Serum C-reactive Protein and Neutrophil/Lymphocyte Ratio After Neoadjuvant Radiotherapy in Soft Tissue Sarcoma

Discussion

Similar to other solid tumors, previous work has clearly demonstrated the prognostic value of CRP and N/L ratio in patients with STS (6). Nakamura et al., for example, observed that an elevated CRP >0.3 mg/dl before initial treatment was an independent predictor of worse event-free survival by approximately 30% in 102 patients with localized STS (2). In a subsequent report, Nakamura et al., analyzed 142 STS patients and found that elevated CRP and an N/L ratio >2.3 independently predicted shorter disease-specific survival (87% 5-year OS for normal CRP and low N/L ratio compared to 46% for elevated CRP and high NLR) (3). More recently, Nakamura explored CRP levels in patients with metastatic STS and also found that CRP was an independent prognostic factor for shorter disease-specific survival (12). Similarly, Szkandera et al. evaluated 304 STS patients, identifying elevated preoperative CRP ≥0.69 mg/dl as an independent predictor of worse survival (5). In a separate report from the same group, Szkandera analyzed 260 STS patients and found that an N/L ratio ≥ approximately 3.5 was associated with worse time to STS recurrence and OS (4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Overall survival among upfront surgery and preoperative radiotherapy cohorts stratified by N/L ratios. Patients were stratified into high and low neutrophil/lymphocyte ratios based on the median value for the entire cohort as the cut-point.

Yet, despite these consistent data regarding the adverse oncological impact of elevated CRP and N/L ratio, the overwhelming majority of studies (in STS and other solid tumors) have evaluated these markers in the setting of either upfront surgery or metastatic disease. In both Nakamura reports, no patients received neoadjuvant RT, and the majority (~80%) was treated with surgery as monotherapy. In the Szkandera studies, all patients were treated with upfront surgery, but a larger fraction (~60%) received adjuvant RT. To our knowledge, our study is, therefore, the first to evaluate the impact of CRP and N/L levels on STS patients receiving neoadjuvant RT, comparing them to patients receiving upfront surgery (and adjuvant RT in approximately 30% of cases). Our results uphold the association between elevated N/L ratio with decreased OS in patients undergoing upfront surgery. However, in patients receiving neoadjuvant RT, we observed a novel finding of no association between elevated CRP and elevated N/L ratio for both OS and DRFS. Therefore, our findings fill a gap in the literature, which although potentially unexpected, are nevertheless novel and clinically relevant.

Although our data do not address the underlying mechanism for why neoadjuvant RT may decouple the association of elevated CRP and N/L ratio with worse STS outcome, one broad hypothesis is that neoadjuvant RT may impact the inflammatory milieu of the TME in a different manner than upfront surgery and thereby alter the interaction of these bio-markers with outcome (12-15). For example, pre-clinical studies suggest that complex inflammatory and cell-signaling cascades secondary to RT may mediate a strong immune response which can counteract and in some cases overcome the local immunosuppression of the TME (16). These observations and further studies to understand the impact of RT on the TME are especially pertinent with the ongoing development of novel radiotherapy regimens such as hypofractionated RT and innovative energy sources such as protons and carbon ions (17). Although the question of how local RT-induced changes translate to systemic changes in inflammatory markers and possible anti-inflammatory mediators remains unanswered, our data support a potential impact of neoadjuvant RT on these inflammatory pathways.

Other studies have suggested that metabolic pathways are significant in sarcoma oncogenesis and progression, and the impact of anti-metabolites and RT on the TME are worthy of further investigation. For example, Igarashi et. al observed decreased survival of osteosarcoma cells and xenografts after administration of a methionine inhibitor (18). As RT is known to produce reactive oxygen species and dysregulate key machinery of cellular metabolism, it is reasonable to hypothesize an interaction causing alterations in the inflammatory environment. Interestingly, the same group studied the therapeutic effects of caffeine and valproic acid and found similar anti-metabolic and anti-tumor effects (19). Other groups such as Ahn et al. have explored the anti-tumorigenic effects of compounds such as sphingosine, a metabolite of sphingolipids, which has known anti-inflammatory properties (20). They showed that administration of sphingosine to alveolar rhabdomyosarcoma cell lines caused a dose dependent inhibition of growth and acceleration of cell death. Therefore, the interaction of CRP and N/L ratio in the setting of other anti-inflammatory mediators requires further study, and comparative proteomics approaches using novel mass spectrometry techniques may reveal insights into these complex protein interactions (21).

Despite the potential novelty of our results, it is important to acknowledge our study's limitations. Firstly, it is a retrospective, single-institution study which has a smaller sample size relative to other studies evaluating CRP and N/L ratios in STS. As is typical of STS studies, our analysis included a heterogeneous cohort of numerous histologic subtypes and tumor locations, and this created an imbalance in baseline characteristics between our neoadjuvant RT and upfront surgery cohorts which may have biased our results. However, despite these differences in clinical/pathological characteristics, the baseline CRP and N/L ratios were statistically similar between the two cohorts. Since these markers were the principal predictor variables for our analysis, the similarity in baseline values between the groups suggests that the associations we observed with the outcomes of interest were not unduly confounded by selection bias between the RT and upfront surgery groups. However, we acknowledge that our study may have been underpowered to detect modest but potentially significant differences in baseline CRP and N/L ratio levels between the cohorts. With a larger sample size, especially among the neoadjuvant RT patients, it is possible that CRP levels and N/L ratios may show potentially significant associations with oncological outcomes as studies of STS patients receiving upfront surgery have shown. Further research of this question appears warranted.

Ultimately, we found that baseline and post-treatment N/L ratio and CRP do not correlate with survival in STS patients receiving neoadjuvant RT. Thus, the utility of N/L ratio and CRP as predictors of poor clinical outcome may not apply in these patients. In fact, patients with elevated CRP and N/L ratio at diagnosis may be good candidates for neoadjuvant RT, although these decisions should remain individualized.