Introduction:Paraneoplastic leukemoid reaction (PLR) comprises 10% of leukemoid reactions among patients with solid tumors, especially those with pulmonary malignancy and metastatic disease. Defined as a white blood cell (WBC) count of >50 x 109/L with mature, non-clonally derived neutrophils, and without tumor involvement in the bone marrow, PLR is associated with a poor prognosis. Systemic inflammation promotes tumor growth and metastasis; however, the mechanisms underlying PLR are not well elucidated. We performed a comprehensive clinico-pathological-molecular analysis of cytokines and gene expression in bone marrow, metastatic tumor, and serum from a 76-year-old man with metastatic poorly differentiated non-small cell lung cancer (NSCLC), hyperleukocytosis (WBC 146.5 x 109/L) and extreme neutrophilia (96% neutrophils; absolute neutrophil count 140.6 x 109/L).Methods:After informed consent, we conducted an extensive clinical evaluation of the patient's neutrophilic hyperleukocytosis. We examined the patient's peripheral blood, serum and bone marrow via histologic examination, flow cytometry, cytogenetics, and fluorescent in situ hybridization (FISH) following established protocols. We evaluated levels of 12 cytokines (G-CSF, GM-CSF, IFN-gamma, IL-1a, -2, -4, -6, -8, -10, -12, and -17a, and TNF-a) in serum by enzyme-linked immunosorbent assay (ELISA; Quantikine, R&D Systems). We compared gene expression of 30 cytokines and their receptors (CSF2/R, CSF3/R, IFN-gamma/R1, IL-1A/B/R1/RN, IL-2/RA/RB/RG, IL-4/R, IL-6/R, CXCL8/CXCR1, IL-10/RA, IL-12A/RB2, IL17/RA, and TNF-a/RSF1A/1B) on paraffin-fixed samples of the patient's NSCLC and on an age- and gender-matched sample of NSCLC from a patient without PLR (Geneticist Inc. Biorepository), using real-time polymerase chain reaction (PCR; RT2 Profiler, QIAGEN).Results:We confirmed the diagnosis of PLR after an extensive evaluation did not show any infectious or clonal myeloproliferative process. The peripheral blood smear showed marked leukocytosis, composed mainly of mature neutrophils and mild absolute monocytosis without circulating blasts or atypical cells; there was also a normochromic, normocytic anemia (hemoglobin 11.9 g/dL, hematocrit 35.7%, and MCV 94 fl) and mild thrombocytopenia (platelets 148 x 109/L). Flow cytometric analysis of peripheral blood showed granulocyte predominance (98.8% of the events) and no blast population. Bone marrow showed 40-50% cellularity, trilineage hematopoiesis, and myeloid: erythroid ratio 14.3, without dysplasia, increased blasts and metastatic cancer. FISH analysis with extended acute leukemia panel probe showed no malignancy. Cytogenetics showed 46, XY, and PCR studies were negative for mutations of JAK2 V617F and CSF3R. Serum levels of IFN-G, IL-2, IL-4, IL-10, IL-12, and IL-17a were modestly elevated relative to normal values (3.1- to 6.3-fold increase), while level of IL-1a was decreased (0.7 normal level). In contrast, the serum levels of GM-CSF (40.06 pg/mL), G-CSF (1880.63 pg/mL), and IL-6 (361.91 pg/mL) were all markedly elevated above normal by 48.2-fold, 40.1-fold, and 72.4-fold, respectively. When compared with control non-PLR NSCLC tissue, the patient's tumor showed 3-fold overexpression of the G-CSF receptor, 13.3-fold overexpression of the GM-CSF receptor, and 1.7-fold overexpression of the IL-6 receptor. However, neither PLR nor control NSCLC samples showed increased expression of genes for those cytokines (Table 1).Conclusion:In this comprehensive mechanistic analysis of PLR, we have shown that the metastatic NSCLC tumor overexpressed genes for receptors for G-CSF, GM-CSF, and IL-6, but did not overexpress the genes for those cytokines. Significantly elevated serum levels of G-CSF, GM-CSF and IL-6, synthesized from non-tumor tissues, caused hyperleukocytosis. We hypothesize that an autocrine positive feedback loop, in which these cytokines led to autostimulation of their respective aberrantly expressed receptors on tumor cells, resulted in tumor proliferation as well as off-target stimulation of granulocytopoiesis and corresponding PLR. DisclosuresNo relevant conflicts of interest to declare.
Read full abstract