If you’ve got a number of products that are using the same cleaning procedure, the representative intermediate or API could be a worst-case product. In defining the “worst case” product you need to take into account difficulty in cleaning, toxicity, and potency. Validation may need to include more that one product in order to fully address “worst case”. It really depends upon what type of product matrix you’ve got running through that piece of equipment.
I’m taking a different approach on this question because I hear that your concern is about using a very expensive product for cleaning and substituting a placebo instead. Why would you consider that wasting product? What you’re trying to do is remove residues of the product from equipment after that product has been processed in the equipment, so you’re not wasting product. You produce a batch of what you consider to be the most difficult material to remove from the equipment, and then you validate that you can remove any residues of that from the equipment surfaces. What’s possible with using a placebo for cleaning validation purposes is that you only prove that you can effectively remove a placebo, not the product.
