Surgical closure of atrial septal defect in patients older than 50 years of age.

Between 1962 and 1986, 51 patients 50 years of age or older (mean, 59.7 years; range, 50 to 77 years) underwent operative closure of atrial septal defect. The mean pulmonary vascular resistance was 1.7 +/- 0.79 wood units and the mean pulmonary blood flow-systemic blood flow ratio was 2.8 +/- 1.4. Patch closure with pericardium or Dacron was performed in 36 patients (70%), while 15 patients (30%) underwent primary closure. There were no operative deaths, and one patient was not available for follow-up. The remaining 50 patients (98%) have been followed up for 0.5 to 25 years (mean, 9.6 years). Mean preoperative New York Heart Association classification was 2.40 +/- 0.70. This improved significantly to 1.21 +/- 0.42 at follow-up, with all patients improving by at least one category. No patient received long-term anticoagulation treatment, and no pulmonary or systemic emboli were identified. Two (13%) of 15 patients who underwent primary closure developed septal dehiscence. Actuarial survival was 93%, 86%, and 79% at 5, 10, and 15 years, respectively. Atrial septal defect closure can be safely performed in older patients with excellent results, provided the ratio of systemic to pulmonary blood flow is greater than or equal to 1.5 to 1.0, the pulmonary vascular resistance is low, and the shunt remains left to right. Primary closure should be discouraged, and postoperative anticoagulation therapy appears unwarranted.

Concomitant valvotomy and subclavian-main pulmonary artery shunt in neonates with pulmonary atresia and intact ventricular septum.

Our current approach to the management of neonates with pulmonary atresia and intact ventricular septum is to perform a transarterial pulmonary valvotomy through a left anterolateral thoracotomy followed by a polytetrafluoroethylene shunt between the left subclavian artery and the pulmonary trunk at the site of the pulmonary arteriotomy. From October, 1983, to December, 1985, 7 consecutive neonates with pulmonary atresia and intact ventricular septum were managed in this fashion. Mean age was 5.1 days (5 patients, less than 48 hours old), and mean weight was 3.3 kg (range, 2.5-4.3 kg). Right ventricular morphology was type I (tripartite) in 4 patients, type II (absent trabecular portion) in 2, and type III (absent trabecular and infundibular portions) in 1. The mean right ventricular to left ventricular peak systolic pressure ratio was 1.5. One patient who initially had valvotomy alone required a left subclavian-pulmonary trunk shunt the next day for hypoxemia. All other patients had a valvotomy and shunt during the same procedure. There were no operative or hospital deaths. Follow-up of 3.5 to 34 months (mean, 17.5 months) confirmed shunt patency in all patients. Three of 4 patients undergoing postoperative catheterization have shown good right ventricular growth; 2 have undergone successful repair at 10 and 23 months. There have been 3 late deaths at 3.5, 4, and 8 months. Two other patients are doing well and are awaiting postoperative catheterization. This procedure permits synchronous valvotomy and shunting without the need for cardiopulmonary bypass in these critically ill neonates.(ABSTRACT TRUNCATED AT 250 WORDS)

Gore-Tex patch repair of the anterior rectus sheath in free rectus abdominis muscle and myocutaneous flaps.

Reconstruction of the breast and other tissues by the free transverse rectus abdominis myocutaneous (TRAM) flap is an accepted, reliable technique with a high success rate. Closure of the anterior rectus sheath defect that results from this flap and the related rectus abdominis muscle-only (RAM) free flap has been the subject of debate, since hernia formation is considered a risk. Some authors prefer direct closure, while others recommend a variety of synthetic mesh reinforcements. We have reviewed 81 patients from a consecutive series of free TRAM and RAM flaps performed by one surgeon (Pennington) over an 8-year period. The majority of patients had repair of the anterior rectus sheath with a 1-mm-thick synthetic patch of polytetrafluoroethylene (Gore-Tex), used in 52 of 71 patients having the TRAM flap and 4 of 10 patients having the RAM flap. There were 5 wound infections (6.2 percent) overall, 3 of which occurred in the Gore-Tex group (5.4 percent). In all 3 of the latter patients, Gore-Tex was removed, although usually only after some months. Even after patch removal, no hernias developed subsequently. The two other infections occurred in patients with a Prolene mesh patch and no patch, respectively. One hernia was found in 71 patients with TRAM flaps, and that was in a patient in whom no patch was used. No hernias occurred in 52 TRAM flap patients with Gore-Tex patches. Hernia formation was noted in 6 of 10 RAM flap patients, although only 1 hernia occurred in the 4 patients with Gore-Tex patches. We conclude that the 1-mm polytetrafluoroethylene (Gore-Tex) patch is a satisfactory method for repair of the anterior rectus sheath after harvesting of either free RAM or TRAM flaps. Even if removal of the patch is required because of infection, the risk of subsequent hernia formation appears minimal.

Free-flap reconstruction of large defects of the scalp and calvarium.

Beyond a certain size, full-thickness defects of scalp are not amenable to local flap repair. Staged distant flaps have now been virtually eliminated by free-flap reconstruction. The authors present 12 patients in whom full-thickness scalp defects with an average area of 275 cm2 were reconstructed utilizing free flaps. Nine patients had corresponding large calvarial defects. Ten patients had reconstruction with free latissimus dorsi muscle flaps and overlying skin grafts, and one patient had reconstruction with a scapular free flap. Of the 12 patients, 8 had extirpative surgery for tumor with immediate reconstruction and the remaining 4 had reconstruction for chronic radionecrosis of the scalp, usually associated with infected osteoradionecrosis of the calvarium. Of this latter group, 2 patients underwent simultaneous acrylic cranioplasty. The technique and results are discussed.

Osseointegration of preformed polymethylmethacrylate craniofacial prostheses coated with bone marrow-impregnated poly (DL-lactic-co-glycolic acid) foam.

Osseointegration of bone marrow-PLGA-coated, preformed polymethylmethacrylate cranioplasties offers the possibility of reducing: operative time, periimplant seroma and infection, metallic fixation, and periprosthetic resorption following surgical skull remodeling. These alloplastic materials are FDA-approved but previously have not been used together to promote cranioplasty incorporation. The objective of this study was to determine whether the use of PLGA foam coating improves host osseointegration of preformed, textured, polymethylmethacrylate prosthetic cranioplasties. A critical-sized cranial defect was created in two groups of 10 and one group of three rabbits. The defect was filled with either a textured, preformed polymethylmethacrylate disc or a textured, preformed polymethylmethacrylate disc coated with poly (DL-lactic-co-glycolic acid). Both implants were immersed in autologous bone marrow for 20 minutes before implantation. Half of each group of 10 were killed at 3 weeks, and the remainder at 6 weeks. A third group of three rabbits with excised periosteum was evaluated at 6 weeks. Histologic analysis of the discs determined relative amounts of cancellous bone formation adjacent to the prostheses. Woven trabecular bone was present at each host bone to implant perimeter interface at 3 weeks, with fine fibrous capsular formation around the implants. Thicker, lamellar trabeculae were present at 6 weeks with an increased fibrous layer surrounding both types of implants. Bone formed on the superficial and deep implant surfaces in a noncontiguous fashion. Two of five measures showed that total bone formation was significantly greater in the PLGA-coated implants. Polymethylmethacrylate discs coated with bone marrow-impregnated PLGA foam demonstrate increased bone formation at 3 and 6 weeks as compared with non-coated preformed polymethylmethacrylate discs. Only implants with preserved periosteum showed bone formation away from the host-implant interface (centrally) on the superficial surface at 6 weeks.

Initial clinical experience with centrifugal pumps coated with the Carmeda process.

From September 5 to November 4, 1991, four consecutive patients placed on centrifugal ventricular assist devices (VADs) for cardiac failure were supported with Biomedicus pumps coated with the Carmeda bioactive surface. The study included three men and one woman aged 52-65 years. Two patients were supported with a right VAD, one with a left VAD, and one with a biventricular VAD. Support ranged from 35.5 to 65.75 hr, and VAD flows ranged from 1.0-5.5 L/min. Three patients were weaned from the VADs, and two survived. At explant, all four systems had clots adherent to the surface of the tubing and connectors on the pump outflow side of the circuit. Two patients had clots in the pump. Some of the clots were firmly adherent, whereas others dislodged easily after being rinsed with saline. All patients received heparin for insertion, and in three patients, heparin was reversed with protamine. Two patients received no further anticoagulation, and two received continuous heparin within 24 hr of implant to maintain activated clotting times of 140-150 sec. All patients had bleeding complications before and after VAD placement, necessitating multiple blood product transfusions. One patient who was weaned and survived had multiple thromboembolic strokes. These data suggest that clots can form on surfaces coated with the Carmeda process, even if a low dosage of heparin is used.