This article is mentioned in:

Oncol Lett 14: [Related article:] 776-786, 2017; DOI: 10.3892/ol.2017.6254

Following the publication of this article, we have realized that certain of the percentages and data were calculated and/or stated incorrectly. The correct percentages and data (highlighted in bold) should have been presented as follows. In the Materials and methods section on p. 777, first paragraph, line 13, the sentence here should have read: “Of the 219 patients, 132 (60.3%), 40 (18.3%) and 43 (19.6%) patients were pathologically diagnosed…”. Secondly, in the Results section on p. 781, the final paragraph, line 9, the sentence here should have read: “Although 27.3% (9/33) and 9.1% (3/33) of patients with borderline malignancy…”. Additionally, on p. 780, the right-hand column, lines 2 and 3, the text should have stated that a serum KL-6 level <515 U/ml was considered low (median, 515 U/ml).

Furthermore, we noted that, in Table I, the P-value for the difference in tumor size (<100 vs. ≥100 mm) should have been written as P=0.3934 (not 0.0034), and in the Results section on p. 779, the sentence commencing on line 2 should have read: “Therefore, the levels of serum KL-6 were compared among localized ovarian tumors, including benign tumors, borderline tumors, ovarian metastases and stage I/II EOCs.” Finally, in the Discussion, on p. 782, line 3, “tumor size” should have been omitted from this sentence; thus, the sentence should have read: “Elevated serum KL-6 levels correlated with FIGO stage, histological type, tumor grade, lymph node metastasis and residual tumor size.”

These errors did not affect the overall conclusions of the work reported in this study. We sincerely apologize for these mistakes, and regret any inconvenience these have caused.