Pii: S1010-7940(01)00607-8

The raison d'eÃtre of total arterial revascularisation (TAR), namely using arterial grafts to non left anterior descending coronary artery (LAD) territories, is the achievement of the patency rates obtained by grafting the pedicled internal mammary artery (PIMA) to the LAD. Grafting the PIMA to the LAD is the single most important technical factor, in experienced hands, determining the long-term success of the coronary bypass operation (CABG). Moreover, the excellent long term patency rates of the pedicled IMA-LAD graft (.90% at 10 years and beyond) have been readily reproducible, by several generations of surgeons, world-wide. This graft therefore, I believe, is sacrosanct. The paper by Royse and colleagues [1], on TAR and Y grafting, is a landmark paper. This paper provides unequivocal, angiographic proof that the cornerstone foundation of those advocating TAR is not only unsound, but also critically, that the practice of the Y graft technique is untenable. In Royse's paper, at a mean angiographic follow up of 1 year, arterial grafts to the right coronary artery territory (RCA) had statistically signi®cant lower patency rates (81.9%) when compared to patients who received a pedicled and importantly non composite (i.e. no Y graft attached) IMA graft to the LAD (100% patent). Furthermore, Royse states that there were no statistically signi®cant differences between the patency rates of the saphenous vein grafts (SVG) used when compared to all arterial grafts (90.5% vs. 91.1% respectively, P ˆ 0:781). Moreover, the patency rates of non-pedicled (free) arterial conduits were signi®cantly inferior to that of the pedicled, non-composite IMALAD graft. It is therefore irrefutable, by this data, that using free arterial conduits produces inferior results than that achieved by the non composite PIMA-LAD graft. Indeed, these results are no better (creditably, Royse accepts that their reported patency rates are lower than previous reports) than the results of similar angiographic studies involving the humble SVG [2]. However, the use of the aortocoronary SVG does not have a profound, adverse and early effect on the patency of the PIMA-LAD graft, but use of the Y graft technique however does! It is this aspect of Royse's paper that I take strong issue with. In Royse's hands (coming from a group of determined and experienced surgical advocates of TAR and Y grafting) when a Y graft (using either radial or free IMA conduits) is anastomosed to a PIMA used to graft the LAD, the patency of the IMA-LAD segment is adversely affected. The patency of the pedicled, non composite (no Y graft) IMA-LAD graft was 100% (80 patients) compared to 89.4% (84 patients) when the IMA-LAD pedicle was used as a composite conduit (Y graft). It is clearly seen in Royse's Table 4, that this technique achieved statistical signi®cance (P ˆ 0:002) even when corrected for multiple hypotheses testing according to the Ryan±Holm step-down Bonferroni correction (0.042). At a mean follow up of 1 year it is unacceptable to have one in ten of the IMA-LAD grafts occluded from the application of a Y graft, which confers no angiographically evident bene®t compared to either a free arterial aortocoronary graft or a SVG. Royse explains that this difference in patency rates is because the degree of the native LAD stenosis was statistically signi®cantly less for the composite IMA Y grafts compared to the native LAD stenoses where pedicled non-composite IMA grafts were used (79.8% vs. 86.3% respectively, P ˆ 0:034). The relevance of such an explanation is, at best, obscure. I believe it is axiomatic that an 80% stenosis is haemodynamically signi®cant. Hence, I ask: what has been the management and outcome of these patients with blocked IMA grafts and with an underlying LAD stenosis of 80%? How does Royse justify using the Y graft technique in 90± 95% of patients when 11% of these IMA grafts are blocked at 1 year? In patients with native LAD stenosis less than 80% their own data does not support this practice. Royse states that aThe concept that the `LIMA to LAD' graft is the key to coronary surgery may be challenged by parts of these datao. This statement aimed at the heart of the practice of European Journal of Cardio-thoracic Surgery 19 (2001) 541±542