Manual integration error in Masshunter quant 10.1

I have a peak cluster that I need to integrate as a single peak (unresolved isomers) in GCMS SIM data. I can manually integrate the cluster for the quant ion, the qualifier often (maybe 1 out of 5 samples) will not manually integrate the peak. It's an ISTD qualifier ion, and I've noticed that upon automatic integration, one portion of the peak cluster is integrated, but it disappears once I try to adjust the baseline. I can move the base back to the approximate location where the manual integration found the peak, but that's it. I've tried different integrators (Agile2 and Spectrum Summation), tried adjusting the integration parameters in Spectrum Summation, and tried adjusting the left and right RT Delta values, thinking it might be how Masshunter finds the baseline. So far I haven't had much luck, does anyone have suggestions about what else I could try?

Thank you!