Faculty Bibliography

This is a retrospective review of all carotid endarterectomies (CEA) (n=91) done from 1993 to 2002 at an inner-city hospital (Group I). This group was compared to a randomly selected group of patients (n=445) treated at a private hospital (Group II). The same high-volume surgeons performed CEAs at both hospitals. The majority of Group I patients (71.4%) were members of racial minority groups. They were also more likely to be younger (p<0.001), hypertensive (p<0.03), diabetic (p<0.001), and current smokers (p<0.001); have contralateral carotid artery occlusion (p=0.04); and present with stroke (p<0.001) than Group II patients. Despite this, the incidence of postoperative myocardial infarction (2.2% vs 0.2%, p=0.08), stroke (1.1% vs 1.6%, NS), and death (1.1% vs 0%, NS) was comparable between the 2 groups. Aggressive preoperative workup for occult cardiac disease in Group I revealed an incidence of 25.9% (n=15). Of these, 5 (33.3%) were found to have coronary artery disease severe enough to warrant intervention before CEA. In an inner-city population with increased medical comorbidities, more severe cerebrovascular disease, and relatively low volume of carotid surgery, the results of CEA were comparable to those in patients treated at a high-volume private hospital. The presence of high-volume surgeons, operating at the low-volume municipal hospital, may contribute to the low complication rate. Finally, aggressive preoperative cardiac workup in this underserved population revealed a meaningful incidence of occult coronary artery disease requiring intervention before CEA

Objectives Limb and pelvic ischemia are known complications after endovascular abdominal aortic aneurysm repair (EVAR). The objective of this paper is to present our experience with the incidence, presentation, and management of such complications. Methods Over 9 years 311 patients with aortic aneurysms underwent EVAR. A retrospective review identified 28 patients (9.0%) with ischemic complications. Results Among 28 patients with ischemic complications, 21 had lower extremity ischemia and 7 had pelvic ischemia: colon (n = 4), buttock (n = 2), and spinal cord (n = 2). Of the 21 patients with lower extremity ischemia, 15 had limb occlusions (71.4%), 3 due to embolization (14.7%) and 3 the result of common femoral artery thromboses (14.7%). Limb occlusions were manifested as severe acute arterial ischemia (n = 6), rest pain (n = 3), intermittent claudication (n = 5), and decreased femoral pulse (n = 1). Limb occlusions were managed with thrombectomy and stent placement (n = 4), femorofemoral bypass (n = 7), eventual explantation because of persistent endoleak (n = 1), and expectant management (n = 3). The 3 patients with occlusions managed expectantly all had intermittent claudication, which has subsequently improved. In the 6 patients with lower extremity ischemia due to embolization or common femoral artery injury presentation was acute, and embolectomy was performed, followed by femoral artery endarterectomy and patch angioplasty or placement of an interposition graft. One patient who had a prolonged postoperative course including cardiac arrest subsequently required distal bypass and ultimately above- knee amputation. Among the 7 patients with pelvic ischemia, 2 patients had unilateral hypogastric artery embolization before the original surgery. Among the patients with colonic ischemia, 3 were seen immediately postoperatively, and required colectomy and colostomy. Two patients who required urgent colectomy subsequently had multiple organ failure, and died in the perioperative period. One patient had abdominal pain 1 week after surgery, which was managed with bowel rest, with subsequent improvement. In 2 patients spinal cord ischemia developed immediately after surgery, w hich resulted in persistent paraplegia. Buttock ischemia developed in 2 patients, 1 of whom required fasciotomy because of gluteal compartment syndrome, and had transient renal failure. Conclusions Ischemic complications are not uncommon after EVAR, and may exceed the incidence with open surgical repair. Limb ischemia is most often a result of limb occlusion, and can be successfully managed with standard interventions. Pelvic ischemia often results from atheroembolization despite preservation of hypogastric arterial circulation. Colonic and spinal ischemia are associated with the highest morbidity and mortality

When peripheral vascular injuries present in conjunction with life threatening emergencies, controlling hemorrhage from a peripheral blood vessel may take initial priority, however, sacrificing a limb to preserve life is a well-established dictum. The use of intravascular shunts has allowed arterial and venous injuries to be controlled and temporized while treating other injuries. Typically, intravascular shunts are used for short time periods while orthopedic injuries are repaired or other life threatening injuries are managed. The following case demonstrates the long-term use of an intravascular arterial shunt to treat a traumatic transection of the common femoral artery and vein in a patient with an open pelvic fracture from blunt trauma. A 20-year-old woman fell between a subway platform and an oncoming train. She sustained a crush injury to her lower extremity and pelvis as she was pinned between the train and platform. The patient presented with active hemorrhage from a groin laceration, quickly became hemodynamically unstable, and was brought to the operating room. In addition to a pelvic fracture with massive pelvic hematoma she sustained a complete transection of the bifurcation of the common femoral artery (CFA), the common femoral vein (CFV), and associated orthopedic injuries. Vascular shunts were placed in the common femoral artery and vein. The patient became hypotensive from an expanding retroperitoneal hematoma. Pelvic bleeding was controlled with angioembolization and the venous injury was repaired. At this time the patient became cold, acidotic, and coagulopathic. It was thought unsafe to proceed with the arterial repair and it was elected to keep her arterial shunts in place and perform a planned reexploration in 24 hours after correcting her physiologic status. The patient returned to the operating room for an elective repair of her CFA the following day. Her shunt had remained patent throughout this time. She underwent a reverse saphenous vein graft from her CFA to her SFA. After a prolonged hospital course she was ultimately transferred to a rehabilitation center with intact pulses in both lower extremities. This case demonstrates the effectiveness of prolonged (>6 hours) use of an intravascular shunt as part of damage control surgery for peripheral arterial and venous injuries. In a patient who would otherwise undergo an amputation for their injury, the risk of shunt thrombosis, or infection, during damage control resuscitation may not be a contraindication for placement

Citing the higher perioperative risk of redo carotid surgery, balloon angioplasty and stenting of the carotid artery (CAS) has been advocated for recurrent carotid stenosis (RCS). To examine the impact of CAS on the management and outcome of recurrent stenosis, a retrospective review of a prospectively compiled database was performed. From a registry of patients treated for carotid disease, 105 procedures were performed from 1992 to 2002 for RCS. For comparison, two study groups were examined. Time I consisted of 77 reoperations performed through 1998, before CAS was introduced at our institution. Time II included 12 reoperations and 16 CAS procedures performed for RCS from 1999 through 2002. Using perioperative stroke as a measure of outcome, the results for time II were poorer than for time I (7.2% vs. 5.2%, p = NS). Overall, the risk of perioperative stroke was the same for reoperation (5/89) and CAS (1/16) (5.6% vs. 6.3%, p = NS). Although not statistically significant, there was a trend toward a higher risk of perioperative stroke for patients treated with reoperation during the latter time period (8.3% vs. 5.2%, p = NS). This probably relates to the finding that during time II, CAS was most likely to be used in asymptomatic patients (68.6% vs. 41.7%, p = NS) with early (<3 years) RCS (87.5% vs. 41.7%, p= 0.01). No patient with asymptomatic, early RCS had a perioperative stroke with either surgery or CAS (0/35 cases, 0%). The presence of preoperative neurologic symptoms was significantly predictive of a perioperative stroke among all procedures performed for RCS (13.6% vs. 0%, p = 0.004). Contrary to suggestions that CAS might improve the management of RCS, a review of our data shows the overall risk of periprocedural stroke to be no better since CAS has become available. The bias for using CAS for asymptomatic myointimal hyperplastic lesions, and reoperation for frequently symptomatic late recurrent atherosclerotic disease, makes direct comparisons of the two techniques for treating RCS difficult. It is expected that the overall risk for redo carotid surgery will increase, as fewer low-risk patients will be receiving open procedures. However, the increased risk among symptomatic patients undergoing reoperation suggests that endovascular techniques should be investigated among this group of cases as well

Adult intussusception in the setting of inflammatory bowel disease (IBD) is a rare phenomenon. Giant pseudopolyps, while generally considered benign, may function as lead points for intussusception. Diagnosis and management of intussusception in the setting of IBD can be fraught with hazards. We report the case of a 27-year-old male, recently diagnosed with ulcerative colitis and giant pseudopolyps, who presented with colocolonic intussusception and obstruction. Diagnosis was confirmed using CT imaging and the patient underwent resection of the colocolonic intussusception without reduction. The following case underscores the challenges in managing adult intussusception in the setting of IBD and allows for a review of the literature to date. Resection of non-reduced intussusception, rather than endoscopic or enema reduction, should continue to be definitive treatment of patients presenting with this unusual problem

OBJECTIVE: Transcatheter embolization with coils and other agents has been described as a treatment method for type II endoleak after endovascular aortic aneurysm repair (EVAR). Type I endoleak has not been treated commonly with such therapies, although most investigators believe they warrant definitive intervention. The liquid adhesive n-butyl 2-cyanoacrylate (n-BCA) is often used to treat congenital arteriovenous malformations. The objective of this study is to report our initial experience in treating type I endoleak with n-BCA and with a variety of other interventions. METHODS: A retrospective review was performed of 270 patients who underwent EVAR at our institution between January 1994 and December 2002. Of these, 24 patients had type I endoleak (8.9%), diagnosed either intraoperatively (n = 13, 52%) or during follow-up (n = 12, 48%). Among these 24 patients, 17 had proximal leaks and the remaining 8 patients had distal leaks. These cases form the focus of this study. RESULTS: Twenty-two leaks required endovascular intervention, with the following success rate: n-BCA, 12 of 13 cases (92.3%); extender cuffs, 4 of 5 cases (80%); coils with or without thrombin, 3 of 4 cases (75%). In one patient with persistent endoleak despite attempted endovascular intervention the device ultimately was surgically explanted, and the patient did well. Of six patients with endoleak initially managed expectantly, two eventually underwent attempts at definitive intervention, both with n-BCA. Three sealed spontaneously before definitive intervention could be performed; and in one 97-year-old patient who refused intervention, the aneurysm subsequently ruptured and the patient died. In total, 13 patients with type I endoleak underwent n-BCA transcatheter embolotherapy. No serious complications were directly related to this therapy. Colon ischemia developed in one patient, and was believed to be a result of thromboembolism during wire and catheter manipulation rather than n-BCA treatment. Twelve of these 13 leaks remain sealed at mean follow-up of 5.9 months (range, 0-19 months). CONCLUSION: Our initial use of n-BCA occlusion suggests that it may be an effective and safe method of treatment of type I endoleak after EVAR. In particular, n-BCA embolotherapy may be especially useful in treating type I endoleak not amenable to placement of extender cuffs. Larger case series and longer follow-up are needed before this treatment is more broadly recommended. Type I endoleak after EVAR can be treated successfully with a variety of endovascular methods, and surgical explantation is rarely required

Successful endovascular aortic aneurysm repair (EVAR) is often defined as complete exclusion of blood flow within the aneurysm sac. Perigraft flow, also known as endoleak, is the most common complication following EVAR. Attachment site related endoleaks (type I) are generally considered to warrant some form of intervention due to the belief that they represent a risk for future rupture. Management of type II endoleaks, also known as branch or collateral endoleaks, is more controversial. Some advocate a policy of watchful-waiting whereas others treat all type II endoleaks as soon as they are discovered. The following review explores the controversies pertaining to the management, diagnosis and surveillance imaging, and treatment of type II endoleaks

The authors previously established an in vitro palate nonfusion model on the basis of a spatial separation between prefusion embryonic day 13.5 mouse palates (term gestation, 19.5 days). They found that an interpalatal separation distance of 0.48 mm or greater would consistently result in nonfusion after 4 days in organ culture. In the present study, they interposed embryonic palatal mesenchymal tissue between embryonic day 13.5 mouse palatal shelves with interpalatal separation distances greater than 0.48 mm in an attempt to 'rescue' this in vitro palate nonfusion phenotype. Because no medial epithelial bilayer (i.e., medial epithelial seam) could potentially form, palatal fusion in vitro was defined as intershelf mesenchymal continuity with resolution of the medial edge epithelia bilaterally. Forty-two (n = 42) palatal shelf pairs from embryonic day 13.5 CD-1 mouse embryos were isolated and placed on cell culture inserts at precisely graded distances (0, 0.67, and 0.95 mm). Positive controls consisted of shelves placed in contact (n = 6). Negative controls consisted of shelves placed at interpalatal separation distances of 0.67 mm (n = 6) and 0.95 mm (n = 7) with no interposed mesenchyme. Experimental groups consisted of embryonic day 13.5 palatal shelves separated by 0.67 mm (n = 11) and 0.95 mm (n = 12) with interposed lateral palatal mesenchyme isolated at the time of palatal shelf harvest. Specimens were cultured for 4 days (n = 19) or 10 days (n = 23), harvested, and evaluated histologically. All positive controls at 4 and 10 days in culture showed complete histologic palatal fusion. All negative controls at 4 days and 10 days in culture remained unfused. Five of six palatal shelves separated at 0.67 mm interpalatal separation distance with interposed mesenchyme were fused at 4 days, and all five were fused at 10 days. At an interpalatal separation distance of 0.95 mm with interposed mesenchyme (n = 12), no palates (zero of four) were fused at 4 days, but seven of eight were fused at 10 days. These data suggest that nonfused palatal shelves can be 'rescued' with an interposed graft of endogenous embryonic mesenchyme to induce fusion in vitro

It is unclear whether cleft palate formation is attributable to intrinsic biomolecular defects in the embryonic elevating palatal shelves or to an inability of the shelves to overcome a mechanical obstruction (such as the tongue in Pierre Robin sequence) to normal fusion. Regardless of the specific mechanism, presumably embryonic palatal shelves are ultimately unable to bridge a critical distance and remain unapproximated, resulting in a clefting defect at birth. We propose to use a palate organ culture system to determine the critical distance beyond which embryonic palatal shelves fail to fuse (i.e., the minimal critical intershelf distance). In doing so, we hope to establish an in vitro cleft palate model that could then be used to investigate the contributions of various signaling pathways to cleft formation and to study novel in utero treatment strategies.Palatal shelves from CD-1 mouse embryos were microdissected on day 13.5 of gestation (E13.5; term = 19.5 days), before fusion. Using a standardized microscope ocular grid, paired palatal shelves were placed on a filter insert at precisely graded distances ranging from 0 (in contact) to 1.9 mm (0, 0.095, 0.19, 0.26, 0.38, 0.48, 0.57, 0.76, 0.95, and 1.9 mm). A total of 68 paired palatal shelves were placed in serum-free organ culture for 96 hours (n = 68). Sample sizes of 10 were used for each intershelf distance up to and including 0.48 mm (n = 60). For intershelf distances of 0.57 mm and greater, two-paired palatal shelves were cultured (n = 8). All specimens were assessed grossly and histologically for palatal fusion.Palatal fusion occurred in our model only when intershelf distances were 0.38 mm or less. At 0.38 mm, eight of 10 palates appeared grossly adherent, whereas six of 10 demonstrated clear fusion histologically with resolution of the medial epithelial seam and continuity of the palatal mesenchyme. None of the 18 palates fused when placed at intershelf distances of 0.48 mm or greater.Using our selected intershelf distances as a guideline, we have established an approximate minimal critical intershelf distance (0.48 mm) at which we can reliably expect no palatal fusion. Culturing palatal shelves at intershelf distances of 0.48 mm or greater results in nonfusion or clefting in vitro. This model will allow us to study biomolecular characteristics of unfused or cleft palatal shelves in comparison with fused shelves. Furthermore, we plan to study the efficacy of grafting with exogenous embryonic mesenchyme or candidate factors to overcome clefting in vitro as a first step toward future in utero treatment strategies

BACKGROUND/PURPOSE: Retinoid signaling plays an important role in many differentiation pathways. Retinoid signaling has been implicated in the induction of differentiation by pancreatic ductal cancer cell lines and in patients with pancreatic cancer. The authors wished to better understand the role of retinoid signaling in pancreatic development. METHODS: Embryonic pancreas was harvested from mice at serial gestational ages and immunohistochemical analysis was performed for retinoic acid receptors (RAR-alpha, RAR-beta, RAR-gamma), and retinoid X receptors (RXR-alpha, RXR-beta, and RXR-gamma). Also, early embryonic pancreases were cultured for 7 days with exogenous 9-cis retinoic acid (9cRA) or all-trans retinoic acid (atRA) and analyzed histologically and immunohistochemically. RESULTS: Retinoid receptors were expressed in a lineage-specific distribution, with stronger expression for many in the exocrine compartment. The receptors were not often expressed until late gestation. Exogenous 9cRA induced predominantly ducts instead of acini, plus more mature endocrine (islet) architecture. Exogenous atRA induced predominantly acini instead of ducts, with no apparent endocrine effect. CONCLUSIONS: Retinoids may have an important role in pancreatic differentiation, with a particular effect on secondary lineage selection between ductal and acinar phenotype. Because the control of ductal versus acinar differentiation has been implicated strongly in the pathogenesis of pancreatic ductal carcinoma, these results may lay the groundwork for studies in the mechanism of induced differentiation of pancreatic ductal cancer by retinoids.