Abstract: We compute the dominant term in the expansion in rho = 1 – M-w/m(t) of the unknown next-to-next-to-leading order nonresonant contributions to the e+ e(-) -> W+ W- b (b) over bar total cross section at energies close to the top-antitop threshold. Our analytic result disagrees with a previous calculation by other authors [A. A. Penin and J. H. Piclum, J. High Energy Phys. 01 (2012) 034]. We show that our determination has the correct infrared structure needed to cancel the divergences proportional to the top width arising in the resonant production of the same final state, and we point to a missing contribution in the computation of Penin and Piclum to explain the discrepancy.
