Effect of anti–TNF-α on the growth of erythroid committed progenitors in FA patients and controls. (A) Each column represents mean values; error bars indicate the SE of the estimate. We also reported in brackets median, minimum, and maximum values. In the patients, CFU-Es (mean, 34.7; median, 19.0; minimum, 0; maximum, 100) grown after the addition of LD anti–TNF-α (1 μg/mL) significantly outnumbered (P = .043) that grown in the absence of anti–TNF-α (mean, 12.0; median, 9.0; minimum, 0; maximum, 37; left side). A similar significant difference was obtained with HD anti-TNF-α (10 μg/mL; P =.028; diagram not shown). The addition of anti–TNF-α had no effect on the growth of CFU-Es in healthy controls (mean number of colonies without LD anti–TNF-α, 362.2; median, 353.5; minimum, 102; maximum, 522; mean number of colonies with LD anti–TNF-α, 362.0; median, 349.0; minimum, 96; maximum, 525; right side). (B) Each column represents mean values; error bars indicate the SE of the estimate. We also reported in brackets median, minimum, and maximum values. In the patients, BFU-Es grown after the addition of LD anti–TNF-α (mean, 19.3; median, 12.0; minimum, 0; maximum, 45) significantly outnumbered (P =.043) that grown in the absence of anti–TNF-α (mean, 7.3; median, 6.0; minimum, 0; maximum, 24; left side). The same held true, with the identical significance (P = .043), also for HD anti–TNF-α (diagram not shown). The addition of anti–TNF-α had no effect on the growth of BFU-Es in healthy controls (mean number of colonies without LD anti–TNF-α, 343.1; median, 336.5; minimum, 66; maximum, 691; mean number of colonies with LD anti–TNF-α, 350.6; median, 332.5; minimum, 142; maximum, 695; right side). indicates no anti–TNF-α; ▦ with anti–TNF-α (1 μg/mL).