Page 40 - Read Online

Briggs et al. J Cancer Metastasis Treat 2021;7:46 https://dx.doi.org/10.20517/2394-4722.2021.84 Page 9 of 13

low
high
Teff vs. Myeloid , Teff GES in the JAVELIN Renal 101 cohort in either the avelumab + axitinib arm, or
high
high
the sunitinib monotherapy arm, suggesting that the Myeloid , Teff subgroup may be most resistant to
high
ICI monotherapy rather than targeted monotherapy.
While the prognostic and predictive value of these GES requires further validation, we found the strongest
consensus for angiogenic GES (IMmotion 150 Angio, JAVELIN Renal 101 Angio) as biomarkers predictive
of improved response to sunitinib and for immunogenic GES (IMmotion 150 Teff, JAVELIN Renal 101
Immuno) as biomarkers predictive of improved response to ICI therapy. Additionally, myeloid
inflammation GES (IMmotion Teff, Myeloid) may predict improved response to combination anti-VEGF +
ICI therapy vs. ICI therapy alone. Table 4 and Supplementary Table 4 list all GES biomarkers associated
with predictive or prognostic outcomes.

PDL1 STATUS AS A PROGNOSTIC AND PREDICTIVE BIOMARKER
As the principal biologic target of many of the ICIs, the expression of PDL1 on renal tumor cells has
received significant attention as a potential prognostic and predictive biomarker. Prognostically, a meta-
analysis in 2020 reported that PDL1 expression of tumor cells was positively associated with both OS (HR =
[49]
1.98, 95%CI: 1.22-3.22) and DFS (HR = 3.70, 95%CI: 2.07-6.62) . These findings are notable because
tumors with high expression of PDL1 have been previously shown to demonstrate aggressive behavior [50-58] .
The improved OS in PDL1-expressing tumors in the era of ICIs possibly occurs because PDL1 expression
may also predict tumor response to immunotherapy.

A recent 2020 meta-analysis included 4635 patients across six randomized controlled trials (RCTs)
published before May 2018 with available PDL1 expression data and compared ICI vs. standard of care
therapy (SOC). Regardless of PDL1 expression level, ICI therapy improved both PFS and OS compared to
SOC. However, in PDL1 positive patients receiving ICI, PFS was improved vs. SOC (HR = 0.75, 95%CI:
0.63-0.89, P < 0.0001) but OS was not (HR = 0.72, CI: 0.63-0.81, P = 0.63). Since this meta-analysis, two of
the included RCTs have published longer-term follow-up data on the effect of PDL1 status on response to
ICI without significant change to earlier-published data. Furthermore, other studies assessing response to
ICI based on PDL1 status report both significant [Table 5] and non-significant [Supplementary Table 5]
associations between differential PDL1 expression and PFS and OS.

Overall, we found the strongest consensus for PDL1 as a prognostic biomarker for OS and PFS. Notably,
PDL1 expression is dynamic. Therefore, the assessed tissues (primary tumor vs. metastasis) and timing of
tissue acquisition (especially if primary tumor resection occurs long before evidence of metastasis) may
impact PDL1 expression, and therefore the accuracy of assessment as a biomarker. Table 5 and
Supplementary Table 5 listed all identified data depicting the values of PDL1 as a prognostic or predictive
biomarker associated with PFS, OS, or CSS.

CONCLUSION
We reviewed the serum, gene mutation, genetic expression, and histologic biomarkers that predict response
to treatment and prognosticate clinical outcomes. Current survival models may be improved by
incorporation of newly proven biomarkers, allowing providers to give more accurate and individualized
prognosis to patients. Future predictive models may be built to allow oncologists to prescribe the most
effective treatment regimens for an individual patient’s tumor and biologic profile. It is clear that patients
with mRCC will benefit from continued measurement of biomarkers in large clinical trials assessing clinical
responses to various treatment regimens in patients with mRCC, and their incorporation into increasingly
personalized predictive tools.