# Randomised phase 2 trial of site-specific treatment based on gene expression profiling versus carboplatin and paclitaxel in cancer of unknown primary

Source: https://onco.cc/key-papers/paper-hayashi-site-specific-vs-empirical-chemotherapy-cup-jco-2019/  
OnCo record `paper-hayashi-site-specific-vs-empirical-chemotherapy-cup-jco-2019` (Key paper). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Using a gene expression test to guess where a cancer of unknown primary came from, and treating it as that cancer, did not help patients live longer than standard carboplatin and paclitaxel.

## Summary

Japanese multicentre randomised phase 2 trial of 130 patients with cancer of unknown primary, assigned to site-specific chemotherapy chosen from a 2,000-gene expression tissue-of-origin assay or to empirical carboplatin plus paclitaxel.

One-year survival was 44.2 percent with site-specific treatment against 54.9 percent with empirical chemotherapy, and progression-free survival did not differ. The trial, like the earlier GEFCAPI 04 study, argued against routine tissue-of-origin testing to direct chemotherapy.

## Fields

- Kind: Key paper
- Last checked: 2026-09-18
- Journal: Journal of Clinical Oncology
- Year: 2019
- DOI: 10.1200/JCO.18.00771
- Authors: Hayashi H, Kurata T, Takiguchi Y, et al.
- Findings: One-year survival 44.2 percent with site-specific therapy versus 54.9 percent with carboplatin and paclitaxel; no improvement.
- What it means: Predicting the primary site by gene expression and treating accordingly is not better than empirical chemotherapy, so guidelines do not recommend it; the field moved to genomic profiling for actionable targets instead.
- Caveats: Randomised phase 2 with survival as the primary endpoint; underpowered for small differences.; Some predicted primaries received chemotherapy of limited efficacy, which may have blunted any benefit.

## Sources

- J Clin Oncol 2019: https://doi.org/10.1200/JCO.18.00771
- PubMed: https://pubmed.ncbi.nlm.nih.gov/30653423/

## Connected records

- cancers: [Cancer of unknown primary, favourable subsets](https://onco.cc/cancers/cup-favourable-subsets/), [Cancer of unknown primary, unfavourable (adenocarcinoma and poorly differentiated carcinoma)](https://onco.cc/cancers/cup-unfavourable/)
- drugs: [Carboplatin](https://onco.cc/drugs/carboplatin/), [Paclitaxel / nab-paclitaxel](https://onco.cc/drugs/paclitaxel/)
- journals: [Journal of Clinical Oncology](https://onco.cc/journals/jco/)

---
JSON: https://onco.cc/api/v1/entities/paper-hayashi-site-specific-vs-empirical-chemotherapy-cup-jco-2019.json