Premium
Importance of non‐regional lymph nodes in assigning risk in primary metastatic prostate cancer
Author(s) -
Ali Adnan,
Hoyle Alex,
Mistry Hitesh,
Clarke Noel W.
Publication year - 2019
Publication title -
bju international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.773
H-Index - 148
eISSN - 1464-410X
pISSN - 1464-4096
DOI - 10.1111/bju.14400
Subject(s) - medicine , prostate cancer , hazard ratio , proportional hazards model , lymph node , propensity score matching , cancer , prostate , cohort , radiology , retrospective cohort study , oncology , confidence interval
Objective To determine the prognostic relevance of non‐regional lymph node ( NRLN ) metastases presenting synchronously with bone metastases in metastatic prostate cancer ( mPC a) for guiding treatment decisions based on oligometastatic definitions. Patients and Methods Patients diagnosed with mPC a between 2004 and 2013 were identified from the Surveillance, Epidemiology and End Results database and were grouped by metastatic sites into only NRLN , only bone, bone + NRLN and other sites ± bone/ NRLN metastases. Multivariate Cox and competing risk regression analyses were performed to compare the risks of all‐cause mortality ( ACM ) and prostate cancer‐specific mortality ( PCSM ) associated with bone + NRLN metastases before and after propensity‐score matching to patients with only bone metastases. This was complemented with landmark and supplementary analyses. Results Of 17 167 patients with mPC a identified, 63.1% presented with only bone metastases, while bone and NRLN metastases co‐occurred in 8.9% of the cohort. On multivariate analyses, after adjusting for potential confounders (clinical and sociodemographic), patients with bone + NRLN metastases had a significantly higher risk of ACM (hazard ratio [ HR ] 1.161, 95% confidence interval [ CI ] 1.084–1.243; P < 0.001) and PCSM (subdistribution HR 1.149, 95% CI 1.067–1.237; P < 0.001) compared with patients with only bone metastases. Landmark analyses limited to survivors of ≥6 and ≥12 months again showed a significantly increased risk of ACM for patients presenting with bone + NRLN metastases compared with patients with only bone metastases. In a subsequent 1:1 propensity‐score‐matched cohort of patients with bone + NRLN metastases and only bone metastases, the bone + NRLN group had higher multivariate‐adjusted hazard rates for ACM ( HR 1.202, 95% CI 1.102–1.311; P < 0.001) and PCSM (subdistribution HR 1.146, 95% CI 1.044–1.259; P = 0.004). Conclusions Patients with concomitant NRLN and bone metastases have a higher risk of death, NRLN and bone metastases therefore representing a high‐risk feature, when compared with patients with bone metastases alone. The current therapeutic stratification of ‘low‐’ vs ‘high‐volume’ disease does not account for this phenomenon, and patients requiring aggressive combination therapy may not receive maximum therapeutic benefit as a consequence.