However, the treatment options for metastatic RCC continue to increase and the role of traditional interventions (cytoreductive nephrectomy) are being re-evaluated. In that setting, we noted that there is little contemporary data regarding the clinical presentation of these patients, specifically with regards to the pattern of metastatic spread.
Only one other group, Bianchi et al,1 had looked at this question in the recent past. They utilized the Nationwide Inpatient Sample database to assess metastatic spread in 11,157 patients with metastatic RCC between 1998 and 2007, but their study was limited by the inpatient nature of the database. As many patients present on an outpatient basis, their study was very selective. The objective of our study was to build on the work of Bianchi and colleagues by utilizing a larger contemporary population-based dataset to provide guidance regarding predictors and patterns of metastatic disease in patients with RCC.
To that effect, we utilized the well-recognized SEER dataset to identified patients with primary kidney RCC between 2010 and 2013; the years were selected based on the availability of data regarding the presence of bone, brain, liver and lung metastases at the time of diagnosis. They also represented the 4 most recent years of data available in the SEER dataset.
We identified 50,815 patients with primary kidney RCC, of whom 6,610 patients (13.0%) had metastases at the time of diagnosis. With a mean follow-up of 19.1 months for the entire cohort, we found that patients with sarcomatoid or collecting duct histology, higher clinical T-stage (cT stage), higher histologic Fuhrman grade (FG), or clinical nodal disease (cN1) were more likely to have metastases at diagnosis. Patients with metastatic disease at diagnosis were much less likely to undergo a radical nephrectomy.
One of the key analyses in our study is the distribution of metastatic disease. First, we looked at all metastases, regardless of synchronous spread. Figure 1A highlights this distribution - metastasis to the lung (51.2%) and regional lymph nodes (41.5%) predominate, followed by spread to the bone (33.5%). Figure 1B identifies the proportion of patients who had isolated metastatic disease to the sites listed. Again, metastatic spread to the lung (16.8%), regional lymph nodes (12.1%), and bone (11.1%) are most common. The distribution of metastases is similar to that found by Bianchi and colleagues. As they alluded to, over one-third of patients with metastatic RCC continue to present with bone metastases. In fact, while 16.8% of patients present with lung metastases alone, 11.1% of patients present with only bone metastases (Figure 1B). While current staging guidelines would capture lung metastases and lymphadenopathy with a chest x-ray and abdominal cross-sectional imaging, bone metastases may potentially be missed. The current guidelines only recommend bone imaging in clinically symptomatic patients or those with elevated alkaline phosphatase. Based on this data and prior data from Bianchi and colleagues, there may be benefit to more routine bone imaging for patients being staged for RCC, particularly higher risk RCC. Studies assessing newer imaging modalities may improve our ability to identify asymptomatic bony metastatic disease. Further prospective studies are warranted to assess the clinical impact of bone imaging as part of routine staging.
Next, we looked at predictors of presenting with metastatic disease at each specific site. Histology had an important impact - papillary RCC was less likely to develop metastases to all four distant metastatic sites; chromophobe RCC was associated with less metastasis to the lung and bone; sarcomatoid RCC was associated with a significantly higher rate of bone metastases at diagnosis; collecting duct RCC was not an independent predictor of any specific site of distant metastases. Importantly, for the most part, patient related variables, such as age, race, SES, insurance status, region, and marital status, were not associated with the site of metastatic disease at the time of diagnosis. The one exception was that patients in the Southeastern US were much more likely to present with bone metastases.
Lastly, we looked at impact on cancer-specific (CSM) and overall mortality (OM). Cancer-specific mortality was assessed using a competing risks regression analysis. There were 3398 cancer-specific deaths (51.4%) in the cM1 cohort and 1384 (4.4%) deaths in the cM0 cohort. The predictors of higher CSM in patients with metastatic disease at the time of diagnosis were higher cT stage, cN+ disease, higher FG, sarcomatoid histology, collecting duct histology, and papillary histology. All four metastatic sites were independent predictors of worse CSM, but brain and liver metastases were associated with the highest CSM. Patients undergoing a cytoreductive nephrectomy in the setting of metastatic disease had improved CSM compared to those without nephrectomy.
Importantly, marital status (widowed or divorced/separated) and lower SES predicted a higher CSM, but were not associated with higher OM. As these factors were not associated with presence of metastatic disease at diagnosis, this discrepancy highlights a deficit between diagnosis and treatment of this population. Despite presenting with a similar clinical stage and metastatic spread, patients with lower SES and patients who have lost a partner have higher CSM. Our results suggest that patients of lower socioeconomic status or patients who are widowed or divorced/separated may not receive the same treatment for their metastatic RCC despite being diagnosed at the same clinical stage. They represent an at-risk population for whom additional healthcare resources should be considered.
Ultimately, understanding the distribution of distant metastases and associated CSS is important to counseling patients with newly diagnosed metastatic RCC. Solitary lung and bone metastases can occur in 10-20% of patients with metastatic disease at the time diagnosis; both should be assessed for during staging evaluations. Reassuringly, disease-specific factors drive the presence of metastases at diagnosis more than patient-related factors. However, there is still a gap between diagnosis and treatment that needs to be rectified.
Read the Abstract
Written By: Thenappan (Thenu) Chandrasekar, MD
Clinical Fellow in Urologic Oncology (SUO), Division of Urologic Oncology, University Health Network and University of Toronto, Toronto, ON, Canada
1. Bianchi M, Sun M, Jeldres C, et al. Distribution of metastatic sites in renal cell carcinoma: a population-based analysis. Annals of oncology : official journal of the European Society for Medical Oncology. Apr 2012;23(4):973-980.