institutional review board, and an independent data and safety monitoring committee. All the patients provided written informed consent.

PATIENTS

Eligible adults (≥18 years of age) had a histologically confirmed diagnosis of advanced, unresectable, or metastatic colorectal cancer with a KRAS G12C mutation. All the patients had no available treatment with curative intent and no available standard-of-care treatment (or were ineligible or declined treatment), along with an Eastern Cooperative Oncology Group performance-status score of 0 or 1. Patients who received adagrasib had measurable disease according to the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1; patients who received adagrasib plus cetuximab had measurable or evaluable disease according to the same criteria, as defined in the Supplementary Appendix, available at NEJM.org. Key exclusion criteria were active CNS metastases, carcinomatous meningitis, and receipt of systemic or radiation therapy within 2 weeks before the planned initiation of adagrasib. Full eligibility criteria are provided in the protocol.

TRIAL DESIGN AND TREATMENT

In phase 2 of the trial, we evaluated the activity of adagrasib monotherapy in patients with metastatic colorectal cancer with mutant KRAS G12C (monotherapy group). In a separate pilot phase 1b cohort, we evaluated the safety and activity of adagrasib in combination with cetuximab (combination-therapy group). In the monotherapy group, patients received oral adagrasib at a dose of 600 mg twice daily, administered in a capsule formulation without food. In the combination-therapy group, patients received a combination of oral adagrasib (600 mg twice daily) and intravenous cetuximab, which was administered either once a week (with an initial loading dose of 400 mg per square meter of body-surface area, followed by a dose of 250 mg per square meter) or every 2 weeks (at a dose of 500 mg per square meter).

In the two cohorts, treatment continued until the occurrence of disease progression, unacceptable adverse events, withdrawal of consent, or death. Patients who had clinical benefit as judged by the investigator could continue therapy beyond the diagnosis of RECIST-defined disease progression. In the monotherapy group, crossover to receive combination therapy was permitted in patients who had stable disease at week 13 or later.

END POINTS

In the monotherapy group, the primary end point was the objective response, which was defined as a complete or partial response (according to RECIST, version 1.1), as confirmed by the investigator. Secondary end points included the duration of response, progression-free survival, overall survival, 1-year survival, and safety. In the combination-therapy group, the primary end point was safety, including dose-limiting toxic effects. Secondary end points included the tumor-response outcomes that were evaluated in the monotherapy group. Details regarding the trial end points are provided in the Supplementary Appendix.

CLINICAL ASSESSMENTS, SAFETY, AND EXPLORATORY ANALYSES

Patients were tested for mutations in tumor tissue or circulating tumor DNA (ctDNA) with the use of preapproved methods (see the Supplementary Appendix). Patients underwent disease evaluation at baseline by computed tomography (CT) of the chest, CT or magnetic resonance imaging of the abdomen and pelvis, and clinical evaluation of any superficial lesions. Disease-evaluation scans were performed every 6 weeks during the trial and were assessed (according to RECIST criteria) by the investigators. Patients with a tumor response had a confirmatory assessment at least 4 weeks after the initial response.

Adverse events were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 5.0). Patients were followed for adverse events for at least 28 days after the last dose of a trial treatment.

Serial plasma samples were collected (before treatment and at the beginning of cycles 2 and 4) and analyzed to assess the effect of adagrasib monotherapy or combination therapy on ctDNA clearance of KRAS G12C alterations. In an additional exploratory post hoc analysis, candidate genomic biomarkers (TP53 or PIK3CA mutations) were evaluated at baseline for any association with tumor response to monotherapy or combination therapy.

STATISTICAL ANALYSIS

In the monotherapy cohort, we assumed that the response in patients who had previously received standard-of-care treatment would be 10%, with a target response rate of 30%. A 95% confidence interval was constructed around the point estimate of the response. Monotherapy with adagrasib was considered to be promising (according to predictive-probability-design stopping rules) if more than 7 responses were observed among 40 patients in the clinically evaluable population. Combination therapy with adagrasib plus cetuximab was considered to be promising if the number of patients with dose-limiting toxic effects among the first 7 patients was low enough to warrant continued enrollment; approximately 30 patients could be enrolled to fully characterize safety. In the two cohorts, we used the Kaplan–Meier method to estimate the median response duration, progression-free survival, overall survival, and 1-year survival.

RESULTS

PATIENTS

As of the data-cutoff date on June 16, 2022, a total of 44 patients with pretreated metastatic colorectal cancer had received adagrasib monotherapy (Fig. S1 in the Supplementary Appendix). The median duration of follow-up was 20.1 months, and the median duration of treatment was 5.9 months (range, 0.6 to 23.8). The median age of the patients was 59 years (range, 29 to 79), and 50% of the patients were women (Table 1). By the same data-cutoff date, 32 patients had received adagrasib plus cetuximab (median follow-up, 17.5 months; median duration of treatment, 7.3 months [range, 0.0 to 24.0]). The median age of the patients was 60 years (range, 41 to 74), and 53% were women.

Table 1.

Characteristics of the Patients at Baseline.^*

Characteristic	Adagrasib Monotherapy (N = 44)	Adagrasib plus Cetuximab (N = 32)
Demographic
Median age (range) — yr	59 (29–79)	60 (41–74)
Female sex — no. (%)	22 (50)	17 (53)
Race — no. (%)†
White	33 (75)	26 (81)
Black	6 (14)	4 (12)
Asian	3 (7)	2 (6)
Other	2 (5)	0
Medical history
ECOG performance-status score — no. (%)
0	23 (52)	14 (44)
1	21 (48)	18 (56)
Previous lines of systemic anticancer therapy
Median no. (range)	3 (1–9)	3 (1–8)
Total no. — % of patients
1	18	9
2	20	25
3	25	34
≥4	36	31
Type of therapy — no. (%)
Fluoropyrimidine	44 (100)	32 (100)
Oxaliplatin	43 (98)	32 (100)
Irinotecan	35 (80)	28 (88)
Fluoropyrimidine–oxaliplatin–irinotecan	34 (77)	28 (88)
Anti-VEGF	36 (82)	28 (88)
Anti-EGFR	1 (2)	0
Regorafenib, trifluridine–tipiracil, or both	10 (23)	6 (19)
PD-1 or PD-L1 inhibitor	5 (11)	4 (12)
Presence of metastatic target lesions — no. (%)
Liver	29 (66)	19 (59)
Lung	25 (57)	16 (50)
Lymph nodes	12 (27)	9 (28)
Other	17 (39)	9 (28)
Molecular status — no./total no. (%)‡
BRAF V600E mutation	0/42	0/30
High microsatellite instability or deficient mismatch repair	1/35 (3)	0/19
EGFR amplification	1/35 (3)	1/28 (4)
TP53 mutation	23/34 (68)	18/26 (69)
PIK3CA mutation	5/36 (14)	3/26 (12)

^*BRAF denotes B-type Raf kinase, ECOG Eastern Cooperative Oncology Group, EGFR epidermal growth factor receptor, PIK3CA phosphatidylinositol 3-kinase catalytic subunit alpha, PD-1 programmed cell death 1, PD-L1 programmed cell death ligand 1, TP53 tumor protein 53, and VEGF vascular endothelial growth factor.

^†Race was reported by the patients.

^‡Included in the molecular-status category are data for patients with conclusively evaluable test results.

The characteristics of the patients at baseline were similar in the two cohorts. All the patients had previously received fluoropyrimidine-based chemotherapy; the majority had also received oxaliplatin, irinotecan, or both. The median number of previous lines of systemic therapy was three; the median interval between the last systemic therapy and trial enrollment was 42 days in the monotherapy cohort and 53 days in the combination-therapy cohort. The patients were representative of the broader population of patients with mutant KRAS G12C colorectal cancer (Table S1).

CLINICAL ACTIVITY

Adagrasib Monotherapy

Of the 43 patients who received adagrasib monotherapy in the clinically evaluable population, the response was 23% (95% confidence interval [CI], 12 to 39) according to blinded independent central review and 19% (95% CI, 8 to 33) according to investigator assessment (Table 2). In the full analysis set (44 patients), the response according to investigator assessment was 18% (95% CI, 8 to 33) (Fig. 1A and Table S2). Details regarding the concordance of response are presented in Table S3.

Table 2.

Overall Summary of Clinical Activity.^*

Variable	Adagrasib Monotherapy (N = 43)†	Adagrasib plus Cetuximab (N = 28)‡
Objective response§
Per blinded independent central review — no. of patients	10	13
% (95% CI)	23 (12–39)	46 (28–66)
As confirmed by investigator — no. of patients	8	13
% (95% CI)	19 (8–33)	46 (28–66)
Best overall response — no. (%)
Complete response	0	0
Partial response	8 (19)	13 (46)
Stable disease	29 (67)	15 (54)
Progressive disease	6 (14)	0
Not evaluable	0	0
Median duration of response — mo	4.3	7.6
95% CI	2.3–8.3	5.7–NE
Median progression-free survival — mo¶	5.6	6.9
95% CI	4.1–8.3	5.4–8.1
Median overall survival — mo¶	19.8	13.4
95% CI	12.5–23.0	9.5–20.1

^*All end points were measured in the evaluable population (investigator assessed), unless stated otherwise. The data-cutoff date for clinical activity end points was June 16, 2022 (median follow-up, 20.1 months for adagrasib monotherapy; 17.5 months for adagrasib plus cetuximab). NE denotes not estimable.

^†One patient withdrew consent before the first scan was performed and was not included in the evaluable population.

^‡Not included in the post-baseline assessment of target lesions were 4 patients (1 who withdrew consent, 2 who had cetuximab-related adverse events, and 1 who had an adverse event unrelated to treatment).

^§An objective response was defined as a complete or partial response.

^¶This end point was evaluated in the full analysis set (44 patients who received adagrasib monotherapy group and 32 patients who received adagrasib plus cetuximab).

Figure 1. Clinical Activity End Points with Adagrasib Monotherapy.

The best tumor change from baseline (Panel A) is shown for 43 patients in the population that could be evaluated for clinical activity. Progression-free survival (Panel B) and overall survival (Panel C) are shown for 44 patients in the full analysis set; hatch marks indicate data censoring. The data-cutoff date for clinical activity was June 16, 2022; the median follow-up was 20.1 months.

The median duration of response was 4.3 months (95% CI, 2.3 to 8.3); the median time to response was 1.5 months (range, 1.2 to 4.4) (Fig. S2A). Among all 44 patients in this cohort, the median progression-free survival was 5.6 months (95% CI, 4.1 to 8.3) (Fig. 1B). The median overall survival was 19.8 months (95% CI, 12.5 to 23.0) (Fig. 1C). At the data-cutoff date, 29 patients (66%) had started subsequent anticancer therapy.

Adagrasib plus Cetuximab

Among the 28 patients who received adagrasib plus cetuximab in the clinically evaluable population, the response was 46% (95% CI, 28 to 66) according to blinded independent central review and to investigator assessment; in the full analysis set (32 patients), the response according to investigator assessment was 41% (95% CI, 24 to 59) (Fig. 2A). Details regarding the concordance of response are presented in Table S4.

Figure 2. Clinical Activity End Points with Adagrasib plus Cetuximab.

The best tumor change from baseline (Panel A) is shown for 28 patients in the population that could be evaluated for clinical activity. Progression-free survival (Panel B) and overall survival (Panel C) are shown for 32 patients in the full analysis set. The median follow-up was 17.5 months.

The median duration of response was 7.6 months (95% CI, 5.7 to not estimable); the median time to response was 1.4 months (range, 1.2 to 19.2) (Fig. S2B). Among all 32 patients in this cohort, the median progression-free survival was 6.9 months (95% CI, 5.4 to 8.1) (Fig. 2B). The median overall survival was 13.4 months (95% CI, 9.5 to 20.1) (Fig. 2C). As of the data-cutoff date, 15 patients (47%) had started subsequent anticancer therapy. Response end points for 6 patients who crossed over from the monotherapy cohort to receive adagrasib plus cetuximab are provided in Table S5.

EXPLORATORY ANALYSES

In an exploratory analysis of ctDNA response, at least 95% plasma clearance of the KRAS G12C mutant allele occurred by cycle 2 in 16 of 29 patients (55%) in the monotherapy cohort and in 14 of 16 patients (88%) in the combination-therapy cohort (Figs. S3 and S4). In addition, no apparent association was found between the frequency of response in either cohort and the presence or absence of a TP53 or PIK3CA mutation at baseline (Table S6).

SAFETY

Adagrasib Monotherapy

Of the 44 patients in the monotherapy cohort, treatment-related adverse events of any grade (as determined by the investigators) occurred in 41 patients (93%) (Table 3). The events that occurred in at least 20% of the patients were diarrhea (66%), nausea (57%), vomiting (45%), and fatigue (45%). Grade 3 or 4 treatment-related adverse events occurred in 15 patients (34%) overall; events that occurred in at least 5% of the patients were anemia (9%) and diarrhea (7%). Treatment-related adverse events led to a dose reduction in 17 patients (39%) (Table S7). No treatment-related adverse events led to treatment discontinuation, and no grade 5 treatment-related events occurred. Adverse events that were observed during treatment are summarized in Table S8.

Table 3.

Summary of Treatment-Related Adverse Events.^*

Adverse Event	Adagrasib Monotherapy (N = 44)
	Any Grade	Grade 1	Grade 2	Grade 3	Grade 4
	number of patients (percent)
Any event	41 (93)	10 (23)	16 (36)	13 (30)	2 (5)
Leading to dose discontinuation	0	—	—	—	—
Leading to dose interruption	20 (45)	—	—	—	—
Leading to dose reduction	17 (39)	—	—	—	—
Most frequent events †
Diarrhea	29 (66)	16 (36)	10 (23)	3 (7)	0
Nausea	25 (57)	15 (34)	10 (23)	0	0
Vomiting	20 (45)	12 (27)	8 (18)	0	0
Fatigue	20 (45)	11 (25)	7 (16)	2 (5)	0
Anemia	7 (16)	2 (5)	1 (2)	4 (9)	0
Prolonged QT interval on ECG	7 (16)	2 (5)	3 (7)	2 (5)	0
Peripheral edema	7 (16)	6 (14)	1 (2)	0	0
Decreased appetite	8 (18)	4 (9)	4 (9)	0	0
Increased ALT	5 (11)	3 (7)	0	2 (5)	0
Increased AST	5 (11)	3 (7)	0	2 (5)	0
	Adagrasib plus Cetuximab (N = 32)
Any event	32 (100)	5 (16)	22 (69)	3 (9)	2 (6)
Leading to dose discontinuation
Adagrasib	0	—	—	—	—
Cetuximab	5 (16)	—	—	—	—
Leading to dose interruption
Adagrasib	14 (44)	—	—	—	—
Cetuximab	10 (31)	—	—	—	—
Leading to dose reduction
Adagrasib	10 (31)	—	—	—	—
Cetuximab	1 (3)	—	—	—	—
Most frequent events †
Nausea	20 (62)	13 (41)	7 (22)	0	0
Diarrhea	18 (56)	11 (34)	6 (19)	1 (3)	0
Vomiting	17 (53)	13 (41)	4 (12)	0	0
Dermatitis acneiform	15 (47)	11 (34)	3 (9)	1 (3)	0
Fatigue	15 (47)	8 (25)	7 (22)	0	0
Dry skin	13 (41)	11 (34)	2 (6)	0	0
Headache	10 (31)	7 (22)	3 (9)	0	0
Dizziness	8 (25)	4 (12)	4 (12)	0	0
Maculopapular rash	8 (25)	7 (22)	1 (3)	0	0
Stomatitis	7 (22)	5 (16)	1 (3)	1 (3)	0
Dyspepsia	6 (19)	4 (12)	2 (6)	0	0
Hypomagnesemia	6 (19)	3 (9)	3 (9)	0	0
Infusion-related reaction	6 (19)	1 (3)	4 (12)	0	1 (3)
Peripheral edema	6 (19)	3 (9)	3 (9)	0	0
Rash	6 (19)	3 (9)	3 (9)	0	0
Increased ALT	5 (16)	3 (9)	2 (6)	0	0
Decreased appetite	5 (16)	2 (6)	3 (9)	0	0
Dyspnea	5 (16)	3 (9)	2 (6)	0	0
Prolonged QT interval on ECG	5 (16)	2 (6)	2 (6)	1 (3)	0
Dehydration	4 (12)	2 (6)	1 (3)	1 (3)	0
Hypokalemia	4 (12)	3 (9)	1 (3)	0	0

^*Adverse events were counted once at their maximum severity in all the patients. Dashes indicate that data were available only for “any grade.” ALT denotes alanine aminotransferase, AST aspartate aminotransferase, and ECG electrocardiogram.

^†The listed adverse events were observed in at least 10% of the patients.

Adagrasib plus Cetuximab

Of the 32 patients in the combination-therapy cohort, treatment-related adverse events of any grade occurred in 100% of patients (Table 3). The events that occurred in at least 20% of the patients were nausea (62%), diarrhea (56%), vomiting (53%), dermatitis acneiform (47%), fatigue (47%), dry skin (41%), headache (31%), dizziness (25%), maculopapular rash (25%), and stomatitis (22%). Five patients (16%) had grade 3 or 4 treatment-related adverse events, with none occurring in more than 1 patient. Treatment-related adverse events that led to a reduction in the dose of either adagrasib or cetuximab occurred in 10 patients (31%) and 1 patient (3%), respectively (Table S9). Treatment-related adverse events leading to cetuximab discontinuation were reported in 5 patients (16%) and included infusion-related reactions (in 3 patients), malaise (in 1 patient), and vascular flushing (in 1 patient). No treatment-related adverse events led to the discontinuation of adagrasib. All 5 patients continued to receive adagrasib after the discontinuation of cetuximab. No grade 5 treatment-related events were reported. Adverse events that were observed during treatment are summarized in Table S10.

DISCUSSION

The therapeutic landscape of metastatic colorectal cancer is moving toward biomarker-selected strategies. However, options are limited to agents directed at tumors carrying BRAF V600E mutations and to antibodies against programmed death 1 (PD-1) in patients with high microsatellite instability or mismatch-repair deficiency.^13,26,27 KRAS G12C mutations occur in approximately 3 to 4% of patients with metastatic colorectal cancer and are associated with shorter survival than are other KRAS mutations and KRAS wild-type tumors.^{10–12,17,28} The emergence of drugs that target mutant KRAS G12C represents a new therapeutic strategy in this population. In this trial, we provide clinical evidence that KRAS G12C can be targeted in metastatic colorectal cancer, with activity observed for adagrasib as monotherapy and in combination with cetuximab in heavily pretreated patients (i.e., in those who had received a median of three previous lines of therapy, the majority of whom had disease that was resistant to previous therapies). These results show the activity of combined KRAS and EGFR inhibition in patients with colorectal cancer, a finding that is consistent with preclinical observations.²⁴

Adagrasib and sotorasib are both selective covalent inhibitors of KRAS G12C, but pharmacologic differences between the two drugs have been reported, including half-life (5 hours for sotorasib and 23 hours for adagrasib), dose-dependent exposure with adagrasib, and potential CNS penetration with adagrasib.^22,29 In a recent phase 2 trial of sotorasib, a response of 10% was observed, with a median response duration of 4.2 months in pretreated patients with metastatic colorectal cancer with mutant KRAS G12C.³⁰ In our trial, adagrasib monotherapy resulted in a response of 19% and a median response duration of 4.3 months. In addition, adagrasib monotherapy was associated with a median progression-free survival of 5.6 months and median overall survival of 19.8 months. These results are clinically meaningful and in line with the activity seen in 220 patients with metastatic colorectal cancer with BRAF V600E mutations who received a regimen of encorafenib plus cetuximab that had been approved by the Food and Drug Administration. Among these patients, the response was 19.5%, with progression-free survival of 4.3 months, overall survival of 9.3 months, and a response duration of 5.5 months.^26,31 Outcomes with adagrasib monotherapy compared favorably with those associated with trifluridine–tipiracil (a current standard-of-care third-line treatment) that were observed in a trial involving 534 patients, who had a response of 1.6%, progression-free survival of 2.0 months, and overall survival of 7.1 months. Findings in the monotherapy cohort also compared favorably with the results of a trial of regorafenib (another third-line treatment) involving 505 patients, who had a response of 1.0%, progression-free survival of 1.9 months, and overall survival of 6.4 months.^15,16

Reactivation of the EGFR pathway in metastatic colorectal cancer as an adaptive response to KRAS G12C inhibition suggests that combining inhibitors of KRAS G12C and EGFR may enhance biologic activity.²⁴ Early clinical data support this hypothesis, and in a recent phase 1b study of sotorasib in combination with the EGFR-targeting monoclonal antibody panitumumab, a confirmed response of 30% was reported in pretreated patients who had colorectal cancer with a KRAS G12C mutation,³² as compared with 10% for sotorasib monotherapy.³⁰ In the current trial, the combination of adagrasib and cetuximab was associated with a response of 46% and median response duration of 7.6 months. In addition, combination therapy was associated with a median progression-free survival of 6.9 months, which suggests a potential for enhanced clinical benefit through a combination approach. The potential improvement in activity observed for the combination is consistent with outcomes reported in 665 patients who had metastatic colorectal cancer with a BRAF V600E mutation, in which combined BRAF and EGFR inhibition led to improved activity over BRAF inhibitor monotherapy.³³ In our trial, 6 patients (14%) had progressive disease as the best response with adagrasib monotherapy; no patient had progressive disease as the best response with adagrasib plus cetuximab. Maintaining disease control may translate into improved quality of life for patients, because disease progression is associated with deterioration in health-related quality of life across several cancer types, including metastatic colorectal cancer.³⁴ The combination of adagrasib plus cetuximab did not produce synergistic toxic effects, and observed treatment-related adverse events were consistent with those reported for each treatment alone.

In an exploratory analysis of ctDNA, at least 95% plasma clearance of the KRAS G12C mutant allele was observed in 55% of evaluable patients in the monotherapy cohort and in 88% in the combination-therapy cohort, which suggests a greater depth of ctDNA response for the combination.

TP53 and PIK3CA are commonly mutated in colorectal cancer and have been postulated to potentially affect treatment response and sensitivity to KRAS inhibition.^35,36 Exploratory analyses of tumor molecular profiles showed that co-occurring mutations in TP53 and PIK3CA had no apparent association with the response to adagrasib, but interpretation was limited by the small sample size. Emerging data suggest that potential mechanisms of acquired resistance to KRAS G12C inhibitors in colorectal cancer largely include reactivation of MAPK signaling, and future analyses from this trial and others will help to clarify these mechanisms.³⁷

Our trial has some limitations. The nonrandomized design does not allow for direct comparisons between cohorts. The small number of patients who were enrolled resulted in wide confidence intervals for median overall survival in the two cohorts. In addition, a larger percentage of patients received subsequent anticancer therapy in the monotherapy cohort than in the combination-therapy cohort (66% and 47% of patients, respectively). The documentation of disease progression while patients were receiving previous therapy was not a specific enrollment criterion for these cohorts; however, this was a patient population with poor prognostic features and for whom no standard-of-care treatment option was available. All the patients had received previous fluoropyrimidine-based therapy with oxaliplatin, irinotecan, or both, and most had received an inhibitor of vascular endothelial growth factor before enrollment. The activity and safety of adagrasib plus cetuximab in patients with colorectal cancer with mutated KRAS G12C is being evaluated in the ongoing phase 3 KRYSTAL-10 trial as second-line treatment as compared with standard chemotherapy (ClinicalTrials.gov number, NCT04793958)²⁵ and in the ongoing phase 2 cohort of the current trial as late-line treatment.

In this phase 1–2 trial of adagrasib as monotherapy and in combination with cetuximab, both treatments produced reversible toxic effects in the majority of patients and resulted in no new safety concerns. Adagrasib showed promising clinical activity in heavily pretreated patients with metastatic colorectal cancer with a KRAS G12C mutation. The biologic activity of adagrasib appeared to be even greater in combination with cetuximab and supports ongoing clinical investigation.