Fig. 3.

CRC is associated with an increase in circulating CD45RO⁺CD30⁺OX40⁺ Tregs. a Representative dot plots showing co-expression of CD30 and OX40 by CD45RO⁺ Tregs. Gates were drawn based on isotype control antibodies. b Treg cells imaged by the DEPArray™ NxT digital sorting system. Each row represents a single cell imaged in brightfield and four fluorescent channels. c–e Comparison of frequencies of circulating CD30⁺OX40⁺, CD30^–OX40⁺, or CD30⁺OX40⁻CD45RO⁺ Tregs between healthy subjects (n = 14) and CRC patients (n = 25). Means ± SD are shown. Statistical analysis was performed using the two-tailed unpaired t test with Welch’s correction. *** denotes P < 0.001. f ROC analysis of % circulating CD45RO⁺CD30⁺OX40⁺ Tregs in discriminating healthy subjects and CRC patients, with AUC = 0.92, sensitivity = 88%, specificity = 100%, positive predictive value = 100%, negative predictive value = 82.35%, accuracy = 92.3% at a trade-off value of 3.44%. g ROC analysis of % circulating CD45RO⁺CD30⁻OX40⁺ Tregs in discriminating healthy subjects and CRC patients, with AUC = 0.90, sensitivity = 85.7%, specificity = 92%, positive predictive value = 92%, negative predictive value = 85.7%, accuracy = 89.7% at a trade-off value of 3.37%. h ROC analysis of % circulating CD45RO⁺OX40⁺ Tregs in discriminating healthy subjects and CRC patients, with AUC = 0.91, sensitivity = 84%, specificity = 78%, positive predictive value = 87.5%, negative predictive value = 73.3%, accuracy = 82% at a trade-off value of 8.87%