KRAS G12C in Non–Small Cell Lung Cancer: A Predictive Biomarker That Rewrote the Rules

For most of my career, KRAS was the mutation you reported and then apologized for — the one that told the clinician a target existed but that nothing could be done about it. That changed on May 28, 2021, when sotorasib became the first approved drug to directly inhibit a KRAS protein, validated by the CodeBreaK 100 trial [1, 2]. Adagrasib followed in December 2022 on the strength of KRYSTAL-1 [3]. The relevance is not academic: KRAS G12C is among the more common actionable alterations in lung adenocarcinoma, and identifying it is now a routine, consequential part of the diagnostic workup.

What the Test Measures

KRAS encodes a small GTPase that toggles between an active, GTP-bound state and an inactive, GDP-bound state to relay growth signals downstream of receptor tyrosine kinases. Oncogenic mutations lock the protein in a signaling-competent conformation, driving proliferation.

The G12C variant is a specific substitution: a single-nucleotide change (c.34G>T) at codon 12 replaces glycine with cysteine. That newly introduced cysteine residue is the entire point. The approved inhibitors are covalent drugs — they form an irreversible bond to the reactive thiol of that cysteine, trapping KRAS in its inactive state [2, 3]. No cysteine, no covalent handle, no drug binding. This is why the biomarker is defined at the level of a single amino acid rather than at the level of the gene or the codon. A KRAS G12D or G12V tumor has a mutation at the very same codon and is emphatically not a candidate for these agents. The chemistry, not the map position, dictates eligibility.

How It's Tested

KRAS G12C is detected on tumor tissue (formalin-fixed, paraffin-embedded, FFPE) or on circulating tumor DNA (ctDNA) extracted from plasma. Two broad assay approaches are used:

• Next-generation sequencing (NGS): Panel-based sequencing reads the KRAS gene and reports the specific variant, ideally at nucleotide-level resolution. NGS has the advantage of interrogating co-occurring alterations in the same run.
• PCR-based assays: Targeted PCR can call the specific c.34G>T change with high sensitivity when the variant of interest is known in advance.

Preanalytic constraints matter. FFPE tissue must contain adequate tumor cellularity; a low tumor fraction can push a true-positive variant below the assay's limit of detection. Fixation artifacts, particularly cytosine deamination, can generate spurious low-frequency C>T changes — a real interpretive nuisance when the variant you're hunting for is itself a G>T transversion. A sequencing pipeline that does not account for these artifacts can, in principle, muddy a low-level call.

Scoring is variant-specific and binary in practice. The report must distinguish G12C from all other KRAS variants. A generic "KRAS mutation positive" result is inadequate for this purpose, because it does not establish eligibility for the drug class.

What Each Result State Means

• KRAS G12C: The tumor carries the targetable cysteine substitution. This is the result that opens the door to the G12C inhibitor class.
• Other KRAS (e.g., G12D, G12V, G13): A KRAS driver is present, but it is not the target of the currently approved covalent inhibitors. These patients are not eligible for the G12C agents on the basis of this finding.
• Wild-type: No KRAS mutation detected. Interpretation depends on the assay context — and this is where specimen type becomes decisive.

The interpretive asymmetry of ctDNA deserves emphasis. A positive plasma result is clinically actionable and can spare a patient a repeat biopsy. A negative plasma result does not exclude the mutation: tumors shed DNA at variable rates, and a tumor that sheds poorly can yield a falsely wild-type plasma result. When plasma is negative and clinical suspicion is high, tissue testing remains the arbiter. The rule of thumb I'd offer trainees: in ctDNA, trust the positive, interrogate the negative.

What It Determines for Treatment Eligibility

A confirmed KRAS G12C result informs eligibility for the KRAS G12C inhibitor class — sotorasib and adagrasib — in the metastatic setting after prior systemic therapy [1, 2, 3]. The approvals were granted in the second-line context, meaning the biomarker gates access to this class of drugs for patients whose disease has progressed on previous treatment.

To be explicit about the boundaries of this article: a positive result establishes eligibility for a therapeutic class. It does not, by itself, dictate that any individual patient should receive any specific drug. That determination integrates performance status, prior therapies, comorbidities, and co-mutations, and belongs to the treating oncologist in conversation with the patient.

Caveats and What's Evolving

1. Variant specificity cuts both ways. The elegance of covalent G12C inhibition is also its limitation. G12D — a far more common variant in some tumor types — substitutes aspartate for glycine and offers no reactive cysteine, so the covalent strategy simply does not apply; distinct chemical approaches are required and remain investigational [4]. A reporting pathologist should resist the temptation to collapse all codon-12 variants into a single actionable bucket.

2. Resistance is emerging and heterogeneous. As with other targeted agents, tumors acquire resistance to G12C inhibitors through diverse mechanisms, and the durability of response is an active area of study. The evidence here is still maturing, and the field is moving quickly.

3. Co-mutations may modulate benefit. The genomic context in which G12C occurs — particularly co-occurring alterations — is under investigation as a determinant of response, which is one argument for broad NGS panels over narrow single-gene assays.

4. Line-of-therapy context is not fixed. The current approvals sit in the second-line setting, but ongoing trials are examining these inhibitors earlier in the treatment course and in combination with other agents [4]. Where the biomarker sits in the treatment sequence may shift as those data mature.

A closing thought aimed squarely at the diagnostic bench: the KRAS G12C story is a reminder that biomarker reporting is not a clerical act of transcription. The difference between "KRAS mutant" and "KRAS G12C (c.34G>T)" is the difference between a dead end and a therapeutic option — and it is the pathologist who guarantees that distinction survives from the sequencer to the report. As more variant-specific drugs enter the clinic, panel design and reporting granularity will need to keep pace. The "undruggable" era of KRAS is over; the era in which precise variant-level reporting carries real therapeutic weight has arrived, and it places the interpretive burden exactly where it should be.

References

1. Hong DS, Fakih MG, Strickler JH, et al. KRAS^G12C Inhibition with Sotorasib in Advanced Solid Tumors. N Engl J Med. 2020;383(13):1207–1217.

2. Skoulidis F, Li BT, Dy GK, et al. Sotorasib for Lung Cancers with KRAS p.G12C Mutation. N Engl J Med. 2021;384(25):2371–2381.

3. Jänne PA, Riely GJ, Gadgeel SM, et al. Adagrasib in Non–Small-Cell Lung Cancer Harboring a KRAS^G12C Mutation. N Engl J Med. 2022;387(2):120–131.

4. Fell JB, et al. FDA Approval Summary: Sotorasib for KRAS G12C-Mutated Metastatic Non-Small Cell Lung Cancer. Clin Cancer Res. 2022. PMCID: PMC9012672.