No abstract

SummaryThe relationship between interleukin 6 (IL-6) levels and clinical parameters was studied in 25 patients with malignant pleural mesothelioma. The serum levels of IL-6, C-reactive protein, a,-acid glycoprotein and fibrinogen were significantly higher in mesothelioma than in lung adenocarcinoma with cytology-positive pleural effusion. Serum IL-6 levels correlated with the levels of the acute-phase proteins. We demonstrated a high incidence of thrombocytosis (48%) and a significant correlation between platelet count and the serum IL-6 level. The level of IL-6 in the pleural fluid of patients with mesothelioma was significantly higher than in the pleural fluid of patients with adenocarcinoma, and was about 60-1400 times higher than in the serum. However, even higher levels of IL-6 in the pleural fluid and of thrombocytosis were found in patients with tuberculous pleurisy. These results indicate that large amounts of IL-6 from the pleural fluid of patients with mesothelioma leak into the systemic circulation and induce clinical inflammatory reactions. These profiles are not specific to mesothelioma as similar profiles are found in patients with tuberculous pleurisy. However, the detection of a markedly increased level of IL-6 in pleural fluid argues against a diagnosis of adenocarcinoma.

No abstract

At the request of the corresponding author, the JCI is retracting this article. Following an institutional investigative review of errors in the article, it was determined that the data reported in Figure 1B were an analysis of E2F1 expression rather than CERS6 expression, as originally reported. The correct data do not show a difference in CERS6 expression with invasion status in human lung adenocarcinomas. In addition, errors were noted in Supplemental Figure 2, Supplemental Figure 3B, and Supplemental Table 1. The institutional review found no evidence of intention to falsify results.

Original citation: J Clin Invest. 2016;126(1):254-265. https://doi.org/10.1172/JCI79775. Citation for this expression of concern: J Clin Invest. 2019;129(8):3464. https://doi.org/10.1172/JCI131245. The corresponding author recently notified the JCI that the patient data presented in Figure 1B were not correct. Analysis of the correct patient data set does not show a significant difference in CERS6 expression in human lung adenocarcinomas with positive invasive growth (definite) compared with those with negligible invasive growth or without invasive growth (focal/none). The Editors have requested an institutional investigation into this matter, and we will inform our readers of the outcome when the investigation is complete. Original citation: J Clin Invest. 2016;126(8):3036-3052. https://doi.org/10.1172/JCI83416.Citation for this expression of concern: J Clin Invest. 2019;129(8):3464. https://doi.org/10.1172/JCI131246. A reader recently alerted the Journal that two images in this JCI article appear similar to images subsequently published in a Neuro-Oncology paper from the same lab as unique samples (1). Specifically, in Figure 9D of the JCI paper, the image for IL13Rα2 staining for the HER2 CAR sample appears to be similar to the image for EphA2 staining of a nontransduced T cell-treated sample published in Figure 6A of the Neuro-Oncology paper. In addition, in Figure 9D of the JCI paper, the image for IL13Rα2 staining for the tumor sample appears to be similar to the image for HER2 staining of a nontransduced T cell-treated sample in Figure 6B of the Neuro-Oncology paper. An institutional investigation into this matter is ongoing, and we will inform our readers of the outcome when the investigation is complete.