A short and troubled entry (taken over unchanged by ps.-Zonaras); perhaps the work of more than one hand.
The headword as transmitted, from the verb e)pi/stamai (cf. epsilon 2606), is e)pi/sth, but in fact it should almost certainly have an iota subscript, i.e. the second person singular form e)pi/sth| (and as such is probably one of the many instances of this in the Septuagint, from Genesis 47.5 onwards: so Latte on Hesychius epsilon5242). The connection between e)pi/sth| and e)pi/stasai [n. 2 below] is a commonplace in grammarians and lexica, including Choeroboscus: see ps.-Herodian, On verbs ending in -mi 839.45-840.5, and elsewhere; Lexica Segueriana 93.8; Etymologicum Magnum 290.33-41.
[1] Correct elsewhere in the Suda (alpha 1264, epsilon 1031, epsilon 2788, pi 1586, pi 1587, pi 2608, pi 3126), but not here: see above. If, as seems likely, the gloss has intruded here from another entry, the obvious one is epsilon 2623.
[2] See LSJ s.v. e)pi/stamai (web address 1): the second person singular present is attested as e)pi/stasai in Attic (e)pi/sta| in poetic Doric, e)pi/sth| in Theognis, and e)pi/steai in Herodotus).
[3] Seemingly an alternative interpretation of the headword as an imperative, with e)piti/qei intended as an analogous form rather than a substantive gloss.