I guess this stongly depends. Are you sure only your metabolites are causing the toxicity? If the drug itself is toxic to some extent then extended (metabolic) stability is of course critical. 

And while it is true that "deuterated" versions of certain drugs are created to reduce the rate of metabolism this is usually not done to avoid toxic metabolites, but rather e.g. to give dosing advantages. In the end the deuterated version will be often metabolized in the same way as the original "hydrogen drug", just slower. This means you will at some point end up with the same toxic metabolites (however probably in a different concentration and with a different dose-time curve). 

I would not recommend deuteration as a strategy to avoid toxic metabolites but rather put in some medchem efforts to avoid the formation of these metabolites e.g. by blocking the site of metabolism or using (bio)isosteric groups.

Enzymes are not capable of distinguishing between hydrogen and deuterium so I am not convinced about any change in toxicity due to stable isotope labelling.

Stable isotope labelling is very expensive and the effect would be minimal to the toxic effect but you would be able to track the kinetics using the isotope dilution effect in vivo if you use mass spectrometry for quantitation.

In no way, you will reduce the toxicity by deuteration. You don't change chemical properties by deuteration, that's why deuterated compounds are used as best internal standards for LC-MS or GC-MS in isotope dilution MS.   

Ya agreed, Thank you Fabian Heider

By deuteration rate of formation of toxic metabolites can be reduced but toxicity remains the same. If the drugs metabolites to alternate pathway then it is possible to reduce the toxic metabolites I think. (Paracetamol metabolism)

The rate of metabolism will not be altered at all for any drugs undergoing enzymatic reactions as enzymes cannot distinguish between the hydrogen and the deuterium atom in a molecule.

All the metabolic products formed in vivo are through enzyme reactions, mostly within cells and the toxic effect of especially paracetamol depends on the rate of conjugate formation (dependent on glucuronic acid and sulphate concentrations and the specific activity of the two different conjugation enzymes) and the availability of GSH to mop up the toxic product NAPQI formed by CYP2E1 and CYP3A4 from excessive concentrations of the parent drug in the liver.

Lowering the dose or maintaining high liver concentrations of GSH will avoid the toxicity far better than trying to deuterate the paracetamol.

The cost is also significantly lower.