The role of microvascular damage in photodynamic therapy: the effect of treatment on vessel constriction, permeability, and leukocyte adhesion.

Intravital microscopy of the rat cremaster muscle was used to evaluate changes in vessel constriction, vessel permeability, and leukocyte adhesion during and after photodynamic therapy (PDT). Animals were given Photofrin doses of 0-25 mg/kg i.v. 24 h before treatment. Cremaster muscles were exposed to 135 J/cm2 light at 630 nm. Animals given 5 mg/kg Photofrin showed no vessel constriction or increase in vessel permeability to albumin. Doses of 10 and 25 mg/kg Photofrin caused a dose-related constriction of arterioles which was observed within the first minutes of illumination at the higher drug dose. After the initial constriction, arteriole response to PDT was biphasic in nature, with some vessels relaxing to nearly control levels while others remained fully constricted. Constriction of venules occurred only at the highest porphyrin dose studied (25 mg/kg) and was delayed in comparison to arteriole constriction. Photofrin doses which produced arteriole constriction also caused an increase in venule permeability to albumin, which occurred shortly after the start of light treatment and was progressive with time. Leakage began at specific sites along the venule wall but became uniform along the entire length of the venule by 1 h after treatment. Changes in the adherence of polymorphonuclear leukocytes to venule endothelium were also observed with PDT. Photofrin doses of 25 mg/kg and 45 J/cm2 light were sufficient to cause polymorphonuclear leukocytes to become adherent to the vessel wall. A second group of animals was given indomethacin trihydrate to examine the involvement of cyclooxygenase products such as thromboxane in vessel response to PDT. Animals given 5 mg/kg indomethacin intraarterially 1 h before light treatment showed no constriction of arterioles or venules at all Photofrin and light doses studied. No increases in venule permeability to albumin were seen in this group of animals. This suggests that cyclooxygenase products including thromboxane are important in causing vessel constriction and changes in permeability during PDT. The initiating event which causes the release of these vasoactive agents remains unknown.

Sentinel lymph node biopsy for breast cancer: a suitable alternative to routine axillary dissection in multi-institutional practice when optimal technique is used.

PURPOSE: Previous studies have demonstrated the feasibility of sentinel lymph node (SLN) biopsy for nodal staging of patients with breast cancer. However, unacceptably high false-negative rates have been reported in several studies, raising doubt about the applicability of this technique in widespread surgical practice. Controversy persists regarding the optimal technique for correctly identifying the SLN. Some investigators advocate SLN biopsy using injection of a vital blue dye, others recommend radioactive colloid, and still others recommend the use of both agents together. PATIENTS AND METHODS: A total of 806 patients were enrolled by 99 surgeons. SLN biopsy was performed by single-agent (blue dye alone or radioactive colloid alone) or dual-agent injection at the discretion of the operating surgeon. All patients underwent attempted SLN biopsy followed by completion level I/II axillary lymph node dissection to determine the false-negative rate. RESULTS: There was no significant difference (86% v 90%) in the SLN identification rate among patients who underwent single- versus dual-agent injection. The false-negative rates were 11.8% and 5.8% for single- versus dual-agent injection, respectively (P <.05). Dual-agent injection resulted in a greater mean number of SLNs identified per patient (2. 1 v 1.5; P <.0001). The SLN identification rate was significantly less for patients older than 50 years as compared with that of younger patients (87.6% v 92.6%; P =.03). Upper-outer quadrant tumor location was associated with an increased likelihood of a false-negative result compared with all other locations (11.2% v 3. 9%; P <.05). CONCLUSION: In multi-institutional practice, SLN biopsy using dual-agent injection provides optimal sensitivity for detection of nodal metastases. The acceptable SLN identification and false-negative rates associated with the dual-agent injection technique indicate that this procedure is a suitable alternative to routine axillary dissection across a wide spectrum of surgical practice and hospital environments.

Atrioventricular discordance: results of repair in 127 patients.

OBJECTIVE: The conventional management of patients with atrioventricular discordance is directed at associated lesions, taking advantage of physiologic "correction"; however, the morphologic right ventricle and tricuspid valve support the systemic circulation. Questions surrounding survival using this approach led us to analyze our institutional results. METHODS: All patients with atrioventricular discordance undergoing biventricular repair were analyzed (n = 127, 1959-1997), excluding those with functionally univentricular hearts. The ventriculoarterial connection associated with atrioventricular discordance varied and was most commonly discordant (87%), but occasionally concordant (6%), double-outlet right ventricle (6%), or double-outlet left ventricle (1%). At initial presentation, the most common lesions associated with atrioventricular discordance were ventricular septal defect (86%), pulmonary stenosis (64%), tricuspid regurgitation (28%), and atrioventricular block (12%). Nine patients underwent a double switch procedure to create ventriculoarterial concordance and the remainder were managed conventionally without correcting discordant connections. RESULTS: Operative mortality was 6% and did not vary by associated lesion. Twenty years after repair, survival was 48%. Within 20 years, 56% of patients required reoperation, usually for atrioventricular valve incompetence (n = 16), pulmonary stenosis (n = 16), or both (n = 3). Pacemakers were required in 50 patients, 4 before repair, 40 within 2 months of repair, and 6 remotely after repair. In early follow-up, the double switch procedure (n = 9) had equivalent mortality and a high pacemaker requirement for atrioventricular block. CONCLUSIONS: Analysis of conventional management of atrioventricular discordance revealed cumulative increases in mortality, systemic atrioventricular valve (tricuspid) replacement, complete atrioventricular block, and incidence of reoperation. Alternative management should be examined.

Results after partial left ventriculectomy versus heart transplantation for idiopathic cardiomyopathy.

OBJECTIVE: Partial left ventriculectomy has been introduced as an alternative surgical therapy to heart transplantation. We performed a single-center, retrospective analysis of all patients with idiopathic dilated cardiomyopathy who underwent partial left ventriculectomy or heart transplantation or who were listed for transplantation to determine operative mortality rate, 12-month survival, freedom from death on the heart transplantation waiting list, and freedom from death or need for relisting for heart transplantation. METHODS: Patients who had partial left ventriculectomy (October 1996 to April 1998) were retrospectively compared with patients who were listed for heart transplantation (January 1995 to April 1998). Survival was assessed after the surgical procedure (partial left ventriculectomy vs heart transplantation) and from time of listing for heart transplantation to assess the additional impact of waiting list deaths. Freedom from death or relisting for heart transplantation was also compared. RESULTS: There was no difference in age or United Network for Organ Sharing status between the 2 groups. Twenty-nine patients with idiopathic dilated cardiomyopathy were listed for heart transplantation; 17 patients underwent transplantation, 6 patients died while on the waiting list, and 6 patients remain listed. One patient died after heart transplantation, and 1 patient required relisting. Sixteen patients had partial left ventriculectomy; 10 patients are in improved condition, 2 patients died (1 death early from sepsis and 1 death from progressive heart failure), and 4 patients required relisting for heart transplantation. Operative survival was 94% after partial left ventriculectomy and 94% after heart transplantation (P =.92). Postoperative 12-month Kaplan-Meier survival was 86% after partial left ventriculectomy and 93% after heart transplantation (P =.90). Twelve-month Kaplan-Meier survival after listing for heart transplantation was 75% due to death while on the waiting list (P =.76). Freedom from death or need for relisting for heart transplantation was 56% after partial left ventriculectomy and 86% after transplantation (P =.063). CONCLUSION: Operative and 12-month survival after partial left ventriculectomy and heart transplantation were comparable. However, despite their initial improvement, many patients who underwent partial left ventriculectomy required relisting for transplantation. Although partial left ventriculectomy is associated with acceptable operative and 12-month survival, it may prove to serve better as a bridge to transplantation in patients with idiopathic dilated cardiomyopathy rather than definitive therapy, given the number of patients who required relisting for transplantation.

Results of partial left ventriculectomy in patients with end-stage idiopathic dilated cardiomyopathy.

BACKGROUND: Partial left ventriculectomy has been recently introduced as a surgical therapy for end-stage heart failure. We performed a prospective study of partial left ventriculectomy in patients with end-stage idiopathic dilated cardiomyopathy (IDCM). METHODS: Patients considered as candidates for partial left ventriculectomy had IDCM, left ventricular end-diastolic diameter greater than 7 cm (LVEDD), refractory New York Heart Association class IV symptoms (NYHA), and severely impaired exercise oxygen consumption. All patients underwent a complete heart transplantation evaluation. RESULTS: Partial left ventriculectomy was performed in 16 patients with a mean follow-up of 11.1 months. Fourteen patients were male with a mean age of 49.6+/-10.5 years. After surgery there were significant changes in NYHA class, left ventricular ejection fraction, LVEDD, and degree of mitral regurgitation at up to 12 months follow-up. The operative mortality rate was 6.25% and 12-month Kaplan-Meier was 86%. Twelve-month freedom from need for listing for orthotopic heart transplantation was 65%. CONCLUSION: Partial left ventriculectomy can be performed in patients with idiopathic cardiomyopathy with acceptable operative and 12 month survival rates. Significant improvements are seen in ejection fraction, LVEDD, and NYHA class at 12 month follow up. Late failures do occur and some patients have required relisting for transplant after partial left ventriculectomy.

Factors that predict the presence of sentinel lymph node metastasis in patients with melanoma.

BACKGROUND: This analysis was performed to identify prognostic factors that are predictive of sentinel lymph node (SLN) metastasis in melanoma. METHODS: Analysis was performed of a multi-institutional, prospective, randomized trial of SLN biopsy for melanoma. Eligibility criteria included age 18 to 70 years, Breslow thickness of 1.0 mm or more, and clinically negative regional lymph nodes. SLNs were evaluated by serial sectioning and immunohistochemistry for S100. Univariate chi-square and multivariate logistic regression analyses were performed to assess factors predictive of the presence of a positive SLN. Probability values of less than.05 were considered significant. RESULTS: SLNs were identified in 99.7% of patients. A total of 1058 patients were evaluated; 961 patients had complete data and were included in the statistical analysis. SLNs were positive for tumor in 208 of 961 patients (22%). Breslow thickness, Clark level, ulceration, and patient age were factors that were found to be independently predictive of the presence of SLN metastasis. CONCLUSIONS: Increasing Breslow thickness, Clark level of more than III, the presence of ulceration, and patient age of 60 years or less are the most important independent prognostic factors associated with the finding of positive SLN in patients with melanoma.

Practical guidelines for optimal gamma probe detection of sentinel lymph nodes in breast cancer: results of a multi-institutional study. For the University of Louisville Breast Cancer Study Group.

INTRODUCTION: Multiple radioactive lymph nodes are often removed during the course of sentinel lymph node (SLN) biopsy for breast cancer when both blue dye and radioactive colloid injection are used. Some of the less radioactive lymph nodes are second echelon nodes, not true SLNs. The purpose of this analysis was to determine whether harvesting these less radioactive nodes, in addition to the "hottest" SLNs, reduces the false-negative rate. METHODS: Patients were enrolled in this multicenter (121 surgeons) prospective, institutional review board-approved study after informed consent was obtained. Patients with clinical stage T1-2, N0, M0 invasive breast cancer were eligible. This analysis includes all patients who underwent axillary SLN biopsy with the use of an injection of both isosulfan blue dye and radioactive colloid. The protocol specified that all blue nodes and all nodes with 10% or more of the ex vivo count of the hottest node should be removed and designated SLNs. All patients underwent completion level I/II axillary dissection. RESULTS: SLNs were identified in 672 of 758 patients (89%). Of the patients with SLNs identified, 403 patients (60%) had more than 1 SLN removed (mean, 1.96 SLN/patient) and 207 patients (31%) had nodal metastases. The use of filtered or unfiltered technetium sulfur colloid had no impact on the number of SLNs identified. Overall, 33% of histologically positive SLNs had no evidence of blue dye staining. Of those patients with multiple SLNs removed, histologically positive SLNs were found in 130 patients. In 15 of these 130 patients (11.5%), the hottest SLN was negative when a less radioactive node was positive for tumor. If only the hottest node had been removed, the false-negative rate would have been 13.0% versus 5.8% when all nodes with 10% or more of the ex vivo count of the hottest node were removed (P =.01). CONCLUSIONS: These data support the policy that all blue nodes and all nodes with 10% or more of the ex vivo count of the hottest SLN should be harvested for optimal nodal staging.

DNA flow cytometry does not predict 5- or 10-year recurrence rates for T1-2 node-negative breast cancer.

BACKGROUND: A small proportion of T1 or T2 node-negative breast cancer tumors will recur in patients by 5 years, and more by 10 years. Results of recent studies have suggested improvement in overall survival with administration of adjuvant chemotherapy to all patients. More sensitive and specific methods are needed to identify patients at highest risk for recurrence who might benefit most from adjuvant therapy, saving others from unnecessary treatment. Some investigators have suggested DNA flow cytometry as a method to discriminate patients at greatest risk for recurrence. HYPOTHESIS: DNA flow cytometry has predictive value for breast cancer recurrence in node-negative patients. METHODS: The cancer registry of a medium-sized university-affiliated hospital was used to identify patients with T1-2 N0 M0 breast cancer treated with a uniform surgical approach and no adjuvant therapy who had completed at least 5 years of follow-up or had recurrence. Flow cytometric analysis was performed on paraffin-embedded specimens. RESULTS: Of 115 patients, 92 (80%) had disease-free survival without recurrence and 23 (20%) had recurrence. Comparison of diploid and nondiploid tumors for likelihood of recurrence revealed no association (P = .79). Furthermore, the DNA index and S-phase fraction were not significantly different between recurrent and nonrecurrent groups. CONCLUSIONS: The likelihood of recurrence of small node-negative breast cancers after mastectomy cannot be accurately predicted on the basis of DNA flow cytometric analysis. Traditional methods for determining risks-such as nuclear and histological grade, lymph node status, and tumor size-seem to be more useful. Sentinel lymph node biopsy techniques may increase the detection of micrometastases.

Predicting the status of the nonsentinel axillary nodes: a multicenter study.

BACKGROUND: Sentinel lymph node (SLN) biopsy is a minimally invasive procedure that provides accurate nodal staging information. The need for completion axillary dissection after finding a positive SLN for breast cancer has been questioned. HYPOTHESIS: The presence of nonsentinel node (NSN) metastases in the axillary dissection specimen correlates with tumor size, the number of SLNs removed, and the number of positive SLNs. DESIGN: Prospective, multi-institutional study. PARTICIPANTS AND METHODS: The University of Louisville Breast Cancer Sentinel Lymph Node Study is a nationwide study involving 148 surgeons. All patients underwent SLN biopsy, followed by level I/II axillary dissection. All SLNs were evaluated histologically at a minimum of 2-mm intervals. Immunohistochemical analysis using antibodies for cytokeratin was performed at the discretion of each participating institution. All NSNs were evaluated by routine histologic examination. RESULTS: An SLN was identified in 1268 (90%) of 1415 patients. Increasing tumor size was significantly correlated with increasing likelihood of positive NSNs: T1a, 14%; T1b, 22%; T1c, 30%; T2, 45%; and T3, 57% (P =.002, chi(2) test). The presence of positive NSNs was not significantly associated with the number of SLNs removed. Patients with more than 1 positive SLN were more likely to have positive NSNs than those with only 1 positive SLN (50% vs 32%; P<.001, chi(2) test). Increasing tumor size and the presence of multiple positive SLNs were also associated with the presence 4 or more positive axillary nodes. Multivariate analysis confirmed that tumor size and the number of positive SLNs were independent factors predicting the presence of positive NSNs. CONCLUSIONS: The likelihood of positive NSNs correlates with increasing tumor size and the presence of multiple positive SLNs. However, even patients with small primary tumors have a substantial risk of residual axillary nodal disease after SLN biopsy. These data will be helpful in counseling patients regarding the need for completion axillary dissection after a positive SLN is identified.

Role of graft flow measurement technique in anastomotic quality assessment in minimally invasive CABG.

BACKGROUND: Anastomotic quality is currently the critical issue in minimally invasive coronary surgery. Although little is known about its effectiveness, surgeons routinely assess grafts intraoperatively using flow probes. This study was designed to determine whether mean flow and the pattern of flow tracing in internal mammary artery grafts obtained with a transit-time flow probe are reliable indicators of anastomotic quality. METHODS: Mongrel dogs (n = 14, 30 to 35 kg) underwent off-pump left, right, or left and right internal mammary artery to left anterior descending artery anastomosis (23 grafts). Moderate to severe degrees of stenosis were created at the anastomosis by an additional suture. Internal mammary artery graft flow was measured before and after the stenosis was created with the left anterior descending artery occluded. Angiography was performed at random postoperatively to validate the degree of stenosis. Mean flow and flow tracing morphology were compared under various degrees of stenosis. RESULTS: There were no significant differences in mean graft flow or the morphology of the flow tracing between patent (<15%), mild (<25%), moderate (<50%), and moderately severe (<75%) stenosis. However, mean graft flow decreased (p < 0.05) with severe stenosis (>75%). CONCLUSIONS: Although differences in mean graft flow and graft flow morphology were detectable in anastomoses with severe stenosis (>75%), they were indistinguishable in anastomoses with mild (<25%) to moderately severe (<75%) stenosis. Flow measurement techniques are valuable tools intraoperatively, but surgeons should exercise caution in their interpretation.

Sentinel lymph node biopsy for breast cancer: impact of the number of sentinel nodes removed on the false-negative rate.

BACKGROUND: Numerous studies have demonstrated that sentinel lymph node (SLN) biopsy can accurately determine axillary nodal status for breast cancer, but unacceptably high false negative rates have also been reported. Attention has been focused on factors associated with improved accuracy. We have previously shown that injection of blue dye in combination with radioactive colloid reduces the false negative rate compared with injection of blue dye alone. We hypothesized that this may be from the increased ability to identify multiple sentinel nodes. The purpose of this analysis was to determine whether removal of multiple SLNs results in a lower false negative rate. STUDY DESIGN: The University of Louisville Breast Cancer Sentinel Lymph Node Study is a prospective multiinstitutional study. Patients with clinical stage T1-2, N0 breast cancer were eligible for enrollment. All patients underwent SLN biopsy using blue dye alone, radioactive colloid alone, or both agents in combination, followed by completion level I and II axillary dissection. RESULTS: A total of 1,436 patients were enrolled in the study from August 1997 to February 2000. SLNs were identified in 1,287 patients (90%), with an overall false negative rate of 8.3%. A single SLN was removed in 537 patients. Multiple SLNs were removed in 750 patients. The false negative rates were 14.3% and 4.3% for patients with a single sentinel node versus multiple sentinel nodes removed, respectively (p = 0.0004, chi-square). Logistic regression analysis revealed that use of blue dye injection alone was the only factor independently associated with identification of a single SLN (p<0.0001), and patient age, tumor size, tumor location, surgeon's previous experience, and type of operation were not significant. CONCLUSIONS: The ability to identify multiple sentinel nodes, when they exist, improves the diagnostic accuracy of SLN biopsy. Injection of radioactive colloid in combination with blue dye improves the ability to identify multiple sentinel nodes compared with the use of blue dye alone.

Application of data mining for examining polypharmacy and adverse effects in cardiology patients.

This article comments upon the use of data mining tools to examine clinical data. Many cardiovascular patients have co-morbid diseases that put them at risk for polypharmacy, or severe adverse reactions from the interactions of multiple medications. Clinical trials typically use too few patients with stringent inclusion/exclusion criteria that prevent an examination of the issue of polypharmacy. However, clinical data collected in the course of patient treatment can be used in conjunction with data mining to find meaningful results.

In vivo and in vitro photodynamic studies with benzochlorin iminium salts delivered by a lipid emulsion.

Benzochlorin iminium salts (BIs) are hydrophobic photosensitizers based on an octaethylbenzochlorin nucleus that absorb in the near-IR region of the visible spectrum. In these studies the photodynamic activities of the zinc, copper and metal-free BI derivatives were compared in vivo in C3H-HeJ mice bearing a mammary adenocarcinoma tumor line. In vitro studies were also performed with the radiation-induced fibrosarcoma tumor line. An argon-pumped Ti-sapphire laser tuned to deliver light between 710 and 800 nm or an Oriel arc-lamp filtered to deliver broadband light above 590 nm were used as light source. A lipid emulsion was used as the delivery system for sensitizers in all studies. A pronounced solvent dependence was observed for the Q band for each of all iminium salts examined. As an example, the metal-free (BI) derivative had an absorption maximum at 798 nm in dichloromethane and at 727 nm in serum. The action spectra showed a greater PDT response at blue-shifted wavelengths for each of the three iminium salts both in vivo and in vitro. Among the three derivatives, the zinc analog (ZnBI) produced the greatest tumor regression at the low drug/light dose of 0.7 (mumole/kg and 200 J/cm2. These results indicate that iminium salts have characteristics that may make them promising third-generation photosensitizers.

Analysis of pulmonary microvasculature changes after photodynamic therapy delivered to distant sites.

Photodynamic therapy (PDT) can exert local damage by direct tumor cytotoxicity, by disruption of the microvasculature or by a combination of these effects. Although systemic effects after PDT of small tissue areas (< 1% total body surface area) are unlikely, treatment of larger areas may result in an accumulated effect leading to toxicity. Several investigators have described animal death after high dose PDT to tumors on the hind limb of animals and hypothesized that a toxic shock syndrome caused by vasoactive agents released after PDT is responsible. Because one of the most vulnerable organs to toxic shock injury is the lung, we studied the systemic effects of local PDT to this organ by intravital microscopy using a pulmonary window chamber. The PDT treatment conditions (25 mg/kg Photofrin, 24 h, 150 J/cm2 630 nm, maximum area 6.28 cm2) were chosen that produce systemic toxicity and lethality in rats. Adhesion of leukocytes in the lung was monitored in vivo using anti-CD-13-labeled microspheres. The progression of pulmonary edema was assessed by monitoring the leakage of rhodamine-labeled albumin and by wet-to-dry lung weight ratios. Although an increased leukocyte adherence was observed and a significant number of animals died after the extensive PDT treatment, no biologically significant lung edema could be demonstrated. These data indicate that lung edema and acute respiratory distress syndrome is not the cause of death in these animals and that the toxicity is related to other mechanisms including circulatory shock after extensive muscle damage.

Resection of the metatarsal head for diabetic foot ulcers.

BACKGROUND: Diabetic foot ulceration is a worldwide health problem. Approximately 15% of the 10 million diabetic patients in the United States will develop a foot ulceration at some time in their lives. The presence of a foot ulcer in this population is extremely debilitating and dramatically increases the risk of lower extremity amputation, accounting for approximately 67,000 lost limbs each year. Additionally, the costs associated with treating foot ulcers in diabetic patients is a major expense in the overall care of this patient group. METHODS: An 11-year retrospective study was conducted to evaluate 101 consecutive patients with diabetic ulcers of the forefoot who were treated using resection of the metatarsal head as the primary means of obtaining wound closure. RESULTS: The results indicate that 88% of the ulcers were healed by using this technique, and relatively more rapidly than would be expected when compared with historical norms. CONCLUSIONS: Resection of the metatarsal head is a safe and relatively inexpensive procedure that facilitates closure of the lesion, helps to control infection, and prevents countless and costly amputations.

Double grafting of the left anterior descending artery: is the distance between the internal mammary artery and supplemental vein graft anastomoses relevant in graft survival?

INTRODUCTION: Under certain conditions (small internal mammary artery (IMA) or large runoff), double grafting of the left anterior descending (LAD) artery system is necessary to avoid the ominous consequences of myocardial hypoperfusion. Previous studies have shown that a saphenous vein (SVG) adjacent to an IMA graft leads to failure of the IMA. This study compares IMA flow patterns when adjacent ( < 1 cm) and separated (3-4 cm) from a SVG placed on a proximally occluded LAD. METHODS: A SVG and right IMA (PIMA) to proximal LAD (2.5-3 mm) coronary bypass were performed in 12 mongrel dogs. The left IMA (DIMA) was anastomosed to the distal LAD (1.5 mm). All anastomoses were carried out without cardiopulmonary bypass. The native LAD was occluded proximally to the PIMA anastomosis, and all graft flows were measured in competitive and non-competitive flow conditions. RESULTS: Isolated graft to LAD flows were similar for the three conduits. There was a drop in flow in both the PIMA and DIMA when placed in competition with the SVG (10.1+/-3.0 vs. 19.1+/-4.6 ml/min; P < 0.05). The total drop in flow was significantly greater in the PIMA (67.6 vs. 39.9%; P < 0.05). Diastolic flow was better preserved in the distal IMA graft (19.6 + 5.6 vs. 10.2+/-3.0 ml/min; P < 0.05). The patterns of flow were much different during competition and there was significant retrograde systolic flow in all PIMA grafts while there was no (n = 5) or minimal retrograde flow (n = 7) in the DIMA grafts. CONCLUSION: An IMA graft, when adjacent to a SVG, sustains a significant decrease in both total and diastolic flows and develops an oscillating pattern of flow in early systole (retrograde then antegrade). Placing the IMA more distally on the LAD improves flow and decreases retrograde flow. In clinical situations requiring double grafting on the LAD, distance between grafts may be an important factor in maintaining IMA patency.

Neural network pattern recognition analysis of graft flow characteristics improves intra-operative anastomotic error detection in minimally invasive CABG.

OBJECTIVE: The intra-operative assessment of the quality of anastomosis in minimally invasive coronary artery bypass surgery (CABG) is critical. Recent investigations demonstrated that flow probes used intra-operatively to assess anastomotic errors may give the surgeon a false sense of confidence as only severely stenotic anastomoses (>90%) could be reliably detected. We developed a neural network system using graft flow data and assessed its potential to improve anastomotic error detection. METHODS: Mammary to LAD grafts (n = 46) were constructed in mongrel dogs off-pump. Continuous beat-to-beat graft flow was recorded using transit-time flow probes. Various degrees of anastomotic stenoses (0-100%) were created by an additional suture. The degree of anastomotic stenosis was confirmed by postoperative angiography. A learning vector quantization neural network was created using heart rate, mean aortic pressure, mean systolic, maximum systolic, minimum systolic, mean diastolic, maximum diastolic, minimum diastolic, and mean graft flows. In addition, a spectral analysis of the flow waveforms was performed and the magnitude and phase of the first five harmonics were used to further develop the neural network. RESULTS: The neural network pattern recognition system was 94% accurate in detecting any stenosis >50%. To validate the model, a testing set was used with 20% of the data values, and the accuracy remained at 100% above chance alone. CONCLUSION: Pattern recognition of transit-time flow probe tracings using neural network systems can detect anastomotic errors significantly better than the surgeon's visual assessment, thereby improving the clinical outcome of minimally invasive CABG.

Spectral analysis of graft flow for anastomotic error detection in off-pump CABG.

OBJECTIVE: Flow probes have been introduced as a non-invasive means of anastomotic quality assessment in off-pump coronary artery bypass graft (CABG). Flow waveform morphology cannot reliably be assessed visually unless severe anastomotic stenosis is present ( > 90%). We applied spectral analysis techniques to determine whether the frequency content of graft flow can improve the surgeon's ability to detect anastomotic errors. METHODS: Forty-six mammary to left anterior descending artery (LAD) anastomoses were created in mongrel dogs during off-pump CABG surgery. Graft flow was measured using transit-time flow probes with the LAD closed, and the mammary graft patent and with varying degrees of stenosis. The degree of anastomotic stenosis was created by an artificial stitch and verified by random postoperative angiography. Spectral analysis of the graft flow waveforms was performed. Differences in the magnitude and phase components of the graft flow for the first five harmonics were determined for the varying anastomosis test conditions. Differences were determined using analysis of variance and least square means techniques. RESULTS: The magnitude of the fundamental (zeroth) harmonic was statistically different in the internal mammary artery (IMA) with 0-25% stenosis compared to IMA with 50-75% stenosis (P < 0.01 ). Further, the magnitude of the first, second, and fourth harmonics were statistically different in IMA with 0-25% compared to IMA with 75% (P < 0.01). The phase of the first harmonic was statistically different in IMA with 25% stenosis than IMA with 50% stenosis (P < 0.01 ). No differences in interaction between the LAD and IMA for all ranges of stenosis were detected (P > 0.50). CONCLUSION: Spectral analysis of graft flow waveforms may be beneficial in detecting lesser degrees of anastomotic stenosis (i.e. < 90%) compared to traditional visual assessment of mean graft flow and/or graft flow waveform morphology.

A nonparametric approach to analyzing large human populations.

There are many statistical techniques that require the assumption that the population being studied is normally distributed--regression analysis, multivariate analysis, time series analysis, and so on. Unfortunately, as the development of survey sampling has long acknowledged, large human populations are usually stratified into several different subpopulations. Since the boundaries between the strata are somewhat blurred, they are not independent, so the overall distribution of the population tends to be multimodal rather than normal. In this paper, a technique is developed to find these multimodal techniques using nonparametric density estimation. Its effectiveness is demonstrated by means of an example.

PIP: The nonparametric kernal density estimation technique based on the survey sampling technique is explained, and its effectiveness is demonstrated by a simulation and by an actual example in order to show its usefulness in capturing multimodel distribution. The universe under investigation is usually stratified into several subpopulations. Incorrect assumptions are made when unimodel techniques are used on multimodel populations. Multimodel populations can also be dealt with by using only large samples with estimates made only on the sample mean, but this approach is not always possible. Another option is to study only one of the subpopulations, which can result in bias or oversmoothing the data. Other methods of multimodel analysis involve use of multiple regression, which does not consider overall distribution in the splintering of the population. Mixing distributions, which attempts to find a linear combination of usually normal distributions is another multimodel method which may result in data which have a nonnormal distribution. The method described in this article estimates the density function on each subpopulation and then assigns weights based on knowledge of the subpopulations before the densities are combined. In density estimation, calculation of window size is based on additional random variables, and then a determination is made of how many modes are required. The ramifications of a theorem which considers that several density estimates of the density functions are continuous and bounded and K(x) is a known density function that is bounded and symmetric and has compact support are described. The simulation example generates sample means and variances on 3 subpopulations P1, P2, and P3. P1 is normal with a mean of 5.01 and variance of .09 squared, and P3 is normal with a mean of 6.23 and variance of .11 squared. Means and variances are generated with and without stratification, and a comparison is made by computing the integrated mean square error. The result is a reduction in error by a factor of 10 using stratification. The example is based on a clinical trial of 100 people to compare safety and efficacy of antidepressants.