Volume 93 Number 5 May 1987 of natural versus prosthetic aortic valve leaflets in vivo. Ann Thorac Surg 1980;30:230-9. 4. Thubrikar M, Aouad J, Nolan SP. Comparison of the in vivo and in vitro mechanical properties of aortic valve leaflets. J THORAC CARDIOVASC SURG 1986;92:29-36. Skeletal muscle in the repair of a left ventricular aneurysm To the Editor Magovern and associates' from the Allegheny General Hospital report the use of the latissimus dorsi as a functioning synchronously paced muscle component in the repair of a left ventricular aneurysm. Two aspects of the paper draw our attention: First, we note the improved ventricular function by Doppler evaluation at 6 weeks by the synchronously paced latissimusdorsi segment. Second, why are the optimal physio- logic muscle changes expected at 3 to 4 months? Our results have shown some controversy. We2.3 have previously reported our experience using skeletal muscle as a pump, and we have also studied changes produced on the muscle by long-term pacing. Our results show that skeletal muscleis capable of maintaining acceptable pressures and can be used as a source of energy in assisting cases of left ventricular failure in the first 2 weeks. The histopathologic and histochemical studies of the striat- ed muscle did not reveal any changes at 3 and 5 days. At 7 days there was a clear increase in the reticulum of the muscle fibers, corresponding to granules with intensely positive oxida- tive enzyme reactions. The granules either were dispersed in the intermyofibrillar spaces or formed thick subsarcolemmal deposits. Adenosinetriphosphatase and phosphorylase methods differentiated two fiber types (the smaller ones were type II). At 14days type II fibers predominated and almost all of them were small. The loss of myelinated fibers was evident in the intramuscular nerve branches. There was a fibrous reaction on the muscle surface and a focal increase in the perimysial and endomysial septa. At 26 days the picture was dominated by an active, intense interstitial fibrous reaction that penetrated the endomysial spaces, isolating small bunches of muscle fibers. Finally, we think that the skeletal muscle could not be used for auxiliary power to the ventricle after the second week, as reported by Mannion, Hammond, and Stephenson.' A. Juffe, M.D. A. Thompson. M.D. J. Lloves, M.D. A. Diluch, M.D. Department of Thoracic and Cardiovascular Surgery Cetryc. Policlinico Central de Bernal Buenos Aires, Argentina REFERENCES I. Magovern GJ, Sang BP, Magovern GJ Jr, et al. Latissimus dorsi as a functioning synchronously paced muscle compo- nent in the repair of a left ventricular aneurysm. Ann Thorac Surg 1986;41:116. 2. Juffe A, Castillo-Olivares JL, Zavanella C, Goitti JJ, Figuera D. Uso experimental del musculo esqueletico como ventriculo artificial. Prensa Medica Argent 1974;61:787- 9. 3. Juffe A, Ricoy JR, Marquez J, Castillo-Olivares JL, Letters to the Editor 795 Figuera D. Cardialization: a new source of energy for circulatory assistance. Vase Surg 1978;12:10-7. 4. Mannion JD, Hammond R, Stephenson LW: Hydraulic pouches of canine latissimus dorsi: potential for left ventric- ular assistance. J THORAC CARDIOVASC SURG 1986;91:534- 44. Reply To the Editor We have read with interest the letter by Juffe and associates. First, they ask how the patient described by Magovern and associates could have had an improved ventric- ular ejection fraction after 6 weeks, since their own experi- ments indicate that "skeletal muscle could not be used for auxiliary power to the ventricle after the second week." Carpentier and Chachques ' have also reported an increase in ventricular ejection fraction in a woman 30 days after replacing a portion of the left and right ventricles with a muscle graft stimulated in synchrony with the heart. Our earlier laboratory studies suggest these clinical findings are possible." We did not measure ventricular ejection fractions. However, we demonstrated that muscle grafts used to replace portions of the canine ventricle that are stimulated in synchro- ny with the heart acutely and chronically remained viable and developed active tension during stimulation as measured directly with a strain gauge. Second, Juffe and associates ask why Magovern's group expects optimal physiologic muscle changes to occur at 3 to 4 months in the chronically stimulated muscle patch. From the references supplied by Magovern and co-workers, we would have to ask the same question. On the other hand, we" have shown, as have other investigators." that muscle grafts placed on or in the heart will develop a collateral blood supply with the myocardium as early as 1 month. In addition, we'?" have studied canine muscles that have been stimulated in vivo from a few weeks to as long as 1 year. The adaptive muscle type transformation that occurs appears to be complete in most respects for a given pattern of electrical stimulation by 6 to 8 weeks. Recently, wei? reported in this JOURNAL on the construction of ventricles fashioned from canine latissimus dorsi muscle. The ventricles were connected to a totally implanted mock circulation and then stimulated to pump chronically with good function for 8 to 63 days. Causes for device malfunction and potential causes of muscle damage were discussed. We have also read the published work by Juffe and associates, which had previously escaped our notice. The features they note in muscles that were stimulated for more than 14 days, namely, predominance and atrophy of type II fibers, loss of myelinated intramuscular nerve branches, and fibrous connective tissue reaction, are consistent with nerve damage. This conclusion is shared by the authors themselves, who state: "Atrophy in the type II fibers ... was similar to that seen in dogs 21 days after total denervation.... We feel that neural factors, accelerated by prolonged and repeated stimula- tion of the nerve may have played some role in its appear- ance.?" Such a result is the more disappointing because the changes that they describe for the shorter periods of stimula- tion are indeed indicative of the early stages of adaptive transformation. With the information available to us, it is difficult to identify with certainty the sources of nerve damage, but we