What it is:A malignant neoplasm of the colon, rectum, or anal canal. About 92% of these tumors are adenocarcinomas. It's one of the few cancers where a screening test can catch and physically remove the precursor lesion (a polyp) before it ever becomes cancer.
The core problem:Colorectal cancer (CRC) develops through a multistep pileup of genetic and phenotypic hits to normal bowel epithelium, genomic instability, oncogene activation, tumor-suppressor silencing, DNA mismatch repair failures, growth factor pathway activation, that together push cells into unregulated growth. Those exact same molecular hits are what later decide which drug will actually work if the disease comes back.
What you do about it:Stage I-III is potentially curable: surgery is the backbone, with chemo (and radiation, for rectal disease) layered on to mop up micrometastatic disease. Stage IV is mostly incurable, but 20-30% of metastatic patients can still be cured if their metastases are resectable. Everything downstream of "is it resectable" runs through a biomarker panel, not just a stage number.
Treat this chapter as two separate decisions made in sequence.First: what stage is it, and does surgery cure it? Second, once you're in metastatic territory: what is the tumor's molecular fingerprint (RAS, BRAF, HER2, MMR/MSI, sidedness), because that fingerprint decides which targeted drug is even eligible to be used. Skip the second question and you'll pick a drug that can't work no matter how appropriate it looks on paper.
Stage is assigned by the standard TNM framework: size/depth of the primary tumor (T1-4), lymph node involvement (N0-2), and distant metastasis (M). It is the single most important independent predictor of survival and recurrence.
| Stage | Criteria |
|---|---|
| Stage I | Tumor invades submucosa (T1) or muscularis propria (T2); negative lymph nodes |
| Stage II | Tumor invades through the muscularis propria into pericolorectal tissues (T3), penetrates to the surface of the visceral peritoneum (T4a), or directly invades/is adherent to other organs (T4b); negative lymph nodes |
| Stage III | Any T (T1-4) with positive regional lymph nodes |
| Stage IV | Any T, any N, with distant metastasis |
Nodal status, not tumor depth, is what separates Stage II from Stage III.A T4b tumor with clean nodes is still Stage II. A tiny T1 tumor with one positive node is already Stage III. Don't let "T4 sounds scary" push you to Stage III or IV automatically, check the N and M before you commit to a stage.
Screening colonoscopy with polyp removal is secondaryprevention, not primary, even though it happens before cancer is diagnosed. Primary prevention stops the polyp from forming (or turning into a threat) in the first place; secondary prevention catches it once it's already there and removes it before it can progress.
CRC tumorigenesis is a multistep process: genomic instability, activation of oncogene pathways, mutational inactivation or silencing of tumor-suppressor genes, defects in DNA mismatch repair, and activation of growth factor signaling pathways all stack up to convert normal bowel epithelium into a proliferating tumor. Adenocarcinomas make up about 92% of large intestine tumors.
The reason this matters beyond biology class: every one of those molecular hits becomes a biomarker that decides drug eligibility once you reach metastatic disease.You can't pick a targeted agent rationally without knowing which hits this specific tumor has.
| Marker | What it means biologically | Therapeutic consequence |
|---|---|---|
| dMMR / MSI-H | Broken DNA mismatch repair lets mutations accumulate unchecked, generating a heavy load of abnormal tumor proteins (neoantigens) | Highly visible to the immune system; this is the subgroup where checkpoint inhibitors (pembrolizumab, nivolumab +/- ipilimumab) actually work |
| RAS mutation (KRAS/NRAS) | RAS sits downstream of EGFR in the same growth-signaling cascade; a mutated RAS is locked in the "on" position regardless of what happens upstream at the receptor | Blocking EGFR (cetuximab, panitumumab) accomplishes nothing if RAS is already firing on its own. Anti-EGFR therapy is restricted to RAS wild-type tumors only |
| BRAF V600E | A downstream signaling mutation associated with aggressive, treatment-resistant biology and worse prognosis | Specifically targetable with the BRAF inhibitor encorafenib, always paired with an EGFR inhibitor rather than used alone |
| HER2 amplification | Overexpression of a separate growth-signaling receptor; more common in tumors that are otherwise RAS and BRAF wild-type | Opens the door to HER2-directed antibody-based regimens |
| Tumor sidedness(right vs left) | Right-sided (cecum to hepatic flexure) and left-sided colon reflect different embryologic origin and correlate with different underlying tumor biology | Left-sided, RAS/BRAF wild-type tumors get an overall survival benefit from EGFR inhibitors; right-sided tumors generally don't, even when wild-type |
Tumor genotyping for RAS and BRAF mutation status, HER2 amplification, and MMR/MSI status is recommended for essentially every patient headed toward systemic therapy for metastatic disease. Ordering these tests isn't optional busywork, it's the step that determines which entire categories of drugs are even on the table.
Signs and symptoms are extremely varied, subtle, and often nonspecific. Early-stage CRC is frequently completely asymptomatic and only found through screening.
| Finding | Notes |
|---|---|
| Blood in the stool | The most common sign |
| Change in bowel habits | New constipation, diarrhea, or altered caliber can be a warning sign |
| Abdominal discomfort or distention | Nonspecific, easy to attribute to something else |
| Nausea, vomiting | Less common |
| Fatigue | Only when anemia from chronic blood loss becomes severe |
Twenty percent of patients present with metastatic disease at diagnosis, most commonly to the liver, lungs, and bones. Asymptomatic disease isn't necessarily early disease, that's exactly why average-risk screening starting at 45 matters so much.
Start with a physical exam and a careful personal and family history, then evaluate the entire large bowel by colonoscopy.
CEA is a marker for monitoring response to treatment, not a screening test. It's too insensitive and nonspecific to catch early-stage CRC. Its value shows up later: after curative resection, a rising CEA on routine follow-up can flag a recurrence before the patient has any symptoms at all.
Stage at diagnosis is the most important independent prognostic factor for both survival and recurrence.
| Extent at diagnosis | 5-year relative survival |
|---|---|
| Localized tumor | ~90% |
| Metastatic disease at diagnosis | ~14% |
Goal: cure.Treatment modalities are surgery, radiation, chemotherapy, and biomodulators, but for operable disease surgery is the curative backbone and everything else supports it.
Complete surgical resection of the primary tumor with regional lymphadenectomy is the curative approach. For rectal cancer specifically, the preferred procedure is total mesorectal excision, removing the tissue containing perirectal fat and the draining lymph nodes. Recognized complications include infection, anastomotic leakage, obstruction, adhesions, sexual dysfunction, and malabsorption syndromes.
RT has a limited role in colon cancer because most recurrences are extrapelvic, occurring in the abdomen rather than the pelvis. A subset of patients with recurrent disease or T4 tumors that have penetrated fixed structures may benefit from neoadjuvant (preoperative) fluorouracil-based chemoradiation specifically to improve resectability.
Patients deficient in dihydropyrimidine dehydrogenase (the enzyme that catabolizes fluorouracil) can develop severe, sometimes fatal, toxicity after fluorouracil. This is rare but worth knowing cold, since it's an idiosyncratic, dose-independent risk.
Continuous IV infusion fluorouracilis generally well tolerated but associated with palmar-plantar erythrodysesthesia (hand-foot syndrome) and stomatitis. IV bolus fluorouracilis associated with leukopenia, which is dose-limiting and can be life-threatening. Both routes carry a similar incidence of mucositis, diarrhea, nausea/vomiting, and alopecia. No single administration schedule has proven superior for overall survival, which is why common regimens like FOLFOX actually use both bolus and continuous infusion fluorouracil in the same cycle.
Oxaliplatin-based regimensare the national guideline first-line option for stage III patients who can tolerate combination therapy, usually paired with fluorouracil/leucovorin. Oxaliplatin causes both acute and persistent neuropathies, including a rare acute pharyngolaryngeal dysesthesia, plus neutropenia and GI toxicity. Regimen selection weighs performance status, comorbid conditions, and patient lifestyle preferences; patients older than 70have not shown a clear benefit from adjuvant oxaliplatin in subset analyses. Fluorouracil/leucovorin alone now has limited use, but remains an acceptable option when a patient can't receive oxaliplatin and can't tolerate oral capecitabine either.
Rectal cancer is harder to resect with wide margins than colon cancer, so local recurrences happen more often. For stages II and III, the standard is either neoadjuvant chemoradiation followed by surgery and adjuvant chemotherapy, or total neoadjuvant therapy(fluoropyrimidine-based chemo followed by chemoradiation, or the reverse order, then surgery). FOLFOX and CAPEOX are the preferred fluoropyrimidine-based regimens here, though fluorouracil/leucovorin combinations and capecitabine alone are also usable.
Neoadjuvant chemoradiation significantly reduces local recurrence, causes fewer toxicities, and improves rates of sphincter-preserving surgery compared to giving chemoradiation after surgery. Patients who can't tolerate chemoradiation should be offered preoperative RT alone instead. After surgery, patients receive adjuvant chemotherapy to bring the total course to 6 months.
Colon cancer:surgery first, then adjuvant chemo if indicated. Rectal cancer:chemoradiation typically comes before surgery. The driver is anatomy, the rectum sits in a tight pelvic space where achieving clean margins is harder, so shrinking the tumor first improves both resectability and the odds of preserving the sphincter.
Metastatic CRC (MCRC) is sorted into three buckets: resectable, potentially resectable, or unresectable.Surgery and RT manage isolated sites of tumor; chemotherapy manages disseminated disease and is the primary treatment modality for unresectable disease. Tumor genotyping (RAS, BRAF, HER2, MMR/MSI) is recommended up front if not already done, since it determines which drug categories apply.
Surgical resection of metastases with curative intent is the primary goal, and 20-30% of patients with metastatic disease can be cured if their metastases are resectable.Neoadjuvant (conversional) chemotherapy, FOLFOX, CAPEOX, FOLFIRI, or FOLFOXIRI, is given for about 2-3 months before surgery to increase complete resection rates with resectable or potentially resectable liver or lung lesions. After surgery, adjuvant chemotherapy (preferably FOLFOX or CAPEOX) completes a total of 6 months of chemotherapy. The optimal sequencing, surgery first versus perioperative chemo bracketing the surgery, remains controversial.
For patients with liver-only or liver-predominant MCRC, consider hepatic-directed therapy in addition to or instead of surgical resection. Hepatic artery infusion (HAI)delivers chemotherapy (floxuridine, fluorouracil) directly into the liver through the hepatic artery. Tumor ablationuses radiofrequency or microwave energy to generate heat and destroy tumor cells; cryoablationis another option. All of these strategies are less successful than surgical resection, they're a fallback when resection itself isn't feasible.
Systemic chemotherapy palliates symptoms and improves survival; RT can control localized symptoms. Most MCRC is incurable, but randomized trials confirm chemotherapy both prolongs life and improves quality of life. Regimens are the same for metastatic colon and rectal cancer. Decision-making weighs goals of therapy, prior chemo history, RAS and MMR/MSI mutation status, performance status/comorbidities, and risk of drug-related toxicity.
These are the biomodulators layered onto chemotherapy backbones in metastatic disease, each gated by a specific biomarker.
Bevacizumab is a recombinant, humanized monoclonal antibody that inhibits vascular endothelial growth factor (VEGF). Adding it to fluorouracil-based regimens modestly increases both PFS and OS compared to chemotherapy alone.
Hypertension is common with bevacizumab and usually easy to manage with oral antihypertensives. Other safety concerns are bleeding, thrombocytopenia, and proteinuria. GI perforation is rare but potentially fatal, promptly evaluate any abdominal pain associated with vomiting or constipation. Bevacizumab also interferes with wound healing: schedule surgery at least 6-8 weeks after the last dose, and wait at least 6-8 weeks after surgerybefore restarting it.
Cetuximab and panitumumab are epidermal growth factor receptor (EGFR) inhibitors used only in patients with wild-type RAS and BRAFtumors, in combination with FOLFOX or FOLFIRI, or given alone. Patients with left-sidedprimary tumors have improved OS on EGFR inhibitors; those with right-sidedtumors (cecum to hepatic flexure) do not, even when wild-type.
Severe infusion reactions, including anaphylaxis, occur in about 3% with cetuximaband 1% with panitumumab. Skin toxicity is common and is a separatephenomenon from the infusion reaction. A papulopustular skin rash correlates with response and survival and most commonly shows up 2-4 weeksafter starting therapy.
A two-drug combination of encorafenib(a BRAF inhibitor) plus an EGFR inhibitor (cetuximab or panitumumab) improves outcomes in second- and subsequent-line therapy for patients with a BRAF V600E mutation. It is not used as monotherapy.
Trastuzumab and pertuzumab are monoclonal antibodies directed against HER2. The combination of trastuzumab plus pertuzumab or lapatinib, or fam-trastuzumab deruxtecan-nxki alone, is an option specifically for patients with HER2 amplification and wild-type RAS and BRAF tumors. HER2 overexpression is rare in CRC overall but is more common in tumors that are otherwise RAS and BRAF wild-type.
Baseline and ongoing left ventricular ejection fraction monitoring is required. Do not substitute conventional or biosimilar trastuzumab with ado-trastuzumab emtansine, fam-trastuzumab deruxtecan, or trastuzumab/hyaluronidase, these are not interchangeable products.
Pembrolizumab, a humanized IgG4 monoclonal antibody that binds PD-L1 with high affinity, is effective in MCRC patients with dMMR who've progressed through 2-4 other regimens. Nivolumab, another humanized IgG4 monoclonal antibody targeting PD-1, has also been studied with or without ipilimumab in patients with dMMR/MSI-H tumors. Neither works outside this biomarker-defined population, they are only appropriate in MMR-deficient or MSI-high disease.
Watch for immune-mediated adverse reactions, most commonly affecting skin, liver, kidney, GI tract, lung, and the endocrine system. Patients need close monitoring and should report any new symptom immediately, since management sometimes requires interrupting treatment or starting corticosteroids.
Patients with metastatic disease may cycle through multiple different regimens over their treatment course. The sequence of drugs appears less important than making sure the patient is exposed to allthe active agent classes available to them somewhere along the way. FOLFIRI and FOLFOX are both considered reference standards in MCRC; their differing toxicity profiles, not superior efficacy of one over the other, is what usually drives which comes first.
Adjuvant regimens treat micrometastatic disease after resection. First-line metastatic regimens are organized by fitness for intensive therapy and by biomarker status, biology picks the lane before dose ever matters.
| Regimen | Agents | Notes |
|---|---|---|
| FOLFOX (mFOLFOX6) | Oxaliplatin 85 mg/m² IV D1 + leucovorin 400 mg/m² IV D1 + fluorouracil 400 mg/m² IV bolus D1, then 1200 mg/m²/day x2 days CIV (total 2400 mg/m² over 46-48 hr). Repeat q2wk x24 wks | Preferred for stage III colon and rectal cancer. Toxicities: sensory neuropathy, neutropenia. No demonstrated survival benefit for age ≥70 |
| CAPEOX | Oxaliplatin 130 mg/m² IV D1 + capecitabine 1000 mg/m² PO BID D1-14. Repeat q3wk x24 wks | Improved DFS vs capecitabine alone or Roswell Park regimen. DLTs: neuropathies, hand-foot syndrome. Preferred regimen for adjuvant rectal therapy |
| Capecitabine monotherapy | 1000-1250 mg/m² PO BID D1-14. Repeat q3wk x24 wks | Equivalent DFS to the Mayo Clinic regimen, better tolerated. Useful with poor vascular access or travel barriers to an infusion center |
| Roswell Park regimen | Leucovorin 500 mg/m² IV D1 over 2 hr + fluorouracil 500 mg/m² IV bolus 1 hr after leucovorin. Weekly for 6 of 8 wks x4 cycles | Leukopenia is the common dose-limiting toxicity; diarrhea and stomatitis also common |
| Simplified biweekly | Leucovorin 400 mg/m²/day IV + fluorouracil 400 mg IV bolus, then 1200 mg/m²/day D1-2 (total 2400 mg/m² over 46-48 hr). Repeat q2wk x12 cycles | Hand-foot syndrome common |
In low-risk stage IIIdisease (T1-3, any N), 3 months of CAPEOX is noninferior to 6 months for DFS, though this hasn't been proven for FOLFOX. In high-risk stage IIIdisease (T4, N1-2, or any T with N2), 3 months of FOLFOX is inferiorto 6 months for DFS, though this hasn't been proven for CAPEOX either. Grade 3 neuropathy is lower with the 3-month course of either regimen, so shortening duration is a real toxicity/efficacy tradeoff, not a free lunch.
| Category | Regimen options |
|---|---|
| Fit for intensive therapy, RAS-mutant | |
| FOLFOX, CAPEOX, FOLFIRI, or FOLFOXIRI, each +/- bevacizumab | FOLFOX is the most commonly used first-line regimen overall. FOLFOXIRI is more intense (more neutropenia, peripheral neurotoxicity) and reserved for fit patients with diffuse, aggressive disease or as conversion therapy |
| Fit for intensive therapy, RAS/BRAF wild-type, left-sided | |
| FOLFOX or FOLFIRI + cetuximab or panitumumab | FOLFIRI-based is often preferred with preexisting neuropathy, since oxaliplatin isn't in the picture |
| Not appropriate for intensive therapy, RAS-mutant | |
| Infusional fluorouracil/leucovorin, or capecitabine, each +/- bevacizumab | Infusional fluorouracil/leucovorin is preferred over bolus dosing here |
| Not appropriate for intensive therapy, RAS/BRAF wild-type, left-sided | |
| Cetuximab or panitumumab alone | Single-agent EGFR inhibitor |
| dMMR or MSI-H (any fitness level) | |
| Pembrolizumab, or nivolumab +/- ipilimumab | Only indicated in MMR-deficient or MSI-high tumors |
| HER2-amplified, RAS/BRAF wild-type | |
| Trastuzumab + (pertuzumab or lapatinib), or fam-trastuzumab deruxtecan-nxki alone | Requires baseline and ongoing LVEF monitoring |
Second-line selection is driven primarily by what the patient already received first-line, plus site/extent of disease and patient preference. There's no single established optimal sequence, but the logic is consistent: don't repeat a drug class that already failed, and don't give a targeted agent outside its biomarker.
| Progressed on | Reasonable next options |
|---|---|
| Oxaliplatin-based regimen (FOLFOX, CAPEOX), no prior irinotecan | FOLFIRI or irinotecan; FOLFIRI + bevacizumab/ziv-aflibercept/ramucirumab; irinotecan + bevacizumab/ziv-aflibercept/ramucirumab; if RAS/BRAF wild-type, add cetuximab or panitumumab to FOLFIRI or irinotecan |
| Irinotecan-based regimen (FOLFIRI), no prior oxaliplatin | FOLFOX or CAPEOX +/- bevacizumab; if RAS/BRAF wild-type, FOLFOX + cetuximab or panitumumab |
| BRAF V600E mutation positive | Encorafenib + cetuximab or panitumumab |
| dMMR/MSI-H | Nivolumab +/- ipilimumab, dostarlimab, or pembrolizumab |
| HER2-amplified, RAS/BRAF wild-type | Trastuzumab + (pertuzumab or lapatinib), or fam-trastuzumab deruxtecan-nxki |
Use bevacizumab, ziv-aflibercept, or irinotecan with caution in patients with elevated bilirubin. Bevacizumab is generally the preferred antiangiogenic agent on cost and toxicity grounds when there's a choice among the three.
| Option | Notes |
|---|---|
| Regorafenib | Oral multikinase inhibitor; used after progression through all other available regimens |
| Trifluridine/tipiracil +/- bevacizumab | Approved for patients previously treated with a fluoropyrimidine-, oxaliplatin-, and irinotecan-containing regimen, an anti-VEGF agent, and an anti-EGFR antibody if RAS wild-type |
| Clinical trial | If available and the patient is eligible |
| Best supportive care | Appropriate when a patient doesn't want further cancer-directed therapy, isn't eligible for it, or quality of life is expected to decline with more treatment |
VEGF inhibitors (bevacizumab, ramucirumab, ziv-aflibercept) and the oral multikinase inhibitor regorafenib remain usable in patients who've progressed on other therapies, angiogenesis blockade isn't a "used it once, done" strategy the way some targeted agents are.
| Parameter | When | Watching for |
|---|---|---|
| General chemo toxicities | Every cycle | Loose stools/diarrhea, nausea/vomiting, mouth sores, fatigue, fever |
| Diarrhea on irinotecan | Each cycle, and any time it occurs | Requires prompt intervention, not a "wait it out" symptom |
| Abdominal pain on bevacizumab | Ongoing | GI perforation, especially with concurrent vomiting or constipation; also hypertension and proteinuria |
| Neuropathy on oxaliplatin | Each cycle, cumulative | Acute cold-triggered dysesthesia and chronic, potentially persistent sensory neuropathy |
| Skin exam on cetuximab/panitumumab | First 2-4 weeks especially | Papulopustular rash; correlates with response, don't mistake it for an allergy |
| Serum CEA | Routine follow-up after curative resection | Recurrence, often detectable before symptoms appear |
| Symptoms of recurrence | Every follow-up visit | Pain syndromes, bowel habit changes, rectal or vaginal bleeding, pelvic masses, anorexia, weight loss (fewer than half of recurrences are symptomatic) |
| Quality-of-life indices | Especially in metastatic disease | Overall treatment burden vs benefit |