The treatment paradigm for metastatic renal cell carcinoma (mRCC) has evolved over the last two decades from an era of cytokine regimens to VEGF targeted therapies (VEGF-TT) [1]. Prognostic models have also been refined over time, from the Memorial Sloan-Kettering Cancer Center criteria, which was initially established based on mRCC patients treated with interferon alfa (IFN-α) on clinical trials [2, 3], to the International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) criteria, which has been validated in the era of oral VEGF-TT [4]. More recently, the treatment landscape has expanded to include contemporary immune checkpoint blockade (ICB). Nivolumab, a programmed cell death protein-1 (PD-1) antibody, became the first checkpoint inhibitor to be approved by the US Food and Drug Administration for second line mRCC in November 2015 [5, 6] and was subsequently approved for use in Europe [7]. Currently, there are several Phase 3 randomized trials investigating PD-1/PD-L1-based therapies in the first-line setting (NCT02420821, NCT02684006, NCT02811861). The first to result – the CheckMate-214 study – was stopped early due to a survival benefit seen with combination ICB versus single agent VEGF-TT [8]. Well-validated biomarkers that can be earlier prognostic or predictive readouts in mRCC treated with conventional ICB are lacking and represents an unmet need.

Inflammation is a recognized hallmark of cancer, however, the mechanism by which inflammation leads to worse outcomes in mRCC is not well described [9]. Neutrophilia is thought to occur as an inflammatory response and may lead to suppression of cytolytic activity of immune cells such as lymphocytes, natural killer cells, and activated T cells [10, 11]. An elevated neutrophil-to-lymphocyte ratio (NLR) has been shown to be associated with a poor prognosis in several solid tumors [12–14]. For example, in mRCC patients treated with VEGF-TT, an elevated NLR (NLR > 3) at baseline and an increase in NLR at 6-weeks were associated with worse outcomes in terms of survival and objective response rates [15]. The prognostic value of NLR in the current era of ICB has been evaluated in small subsets of patients, for example, with lung, melanoma, and bladder malignancies [16–18]. However, the utility of NLR in the context of contemporary immunotherapy for mRCC has not been well-defined. In this study, we investigated the association of baseline NLR, and changes during treatment, with outcomes in mRCC patients treated with conventional ICB.

In this analysis, we show that for mRCC patients treated with contemporary PD-1/PD-L1 ICB, higher 6-week NLR was independently associated with worse outcomes in terms of reduced ORR and shorter PFS and OS. In landmark analyses, we also demonstrate that early decline (decrease ≥25%) of NLR at 6-weeks was associated with an improved PFS and significantly better OS, whereas a relative increase by ≥25% was associated with poorer PFS and OS, regardless of baseline levels. Interestingly, while the results seen at baseline (pretreatment) NLR levels were statistically significant for PFS, the numerical values for OS and ORR in this study were nearly identical to results seen in a larger NLR analysis in mRCC patients treated with VEGF-TT [15]. Taken together, our data suggests that NLR appears to be a readily-available, prognostic marker in mRCC patients treated with conventional ICB, and warrants larger, prospective validation.

Our findings are consistent with and build upon previous reports evaluating NLR in solid tumors, including RCC [12–15]. In localized RCC, higher NLR at diagnosis (> 2.7, typically pre-nephrectomy) has been shown to be associated with an increased risk of recurrence [20]. However, a review by Boissier et al. suggests that in this localized RCC setting, NLR has not been shown to be significant for overall survival based on pooled data [21]. In locally-advanced or mRCC, higher NLR (typically >3) has been shown to be an independent prognostic factor for PFS and OS [21]. However, many of these studies were performed in the context of interleukin or IFN-based regimens. In subgroup analysis of the phase III S-TRAC study, which evaluated sunitinib versus placebo in patients with high-risk locoregional RCC post nephrectomy, baseline NLR ≤ 3 was associated with improved disease-free survival (DFS) with sunitinib compared to placebo (HR 0.72, 95% CI 0.54–0.95, p = 0.02), whereas NLR > 3 was not (HR 1.01, 95% CI 0.58–1.77, p = 0.96) [22]. Templeton et al. evaluated the utility of NLR in mRCC patients primarily treated with VEGF-TT and showed that, compared with no change, increase in NLR (≥25%) at week 6 was associated with poorer OS, PFS and reduced ORR whereas an early decline (decrease ≥25%) was associated with improved outcomes [15]. Regarding treatment with contemporary ICB, studies evaluating patients with melanoma or advanced non-small-cell lung cancer (NSCLC) have shown that higher pretreatment NLR is associated with inferior OS and PFS [16–18]. However, we have performed the first analysis of the utility of NLR in mRCC patients treated in the current ICB era. Further, we similarly demonstrate the importance of changes in NLR during treatment and the prognostic relevance of measurements at week 6 independent of other factors. Given the expanding landscape and ongoing studies of PD-1/PD-L1-based therapies in mRCC [8, 23], accessible and affordable prognostic or predictive markers will continue to be a growing need.

There are important clinical implications of our data particularly in the context of the unique and heterogenous radiographic findings in this patient population [24, 25]. While these results would benefit from prospective validation, the readouts at 6-weeks on ICB therapy are informative for both patients and physicians given that this time point typically coincides with the first set of re-staging scans after initiation of treatment. For example, if a patient presents at 6-weeks on therapy with stable or slightly progressive disease on imaging and a simultaneous decline in NLR, this may be reassuring to continue treatment assuming it is otherwise clinically suitable (Fig. 5, upper panels). Similarly, one may be more cautious regarding prognosis in a situation where a patient returns at 6-weeks with slightly progressive disease on imaging and a significant increase in NLR (Fig. 5, lower panels). Ultimately, the NLR is a helpful and available prognostic marker but should be considered in the context of other relevant clinical details when assessing the risk-benefit ratio of continuing ICB treatment at the individual patient level.

Biologically, the NLR is a marker of systemic inflammation and potentially reflects the balance of the immune system in the context of a malignancy. The neutrophil count is thought to reflect the inflammatory microenvironment that in turn has tumor-promoting activity, including cancer cell survival and proliferation, angiogenesis and metastasis, as well as subversion of adaptive immune responses [26]. Lymphocytes are effective suppressors of cancer progression and their presence, particularly in the tumor microenvironment, is thought to reflect host immunity [27]. In a study of 35 advanced RCC patients treated predominantly with IFN, Sejima et al. showed an association of Fas ligand (FasL) expression in nephrectomy tumor cells with reduced lymphocyte count and thus higher NLR [28]. They explained this association by the concept of “FasL tumor counter-attack”, whereby FasL in tumor cells mediates tumor cell immune privilege by inducing apoptosis of cytotoxic T lymphocytes in the microenvironment. On the other hand, this hypothesis has also been challenged, for example, with in vivo data [29]. Ultimately, this highlights the need for further prospective studies in patients treated with contemporary ICB, particularly with the added depth of next-generation sequencing and other informative technologies, to add more granularity to the biological underpinning of NLR in this setting.

These data should be interpreted in the context of the study design. First, this was a retrospective analysis which has the potential for selection bias and confounders. We attempted to control for this by utilizing a multivariable analysis to adjust for mRCC-specific prognostic variables that may impact analysis, including histology, line of therapy and IMDC risk factors. Our cohort included patients who were treated with PD-1 or PD-L1 ICB and there may be subtle differences between these drug pathways. For example, PD-1 inhibitors target PD-1:PD-L1 and PD-1:PD-L2, whereas PD-L1 inhibitors target PD-1:PD-L1 and PD-L1:B7.1 [30, 31]. While these slight mechanistic differences did not significantly affect our overall findings, prospective data would be informative particularly when evaluating single versus combination ICB therapy. Further, we could not control for concomitant medications that may have influenced white blood cell counts. Additionally, PD-L1 expression was not known in this retrospective analysis and may be a worthy point of future prospective study given the utility of this tissue biomarker continues to evolve in mRCC. Finally, similar to previous work investigating NLR in mRCC patients treated with VEGF-TT [15], data from untreated patients were not available in our analysis and thus it was not possible to assess the potential predictive value of NLR at baseline or on therapy.

In our cohort of mRCC patients treated with PD-1/PD-L1 based immune checkpoint blockade, higher 6-week NLR was independently associated with a worse ORR and shorter PFS and OS. Early decline of NLR was associated with an improved PFS and significantly better OS, whereas a relative increase of NLR was associated with poorer PFS and OS, regardless of baseline levels. The NLR appears to be a readily-available, affordable, prognostic marker in mRCC patients treated with immune checkpoint blockade and warrants larger, prospective validation.