A remark on the stability of C∞-hypoellipticity under lower-order perturbations