The perfect middle/passive of kaqi/zw is attested in a few Byzantine authors, just before Suda was compiled: Methodius I (C9) uses kekaqi/sqai in the Vita of St Euthymius of Sardis (37.777, 37.782, as well as keka/qiso 14.279), and the Vita of SS David, Symeon and George (also C9) uses e)keka/qisto (14.29).
Nevertheless, it is more likely the lexicographer is copying the instance of kekaqi/sqai in Herodian's Partitions, where he points out that the perfect middle-passive infinitive of kaqi/zw (if there is one) is spelled with iota, while the infinitive kaqh=sqai (perfect with present sense, formed from ka/qhmai) is spelled with eta (see kappa 83, kappa 88). The Suda's gloss makes explicit what Herodian leaves implicit: the iota in kaqi/zw comes from i(/zw.
For kaqi/zw, see also kappa 98.