Macrophage profile in primary versus secondary liver tumors

Macrophages are important components of the tumor-associated infiltrate and are qualified as one of the major players of the cancer-related inflammation. It was shown that tumor cells can either stimulate or mediate apoptosis of tumor-associated macrophages (TAMs). To date, there is no general agreement regarding the influence of TAMs and their numbers on the progression of hepatocellular carcinoma (HCC) and hepatic metastases (HM). To analyze the presence of TAMs and compare their numbers in intratumoral (IT) and peritumoral (PT) areas with the clinical outcome of HCC and HM patients. Biopsies from 35 HCC and 39 HM cases were analyzed. Clinical and follow-up data was enrolled for each case; the colorectal cancer was the origin of 26 HM patients. TAMs were identified by immunohistochemistry using anti-CD68 monoclonal antibody. The quantitative assessment was performed by determining the mean number of CD68-positive cells in IT and PT areas in HCC and HM. Two threshold methods were applied: threshold 1 (T1) was calculated with the use of (-log) Cox method; threshold 2 (T2) was considered as 1/3 TAMs number of group's mean. For statistical analyses Mann-Whitney U-test, Spearman's correlation, Cox proportional hazard and Kaplan-Meier tests were applied. To date, 36.12% HCC and 27.78% HM patients were alive, median survival was 5 and 17 months for HCC and HM, respectively (P = 0.05). We found significant two-fold decrease of TAMs numbers between IT vs. PT territories in both HCC and HM. A positive correlation between numbers of PT and IT TAMs was observed in HM group (rs = 0.48, P < 0.05) but not in HCC. The number of TAMs was not associated with any studied clinical factor. Univariate Cox regression analysis showed that tumor stage ≤ II (P = 0.01) and increased number of PT TAMs (P = 0.06, only when T2 value was applied) were associated with favorable prognosis in HCC (HR = 2.614 and 2.457, respectively). Univariate and multivariate Cox analyses in HM revealed favorable prognosis for histological grade ≤ G2 and one lobe tumors (P = 0.021 and 0.045; HR = 0.395 and 0.438, respectively). Survival analysis retained the impact of increased TAMs numbers in peritumoral areas (P = 0.03), tumor stages in HCC (P = 0.007), lobes' number (P = 0.007) and histological grade (P = 0.005) on HM patients' outcome. In HCC and HM the low number of TAMs in intratumoral areas was related to the tumor cell microenvironment. The increased peritumoral TAMs number in primary liver tumors was associated with better prognosis.