What does the CCFR measurement of the Gross-Llewelyn Smith sum rule tell us?

Recently the CCFR Collaboration reported the measurement of the Gross-Llewellyn Smith sum rule ∫10dFvp+vp3 (x, Q = 3 GeV2) = 2.50 ± 0.018 stat(x, Q2 = 3 GeV2) = 2.50 ± 0.018 (stat). Subsequently Kataev and Sidorov analyzed the Q2-dependence of this sum rule and pointed out a discrepancy between the results obtained via integration of the NLO fits to xF3(x, Q2) and the purely perturbative prediction. We suggest an explanation of this disagreement and show that the result of the CCFR measurement of the GLS sum rule integral can be determined from the knowledge of the value of ΛMS only.