Introduction
Cancer of Unknown Primary (CUP) represents a significant clinical challenge, accounting for 3–5% of all cancer diagnoses. Defined as metastatic disease where the primary tumor site remains unidentified despite extensive diagnostic efforts, CUP typically carries a poor prognosis, with median survival often falling under one year. Because standard oncology relies heavily on site-specific protocols, patients with CUP frequently exhaust standard treatment options quickly. Precision oncology—the use of Comprehensive Genomic Profiling (CGP) to identify molecular drivers and match them with targeted therapies—offers a potential pathway for these patients. This study evaluates the real-world implementation and efficacy of CGP-guided treatment in a tertiary care setting.
Methods
Researchers conducted a retrospective observational study at the Comprehensive Cancer Center Munich LMU, analyzing 61 CUP patients presented to the Molecular Tumor Board (MTB) between June 2016 and February 2022.
Key methodologies included:
- Comprehensive Genomic Profiling (CGP): Tumor DNA and RNA were sequenced using panels such as Oncomine (50–161 genes) and TSO500 (525 genes).
- Molecular Tumor Board (MTB): An interdisciplinary team of oncologists, pathologists, and geneticists reviewed CGP results alongside clinical histories to formulate personalized treatment recommendations.
- Evidence Ranking: Recommendations were categorized using the ESMO Scale for Clinical Actionability of Molecular Targets (ESCAT) to ensure evidence-based decision-making.
Results
Genomic Findings and Recommendations
CGP was highly effective at identifying molecular alterations. Clinically relevant variants were detected in 74% (45/61) of patients. Based on these findings, the MTB issued specific therapeutic recommendations for 64% (29/45) of those with alterations. The most frequent actionable targets identified were high Tumor Mutational Burden (TMB-high) and mutations in PIK3CA and ERBB2 (HER2).
The Implementation Gap
Despite the high rate of recommendations, a significant "actionability gap" was observed. Only 4 out of 29 patients (14%) actually received the recommended targeted therapy. For the remaining patients, recommendations were not implemented due to:
- Worsening general clinical condition (36%)
- Loss to follow-up (28%)
- Death prior to treatment initiation (24%)
Clinical Outcomes
Of the four patients who received targeted therapy, none derived a measurable clinical benefit; all four succumbed to their disease within three months of starting the new treatment. While the cohort's median overall survival from initial diagnosis was 18.5 months, the median survival after the MTB discussion was only 3.7 months.
Discussion and Conclusion
The findings reveal a stark contrast between biological actionability and clinical utility. While CGP frequently identifies targets, the timing of these interventions is often too late. The median time from initial diagnosis to MTB referral ranged from 79 days for internal patients to 249 days for external referrals. By the time genomic data was processed and reviewed, many patients were too medically fragile to tolerate further therapy.
This study underscores that for precision oncology to be effective in CUP, CGP must move from a "last-resort" diagnostic to an early-stage tool. Integrating genomic profiling into the initial diagnostic workup may allow for the identification of therapeutic targets while patients are still physically robust enough to benefit from targeted interventions. Future prospective trials, such as the CUPISCO trial, remain essential to determining if earlier intervention can successfully bridge the gap between genomic discovery and patient survival.
Summary updated