According to Gram Research analysis, two blood markers—MTHFR and CEA—can predict how well advanced colorectal cancer tumors will shrink after CAPEOX chemotherapy with 71% accuracy. A 2026 study of 36 patients found that combining these blood tests created a predictive formula with a C-index of 0.836, meaning doctors could use a simple blood test before treatment to estimate whether a patient will respond well to chemotherapy.
Researchers in Indonesia studied 36 patients with advanced colorectal cancer to see if two blood markers could predict how much their tumors would shrink after chemotherapy. They measured MTHFR and CEA levels before treatment and tracked tumor changes using imaging scans. The study found that these two blood markers together were quite good at predicting which patients would respond well to chemotherapy. This discovery could help doctors personalize cancer treatment plans and avoid giving ineffective drugs to patients who won’t benefit from them.
Key Statistics
A 2026 prospective study of 36 advanced colorectal cancer patients found that serum MTHFR levels showed a strong correlation (0.764) with tumor size reduction following CAPEOX chemotherapy, with statistical significance of p < 0.001.
According to research reviewed by Gram, a predictive model combining MTHFR and CEA blood markers explained 71.4% of the variation in tumor shrinkage (adjusted R² = 0.714) in 36 colorectal cancer patients treated with neoadjuvant CAPEOX therapy.
A 2026 study of 36 advanced colorectal cancer patients demonstrated that a nomogram derived from MTHFR and CEA levels achieved a Harrell’s C-index of 0.836, indicating high discriminative ability in predicting therapeutic response to CAPEOX chemotherapy.
Research from 2026 involving 36 patients with stage III-IV colorectal cancer found that CEA levels correlated with tumor reduction at 0.654 (p < 0.001), making it a complementary biomarker alongside MTHFR for predicting chemotherapy response.
The Quick Take
- What they studied: Can two specific blood markers (MTHFR and CEA) predict whether a patient’s colon cancer tumor will shrink after receiving a specific chemotherapy combination called CAPEOX?
- Who participated: 36 patients with advanced stage 3 or 4 colorectal cancer in Makassar, Indonesia. Average age was 45.6 years, with about two-thirds being male patients.
- Key finding: Both blood markers together predicted tumor shrinkage with 71% accuracy. MTHFR showed a stronger connection (correlation of 0.764) than CEA (0.654), meaning higher MTHFR levels were more closely linked to better tumor response.
- What it means for you: If these findings hold up in larger studies, doctors could use a simple blood test before starting chemotherapy to predict whether the treatment will work well for an individual patient. This could help avoid unnecessary side effects from ineffective treatments, though this research is still preliminary and limited to one location.
The Research Details
This was a prospective observational study, meaning researchers followed patients forward in time and collected data as events happened naturally, rather than randomly assigning them to different treatments. The 36 patients with confirmed advanced colorectal cancer had their blood drawn before starting CAPEOX chemotherapy (a combination of two cancer drugs: capecitabine and oxaliplatin). Researchers measured two specific proteins in the blood: MTHFR (an enzyme involved in how the body processes certain nutrients) and CEA (a protein that cancer cells often produce in higher amounts).
After chemotherapy treatment, doctors used CT scans and other imaging to measure whether the tumors had shrunk. They compared the tumor size before and after treatment using standard medical criteria called RECIST, which is the established way doctors measure cancer response worldwide. The researchers then used statistical tests to see if the blood marker levels correlated with how much the tumors shrank.
Finally, they created a mathematical formula (called a regression model) that combines both blood markers to predict tumor shrinkage percentage. They also created a visual tool called a nomogram that doctors could use to quickly estimate a patient’s likely response based on their blood test results.
This research approach matters because it bridges the gap between laboratory science and real-world patient care. Rather than just studying how these proteins work in test tubes, researchers followed actual cancer patients through their treatment journey. This makes the findings more relevant to clinical practice. The study also tested whether combining two biomarkers works better than using just one, which reflects how modern precision medicine actually works—using multiple pieces of information to make better predictions.
Strengths: The study used standardized imaging criteria (RECIST) to measure tumor response, which is the gold standard in cancer research. The statistical correlations were very strong (p < 0.001), meaning the results are unlikely due to chance. The predictive model showed good performance with a C-index of 0.836, which indicates the formula correctly predicted outcomes most of the time. Limitations: The sample size of 36 patients is relatively small, which limits how much we can generalize these findings. The study was conducted in only one location in Indonesia, so results may not apply to other populations. This was an observational study, not a randomized controlled trial, so we cannot prove these markers directly cause better outcomes—only that they’re associated with them.
What the Results Show
Both blood markers showed strong statistical relationships with how much tumors shrank. MTHFR had a correlation of 0.764 (on a scale where 1.0 is perfect correlation), while CEA had a correlation of 0.654. Both relationships were highly statistically significant (p < 0.001), meaning there’s less than a 0.1% chance these results occurred by random chance.
When researchers combined both markers into a single predictive formula, the results were even more impressive. The formula explained 71.4% of the variation in tumor shrinkage (adjusted R² = 0.714). This means that if you know a patient’s MTHFR and CEA levels, you can predict their tumor response quite accurately. The formula was: Tumor size reduction (%) = -165.68 + (1.10 × CEA) + (15.38 × MTHFR).
The researchers also created a visual nomogram (a prediction chart) based on this formula. When they tested how well this nomogram worked, it achieved a Harrell’s C-index of 0.836, which indicates excellent discriminative ability—meaning it correctly separated patients who would respond well from those who wouldn’t respond well about 84% of the time.
The study found that MTHFR was a slightly stronger predictor than CEA alone, suggesting that the enzyme involved in nutrient metabolism may be particularly important in determining how well patients respond to this specific chemotherapy combination. The fact that both markers together worked better than either one alone demonstrates the value of using multiple biomarkers in precision medicine. The patient population was predominantly male (66.7%) with an average age of 45.6 years, which is relatively young for advanced colorectal cancer.
CEA has been used as a colorectal cancer marker for decades, but this appears to be one of the first studies examining MTHFR as a predictor of chemotherapy response in this population. The strong predictive performance (C-index of 0.836) is comparable to or better than many other biomarker prediction models reported in cancer literature. However, most previous studies have been larger and conducted in multiple centers, so direct comparisons are limited. The combination approach aligns with the current trend toward precision oncology, where multiple factors are considered rather than relying on single markers.
The most significant limitation is the small sample size of 36 patients from a single center in Indonesia. Results from small studies often don’t replicate in larger populations. The study didn’t include patients from diverse ethnic backgrounds or geographic regions, so the findings may not apply equally to all populations. The researchers didn’t examine whether other factors (like age, gender, or specific tumor characteristics) might affect the relationship between blood markers and treatment response. The study was observational, meaning researchers couldn’t control for all variables that might influence outcomes. Finally, the study didn’t follow patients long-term to see whether tumor shrinkage predicted overall survival or long-term outcomes.
The Bottom Line
Based on this preliminary research, MTHFR and CEA blood tests show promise as tools to predict chemotherapy response in advanced colorectal cancer (moderate confidence level). However, these findings should not yet be used to make treatment decisions without confirmation in larger studies. Patients with advanced colorectal cancer should discuss with their oncologist whether blood biomarker testing might be appropriate as part of their treatment planning, but standard treatment protocols should remain the primary guide. Healthcare providers in resource-limited settings may find this approach particularly valuable if larger studies confirm these results, as it uses simple, cost-effective blood tests rather than expensive genetic testing.
This research is most relevant to: (1) Patients with stage 3 or 4 colorectal cancer considering CAPEOX chemotherapy, (2) Oncologists in developing countries seeking cost-effective ways to personalize treatment, (3) Researchers studying cancer biomarkers and precision medicine. This research should NOT yet be used to make treatment decisions for individual patients, as it requires validation in larger, multi-center studies. Patients should continue following their oncologist’s standard treatment recommendations.
If these findings are confirmed in larger studies, blood biomarker testing could potentially be incorporated into clinical practice within 2-5 years. However, tumor response to chemotherapy is typically assessed 6-12 weeks after starting treatment through imaging scans, so the blood test would provide early prediction of this outcome rather than speeding up the overall treatment timeline.
Frequently Asked Questions
Can blood tests predict if chemotherapy will work for colorectal cancer?
A 2026 study of 36 patients found that MTHFR and CEA blood markers together predicted tumor shrinkage with 71% accuracy. While promising, these findings need confirmation in larger studies before doctors use them routinely to guide treatment decisions.
What do MTHFR and CEA blood tests measure in cancer patients?
MTHFR is an enzyme involved in nutrient metabolism, while CEA is a protein that cancer cells often produce in higher amounts. Both appear in the bloodstream and may indicate how well a patient’s body will respond to CAPEOX chemotherapy.
Is this blood test available for colorectal cancer patients now?
CEA testing is already standard in colorectal cancer care, but using MTHFR as a predictive biomarker is still experimental. Patients should discuss with their oncologist whether biomarker testing might be appropriate, as this research is preliminary and limited to one location.
How accurate is the prediction formula for chemotherapy response?
The predictive model achieved a C-index of 0.836, meaning it correctly predicted tumor response about 84% of the time in this 36-patient study. However, larger studies are needed to confirm whether this accuracy holds in different patient populations.
Who should consider getting MTHFR and CEA blood tests?
Patients with stage 3 or 4 colorectal cancer considering CAPEOX chemotherapy might benefit from discussing biomarker testing with their oncologist. However, standard treatment protocols should remain the primary guide until larger studies confirm these preliminary findings.
Want to Apply This Research?
- Users could track their MTHFR and CEA blood test results before and after chemotherapy cycles, recording the specific values and dates. The app could calculate the predicted tumor response percentage using the study’s formula and compare it to actual imaging results over time.
- Patients could use the app to: (1) Record blood test dates and results to share with their oncology team, (2) Set reminders for scheduled blood draws and imaging scans, (3) Document chemotherapy side effects and correlate them with biomarker levels, (4) Track questions to ask their doctor about whether biomarker-guided treatment planning is appropriate for their situation.
- Long-term tracking could include: (1) Quarterly blood test results for MTHFR and CEA, (2) Imaging scan results and tumor measurements, (3) Treatment response notes from oncology visits, (4) Overall health outcomes and survival milestones. This data could help patients and doctors identify patterns and adjust treatment strategies if needed.
This research is preliminary and based on a small study of 36 patients from a single location. These findings should not be used to make individual treatment decisions without consultation with a qualified oncologist. MTHFR and CEA blood testing for chemotherapy prediction is not yet standard clinical practice and requires validation in larger, multi-center studies. All cancer treatment decisions should be made in consultation with your healthcare team based on established clinical guidelines and your individual medical situation. This article is for educational purposes only and does not constitute medical advice.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.