Liquid and Tissue Biopsies Should Be Viewed as Complementary Tools for Molecular Characterization in mCRC

The explosion of genomic testing options in metastatic colorectal cancer (mCRC) has produced an unprecedented array of effective tools, but that abundance heightens the risk of their misapplication, according to Marwan G. Fakih, MD. Likewise, integrating liquid biopsy with standard tissue-based next-generation sequencing (NGS) can provide a more comprehensive picture of the dominant metastatic disease and serve as a critical hedge against scenarios in which tissue results are unavailable or uninformative.

At the 11th Annual School of Gastrointestinal Oncology^®, an event held by Physicians’ Education Resource®, Fakih delivered a presentation on concordance of liquid and tissue biopsy in CRC and provided practical examples to illustrate the challenges and benefits of using both methods. His presentation aimed to explore a framework for understanding when circulating tumor DNA (ctDNA) and tissue NGS align or conflict, along with how to respond in each scenario.

"We are privileged in the United States to have a lot of tools to work with in trying to find the best treatment for our patients," Fakih said in an exclusive postmeeting interview with OncLive^®. "A lot of these are not accessible in other places, including in many countries in Europe. The challenge is not just having the tools—it is knowing how to use them and, critically, knowing when not to rely on them.

“Serial ctDNA retesting, in particular, must be driven by specific clinical questions rather than reflex practice, Fakih emphasized. He added that "making sense of both sets of data together, analyzing them in context, is what is important—not relying on one rather than the other.”

Fakih is a professor in the Department of Medical Oncology and Therapeutics Research; associate director of clinical sciences; medical director of the Briskin Center for Clinical Research; division chief of GI medical oncology; and codirector of the Gastrointestinal Cancer Program at City of Hope in Duarte, California.

Should clinicians perform both tissue NGS and liquid biopsy at diagnosis?

Fakih began by making the case for obtaining both tissue NGS and ctDNA at diagnosis, rather than treating them as interchangeable alternatives. The practical argument for this, he explained, hinges on a scenario that occurs more often than oncologists may appreciate.

"You have a patient coming in the door with metastatic disease, and you don't have the NGS on the tissue yet. You ordered the tissue NGS, you started chemotherapy, and the NGS comes back indeterminate," Fakih said. "Now you draw a liquid biopsy, and since the tumor is responding, the ctDNA is not shedding. You do your liquid biopsy, and you don't get any results. You decide to biopsy again and 4 weeks later—because the tumor is responding—you get necrotic tissue and you have no answer. You have to wait until the patient progresses to figure it out."

The solution, he argued, is to draw a liquid biopsy at the same time as the tissue specimen prior to systemic therapy. The 2 platforms offer distinct but complementary views of tumor biology: Tissue NGS captures the microenvironment of a specific biopsy site, while ctDNA predominantly reflects the dominant metastatic clone shedding the most DNA into the bloodstream.

A key driver of discordance, Fakih noted, is ctDNA tumor fraction. Findings from a retrospective analysis of 75 patients with mCRC who underwent genomic profiling with Guardant360 (G360) ctDNA and FoundationOne tumor biopsy showed a median ctDNA level of 10.45% with concordant alterations (n = 58/83) detected by G360 vs 0.3% in discordant cases (n = 28/128; P < .001), underscoring that the accuracy of liquid biopsy is dependent on tumor shedding.²

"Making sense of both sets of data together and analyzing them in context is what is important, [rather than] relying on one [or] the other," Fakih said.

Does the site of metastatic disease affect ctDNA reliability?

One of the major practical limitations Fakih raised was variable ctDNA sensitivity based on metastatic site. Drawing on data from a cohort study published in JAMA Network Open, in which 184 patients with resected stage II to IV colorectal cancer (CRC) were monitored with serial Signatera testing, 31% of patients who recurred by imaging (n = 45) had negative ctDNA at the time of imaging-confirmed relapse.³ Most patients had lung-only metastases. Peritoneal and soft tissue involvement were also associated with poor ctDNA sensitivity.

Fakih illustrated the clinical stakes via a case study of a patient with T4b colon cancer who had retroperitoneal lymph node (RPLN) soft tissue involvement.¹ The patient had a negative baseline ctDNA, then became ctDNA-positive during adjuvant therapy, and subsequently returned to ctDNA-negative status after surgery. However, PET imaging continued to show persistent activity in the RPLN region that could not be biopsied. The case exemplifies how soft tissue disease can evade liquid biopsy detection entirely.

Conversely, a second case made a different point. A patient with synchronous liver metastases treated with neoadjuvant chemotherapy, primary/liver surgery, and adjuvant FOLFIRI/hepatic arterial infusion showed ctDNA positivity that reliably tracked liver recurrences. However, a 1-cm lung metastasis that appeared after ablation of a liver lesion registered as ctDNA negative, only to have ctDNA rise again with a subsequent liver recurrence.

"Lung-only metastases are predominantly negative by ctDNA," Fakih said. "Peritoneal and soft tissue involvement are also associated with poor sensitivity. When you see a negative liquid biopsy, you need to ask: Where is this patient's disease? Because a negative result in someone with lung or peritoneal metastases may simply reflect low shedding, not absence of disease."

Could liquid biopsy alone be used to assess microsatellite instability (MSI) status or tumor mutational burden (TMB)?

The bulk of Fakih’s presentation concerned the use of ctDNA to assess MSI and TMB. To illustrate this point, Fakih detailed 2 main cases in which liquid NGS returned results that did not reflect the patients' true tumor biology. In both patients, tissue-based testing demonstrated microsatellite stability (MSS). One was a 50-year-old patient with metastatic left-sided CRC who had progressed on 2 prior lines of therapy, including anti-EGFR therapy; the liquid assay flagged the patient’s tumor as MSI-high (MSI-H) while the original tissue NGS showed stable MSI and a TMB of only 2.1 mutations per megabase. In the second case, a 59-year-old woman who had received 8 years of treatment for metastatic disease, including ablation, radiation, and surgery, also received an MSI-H designation per liquid biopsy that contradicted prior tissue findings of MSS.

“Across the board, liquid TMB is typically higher than tissue TMB,” Fakih cautioned. “We have patients where we have seen a liquid TMB of 30 [mutations per megabase], but when you biopsy the tissue, the TMB is still 5 [mutations per megabase]. Those patients do not respond to immunotherapy and should not be treated based on such findings."

Fakih explained that prolonged chemotherapy and targeted therapy exposure often drive subclonal evolution that inflates liquid TMB further. Therefore, MSI and TMB assessment from ctDNA must always be confirmed with tissue before treatment decisions are made.

Of note, Fakih identified HER2 (ERBB2) amplification as a biomarker where ctDNA can provide information that tissue alone may miss. He described a 50-year-old man with extensive liver metastases from a rectal primary whose primary HER2 immunohistochemistry score was 0, yet pretreatment liquid NGS showed high-level ERBB2 amplification. The team withheld anti-EGFR therapy and proceeded with FOLFOXIRI; at hepatectomy, the liver metastases showed HER2 immunohistochemistry positivity of 3+ in more than 95% of cells.

"This is a situation where the primary tumor did not reflect the HER2 amplification present in distant metastases. We benefited from knowing the HER2 status through the liquid biopsy."

When should repeat ctDNA testing actually be conducted?

The final section of Fakih's presentation addressed the question of repeat liquid biopsy in patients with known metastatic disease. "I think we overdo repeat liquid biopsies in the community," he said. "I see a very strong trend of repeating ctDNA in somebody who has a driver KRAS G12D mutation. If [a patient] has a KRAS mutation, no matter how much you repeat that ctDNA when the patient progresses on chemotherapy, it is not going to uncover new biomarkers. That drives [up] the cost, but more importantly, it has no benefit."

In response, he outlined 3 evidence-supported indications for repeat liquid biopsy: at the time of resistance to anti-EGFR therapy; at the time of resistance to HER2-directed therapy; and at the time of resistance to encorafenib (Braftovi) plus cetuximab (Erbitux) in BRAF V600E–mutated CRC.

The strongest data support anti-EGFR rechallenge strategies. The clearest evidence base exists for anti-EGFR rechallenge strategies in RAS/BRAF wild-type tumors. In the phase 2 PARERE study (NCT04787341), patients who had progressed on prior anti-EGFR therapy and then received anti-EGFR–free chemotherapy were rechallenged with panitumumab (Vectibix) vs regorafenib (Stivarga; arm A; n = 106) or the reverse sequence (arm B; n = 107).⁴ Data from PARERE shared at the 2025 ESMO Annual Meeting showed that retreatment with panitumumab was associated with improvements in objective response rate (ORR), with a first ORR of 16% in arm A vs 2% in arm B (P = .003) and a second ORR of 18% vs 0% (P = .013). A similar trend was observed for disease control rate (DCR; first DCR: 61% vs 36% [P < .001]; second DCR: 62% vs 38% [P = 0.003]), and progression-free survival (PFS; first PFS: 4.2 vs 2.4 months [P = 0.103]; second PFS: 3.9 vs 2.7 months [P = .019]) regardless of the treatment sequence.

The biological basis for the strategy rests on the known half-life of RAS/EGFR resistance mutations after anti-EGFR discontinuation. One analysis showed that RAS and EGFR relative mutant allele frequency decays exponentially (r² = 0.93 for RAS; r² = 0.94 for EGFR) with a cumulative half-life of 4.4 months.⁵ An external data set of 73 patients with a ctDNA profile suggestive of prior anti-EGFR exposure and serial sampling validated these findings, showing an estimated half-life of 4.3 months. This washout period allows sensitive ctDNA monitoring to identify patients who may again be candidates for EGFR inhibition.

"Repeating ctDNA in somebody who progressed on anti-EGFR [therapy] makes sense," Fakih said. "You want to know the cause of the resistance, and 6 or 8 months later, if those mutations are gone, you want to know if rechallenge is an option. There is a role for repeat ctDNA in CRC, but it has to be a rational, data-driven reason."

References

1. Fakih, M. Concordance of liquid and tissue: where does it matter? When do we recheck ctDNA?. Presented at: 11th Annual School of Gastrointestinal Oncology; April 11, 2026; Los Angeles, CA.
2. Gupta R, Meric-Bernstam F, Rothe M, et al. Concordance between tissue-based and circulating tumor DNA next-generation sequencing in colorectal cancer. Oncologist. 2020;25(3):235-243. doi:10.1634/theoncologist.2019-0413
3. Ji J, Wang C, Goel A, et al. Circulating tumor DNA testing in curatively resected colorectal cancer and salvage resection. JAMA Netw Open. 2024;7(12):e2452661. doi:10.1001/jamanetworkopen.2024.52661
4. Ciracì P, Germani MM, Pietrantonio F, et al. Re-treatment with panitumumab followed by regorafenib versus the reverse sequence in chemorefractory metastatic colorectal cancer patients with RAS and BRAF wild-type circulating tumor DNA: the PARERE study by GONO. Ann Oncol. 2026;37(1):79-91. doi:10.1016/j.annonc.2025.10.002
5. Parseghian CM, Loree JM, Morris VK, et al. Anti-EGFR-resistant clones decay exponentially after progression: implications for anti-EGFR rechallenge. Ann Oncol. 2019;30(2):243-249. doi:10.1093/annonc/mdy509