Anodic alumina membranes have straight pores and a very uniform pore size distribution. Data on the permeance for the flow of isobutane vapors through anodic alumina membranes with pore diameters between 20 nm and 90 nm has been reported recently [Petukhov et al., J. Phys. Chem. C 120, 10982-10990, 2016]. For the upstream pressure approaching the saturation pressure, a sudden increase of the permeance was observed. Taking into account capillary condensation and assuming isothermal conditions within the entire flow field, the permeance data was used to compute the radii of curvature of the menisci separating the liquid condensate from the vapor. In the present work, the radius of curvature of the meniscus within the membrane is assumed to be known and kept fixed. From that, the expected permeance is computed employing three different descriptions. These are (i) a well known isothermal description, as also used by Petukhov et al. (2016), and two non-isothermal descriptions for which the energy balance is taken into account and (ii) an adiabatic or (iii) a diabatic boundary condition is applied at the downstream front of the membrane. These two latter descriptions correspond to no heat flux or arbitrarily large heat flux from downstreams to the downstream front of the membrane, respectively. The predicted permeances depend significantly on the chosen description. A comparison with experimental data indicates that a description with heat flux towards the downstream front of the membrane could agree best with data. However, the conclusion is that a determination of the radius of curvature of the meniscus within the membrane from permeance data alone is uncertain. Accurate temperature measurements at the upstream and downstream fronts of the membrane would increase the reliability of the determination of the radius of curvature of the meniscus within the membrane.