Introduction and objectives
Prognostic models have been developed and validated to estimate survival in the setting of metastatic renal cell carcinoma (mRCC) using patient specific data, including the Memorial Sloan Kettering Cancer Center and International Metastatic RCC Database Consortium Risk Models, which prognosticate based on counts of neutrophils, platelets or hemoglobin, Karnovsky performance status, presence of metastatic disease, and other similar metrics. Since the development of these models, there has been rapid advancement in biomarker research, identifying additional prognostic biomarkers predictive of clinical outcomes. Incorporating newer biomarkers into these models could yield even more accurate outcome predictions. We therefore reviewed the prognostic value of common genetic mutations in mRCC to determine suitability for future model inclusion.
Methods
We included human studies reporting associations between genetic tumor mutations and outcomes including overall (OS), cancer specific (CSS), progression-free survival (PFS), or metastasis in mRCC. Data were abstracted via standardized form, then reported using hazard ratios with confidence intervals and/or p values, as available.
Results
Mutated alleles that may be prognostic in some patient populations for poorer OS or PFS compared to wild type (WT) include loss of function (LOF) mutations or alterations in BAP1 (p=.02), CDKN2A (p<.0001), CIMP/FH (p<.0001), and TERT (p=.03). Substantial but conflicting evidence exists regarding the prognostic value of PBRM-1 LOF mutations, with limited evidence of longer OS or PFS in patients receiving nivolumab (p<.05), but no benefit seen in patients receiving everolimus, sunitinib, or combination therapy (p>.05). This suggests that PBRM-1 mutations may prove useful as a treatment-response predictive biomarker. However, increased OS has been shown in patients with PBRM-1 LOF mutations and pancreatic metastasis versus without ( HR .34, p=.007). Additionally, PBRM-1 LOF mutations have been associated with less immunogenic, more angiogenic tumor microenvironments, which may portend poorer prognosis. Conflicting or inconclusive evidence exists regarding whether mutated alleles in ERV, mTOR, or VHL are prognostic for OS or PFS. Finally, SETD2 mutations may be associated with metastatic spread to bone.
Conclusions
Overall, we found the strongest consensus for gene mutations in BAP1, CDKN2A, CIMP/FH, and TERT as biomarkers prognostic for OS or PFS, while the prognostic value of PBRM-1, ERV, mTOR, and VHL mutations has yet to be fully elucidated.
Source of Funding
None