In Chapter 4 syntheses were described in which ring D was attached to an AB moiety, following which ring C was closed producing the steroid. In theory it should be just as feasible to connect B and D by one of the many routes delineated in Chapter 4, close ring C, then construct ring A by one of several available annotation methods. In practice, however, this approach has enjoyed only very limited success.One of the few successful syntheses began with the triketone 1 (Scheme 11-1), which was obtained by Michael addition of 1,3-cyclohexanedione to methyl vinyl ketone [1], The next step was to add carbons 11 and 12 by a Grignard reaction with vinylmagnesium bromide. To protect oxo groups at C-3 and C-5 (steroid numbering) from attack by the Grignard reagents, 1 was converted to the corresponding chromone (temporary closure of ring A) by heating in acetic anhydride. The Grignard reaction and subsequent hydrolysis gave 2 in 24% yield [2,3]. Michael addition of 2-methyl-l,3-cyclohexanedione gave BD intermediate 3, which was cyclized (low overall yield) to the D-homosteroid 4 [2,3]. The latter was identical with the product synthesized by an AB + D-+ABD-+ABCD route [4]. In an effort to obtain a 10-methyl homolog, 5 [5-7], was condensed with 2-methyl-,3-cyclohexanedione and the BD intermediate was cyclized to give 6 in moderate yield [6][7][8], All attempts to construct ring A onto 6, however, were unsuccessful [6,7].