Comprehensive Clinical Analysis: CEA Kinetics, ctDNA, and Treatment Strategy

Prepared for Discussion with Dr. Benjamin Tan — February 2026

Patient: Alexander R. Towell, age 50, male Primary Oncologist: Benjamin R. Tan, MD — Washington University / Siteman Cancer Center Hepatobiliary Surgeon: William Chapman, MD — Washington University

Clinical data extracted from three EHR systems exported 01/30/2026: (1) Siteman/BJC — 76 Epic CDA XML documents (DOC0001–DOC0076), (2) Anderson Hospital — 80 MEDITECH CCDA XML files + FHIR Bundle (691 resources), (3) SIHF Healthcare (athenahealth) — 16 PCP encounters (Sep 2021–Jan 2026), 260 lab results, 125 vitals, 29 medications, 12 conditions, PHQ-9 screenings, immunizations, and behavioral health notes. Guardant 360 MRD and Guardant Tissue results received via MyChart message (both negative). All claims are cross-referenced against the actual medical record unless otherwise flagged.

1. CEA Kinetics Analysis

The 17-Point Time Series

Cross-Laboratory Assay Note

Anderson Hospital uses the Ortho Clinical Diagnostics immunoassay (ref range 0.0–3.0 ng/mL). Siteman/BJC uses a different assay (ref range 0.0–5.0 ng/mL). The Anderson FHIR data explicitly notes: “CEA results determined by assays using different manufacturers or methods may not be comparable.”

This means the Anderson values and Siteman values are not directly interchangeable. However, the relative trajectory within each laboratory’s series is informative. The CEA 5.8 on 06/30/2025 followed by 4.7 at Siteman on 07/08/2025 (8 days later) could reflect: (a) a genuine 1.1-point decline, (b) assay-to-assay variability, or (c) a combination. What is clear is that the true post-resection CEA peak was higher than Siteman data alone suggested — the patient’s CEA exceeded the upper normal limit at both laboratories in late June/early July 2025.

Trend Analysis by Phase

Phase 0 — Baseline through adjuvant (Nov 2021–Aug 2022): CEA 1.4 at diagnosis (pT4aN1a sigmoid cancer), rising to 2.9 twelve days after completing 12 cycles of adjuvant FOLFOX. The post-adjuvant value of 2.9 at Anderson (within their 0.0–3.0 normal range) suggests no residual macroscopic disease at that point — the slight elevation may reflect post-chemotherapy hepatic stress rather than tumor activity.

Phase 0.5 — Surveillance gap (2023–2024): CEA 1.9 on 04/26/2024 is particularly informative — this was drawn 2 weeks before the terminal ileum mass was discovered (05/2024). A CEA of 1.9 in the face of a clinically apparent ileal tumor confirms this tumor’s low CEA secretion phenotype. Then 3.0 on 07/01/2024 (day of hemicolectomy admission) — at the exact upper limit of Anderson’s normal range. The PET at this time was already showing liver metastases. The tumor grew from ileal recurrence through hepatic metastasis while CEA stayed within or barely at the upper normal limit.

Phase 1 — Post-first-liver-resection peak (Jun–Sep 2025): The CEA 5.8 on 06/30/2025 changes the interpretation of this phase. The true CEA peak was not 4.7 (Siteman, 07/08) but 5.8 (Anderson, 06/30) — an above-normal value even by Siteman’s more permissive reference range. This occurred ~6 weeks after the segment 2 liver resection (05/14/2025), with the positive margin and perineural invasion. The elevated CEA likely reflects either (a) residual disease from the positive margin, (b) occult micrometastatic disease already present, or (c) both. CEA then oscillates between 3.8 and 4.7 at Siteman during cycles 1–5.

Phase 2 — Decline through active disease (Oct–Nov 2025): CEA drops from 3.3 to 2.9 to 2.6 even as the November 9 MRI reveals two new liver metastases (1.2 cm and 0.6 cm by radiology; segments 4A/8 per imaging, segments 5/8 and 4B per surgery). This is the critical discordance.

Phase 3 — Post-resection nadir (Jan 2026): CEA reaches 1.8, approaching the 2021 baseline of 1.4 for the first time. This is after 10 weeks without any chemotherapy (last dose was 11/06/2025, surgery 12/18/2025). The January 2026 MRI and CT chest confirm NED: “No new enhancing hepatic lesion” and “Stable pulmonary nodules. No evidence of new or increasing metastatic disease in the chest.”

The CEA/Imaging Discordance

The 17-point series strengthens the evidence that CEA cannot reliably detect disease progression in this patient. Three specific examples:

• April 2024 (CEA 1.9): Normal CEA with a clinically apparent ileal tumor already present
• July 2024 (CEA 3.0): At the upper limit of Anderson’s normal despite active hepatic metastases on PET
• October–November 2025 (CEA declining 3.3→2.6): New liver metastases forming while CEA falls

The November 2025 events show that CEA was falling while new metastases were forming. Several interpretations, not mutually exclusive:

1. CEA is a poor secretor marker in this patient’s tumor biology. His initial sigmoid cancer had a CEA of only 1.4 at diagnosis when the tumor was pT4 (through the serosa), meaning the tumor has always produced relatively little CEA per unit of disease burden. The Anderson data confirms this pattern held consistently across 4+ years. All Siteman serial values remained within normal range (<=5.0 ng/mL), though the Anderson CEA of 5.8 exceeded that lab’s reference range.

2. 5-FU maintenance suppressed CEA expression without controlling disease. Fluoropyrimidines can downregulate CEA promoter activity in some cell lines. The falling CEA may reflect a pharmacological effect on CEA gene expression rather than cytoreduction. This is supported by the rapid continued decline after stopping 5-FU (2.6 to 1.8 without any chemotherapy, suggesting the post-resection tumor burden reduction, not the drug, was driving the decline).

3. The new liver lesions were metabolically quiescent. The PET/FDG on 11/19/2025 showed only “subtle uptake” in segment 4A — the lesions were present but not aggressively metabolic. Low metabolic activity correlates with low CEA secretion.

Clinical implication: CEA is useful for trend monitoring in this patient but cannot be trusted as an all-clear signal. An MRI-based surveillance strategy (every 3 months) is likely more reliable than CEA-driven decision-making. The current value of 1.8 is reassuring in context but does not exclude microscopic residual disease — which is exactly why the Guardant MRD assay matters.

Key takeaway: The 17-point CEA series, spanning two independent laboratory systems over 4 years, conclusively establishes this patient as a low-moderate CEA secretor. CEA trends may reflect disease trajectory over months but cannot detect discrete metastatic events — as proven by the April 2024 normal CEA with active ileal tumor, the July 2024 borderline CEA with hepatic metastases, and the November 2025 falling CEA with new liver lesions. This makes the Guardant ctDNA assay especially important as a complementary molecular surveillance tool.

2. Molecular Profile

Confirmed Profile (multiple specimens, consistent across sigmoid, ileum, and liver)

Clinical Significance

This is the best molecular profile you can have for mCRC systemic therapy options. Specifically:

RAS wild-type + left-sided primary (sigmoid colon): This combination represents the strongest predictor of anti-EGFR response in metastatic colorectal cancer. The PARADIGM trial (2022, Yoshino et al.) confirmed the superiority of panitumumab + mFOLFOX6 over bevacizumab + mFOLFOX6 as first-line therapy for left-sided RAS WT mCRC (OS 37.9 vs 34.3 months, HR 0.82).

BRAF wild-type: BRAF V600E mutation is present in ~8–12% of mCRC and carries the worst prognosis (median OS ~12 months with standard therapy). Its absence eliminates a major adverse prognostic factor.

MSS with low TMB: Checkpoint immunotherapy is not a standard option. However, combinatorial approaches are emerging (anti-EGFR + immunotherapy, bispecific T-cell engagers, adoptive cell therapies). This is a “not now, but watch the landscape” situation.

DPYD normal: No increased toxicity risk from fluoropyrimidine-based regimens. The dose-limiting concern for FOLFOX re-challenge is oxaliplatin neuropathy, not 5-FU metabolism.

CDKN2A/CDKN2B deletions: Not directly actionable in CRC with standard therapies. CDK4/6 inhibitors (palbociclib, ribociclib) target this pathway — worth filing away as a future-line consideration.

Key takeaway: The distinction between left-sided and right-sided primary matters enormously for anti-EGFR therapy. This patient’s sigmoid primary is squarely left-sided, maximizing the expected benefit of panitumumab. The PARADIGM trial data applies directly to his clinical situation.

3. December 2025 Surgery — Pathology and Operative Analysis

Operative Details (12/18/2025)

Key Operative Findings (Dr. Chapman’s words): “The patient had a normal-appearing background liver. Two surface-based tumors were easily identified and as anticipated with 1 in segment 4B and a second tumor site in segments 5 and 8. Nonanatomic excisional resection was performed of both tumor sites. There was no evidence of bile leakage and no significant bleeding at the time of the resection.”

Clinical significance:

• “Normal-appearing background liver” — no sinusoidal obstruction, steatohepatitis, or cirrhosis despite prior oxaliplatin and chemotherapy. This preserves future hepatic resection or ablation as an option.
• “Surface-based tumors” — consistent with the pathology showing capsular/subcapsular locations, supporting the possibility of local (peritoneal) seeding rather than hematogenous spread.
• “Easily identified and as anticipated” — tumors matched pre-operative imaging, confirming these were true recurrences, not false-positive imaging findings.
• No bile leak and no blood products required — technically straightforward resection.

Segment Nomenclature Clarification

The MRI radiologist described the lesions as “segment 4A/8,” while the surgeon and pathologist labeled them as “segments 5/8” and “segment 4B”. This is a common discrepancy — radiologists describe lesion location by imaging anatomy, while surgeons name the parenchyma they actually resect. The full procedure encompassed segments 4B, 5, and 8, reflecting the nonanatomic wedge excision approach.

Pathology — Full Report (12/24/2025)

Pathologist: Changqing Ma, MD, PhD (Barnes-Jewish Hospital / Washington University)

Specimen A — Segment 5/8:

• Diagnosis: Metastatic moderately-differentiated adenocarcinoma with necrosis, morphologically compatible with colorectal primary. Margin of resection: negative for carcinoma.
• Gross: Liver excision 2.3 x 2.3 cm, 1.4 cm deep. White, circumscribed, firm mass 0.9 x 0.5 x 0.7 cm. Maximum depth of invasion 0.5 cm. Mass abuts the nearest margin. Capsular surface puckering 0.2 cm from nearest margin.
• Key observation: Although the margin is negative, the mass “abuts” it — this is the narrowest possible negative margin, effectively a zero-margin clearance.

Specimen B — Segment 4B:

• Diagnosis: Metastatic moderately-differentiated adenocarcinoma with necrosis, morphologically compatible with colorectal primary. Margin of resection: negative for carcinoma.
• Gross: Liver excision 2.9 x 2.4 cm, 1.5 cm deep. White, circumscribed, firm mass 1.5 x 0.7 x 1.5 cm. Maximum invasion 0.9 cm. Mass is 0.4 cm from the nearest margin. Capsular surface with ulcerated lesion 1.6 x 1.4 cm, 0.3 cm from nearest margin.

Analysis

1. Negative margins — but with important nuance:

• Specimen A “abuts” the margin — technically R0 but the narrowest possible clearance. In the context of the May 2025 resection that had a positive margin (segment 2), getting negative margins this time is a clear improvement. However, the “abutting” pattern on specimen A warrants discussion about whether adjuvant local control (e.g., ablation of the resection bed) should be considered if there is future recurrence in this area.
• Specimen B has 0.4 cm clearance — a more comfortable margin.

2. Necrosis in both specimens — treatment effect:

• Both specimens show adenocarcinoma with necrosis. In the context of prior 5-FU therapy (last dose 11/06/2025, 6 weeks before surgery), necrosis likely represents treatment effect — the 5-FU was partially killing tumor cells even while failing to prevent their appearance. This is consistent with the “CEA falling while disease progressed” pattern: the chemotherapy was exerting an anti-tumor effect (necrosis, CEA suppression) without achieving disease control.

3. Size discrepancy — MRI vs. pathology:

• The November 2025 MRI measured the lesions at 1.2 cm and 0.6 cm. The pathology found them at 0.9 cm (specimen A) and 1.5 cm (specimen B). The larger specimen (B) was significantly bigger on pathology than on MRI (1.5 vs 0.6 cm), suggesting the MRI underestimated the segment 4B lesion. This is relevant for surveillance imaging interpretation.

4. Comparison with May 2025 surgery:

Post-Operative Lab Trends

Key findings:

• The ALT/AST spike to ~2x ULN is a normal hepatic response to parenchymal resection. The rapid normalization by 01/15/2026 (ALT 19, AST 22) confirms excellent hepatic reserve.
• The albumin recovery from 3.6 to 4.3 demonstrates good nutritional status and synthetic function.
• Hemoglobin was essentially unchanged (15.3 → 14.3 → 15.8), confirming “minimal EBL” from the operative note.

Chemotherapy-Related Lab Trends (Jul–Nov 2025)

Notable findings:

• Cycle 1 myelosuppression: WBC nadir of 3.31 (L) with ANC 1.11 (L) on 07/31/2025 — grade 1 neutropenia at the first 5-FU cycle. Recovered by cycle 2.
• Intermittent ALT elevations: Spikes to 67 (08/28) and 62 (09/25) during 5-FU therapy suggest possible hepatotoxicity. These resolved between cycles, indicating the liver tolerated the treatment but was under stress.
• Creatinine peak of 1.31 (H) on 07/08/2025: This was the nadir of renal function (eGFR 67), occurring shortly after the first liver resection. It improved steadily, suggesting a transient post-surgical effect rather than intrinsic renal disease.
• Platelet trend: Gradually declining during chemotherapy (200 → 131), with partial recovery post-cessation (170 on 01/15). This is consistent with cumulative 5-FU myelosuppressive effect.

Post-Operative Course and Discharge

• Hospital stay: One night only (admitted 12/18, discharged 12/19)
• Discharge condition: Good. Vitals: Pulse 50, BP 108/59, afebrile
• Post-op follow-up (12/29/2025, Sarah Matson, PA): “He denies complaints.” Incisions healing well. Weight 107.7 kg. No abdominal pain, fever, or jaundice. Surgical pathology reviewed — both specimens margin-negative.
• ECOG Performance Status: 0 throughout all encounters from 07/2025 through 01/2026 — fully active

4. Guardant ctDNA Results — Both Negative

What Was Ordered (Jan 15, 2026)

Dr. Tan ordered both Guardant Tissue (tumor-informed) and Guardant 360 (liquid biopsy). Per the nursing note, this was a switch from the originally planned Tempus assay — Dr. Tan changed to Guardant during the visit.

• Guardant Tissue: Created a personalized genomic fingerprint from the December 2025 liver resection specimens. This establishes the specific mutations/structural variants unique to this patient’s tumor.
• Guardant 360 MRD: Searched for those specific tumor-derived mutations in circulating plasma DNA. Result: Negative for signature alterations.

What Double-Negative Means

The concordance of both assays being negative strengthens the result:

Favorable interpretation:

• No detectable circulating tumor DNA from a tumor-informed assay — the most sensitive MRD detection approach available
• In the post-metastasectomy CRC setting, ctDNA-negative patients have significantly better outcomes:
  • 3-year DFS: ~70-80% (ctDNA-negative) vs ~20-30% (ctDNA-positive)
  • ctDNA-negative status after metastasectomy is one of the strongest prognostic biomarkers available
• Combined with NED on imaging, normal CEA, and excellent performance status — this is the most favorable position this patient has been in since the metastatic diagnosis

Cautionary interpretation:

• The false-negative rate is ~5-15%, and may be higher in this specific clinical context:
  • This patient’s tumor has demonstrated low biomarker shedding throughout: CEA of only 1.4 at diagnosis of a pT4 tumor, CEA 1.9 with an active ileal mass, CEA declining while new liver metastases formed
  • The PET/FDG in November 2025 showed only “subtle uptake” — metabolically quiescent tumors may also shed less ctDNA
  • The blood draw timing (~4 weeks post-resection) may reflect successful surgical clearance rather than absence of all micrometastatic disease
• Two prior recurrences — the statistical base rate for this patient having truly no residual disease is lower than for a first-time post-resection scenario
• Disease escaped maintenance therapy — the November 2025 recurrence occurred while on active 5-FU treatment, meaning the tumor can progress through fluoropyrimidine monotherapy

The Clinical History vs. The Molecular Result

This is the central tension for the treatment decision:

5. Treatment Decision Framework

Dr. Tan’s Three Options (From Jan 15 Visit)

At the January 15 encounter, Dr. Tan explicitly outlined three treatment paths:

1. Observation with frequent CEA and scans
2. “Adjuvant” chemotherapy (5-FU-based)
3. FOLFOX + panitumumab — given wild-type status and possible lung metastasis

The patient elected to await Guardant results before deciding. The results are now available and the decision must be made.

The Updated Matrix With Confirmed Negative ctDNA

Comorbidity considerations from PCP records: The decision framework should also weigh the patient’s non-oncologic medical context. His renal function has recovered from CKD Stage 3a (eGFR 56 in Oct 2021) to normal (eGFR 88 in May 2025), but oxaliplatin is renally cleared and dose adjustment may be warranted if creatinine rises. His thyroid management requires stable levothyroxine dosing — major metabolic shifts from aggressive chemotherapy could destabilize TSH control. Neither of these contraindicate escalation, but both require monitoring.

The Patient’s Position

The patient’s expressed preference, communicated to Dr. Tan via MyChart:

“After two recurrences, including one that came through on active maintenance, my priority is more time. Not just because the treatment landscape keeps improving, but because every year matters to me regardless of the long-term picture. I’m willing to deal with side effects if it means more time.”

The patient does not want observation alone. The patient is willing to accept side effects for therapeutic benefit. The patient defers to Dr. Tan on the specific regimen (FOLFOX + panitumumab vs. 5-FU/LV + panitumumab) but wants some form of active treatment.

The Argument for the Middle Path (Option 2)

A strong case can be made for 5-FU/LV + panitumumab (without oxaliplatin) as the optimal starting strategy:

1. Anti-EGFR is the untapped resource: Panitumumab has never been used in this patient. For left-sided RAS WT mCRC, anti-EGFR therapy has the strongest evidence base (PARADIGM trial: OS 37.9 vs 34.3 months favoring panitumumab over bevacizumab).

2. Preserves oxaliplatin for escalation: If ctDNA turns positive during 5-FU/LV + panitumumab, oxaliplatin can be added immediately. This strategy maximizes optionality.

3. Avoids permanent neuropathy in a ctDNA-negative patient: The neuropathy cost is real and irreversible. Paying that cost when there is no detectable molecular disease to target is a harder trade-off to justify than when ctDNA is positive.

4. 5-FU resistance is uncertain: While the disease recurred on 5-FU monotherapy, the December pathology showed necrosis in both specimens — the 5-FU was having some effect. Adding panitumumab (a completely new mechanism) may overcome whatever resistance existed.

5. The skin rash biomarker: Panitumumab rash severity correlates with response and survival. Starting panitumumab first allows assessment of this biomarker before committing to the full neuropathy burden of oxaliplatin.

The Argument for Full Escalation (Option 3)

1. This is the minimal disease window. If there are any residual cells, the best time to hit them is now — before they establish, proliferate, and develop additional resistance.

2. The disease’s track record demands respect. Two recurrences, shortening disease-free intervals (18 months → 6 months), and progression through maintenance therapy. Observing or half-measures may not match the tumor’s biology.

3. Published data supports combination over monotherapy. FOLFOX + panitumumab ORR ~80% in first-line (PARADIGM). While this is not first-line, the combination provides the most potent regimen available.

4. Stop-and-go mitigates the neuropathy concern. OPTIMOX1 strategy: 6 cycles FOLFOX + panitumumab, then 5-FU/LV + panitumumab maintenance, with oxaliplatin re-introduction only if needed. Starting dose at 75-80% further limits cumulative exposure.

Panitumumab-Specific Considerations

• Skin toxicity as a biomarker: Acneiform rash occurs in ~90% of patients and its severity paradoxically correlates with response and survival. Grade 2+ rash is associated with significantly longer OS (STEPP study, ASPECCT trial). If treatment proceeds, the development of rash should be framed as a positive signal, not just a side effect to manage.
• Hypomagnesemia: Panitumumab causes renal magnesium wasting in ~30–40% of patients. Current magnesium is 1.8 mg/dL (normal 1.4–2.5) — baseline is adequate but should be monitored closely.
• No bevacizumab overlap: Panitumumab and bevacizumab should not be combined (PACCE trial showed worse outcomes). Since the patient was previously on 5-FU/bevacizumab maintenance, bevacizumab would be stopped.

Stop-and-Go Strategy Details (If FOLFOX Chosen)

OPTIMOX1 model:

1. Start FOLFOX + panitumumab for 6 cycles (3 months)
2. Transition to 5-FU/LV + panitumumab maintenance
3. Re-introduce oxaliplatin if progression or rising ctDNA
4. This limits cumulative oxaliplatin exposure per cycle block

Practical dose modifications:

• Starting at 75-85% oxaliplatin dose (65-72 mg/m² instead of 85 mg/m²)
• Lengthening infusion time (6 hours instead of 2) — reduces acute cold-triggered neuropathy
• Pre-medication with calcium/magnesium infusions
• Close neurological monitoring with dose reduction or discontinuation at Grade 2 neuropathy

De-escalation landmarks:

1. After 4–6 cycles: If ctDNA remains negative and imaging clean, stop oxaliplatin → 5-FU/LV + panitumumab maintenance
2. After 8–12 cycles total (including maintenance): If ctDNA remains negative x 2 serial draws, consider stopping all therapy → active surveillance
3. The “buying time” calculus: Every month of disease control is a month closer to potential approvals of botensilimab/balstilimab (anti-CTLA-4 + anti-PD-1 for MSS mCRC, phase III in progress), CEA-targeted bispecifics (cibisatamab), and ctDNA-guided adaptive therapy (DYNAMIC-III, CIRCULATE trials)

6. Treatment Sequencing — Lines Used and Preserved

PARADIGM Trial Evidence Base

PARADIGM Trial (Yoshino et al., Lancet Oncol 2023):

• Phase III, first-line mCRC, RAS WT only
• Panitumumab + mFOLFOX6 vs bevacizumab + mFOLFOX6
• Left-sided subgroup: OS 37.9 vs 34.3 months (HR 0.82, p=0.03)
• ORR: 80.2% vs 68.6% (significant)

Caveats: PARADIGM studied first-line treatment. This patient has received prior FOLFOX (adjuvant), prior FOLFIRI/bev (metastatic), and prior 5-FU maintenance. Expected response rate to FOLFOX re-challenge + panitumumab in the multi-line setting is lower but still meaningful (~40–60% ORR based on retrospective data).

7. Recurrence Pattern and Surgical History

Timeline of Disease Events

The Anemia That Found the Cancer

The SIHF PCP records reveal the complete diagnostic cascade that led to the cancer discovery — a narrative not captured in either the Anderson Hospital or Siteman systems:

1. Sep 27, 2021: Patient establishes PCP at SIHF with Dr. Suthan. Presenting complaints: lightheadedness, two chest lumps (lipomas), sleep apnea. No mention of GI symptoms. PHQ-9 score: 5 (mild).
2. Oct 2, 2021: Comprehensive labs drawn. Hgb 7.1 g/dL (ref 13.0–17.7), ferritin <1 ng/mL (ref 30–400), iron 13 ug/dL (ref 38–169), iron saturation 3% (ref 15–55), TIBC 498 (ref 250–450). This is profound iron-deficiency anemia — ferritin <1 is effectively undetectable, indicating complete depletion of iron stores.
3. Oct 11, 2021: Fecal immunochemical test (FIT): positive for occult blood. This confirms the clinical suspicion of GI blood loss.
4. Oct 12, 2021: PCP follow-up: previous hemoglobin 7.1, blood in stool, low iron. Dr. Suthan initiates GI referral.
5. Nov 12, 2021: PCP visit: “pt was seen by GI yesterday and awaiting colonoscopy/egd”
6. Nov 22, 2021: EGD + colonoscopy at Anderson Hospital → sigmoid mass discovered, biopsied → adenocarcinoma.

Clinical significance: The PCP’s systematic anemia workup — labs, fecal occult blood testing, and timely GI referral — was the initiating event of the entire cancer narrative. The ferritin <1 finding is particularly striking: it indicates complete iron store depletion, consistent with months to years of chronic occult blood loss from the 6 cm pT4 sigmoid tumor. By the time Anderson Hospital saw the patient for the EGD/colonoscopy (11/22/2021), hemoglobin had risen slightly to 9.4 (from 7.1 on 10/2 to 8.0 on 11/6 to 9.4 on 11/23), likely due to iron supplementation started by the PCP (ferrous sulfate 325 mg daily, documented in SIHF medication list).

Anderson Primary Source Details — Initial Diagnosis and Surgery

Surgical Team (Anderson Hospital, Maryville IL):

• Surgeon: Pei Chang Chung, MD — performed both the initial laparoscopic sigmoid colectomy (12/29/2021) and the port-A-cath insertion (02/08/2022)
• GI Specialist: Edmundo A. Rodriguez-Frias, MD — performed the initial EGD/colonoscopy (11/22/2021) that discovered the sigmoid mass

Procedures Confirmed from Anderson Primary Records:

Pre-operative Baseline Labs (11/23/2021 — Anderson Hospital Lab):

Clinical significance of the Hgb 9.4: The marked anemia at initial presentation is typical for large colorectal tumors causing chronic occult blood loss. The SIHF labs from October 2, 2021 — 3 weeks before the Anderson pre-op labs — show Hgb was even lower at 7.1 g/dL (vs. 9.4 on 11/23), placing the true anemia nadir earlier than previously known. The PCP had already started iron replacement (ferrous sulfate 325 mg daily), which accounts for the improvement from 7.1 → 8.0 (11/6 at SIHF) → 9.4 (11/23 at Anderson). By January 2022 (post-colectomy), SIHF labs show Hgb had recovered to 12.4, and by January 2023 it reached 15.8 — full recovery.

Oligometastatic Phenotype Analysis

Oligometastatic, liver-dominant phenotype:

• Initial liver mets: 4 lesions, segments II, VII, VIII — complete response on FOLFIRI/bev
• First recurrent liver met: segment 2 (single lesion)
• Second recurrence: 2 lesions in segments 4A/8

This is prognostically favorable. Published 5-year survival rates for resected oligometastatic CRC to liver range from 25–58%.

Surgical resectability as a strategic asset: Two liver resections completed with hepatic reserve preserved. Dr. Chapman’s intraoperative finding of “normal-appearing background liver” confirms no chemotherapy-induced hepatotoxicity. Future hepatic resection or ablation remains technically feasible.

The first resection’s positive margin (05/2025) vs. second resection’s negative margins (12/2025): The improvement in margin status is significant. The use of intraoperative ultrasound and robotic assistance in the second surgery may have contributed to better tumor visualization and resection planning.

The Anderson CEA trajectory reinforces the oligometastatic narrative: CEA 1.9 in April 2024 (with an ileal mass already present) and 3.0 in July 2024 (with hepatic metastases on PET) demonstrate that even with multi-site disease, the total disease burden produced only borderline-elevated CEA. Combined with the consistently liver-dominant recurrence pattern, this supports a tumor biology that is amenable to local-regional control strategies.

8. Current Clinical Status (January 15, 2026)

Laboratory Summary

All labs entirely within normal limits. Complete hematologic and hepatic recovery from both the December surgery and the prior 5-FU chemotherapy course.

Imaging Summary (January 15, 2026)

Performance Status

ECOG 0 throughout all encounters from 07/2025 through 01/2026. Weight recovered to 107 kg (near pre-diagnosis baseline of 109 kg). The patient is pursuing a PhD in computer science and maintains a fully active lifestyle.

9. The Lung Nodules

• RLL solid nodule 9 mm — stable >15 months (Oct 2024 → Jan 2026)
• LLL groundglass nodule — stable, partially obscured by motion artifact on Jan CT
• No FDG uptake on any PET
• Patient resides in the histoplasmosis endemic area (Wood River, IL)
• Tumor board previously recommended diagnostic wedge resection referral to Dr. Chapman

Imaging History

The RLL nodule is now reported at 9 mm (vs. 8 mm on 11/06/2025). This 1 mm change is within measurement variability and does not represent true growth — the radiologist read it as “unchanged” and “stable.”

Current assessment: Benign granuloma remains the most likely diagnosis given >15 months of stability without metabolic activity. If systemic therapy is initiated, the lung nodules provide a secondary endpoint — stability on treatment would further support the benign interpretation; growth would argue for tissue diagnosis.

Note on the LLL groundglass nodule: GGNs in the lung are more commonly associated with pre-invasive lung adenocarcinoma than with CRC metastasis. CRC mets to the lung are almost always solid.

10. PCP Perspective: Whole-Patient Context

Comorbidity Management During Active Cancer Treatment

The PCP records reveal a parallel track of non-oncologic medical management that runs alongside the cancer story:

Hypothyroidism (post-thyroidectomy):

• Partial thyroidectomy performed ~September 2022 (documented at 01/18/2023 PCP visit: “partial thyroidectomy 9/2022”)
• This was for a thyroid nodule first identified in the PCP records (02/2022), with persistently elevated TSH (7.17 in Oct 2021, 6.62 in Nov 2021) and positive TPO antibodies (91 IU/mL, ref 0–34), indicating autoimmune thyroiditis (Hashimoto’s)
• Levothyroxine dose escalation: 50 mcg → 75 mcg → 88 mcg → 100 mcg → 112 mcg → 125 mcg (current)
• TSH has been persistently above reference range at most measurements

Renal function:

• Entered PCP care with CKD Stage 3a (eGFR 56, creatinine 1.48 in Oct 2021)
• Remained at eGFR 56 through January 2023
• Recovered to eGFR 74 by December 2023, then eGFR 88 by May 2025

Thrombocytopenia:

• Listed as active condition since 01/20/2023 (platelets 145, borderline low)
• Likely chemotherapy-related (post-FOLFOX adjuvant)
• PCP tracked this alongside oncology monitoring

Hyperlipidemia:

• Total cholesterol trending: 149 (10/2021) → 220 (01/2023) → 205 (12/2023) → 198 (05/2025)
• LDL elevated: 147 (01/2023), 146 (12/2023), 118 (05/2025)
• No statin documented — reasonable given competing priorities of cancer treatment

Provider Transitions

The transition from Dr. Suthan to Dr. Steele occurred between May 2024 and February 2025 — during the most active phase of the patient’s metastatic disease treatment.

11. Mental Health Trajectory

PHQ-9 Trend: 8 Scores Over 4.3 Years

Interpretation:

The PHQ-9 scores never breach the moderate threshold (10+), indicating the patient has maintained functional resilience throughout his cancer journey. However, the pattern reveals meaningful clinical correlations:

• Peak scores of 8 align with the two worst cancer milestones: the ileal recurrence discovery (05/2024) and the period of active treatment with ongoing disease (10/2025). The October 2025 PHQ-9 was administered just 2 weeks before the November MRI revealed new liver metastases.
• The 11/25/2025 score of 5 — taken after learning about new liver metastases and facing a second liver surgery — is notably lower than the pre-discovery score of 8. This suggests the patient copes better with known challenges than with uncertainty.
• The 01/26/2026 score of 0 is remarkable: the patient reports no depressive symptoms one month after major surgery, with the reassurance of negative margins, NED imaging, and negative ctDNA.
• No score ever reaches 10 (moderate depression threshold), and no score includes endorsement of suicidal ideation — with one notable exception: the 10/27/2025 screening included “Several days” for “Thoughts that you would be better off dead or of hurting yourself,” the only non-zero response to this item across all 8 screenings. This coincided with the behavioral health intake where existential anxiety about mortality was explicitly discussed.

Behavioral Health Integration

The patient was referred to SIHF behavioral health in October 2025, establishing care with Dr. Christopher Loynd, DO. Key excerpts from the clinical documentation:

From the 10/27/2025 intake assessment (Dr. Loynd):

“1.5 years ago was diagnosed with stage 4 cancer and given 1 year to live. Was rx. Lorazepam and Ambien by oncology. States medications work well to ease anxiety and sleep disturbance.”

“Anxiety — Onset: has been dealing with cancer for 5 years. Was in remission for 1 year, then had a reoccurrence at stage 4. He was given a year to live at the time. Characteristics: has anxiety going in to chemo. Effects him mentally. Questions about mortality and existential thoughts, he obsesses over this. He is still doing research for his PhD.”

“Triggers: physical pains (mouth sores, diarrhea, inflammation), approaching chemo. Illness, thinking may be worse than what he thought. Questioning mortality.”

Risk/Safety (consistent across all 3 BH visits):

• Denies suicidal/homicidal ideation (SI/HI/SIB)
• “Endorses having many things to accomplish” — a protective factor
• “Endorses history of alcohol use in remission since 2015”

Medication management (BH):

• Lorazepam 0.5 mg PO daily PRN for breakthrough anxiety — continued
• Zolpidem 5 mg PO daily PRN for sleep — continued

Clinical significance for the oncology team: The behavioral health notes reveal a patient who:

1. Is psychologically processing existential anxiety about a terminal diagnosis while maintaining functional performance (ECOG 0, pursuing PhD in computer science)
2. Has substance use history (alcohol, in remission since 2015) that warrants awareness during pain management and benzodiazepine prescribing
3. Uses physical symptoms as anxiety triggers — suggesting that treatment side effects carry a psychological burden beyond their physical impact
4. Has a strong cognitive coping framework (“many things to accomplish”) that serves as a protective factor

12. Physiologic Trends

Weight and BMI Trajectory

Pattern: The weight trajectory tells a story of treatment impact and recovery:

• Chemotherapy-induced weight loss: −13 kg (109 → 96 kg) over ~18 months spanning FOLFOX adjuvant and early surveillance. The nadir coincided with the ileal recurrence discovery.
• Plateau during active metastatic treatment: Weight stabilized at ~96 kg during FOLFIRI/bev (Feb 2025) through first liver resection (Apr 2025).
• Recovery arc: From the post-resection period (Jun 2025) onward, the patient regained ~12 kg, returning to near-baseline weight of 107 kg by November 2025.
• Post-op weight (Siteman, 12/29/2025): 107.7 kg — consistent with the PCP trend.

Blood Pressure Trajectory

Notable findings:

• The 10/12/2021 tachycardia (HR 105) occurred at the nadir of anemia (Hgb 7.1). Classic compensatory response to severe anemia.
• Heart rate trend shows propranolol effect: HR dropped from 91 at baseline to the 50–65 range during 2023–2025, consistent with propranolol 20 mg BID. The bradycardia to 50 on 11/30/2023 is clinically significant, though the patient was asymptomatic.
• Overall BP pattern: Predominantly in the Stage 1 hypertension range (120–139/80–89) without antihypertensive treatment other than propranolol (prescribed for anxiety, not hypertension).

Renal Function Recovery

Interpretation: The patient had CKD Stage 3a at baseline (eGFR 56) when he first presented to the PCP. This persisted unchanged through 14 months of FOLFOX adjuvant chemotherapy and surveillance. The recovery began during the treatment-free interval (2023–2024), reaching normal by May 2025 (eGFR 88). The post-surgical dips in renal function are transient and expected. The pre-existing CKD was likely related to the severe iron-deficiency anemia and possibly chronic dehydration. The full recovery to eGFR 88 is reassuring for oxaliplatin re-challenge, but the historical CKD 3a should prompt more frequent creatinine monitoring if FOLFOX is initiated.

Thyroid Management

Clinical significance for oncology:

• Hypothyroidism can contribute to fatigue, constipation, and weight gain — symptoms that overlap with chemotherapy side effects
• Levothyroxine absorption can be affected by GI changes from chemotherapy (diarrhea, mucositis)
• Checkpoint immunotherapy (if ever considered for future trials) can cause thyroiditis — the pre-existing autoimmune thyroid disease would increase this risk

13. Medication Reconciliation

Current Active Medications (Cross-System Inventory)

Oncology Medications (Siteman/Anderson):

PCP-Managed Medications (SIHF):

Behavioral Health Medications (SIHF BH / originally oncology):

Supplements (Anderson):

Discontinued/Completed Medications of Note:

Drug Interaction Considerations

1. Lorazepam + 5-FU: No direct pharmacokinetic interaction. However, both can cause fatigue — additive sedation is possible during infusion weeks.

2. Propranolol + oxaliplatin: Propranolol can mask tachycardia that may be a sign of oxaliplatin hypersensitivity reactions. If FOLFOX is initiated, the infusion team should be aware of baseline beta-blocker use. Propranolol also contributes to the patient’s bradycardia (HR documented as low as 50 at PCP visits) — oxaliplatin-induced bradycardia is rare but reported.

3. Levothyroxine + chemotherapy: Chemotherapy-induced mucositis, diarrhea, and altered GI motility can reduce levothyroxine absorption. TSH should be checked more frequently during active chemotherapy (every 6–8 weeks rather than the usual 6–12 months) and levothyroxine dose adjusted as needed.

4. Zolpidem + lorazepam: Both are CNS depressants. The behavioral health provider is managing both, which is appropriate. In the context of potential chemotherapy-related fatigue, the combined sedative load warrants monitoring.

5. Panitumumab + magnesium: If panitumumab is added, it causes renal magnesium wasting. Baseline magnesium (1.8 mg/dL at Siteman, 01/15/2026) is normal but should be monitored closely, with supplementation as needed.

Key takeaway: The three-system medication reconciliation reveals no absolute contraindications to the proposed FOLFOX + panitumumab regimen. The main considerations are: (a) propranolol may mask hypersensitivity tachycardia during oxaliplatin infusions, (b) levothyroxine absorption should be monitored during chemotherapy, and (c) the combined sedative load of lorazepam + zolpidem + chemotherapy fatigue should be discussed with the behavioral health team.

14. Comorbidity Summary for Treatment Planning

• Renal: CKD Stage 3a at baseline (eGFR 56, Oct 2021) → recovered to eGFR 88 by May 2025 → currently eGFR 77. Adequate for oxaliplatin dosing but warrants monitoring.
• Thyroid: Post-thyroidectomy hypothyroidism on levothyroxine 125 mcg. TSH persistently elevated. Chemotherapy may affect absorption.
• Neuropathy: Prior FOLFOX neuropathy (2022), largely resolved. Currently on gabapentin 100 mg BID + lidocaine patches. The PCP records document neuropathy-related nail changes as late as April 2025. The patient’s PCP has managed pain medications including tramadol (stopped 05/2024) and hydrocodone (stopped 02/2025), suggesting neuropathy-related pain management occurred in the primary care setting. The behavioral health notes document that “physical pains (mouth sores, diarrhea, inflammation)” trigger anxiety — the neuropathy burden is not merely physical but contributes to psychological distress.
• Behavioral health: Anxiety and insomnia managed by BH provider (Dr. Loynd). Lorazepam 0.5 mg PRN, zolpidem 5 mg PRN. PHQ-9 score 0 at most recent assessment (01/26/2026).
• Propranolol 20 mg BID — may mask infusion reaction tachycardia; baseline HR 50-65.
• Magnesium 1.8 — normal baseline, relevant if panitumumab is started (causes renal Mg wasting).

The Neuropathy Tradeoff (If Oxaliplatin Re-Challenge)

From the medical record:

• FOLFOX adjuvant (Feb–Aug 2022): 12 cycles with dose reduction, documented neuropathy. Anderson records confirm FOLFOX cycle encounters on 02/01, 02/09, 03/16, 05/16/2022; port-A-cath placed 02/08/2022 by Dr. Chung. Post-completion CEA 2.9 on 08/29/2022.
• Current status: Peripheral neuropathy on active problem list, gabapentin 100mg BID + lidocaine patches

Baseline neuropathy assessment from PCP records: The patient’s neuropathy is managed at two levels — oncology (gabapentin, lidocaine patches) and PCP (pain medications as needed). The PCP encounter on 04/01/2025 documents “nail changes” as a presenting concern, consistent with residual oxaliplatin-induced nail dystrophy.

The re-challenge calculus:

• Cumulative oxaliplatin neuropathy is dose-dependent. After 12 cycles of adjuvant FOLFOX in 2022, plus 3.5 years of recovery, some neural regeneration has occurred.
• A re-challenge with oxaliplatin would carry ~60-80% risk of worsening neuropathy.
• Stop-and-go strategies (OPTIMOX1/2 approach): 6–8 cycles of FOLFOX, then maintenance with 5-FU/LV (+/- panitumumab), with oxaliplatin re-introduction if needed. This limits cumulative exposure.
• Dose modifications (75–80% dose intensity) at initiation could reduce risk while maintaining efficacy.

15. Questions for Dr. Tan — February 5, 2026

Primary Treatment Decision

1. “Both Guardant assays came back negative. Given my recurrence history — two recurrences, the second through active maintenance, and a prior positive margin — does this change your recommendation from the three options you laid out on January 15?”

2. “If you lean toward treatment, which regimen? Would you start with FOLFOX + panitumumab (option 3), or would 5-FU/LV + panitumumab (a variation of option 2, adding anti-EGFR) make more sense as a first step?”

3. “If we start 5-FU/LV + panitumumab, what would trigger escalation to full FOLFOX? A positive ctDNA? Imaging findings? Or would you plan a time-limited course regardless?”

Oxaliplatin-Specific (If Chosen)

4. “Can we use a stop-and-go approach — 4-6 cycles with a planned oxaliplatin holiday?”

5. “Would you start at reduced dose (75-80%) given my prior neuropathy?”

6. “My neuropathy from 2022 FOLFOX has largely resolved — no functional impairment. Does that affect your assessment of re-challenge risk?”

Monitoring Strategy

7. “How often should we repeat the Guardant liquid biopsy? Every 8 weeks? 12 weeks?”

8. “Given that CEA failed to detect my November recurrence, should ctDNA and MRI be the primary surveillance tools?”

9. “What imaging schedule during treatment — MRI abdomen every 3 months? CT chest?”

Lung Nodules

10. “With 15+ months of stability and no FDG uptake, is the diagnostic wedge resection still worth considering? Or has time settled this?”

Colonoscopy Surveillance

11. “I received a call from Anderson about a colonoscopy being due. My last full colonoscopy was the diagnostic scope in November 2021, and the July 2024 hemicolectomy included intraoperative bowel evaluation. Is now the right time, or should we coordinate timing with any planned treatment?”

12. “Would you prefer I have the colonoscopy done at Siteman so your team can review the findings directly, or is Anderson fine?”

PCP-Informed Considerations

13. “My kidney function recovered from eGFR 56 to eGFR 88 over the past 3 years — does this affect your oxaliplatin dosing calculation?”

14. “I’m being managed by a behavioral health provider for anxiety and insomnia (lorazepam, zolpidem) — are there any interactions with the proposed regimen I should be aware of?”

15. “My TSH has been persistently elevated despite levothyroxine dose escalation to 125 mcg — could chemotherapy affect thyroid hormone absorption or metabolism?”

Long-Term

16. “If I achieve sustained molecular clearance through treatment, what’s your threshold for stopping therapy?”

17. “Any relevant clinical trials at Siteman for MSS mCRC? Botensilimab/balstilimab, bispecifics, adaptive ctDNA-guided approaches?”

16. Methodological Notes

What I am confident about:

• All 11 CEA values from the BJC/Siteman system are extracted directly from structured lab results in the medical record XML
• 5 additional CEA values from Anderson Hospital are extracted from primary CCDA lab tables (4 values) and FHIR JSON (1 value: the 06/30/2025 CEA 5.8)
• The CEA 1.4 (11/23/2021) is confirmed from Anderson’s primary lab data
• The initial sigmoid colectomy (12/29/2021) by Dr. Pei Chang Chung at Anderson Hospital is confirmed from Anderson’s own procedure records
• The molecular profile (RAS WT, BRAF WT, MSS, PD-L1 negative) is confirmed across multiple specimens and assays
• All pathology findings, operative details, and lab values are verbatim from the medical records
• The Guardant negative result (both assays) is confirmed via MyChart message
• All SIHF PCP lab values (PHQ-9 scores, vitals, thyroid labs, anemia labs, renal labs) are extracted from structured data in the athenahealth EHR export loaded into the clinical database (260 lab results, 125 vitals, 98 mental status records)
• The PHQ-9 total scores (5, 3, 6, 8, 6, 8, 5, 0) are verified from the mental_status table with individual item responses confirming the totals
• The anemia discovery narrative (Hgb 7.1, ferritin <1, FIT positive, GI referral) is confirmed from SIHF structured labs (10/02/2021, 10/05/2021, 10/11/2021) and clinical notes documenting the visit-by-visit progression
• Behavioral health assessment quotes are verbatim from SIHF clinical notes

Where I am applying clinical knowledge that should be verified:

• Response rates for FOLFOX + panitumumab in the re-challenge setting (drawn from published literature, not specific to this patient)
• Guardant sensitivity/specificity numbers (based on published validation studies; Guardant’s specific MRD product specifications may differ)
• The PARADIGM trial results and their applicability to this non-first-line setting
• Botensilimab/balstilimab trial status (based on conference presentations; trial landscape may have evolved)
• Drug interaction assessments (propranolol + oxaliplatin, levothyroxine + chemotherapy) are based on general pharmacologic principles and published case reports, not patient-specific pharmacokinetic data

Areas of genuine uncertainty:

• Whether the two primary tumors (sigmoid and terminal ileum) share a clonal origin or are truly independent primaries
• Whether the positive margin on the first liver resection is causally related to the subsequent segment 4A/8 metastases (segments 5/8 and 4B per surgical nomenclature are in the same hepatic lobe but not immediately adjacent to segment 2)
• The true false-negative rate of Guardant’s MRD assay in a patient with a demonstrated low-shedding tumor phenotype
• Whether this patient’s neuropathy has recovered sufficiently to tolerate even modified-dose oxaliplatin — this is a clinical assessment Dr. Tan will need to make at the bedside
• The significance of the necrosis finding — is it treatment effect from 5-FU, or intrinsic tumor biology?
• Whether the CEA 5.8 (Anderson, 06/30/2025) vs 4.7 (Siteman, 07/08/2025) difference is due to assay variability, genuine decline, or both — the different reference ranges (0.0–3.0 vs 0.0–5.0) suggest different assay calibrations
• Whether the pre-existing CKD Stage 3a (eGFR 56, Oct 2021) was entirely related to the severe anemia/dehydration or had an independent component — the full recovery to eGFR 88 suggests the former, but the persistence at eGFR 56 through January 2023 (despite anemia resolution) leaves some ambiguity
• Whether the family history (father: hypertension; mother: COPD, deceased; sister: depression) has any bearing on treatment tolerance or cancer risk
• The accuracy of the “given 1 year to live 1.5 years ago” statement in the behavioral health notes — the clinical record does not document this specific prognosis being communicated, though the stage IV diagnosis in mid-2024 is consistent with the timeline

Data source reconciliation:

• Anderson CCDA files provide cumulative lab snapshots: the same lab result appears across multiple CCDA files (4,321 raw rows deduplicated to 691 unique results)
• Anderson FHIR Bundle provides structured typed data (73 Observations, 17 Conditions, 92 DiagnosticReports, 25 MedicationRequests, 5 Encounters)
• Where Anderson and Siteman data overlap (e.g., diagnosis codes, medication lists), they are consistent
• Anderson records span 11/2021–07/2025; Siteman records span 07/2024–01/2026 — the overlap period (07/2024–07/2025) provides cross-validation
• SIHF athenahealth records provide 16 encounters (09/2021–01/2026), 260 structured lab results, 125 vital sign measurements, 29 medications, 12 conditions, 98 mental status entries (PHQ-9 screenings), 10 immunizations, 20 social history entries, 3 family history entries, and 4 clinical notes
• The SIHF data overlaps with Anderson records during 10/2021–07/2025 and with Siteman records during 02/2025–01/2026. Where overlapping data exists, the three systems are consistent
• SIHF vitals data is exclusive to the PCP system — neither Anderson nor Siteman captured routine weight, BMI, or blood pressure at the frequency provided by the PCP

Data sources: (1) Siteman/BJC EHR export (01/30/2026) — 76 Epic CDA XML documents; (2) Anderson Hospital EHR export (01/30/2026) — 80 MEDITECH CCDA XML files + FHIR Bundle (691 resources); (3) SIHF Healthcare (athenahealth) EHR export (01/30/2026) — 16 encounters, 260 lab results, 125 vitals, 29 medications, 12 conditions, 98 mental status entries; (4) Guardant 360 MRD and Guardant Tissue results (MyChart). Clinical database: chartfold SQLite (1,600+ lab results, 384 source assets, 16 clinical tables). Analysis generated February 4, 2026.

This analysis is intended to support an informed discussion with Dr. Tan. It does not constitute medical advice and should not replace the clinical judgment of the treating oncology team. All treatment decisions should be made in consultation with Dr. Tan and the multidisciplinary team at Siteman Cancer Center.