VITREOMACULAR OBSERVATIONS .2. DATA ON THE PATHOGENESIS OF IDIOPATHICMACULAR BREAKS

Background: The pathogenesis of idiopathic macular breaks is still unc ertain. Their formation has been ascribed to anteriorly oriented intra vitreous traction and to shrinkage of the prefoveal cortical vitreous. The validity of both hypotheses is considered in this paper. Methods: In order to clarify the pathogenesis of idiopathic macular breaks 127 consecutive patients had their vitreous examined and photographed wit h the El Bayadi-Kajiura precorneal lens and a slit-lamp microscope. Re sults: A comparison with 127 matched controls demonstrated that the vi treous was significantly more often attached in eves with a macular br eak than in controls (P<0.01). In eyes with a macular break the vitreo us was significantly more often attached in early cases (Gass stage 1) than in Gass stages 3 and 4 (P<0.01). Still photographs and observati on of the movements of the operculum demonstrated that, in some cases of stage 3 and also in stage 4, it moved inside the partially liquefie d posterior vitreous, anteriorly to the retinal surface and frequently without evidence of posterior vitreous detachment over the macular ar ea. The following anatomical features characterize the vitreomacular a rea extremely thin hyaloid membrane (<100 mu m) and inner limiting lam ina (10 nm) that adhere strongly to each other and to the underlying M ueller cells. There is no evidence that these structures can shrink se lectively to cause a macular break. The premacular vitreous gel contai ns collagen fibers that attach posteriorly to the macula and anteriorl y to the vitreous base. Conclusions: Our working hypothesis is that wh en detachment of the posterior vitreous is abnormally delayed, anterop osterior traction collagen fibres may pull a foveal operculum off the retina. Our observations make this hypothesis attractive. However, the generally accepted hypothesis of Johnson and Gass cannot be entirely dismissed. In reality, since the two hypotheses are not mutually exclu sive, they may both the partially correct.