I've always done this in launch mode, just because it's quickest for me, but you can probably do something similar in the gui.
In the past I've used a Voigt plus a simple asymmetry term to fit the LaB6 standard. I've then fixed those quantities and just used size/strain terms to describe the sample. For my samples that gave a great fit with just 4 parameters refined. If you're not interested in size/strain you can probably get an equally good fit by just using the Voigt and refining pr1-4 and asymmetry on your sample. All this information can go in the "str" section of an INP file.
'Sample contribution
LVol_FWHM_CS_G_L(1, 147.14258, 0.89, 144.91824, csgcr, 170.05173, cslcr, 2019.50877)
e0_from_Strain( 0.00085, sgcr, 0.37408, slcr, 0.03200)
'John's LaB6 peak shape 3/10/18
prm !pr1 0.00105_0.00006 min 1e-9 max 0.02 val_on_continue = Rand(0.00001,0.01); del 0.0001
prm !pr2 0.00083_0.00001 min 1e-9 max 0.02 val_on_continue = Rand(0.00001,0.01); del 0.0001
prm !pr3 0.01992_0.00008 min 1e-9 max 0.02 val_on_continue = Rand(0.00001,0.01); del 0.0001
prm !pr4 0.00009_0.00001 min 1e-9 max 0.02 val_on_continue = Rand(0.00001,0.01); del 0.0001
lor_fwhm = pr1 Tan(Th) + pr2/Cos(Th) ;
gauss_fwhm = pr3 Tan(Th) + pr4/Cos(Th) ;
Simple_Axial_Model(!axial, 0.65560_0.00125)
The guys on the beamline may have a more sophisticated approach (e.g. fundamental parameters) but this has always been good enough for me.
You can probably also use a TCHz function. I think in practice you just have to be a bit careful as some of the values end up being very small due to the sharp peak shape. You may need to start with smaller values than usual for it to converge.
There is a Y2O3 id31 (not quite id22) example at: https://topas.webspace.durham.ac.uk/tutorial_synchy2o3/