The first patient dosed AMG 510 in August 2018. So that sentence doesn't sound like much on paper. Practically speaking, a date. A code name. A single person in a clinical trial. But if you work in oncology — or if you've watched someone you love run out of options — that moment matters more than most headlines.
It was the first time anyone had ever targeted KRAS G12C in a human being.
For forty years, KRAS was the target everyone wanted and no one could hit. Now, the mutation sits at the center of some of the nastiest cancers: non-small cell lung, colorectal, pancreatic. "Undruggable" became its nickname. And for decades, the best we could do was chemotherapy and hope.
Then came that August day. One patient. One dose. And the whole field shifted.
What Is AMG 510
AMG 510 is the development code for sotorasib — the first KRAS G12C inhibitor to reach the clinic. Developed by Amgen, it's a small molecule designed to do something that looked impossible on paper: covalently bind to a specific cysteine residue on the mutant KRAS protein, locking it in its inactive GDP-bound state.
Let me say that in English. KRAS is a molecular switch. Also, in healthy cells, it toggles between "on" (GTP-bound) and "off" (GDP-bound). Worth adding: the G12C mutation — a glycine-to-cysteine swap at position 12 — traps the switch in the "on" position. In practice, cells keep dividing. Tumors grow.
AMG 510 exploits that very mutation. In real terms, the cysteine created by G12C becomes a handle. The drug grabs it. The switch gets stuck "off.
The chemistry that made it possible
Most drugs work by fitting into a pocket like a key in a lock. KRAS didn't seem to have a pocket. That said, it's a smooth, featureless GTPase — or so we thought. Which means in 2013, Kevan Shokat's lab at UCSF found a cryptic pocket near the switch-II region, accessible only when a specific cysteine was present. G12C provided that cysteine.
Amgen's medicinal chemists spent years optimizing molecules to reach that pocket. So they needed potency, selectivity, oral bioavailability, and a clean safety profile. AMG 510 — later named sotorasib — was the candidate that threaded every needle.
Why It Matters
KRAS mutations drive roughly 25% of all human cancers. G12C specifically accounts for about 13% of non-small cell lung cancer (NSCLC) adenocarcinomas and 1-3% of colorectal cancers. In the US alone, that's over 30,000 new NSCLC patients a year with this mutation.
Before AMG 510, these patients had no targeted option. They got chemo. Immunotherapy if their PD-L1 was high. Maybe a clinical trial. Median survival for previously treated KRAS G12C-mutated NSCLC hovered around 6-8 months.
The first patient dosed AMG 510 in August 2018 changed the conversation from "can we drug KRAS?" to "how far can we take this?"
The ripple effect
That single dosing event triggered a cascade. On the flip side, within two years, multiple companies had KRAS G12C inhibitors in clinic: MRTX849 (adagrasib), JNJ-74699157, GDC-6036, LY3499446. The "undruggable" target suddenly had a crowded pipeline.
More importantly, it proved covalent targeting of a mutant cysteine could work in humans — opening the door for allele-specific inhibitors across oncology. Practically speaking, kRAS G12D, G12V, G12R — different mutations, same logic. The field is now hunting cryptic pockets on every "undruggable" target.
How It Works
The mechanism is elegant. But the clinical development? That's where it gets messy — and interesting.
Phase 1: CodeBreaK 100
The first-in-human trial (NCT03600883) was a classic 3+3 dose escalation followed by expansion cohorts. Primary endpoints: safety, MTD, RP2D. Secondary: PK, ORR, DOR, PFS.
Dose escalation went fast. 180 mg, 360 mg, 720 mg, 960 mg daily. On top of that, no DLTs. The MTD wasn't reached. They settled on 960 mg once daily — a dose that achieved steady-state concentrations well above the IC90 in preclinical models.
Expansion cohorts enrolled heavily pretreated NSCLC, colorectal, and other solid tumors with confirmed KRAS G12C. By the 2020 ASCO presentation: 126 NSCLC patients, ORR 32.2%, DCR 88.1%, median PFS 6.3 months. In colorectal: 62 patients, ORR 7.1%, DCR 74.2%.
Why the colorectal numbers looked different
This is where biology humbles you. Here's the thing — same drug. Same mutation. Different outcomes.
Colorectal tumors have higher baseline EGFR signaling. When you inhibit KRAS G12C, feedback reactivation of EGFR bypasses the blockade. Still, the pathway turns back on through a different door. NSCLC doesn't do this as aggressively.
The fix? Combination therapy. Sotorasib + panitumumab (anti-EGFR) showed ORR 30% in refractory CRC. That's the power of understanding resistance mechanisms — something the first patient dosed in August 2018 started teaching us.
If you found this helpful, you might also enjoy what are the charges of protons or journal of applied materials and interfaces.
Pharmacokinetics that matter
Sotorasib has a short half-life (~5 hours) but prolonged target engagement. On the flip side, the covalent bond means the drug doesn't need to stay in plasma to keep working. Once bound, KRAS stays inhibited until the protein degrades and resynthesizes — roughly 24-48 hours.
This allows once-daily dosing despite rapid clearance. It also means food effect matters. On the flip side, high-fat meal increases AUC 1. On the flip side, 5x. The label says take with or without food — but consistency helps.
Common Mistakes / What Most People Get Wrong
"It's a cure for KRAS cancer"
No. It's a targeted therapy for one specific KRAS mutation (G12C) in certain tumor types. In real terms, kRAS has many variants: G12D, G12V, G12A, G12R, G13D, Q61H... Sotorasib doesn't touch them. Different drugs are needed for each.
"All G12C patients respond"
ORR ~37% in NSCLC means ~63% don't have objective shrinkage. Think about it: many get stable disease — which matters — but primary resistance is real. Mechanisms include: concurrent STK11/KEAP1 mutations (associated with poor IO response and possibly sotorasib resistance), MET amplification, BRAF mutations, histologic transformation.
We're still learning who benefits most.
"Resistance means the drug stopped working"
Resistance is heterogeneous. Some patients have on-target resistance (secondary KRAS mutations like Y96D, R68S, H95D/Q/R that block drug binding). Consider this: others have off-target bypass (MET amp, EGFR upregulation, FGFR signaling). Some have both.
Bi
opsy samples from progressing patients show that the tumor evolves. It doesn't just "wake up"; it adapts. This evolutionary pressure is why the field is moving toward "adaptive" or "sequential" dosing strategies, attempting to hit the bypass pathways before the tumor can pivot.
The Safety Profile: The Liver and the Gut
The most significant clinical concern with sotorasib is hepatotoxicity. Elevated ALT/AST are common, and in a small percentage of patients, these can escalate to severe drug-induced liver injury (DILI). This is often an idiosyncratic reaction rather than a dose-dependent one, making vigilant monitoring essential during the first two months of therapy.
Gastrointestinal side effects—diarrhea and nausea—are frequent but generally manageable with supportive care. Unlike traditional chemotherapy, the toxicity isn't myelosuppressive, which allows for a better quality of life, but the hepatic risk remains the primary reason for dose interruptions or permanent discontinuation.
The Bigger Picture: The G12C Landscape
Sotorasib was the "proof of concept" that KRAS was druggable, but it is no longer the only player. Adagrasib arrived shortly after, offering a different pharmacokinetic profile and potentially better CNS penetration, which is critical for patients with brain metastases.
The current clinical focus has shifted from "can we inhibit KRAS?" to "how do we keep it inhibited?" This has led to the development of:
- Pan-KRAS inhibitors: Drugs that target multiple KRAS variants regardless of the specific mutation.
- KRAS(ON) inhibitors: Sotorasib targets the inactive (GDP-bound) state; new agents are targeting the active (GTP-bound) state to ensure total pathway shutdown.
- Combination strategies: Pairing G12C inhibitors with SHP2 inhibitors or MEK inhibitors to block the feedback loops that lead to resistance.
Conclusion
The journey of sotorasib—from the early discovery of the "switch II" pocket to its FDA approval—represents one of the most significant triumphs of rational drug design in oncology. It transformed KRAS from an "undruggable" target into a manageable vulnerability.
Even so, the clinical reality remains a lesson in the complexity of cancer. On top of that, the disparity between NSCLC and CRC responses proves that a mutation is not a destiny; the cellular context (the "background" signaling) determines the outcome. By defining the mechanisms of resistance and the necessity of combination therapy, it has paved the way for a new generation of precision medicines that aim not just to shrink tumors, but to prevent them from adapting. While sotorasib provides a vital lifeline for a subset of patients, its true legacy is the roadmap it provided. The era of KRAS inhibition has begun, and while the first-generation inhibitors are the foundation, the future lies in the ability to anticipate the tumor's next move.
