Systemic inflammation is one of the hallmarks of cancer, and elevated markers of inflammation have been associated with poor prognosis in patients with solid tumors. In this retrospective analysis of the GEICAM/9906 trial, the prognostic role of the derived neutrophil-to-lymphocyte ratio (dNLR) was evaluated in subgroups of women with early breast cancer.
The GEICAM/9906 trial was a randomized phase 3 study evaluating adjuvant 5-fluorouracil, epirubicin, and cyclophosphamide (FEC) compared with FEC followed by weekly paclitaxel in breast cancer patients with axillary involvement. The primary objective of the study was the 5-year disease-free survival (DFS), with overall survival (OS) as a secondary end point. For the analysis presented here, dNLR was calculated as the ratio of neutrophils divided by the difference between total leukocytes and neutrophils measured in peripheral blood before the start of chemotherapy. Patients with total leukocyte counts >15 (109/L) were excluded. A univariable Cox regression was used to evaluate the prognostic and predictive value of dNLR, which was explored as a median cutoff and grouped into quartiles. Patient subgroups were defined by a 50-gene qPCR assay (PAM50) and by their hormonal receptor expression, as measured by immunohistochemistry. Notably, human epidermal growth factor receptor 2 (HER2)-positive patients did not receive adjuvant trastuzumab, which had not been approved at the time of study design.
Overall, 1243 patients from 65 sites in Spain were evaluated in this analysis. Median follow-up was 10 years, and most patients had received some form of endocrine therapy. PAM50 subtype determinations were available for 66% of the included patients, whose subtypes were as follows: Luminal A (22%), Luminal B (21%), HER2-enriched (14%), basal-like (6%), and normal-like (3%). Tumors with subtypes that were estrogen receptor–positive (ER+)/progesterone receptor–positive (PgR+) and ER–/PgR– comprised 47% and 13% of evaluable tumors, respectively.
The median dNLR in this study was 1.35 (interquartile range, 1.08-1.71), and was less than that described in metastatic scenarios, likely because of the increased inflammation seen with enhanced disease. For HER2-enriched patients by PAM50, a dNLR greater than the median was associated with a significantly worse DFS (P = 0.03; hazard ratio, 1.63; 95% confidence interval, 1.04-2.54). For nonluminal patients by PAM50, a dNLR greater than the median and a high dNLR explored by quartiles were associated with worse DFS (P = 0.02 and P = 0.03, respectively). For patients with ER–/PgR– tumors, a high dNLR grouped into quartiles was associated with worse DFS and OS (P <0.001 and P = 0.007, respectively).
These results indicate that the presence of high levels of systemic inflammation, as defined by dNLR, are associated with worse DFS in women with HER2-enriched and nonluminal intrinsic subtypes. Additionally, a high dNLR is also associated with worse DFS and OS in women with ER–/PgR– tumors. The authors noted that multivariate analysis should be performed, and is currently underway. Regarding the use of dNLR in current practice, Dr Nadia Harbeck of Munich, Germany, who was unaffiliated with the study, commented, “the clinical utility—whether as a static or dynamic biomarker, or therapeutic target—remains to be determined.”
Ocana Fernandez A, et al. ESMO 2016. Abstract 145O.