Yes, it is possible for topical Tacrolimus to cause systemic toxicity, in certain situations.
Evidence from a case series
This paper is a case report of two patients who were prescribed topical Tacrolimus (and other therapy) for post-transplant cutaneous graft versus host disease (GVHD).
Their Tacrolimus trough levels (a way of measuring the blood level of the drug just prior to the next dose) rose sharply from 7.1ng/ml to 22.1ng/ml. Treatment with topical Tacrolimus was abandoned due to this unpredictable systemic absorption. There was no evidence of harm resulting.
The authors note the following:
These case reports suggest that substantial use of topical tacrolimus with occlusive dressings in patients with cutaneous GVHD may contribute to increased systemic absorption resulting in toxic tacrolimus levels.
However, the are some factors that would predispose these patients to increased absorption; the use of occlusive dressings, skin barrier dysfunction (due to GVHD) etc.
Based on the findings from our two patients as well as published case reports, systemic absorption appears to increase with greater skin permeability, skin barrier dysfunction, amount of body surface area applied, and use of occlusive dressings. When one or more of these factors are present, it may be prudent to monitor tacrolimus levels.
About tacrolimus
Tacrolimus is a calcineurin-inhibitor immunosuppressant. It is used systemically to prevent or treat transplant rejection. It is used topically is psoriasis and eczema and other conditions. It is not currently licensed to treat vitiligo in the UK, but it is presumably being used as vitiligo is an autoimmune destruction of melanocytes in the skin.
I cannot find any evidence so far on predicting what level of exposure to topical tacrolimus is likely to cause toxicity. It is recommended to monitor the levels if the risk factors above apply. In vitiligo the skin is intact and not inflamed, so toxicity in this setting would be less likely, especially if occlusive dressings are not being used.
Source: Olsen et al. Pharmacotherapy, 2014. DOI: 10.1002/phar.1418