CUP-COMP: blood profiling beats tissue in CUP
CUP-COMP shows blood profiling can outpace tissue testing in cancer of unknown primary. The British Journal of Cancer report on NCT04750109 found liquid biopsy succeeded more often than tissue NGS, fed a national Molecular Tumour Board, and supported precision-medicine choices for about one in four enrolled patients—an operational signal for APAC CUP pathways short on biopsy material.
Contents10 sections
Key Takeaways
- 117 CUP patients were recruited at seven UK sites (June 2021–February 2023); 113 completed 12 months of follow-up.
- Molecular profiling succeeded in 115/117; blood success was 93% versus 61% for tissue; baseline blood success was 109/117 (93%).
- MTB identified potentially actionable alterations (PAAs) in 63% (73/115); 26% (31/117) received a precision-medicine approach.
- Median overall survival was 9.5 months; 33% of patients had tumour fraction <1%, limiting blood PAA detection.
CUP-COMP at a glance
| Field | Detail |
|---|---|
| Study | CUP-COMP — Carcinoma of Unknown Primary: comparison across tissue and liquid biomarkers |
| Registry | NCT04750109 (ClinicalTrials.gov); status Completed |
| Design | Prospective, multi-centre, non-randomised feasibility; 7 UK sites |
| Sponsor | The Christie NHS Foundation Trust |
| Enrollment | 117 participants (registry); recruitment Jun 2021–Feb 2023 |
| Blood assay | FoundationOne Liquid CDx (baseline ± progression) |
| Tissue assay | FoundationOne CDx on residual FFPE where suitable; optional fresh-tissue WGS |
| Publication | Conway et al., Br J Cancer (2026), DOI 10.1038/s41416-026-03519-6 |
What did the Nature portfolio paper report?
According to Conway and colleagues’ British Journal of Cancer article on CUP-COMP, cancer of unknown primary accounts for about 2–3% of new cancer diagnoses yet remains a leading cause of cancer death. Most patients fall in the unfavourable subset treated with non-specific chemotherapy and median survival under one year.
CUP-COMP tested whether integrating molecular profiling with a National Molecular Tumour Board is feasible in routine CUP care. Of 117 recruits, 113 completed 12 months of follow-up. Among 115 patients with successful profiling, blood-based testing was more successful (93%) than tissue-based testing (61%). The MTB identified PAAs in 63% (73/115). A precision-medicine approach was used in 26% (31/117). Primary tumours were subsequently confirmed or suspected in 41 patients. Median overall survival was 9.5 months.
The same findings are citable via DOI 10.1038/s41416-026-03519-6. Authors conclude liquid biopsies are highly feasible in CUP and that blood-based profiling is a timely alternative when tissue is unavailable or delayed, while tissue remains preferable when tumour fraction is low.
How is NCT04750109 registered?
ClinicalTrials.gov study NCT04750109 lists the official title “Carcinoma of Unknown Primary (CUP): a Comparison Across Tissue and Liquid Biomarkers,” observational design, completed status, 117 participants, sponsor The Christie NHS Foundation Trust, start 9 June 2021, and completion 28 June 2024.
Eligibility required age 16 or older, ECOG performance status 0–2, and CUP confirmed per ESMO guidelines with local MDT discussion. Both favourable and unfavourable CUP subsets were eligible. Primary objectives included sequencing tumour tissue and circulating tumour DNA to improve treatment stratification and establishing a genomic reporting mechanism for potentially actionable abnormalities. Seven UK sites participated, including The Christie in Manchester, UCL Cancer Institute, Velindre, Edinburgh Cancer Centre, and Royal United Hospitals Bath.
Which assays and MTB process did investigators use?
The journal methods state all consenting patients provided baseline blood for FoundationOne Liquid CDx. After amendments, residual FFPE was retrospectively profiled with FoundationOne CDx when suitable tissue remained. Progression blood draws and optional fresh-tissue whole-genome sequencing with germline control were collected where clinically appropriate.
A monthly virtual CUP MTB used the eTARGET platform with clinical scientists from the North West Genomic Laboratory Hub. Variants were classified for pathogenicity/oncogenicity and assessed for possible germline origin or CHIP. The board flagged alteration patterns suggesting a primary site and PAAs predictive of standard-of-care or experimental therapy. Treatment changes remained local clinical decisions, not trial-mandated assignments.
Why does tumour fraction matter for liquid biopsy?
CUP-COMP reports that detecting a PAA in blood depended on tumour fraction, with 33% of patients below 1%. That subgroup risks false-negative liquid results even when the assay “succeeds” technically. The paper therefore positions tissue profiling as still preferred when adequate biopsy is available, and blood as the practical default when tissue fails quality checks or turnaround is too slow for a progressing patient.
For APAC centres building CUP molecular pathways—especially where repeat biopsy logistics are difficult—these numbers argue for dual workflows: reflex ctDNA at baseline plus tissue salvage when tumour fraction is low or results are discordant.
What should APAC oncology and diagnostics teams change?
Map CUP MDTs to a standing molecular board with clear PAA criteria, turnaround SLAs, and documentation of when blood substitutes for tissue. Budget for assays with reported tumour-fraction metrics, not only binary “pass/fail” reports. Track how often profiling changes primary-site suspicion or trial matching, mirroring CUP-COMP’s 41 confirmed/suspected primaries and 31 precision-medicine treatments.
Related NovaPharmaNews APAC reading includes Leqembi real-world LEADER study coverage, the Insilico–Bora alliance, and Nektar clinical-program context.
What remains unproven for Asia-Pacific practice?
CUP-COMP was UK-based and non-randomised; survival contrasts versus chemotherapy are observational and should not be read as a registrational efficacy claim for any drug. Assay access, reimbursement, and MTB staffing differ across APAC markets. The paper supports feasibility and utility of blood NGS in CUP pathways, not a mandate to abandon tissue histology or immunohistochemistry workups required by ESMO-aligned guidelines.
Frequently Asked Questions
What is the CUP-COMP trial?
CUP-COMP (NCT04750109) is a prospective UK multi-centre study of blood- and tissue-based molecular profiling plus a national Molecular Tumour Board in patients with cancer of unknown primary, published in the British Journal of Cancer in 2026.
How often did blood profiling succeed versus tissue?
Among patients with successful molecular profiling, blood-based profiling succeeded in 93% of cases compared with 61% for tissue-based profiling; 109 of 117 patients had successful baseline blood profiling.
What should APAC oncology teams take from CUP-COMP?
When tissue is scarce or delayed, a validated ctDNA panel plus MTB review can still surface potentially actionable alterations in many CUP patients, but low tumour fraction (under 1% in 33% of the cohort) can miss blood PAAs and tissue remains preferred when available.
Primary Sources
This article follows our editorial standards. Report a correction via editorial contact.
Deeper reading
Industry reports & whitepapers
- La Negoziazione del Prezzo dei Farmaci Oncologici in Italia — This whitepaper analyzes the market benchmark role in negotiating oncology drug prices in Italy, hig…
- Cytological Specimens in the Molecular Era of Metastatic Melanoma — This whitepaper discusses the role of cytological specimens in diagnosing metastatic melanoma and th…