1. Once opened the left ventricle, boundaries between normal and fibrotic tissue, if present, are identified; in case of akinetic, non-fibrotic area, thinning of the area and information derived from echocardiographical and/or magnetic resonance studies are used to identify the limits from normal myocardium.

2. The Fontan purse-string is not used, to avoid a circular disposition of the border zone between normal and fibrotic muscle that alters the spatial geometry of myocardial walls and forces myocardial fibers to a non-physiologic orientation; in this way, we let the suture to redirect fibers, without any pre-conditioned deformation.

3. The limit of the fibrosis near the apex is then identified: this is the distal end of the patch and will form the new apex; it corresponds to the angle between the end of necrosis and the septum and is defined as "Apex stitch" (Figure 1).

4. The limit of fibrotic tissue in the interventricular septum (near the aortic valve) is identified: this is the upper end of the patch and is defined as "High stitch" (Figure 2).

5. The distance between these two points (the high and the apex stitches) along the interventricular septum is the length (major axis) of the patch that is then tailored according to this measure with arrow-shaped ends (Figure 3).

6. The upper end of the patch is secured to the High stitch by a 4.0 polypropylene monofilament (Prolene, Ethicon), tied at its half length with few knots (Figure 2).

7. All the septal rim of the patch is then sutured along the basal border of the fibrotic interventricular septum, from the high to the apex stitch, with a continuous running suture using the 4.0 polypropylene monofilament: in this suture there is almost a patch-tissue equivalence, that is the length of the patch and the length of the myocardial wall are quite equal (Figure 4).

8. The lateral rim of the patch is then secured to the lateral border zone between fibrotic and normal tissue of the ventricular wall with the other half of 4.0 monofilament: this line is widened by the dilation and the curvature of the left ventricle, so the length of the patch is smaller than the length of the ventricular wall (patch-tissue mismatch) (Figure 5A, B): surgeon must fit the length of lateral wall to the length of the patch, shrinking in this way the ventricular wall to a shorter line, that is, forcing the orientation of residual myocardium (and myofibers' lamina) towards a tighter disposition, compensating the dilation and spacing due to akinetic/dyskinetic portion of the wall.

9. The two sutures end at the Apex stitch and are tied together (Figure 6).

10. Ventriculotomy was closed by overlapping the free edges in a vest-over-pants fashion, in order to occlude the excluded chamber, with or without a felt strip to reinforce the suture (Figure 7).

• The first running suture ("septal") has quite the same length as the patch: this part of the ventricular wall usually becomes thinner rather than dilated, due to the anatomy of the septum, embedded between the two ventricles, with a dual layers structure vascularized by two coronary systems; in some cases is possible to compensate a greater dilation of this part with the suture, stretching the residual myocardium to a shorter patch, but in most of the cases there is a patch-tissue equivalence.

• Position of the new apex is often determined by the end of the necrotic tissue; otherwise, its relocation should be guided by normal geometry; normal values of diastolic left ventricular length are 80 ± 9 mm 25, so this should be the maximum value of the new chamber, adding patch length to the length of the normal upper part of interventricular septum (from aortic valve to akinetic/fibrotic septum); new apex should also be near the level of right ventricular apex, as is in the normal heart.

• Given that usually we can reach the upper septum at 2–3 cm from the aortic valve, it follows that the length of the patch should be 5–6 cm in most cases; the end of right ventricle is another reference point to set the new apex.

• The second suture ("lateral") forces the lateral wall to fit to the patch, stretching myocardial fibers towards a shorter line, thus realigning them in a narrower setting, representing the key characteristic of this technique. For this reason the lateral sutured myocardial border appears "goffered" at the end of the procedure, to compensate the discrepancy between patch and myocardial suture rims (Figure 7).

• Converting the setting to a geometrical figure, as previously described 26, we have a circular segment in which the chord is the basal line of the fibrotic interventricular septum and the arc is the semicircular line of lateral border zone between fibrotic and normal tissue of the ventricular wall; the surgical suture approaches the arc to the chord, reducing the arc's length of about one half.

• The gross realignment of residual myocardial fibers at the border zone in order to have a recovery in their oblique spatial orientation can only be obtained by a narrow (small minor axis) patch, which forces the realignment, while a wider (great minor axis) patch tends to maintain fibers direction in the setting of a dilated left ventricle, eliminating the length compensation induced by this technique.

• The use of a patch is of primary importance for fibers' realignment: despite this technique could seem like a linear suture, only the use of the patch makes to grade the stretching of the two sutures possible, differentiating the fitting of the two rims (septal and lateral) to the relative border of the patch; otherwise, it would be impossible to adjust different fibers' orientation along inferior septum (less dilated = fibers less divergent) and lateral wall (more dilated = fibers more divergent).

• The key features of this surgical technique are the small short axis of the patch and the asymmetrical suture at the two sides of the patch: the first one approaches residual myocardium as close as possible; the second differentiates the reconstruction of the dilated border zones to re-establish a physiologic contiguity of myocardial lamina.

Twenty-one patients were operated on with this technique. Mean age was 60.8 ± 12.5 years. Eleven of them were male. Eight patients had a mitral regurgitation ≥ grade 3 and were contemporarily treated with mitral ring anuloplasty. All patients received complete coronary revascularization. Patch dimensions were 1.0 ± 0.3 × 5.9 ± 1.0 (short × long axis, cm). No hospital deaths were reported. At a mean follow-up time of 1.95 ± 0.96 years, all patients were alive and in functional NYHA class = 1.3 ± 0.5. Two-dimensional speckle tracking imaging, a reliable ecocardiographical tool for rotation movements of the heart, was used to assess left ventricular torsion in presence of a good acoustic window (16 out of 21 patients). Echocardiographical data are reported in Table 1.

A 52-years-old male patient, ZD, had the acute episode of myocardial infarction on August 2001, treated with thrombolysis. On September 2001 he performed a coronary angiography showing critical stenoses of left anterior descending (LAD) and right coronary arteries. He then underwent off-pump surgical revascularization by means of three coronary grafts (diagonal, anterior descending and right coronary arteries). The left ventricular function was already compromised (EF = 30%). Postoperative course was uneventful. Since 2004 he suffered from several episodes of heart failure. On July 2004, coronary re-angiography showed a mild stenosis of the graft on LAD (50%) and left ventricular dysfunction (EF = 28%) with anterolateral aneurysm. Grade 4 mitral regurgitation was evident. Ventriculography showed a completely upset LV geometry with consequent loss of ventricular apex and torsion movement (Additional file 1). On May 2005, the patient was reoperated (remodeling time from acute infarction to surgery = 49.7 months). Endoventricular reshaping was performed as described. Patch dimensions were 1 × 6 cm. A 30-mm rigid ring mitral anuloplasty was associated. Left anterior descending artery was grafted by a saphenous vein. Once weaned from extracorporeal circulation, left ventricle showed the renewal of apical rotation and, as a consequence, of left ventricular torsion, as evident in Additional file 2. Echocardiographical data are reported in Table 2. Patient was discharged to a rehabilitation hospital eight days after the operation and is now in first functional NYHA class. A 3D full volume echocardiography was performed at late follow-up (19 months), showing a still present apical rotation at several apical slices (Additional file 3).

Due to a poor echocardiographical window, this patient was not studied by 2D speckle tracking echocardiography. Some of patients studied by this method were reported as single clinical cases 2728, demonstrating the renewal and the persistence during time of apical rotation and LV torsion.

The left ventricle is characterized by a unique myocardial tissue structure 2930. The necrosis induced by myocardial infarction alters the myocardial continuum, forcing the whole ventricle to work at a lower efficiency level. Necrotic area spaces out the residual normal tissue and negative remodeling following the acute episode 31 completely alters ventricular geometry and mechanic. Nevertheless, it is demonstrated from in vivo studies that myofibers orientation in the thickness of normal myocardium is not changed after myocardial infarction 32 and that transmural courses of fiber orientation angles near infarct zones were similar to those of normal myocardium 33. This means that infarct site demarcates at the end of necrotic process, with an extension related to the tributary area of the occluded coronary vessel and to the wavefront phenomenon of ischemic cell death 34: definitive boundaries are characterized by a still normal myocardium interlaced to the necrotic region. This is shown also in gross histological samples 35: a normally structured myocardium ends without graduation where starts the fibrotic infarcted tissue. It is also evident that alterations occurring in myocardium adjacent to an infarction consists with myocytes elongation and myofiber rearrangement (slippage), as well as changes in orientation of the laminar structure of the ventricular wall 36.

The key point to understand how myocardial structure is altered by coronary vessel occlusion is the relative position of epicardial coronary vessel and fiber orientation.

Coronary arteries have an epicardial course not to be squeezed during systole, while fiber orientation is changing throughout myocardium thickness, from epicardium to endocardium, following a continuous angular gradient which is crossed more or less perpendicularly by vessels direction (Figure 8A). The transmural necrosis induced by coronary vessel occlusion concerns a sector of myocardium made by several strata of differently oriented myocardial fibers and ends where the territory of another coronary vessel begins (Figure 8A, shaded area). At the border zone, still normal myocardium with still normal fiber orientation interlaces with the necrotic tissue that lost its 3D fiber structure. Necrotic tissue discontinues continuity and contiguity of myocardial fibers, although they remain normally oriented in the thickness of normal myocardium (Figure 8B, correspondence of points "a" and "a¹", "b" and "b¹" etc).

The infarcted area alters regional and then global function of the left ventricle, expands and dilates the ventricular chamber and dislocates adjacent, still anatomically normal, non-infarcted myocardium. Consequently, fibers in the normal myocardium are displaced by the necrotic, thin, akinetic/dyskinetic portion of ventricular wall and (although normally oriented inside the residual wall) their orientation and spatial disposition is completely upset.

On the other hand, it has been demonstrated that surgical intervention could effectively alter (positively or negatively) the helix angles in regions adjacent to repair 32.

Surgical restoration replaces the infarct scar, and tries to modify ventricular geometry in such a way as to recreate a nearly normal ventricle. The described procedure reduces the distance from the contractile myocardium thanks to the narrow patch that absorbs the stress of the suture, acts as a support for the suture lines and guides the restoration of myocardial wall contiguity and fibers' orientation. A narrow patch also helps to rebuild near-normal ventricular volumes because it does not add any volume itself. Comparing the left ventricle to a prolate ellipsoid, small variations in the minor axis correspond to great changes in ventricular volume (1 cm less in the minor axis produces a reduction in volume of about 30 ml). In this way, major determinants of normal physiology are restored: elliptical shape, near-normal volume, small residual akinesia, contiguity of contractile myocardium, spatial geometry and fibers orientation. Complete revascularization and, when required, the competence of mitral valve crown the physiologic restoration.

Aiming the ventricular restoration at the realignment of residual fibers could overcome the endless contest about akinetic and dyskinetic ventricular walls. We should assume that akinesia and dyskinesia are different stages/states of the same disease and both result in a non-contractile zone of the myocardium. From a strictly functional point of view, both interrupt myocardial continuum, lead to fibers disarrangement and ventricular remodeling, so both are responsible of the same pathophysiology. We should then reconsider that the border zone of myocardial infarction is the limit between fibers that lost and fibers that preserved their anatomical and physiological arrangement. This is the consequence of the ischemic disease, and in case of akinetic areas, in which the border zone is not marked by fibrotic degeneration, it is visually less evident, but functionally has the same effect.

Finally, also a non-trasmural infarction can damage myocardial continuum in such a way to alter its functionality, though zones of normal fibers remain in the thickness of the infarcted region. These residual fiber islands, embedded in the fibrotic tissue or simply adjacent to it, can not work properly because they are blocked by the non-contractile cells, and the physiologic fiber interlacement that produces the contraction is lost: they do not act on normal adjacent fibers, propagating the contraction, but their action extinguishes against non-contractile tissue. In these cases, function of myocardial wall is the result of the delicate balance between contractile and non-contractile fibers.

Since 1911 37 left ventricular torsion was recognized to be essential for ventricular ejection and many studies confirm its importance as an index of normal ventricular function in physiologic and pathologic settings 38394041. The twisting motion of the left ventricle about its long axis results from the contraction of the opposite, obliquely oriented epicardial and endocardial fibers that produce the contemporary clockwise rotation of the base and counterclockwise rotation of the apex, obtaining the torsion of the heart along its long axis. Loss of myocardial structure in anterolateral region, chamber dilation, volume overload and loss of the ventricular apex, lead to the loss of apical rotation and, consequently, of left ventricular torsion.

The evidence of the recovery of left ventricular torsion is an indirect but striking demonstration that fibers should be reset to a physiologic orientation with the described endoventricular reshaping. The evidence of torsion recovery corresponds to a more physiologic rearrangement of myocardial fibers. We observed the renewal of ventricular torsion ever since the end of the surgical procedure in the operating room, immediately after the surgical correction: so it must be due to a mechanic correction of fibers orientation and not only to a progressive positive remodeling after surgery.

At present, we have only indirect, "gross" evidence of fiber realignment, but we can not correct fiber disposition in order to obtain a "histological" realignment. Fast improvement in medical imaging techniques able to visualize fiber orientation (diffusion tensor magnetic resonance, limited until now by heart movement) will be very useful in this topic and will probably lead to a preoperative assessment full of key data for the surgeon. Some advanced technology laboratories 4243 are already studying the use of integrated imaging techniques for accurate planning of surgical procedures (for example, tumor removal in the brain with 3D visualization of cerebral fibers, tissue and structures). It would be reasonable to foresee a 3D surgical planning of ventricular reshaping with reference points for fiber disposition that will guide the surgeon during the reconstruction. A so called "fiber-based" therapy could open unexpected improvements in the surgical treatment of the failing heart. For example, the Batista operation and the surgical reshaping in the setting of dilated cardiomyopathy had not the expected diffusion, perhaps because volume reduction did not take care of fiber alignment at the suture site, but caused itself a fiber disarrangement, suturing together non-contiguous parts of ventricular walls with different fiber disposition. A fiber-based physiological view could also help us in the evaluation and/or optimization of medical therapies, mainly in case of new drugs or of non-responder patients.

Finally, basic science implications of fiber-based left ventricular function could let us use an old knowledge that was forgotten until now, due to the imaging limitations and to the complex 3D anatomy of myocardium, but that could represent the basis for a new approach to the failing heart.