Quite a few papers cross-quoted a concentration of 10mIU/mL. But this concentration exceeds over 2000 times of the human physiological upper limit. The original paper (doi:10.1073/pnas.77.6.3455) that first introduced the 6H medium provided a concentration of 10milliunits/ml. Paradoxically, it also stated that "the final hormone mixture was arrived at by testing various hormones or factors at physiological or nearly physiological concentrations".
The papers published before or near the year of the first 6H medium paper almost universally used "μU/ml" as the unitage of TSH concentration.
Therefore, is there any possibility that the original paper misspelled the unitage which should be microunits/ml(μU/ml) in real?
If the concentration is 10mIU/ml indeed, that requires large amounts of research fund to undertake TSH related research, for the price of TSH is nearly 200 U.S. dollars per 10μg(that's only 60mIU according to Sigma immunopotency data).
I found a reasonable answer from the paper "Humanized Medium (h7H) Allows Long-Term Primary Follicular Thyroid Cultures From Human Normal Thyroid, Benign Neoplasm, and Cancer" (doi: 10.1210/jc.2012-3812).
The affinity of bovine TSH (most likely used in the original article) for the rat TSH receptor is low. Therefore, the concentration of 10 milliunits/ml recommemded by the original paper seems right and not misspelled.
However, bovine TSH has equal potency to human TSH on stimulating the human TSH receptor, which indicates a much lower concentration need and adjustment for human thyroid-derived primary cell cultures.