Southern Hemisphere climate variability forced by Northern Hemisphere ice-sheet topography.

The presence of large Northern Hemisphere ice sheets and reduced greenhouse gas concentrations during the Last Glacial Maximum fundamentally altered global ocean-atmosphere climate dynamics. Model simulations and palaeoclimate records suggest that glacial boundary conditions affected the El Niño-Southern Oscillation, a dominant source of short-term global climate variability. Yet little is known about changes in short-term climate variability at mid- to high latitudes. Here we use a high-resolution water isotope record from West Antarctica to demonstrate that interannual to decadal climate variability at high southern latitudes was almost twice as large at the Last Glacial Maximum as during the ensuing Holocene epoch (the past 11,700 years). Climate model simulations indicate that this increased variability reflects an increase in the teleconnection strength between the tropical Pacific and West Antarctica, owing to a shift in the mean location of tropical convection. This shift, in turn, can be attributed to the influence of topography and albedo of the North American ice sheets on atmospheric circulation. As the planet deglaciated, the largest and most abrupt decline in teleconnection strength occurred between approximately 16,000 years and 15,000 years ago, followed by a slower decline into the early Holocene.

Inability of cytomegalovirus infection of cultured endothelial cells to induce HLA class II antigen expression.

Cytomegalovirus infection in the renal allograft recipient has been associated with the initiation of acute rejection. The mechanism of this induction is unknown. It may be related to renal tubular epithelial and endothelial expression of HLA class II antigens or a CMV immediate-early antigen that exhibits immunologic crossreactivity with HLA DR. In this study the ability of CMV to both infect and subsequently induce class II antigen expression on cultured human umbilical-vein endothelial cells (HUVEs), in the absence of cytokines, particularly gamma interferon, was tested. Individual HUVE cell lines were first proven to express HLA class II antigens in the presence of 10, 100, and 200 units of recombinant IFN-gamma as early as 24 hr postincubation by an immunohistochemical technique and by flow cytometry. These cell lines were successfully infected with CMV strains AD169 and CMV3 as determined by the presence of early and late viral antigens and CMV DNA. The degree of infection was dose and incubation-time dependent. Infection of HUVEs with these CMV strains and a nonattenuated clinical isolate failed to induce HLA DR, DP, or DQw1 expression in the absence of IFN-gamma. These findings support the hypothesis that endothelial cells in vivo may serve as reservoirs of CMV infection. They do not support the hypothesis that CMV produces an immediate-early antigen that has immunologic cross-reactivity with HLA DR. Furthermore, there is no support for the hypothesis that CMV independently induces HLA class II antigens in the absence of IFN-gamma.

Infection of endovascular stents in a swine model.

BACKGROUND: Percutaneous balloon angioplasty with intravascular metallic stent placement has rapidly gained popularity for the treatment of arterial occlusive disease. Although the incidence of vascular prosthetic infection is well described, the risk of infection following metallic stent placement is unknown. The purpose of this study was to determine if intravascular metallic stents could become infected following systemic bacterial challenge. METHODS: Balloon expandable metallic stents were implanted in the iliac arteries of 10 swine following balloon catheter angioplasty. A second angioplasty, without stent placement, was also performed in the contralateral iliac artery. A bacterial challenge with Staphylococcus aureus was then infused into the aorta immediately after stent placement. Group 1 (n = 5) was killed at 72 hours, and group 2 (n = 5) at 3 weeks. A third group (n = 5) underwent stent placement without bacterial challenge and was killed at 3 weeks. At the time of death, the stents were cultured, and both iliac arteries were submitted for pathologic examination. Arterial patency and evidence of systemic infection were also assessed. RESULTS: In the animals sacrificed at 72 hours (group 1), 80% had stent cultures with significant growth of S aureus; while at 3 weeks (group 2), 60% of cultures were positive. Of the stents placed without bacterial challenge (group 3), none had a positive culture at 3 weeks. In group 2, 40% of the stented arteries remained patent, while 100% of group 3 remained patent until sacrifice at 3 weeks. All of the stented arteries which were patent at 3 weeks were culture negative, while all those which were thrombosed were culture positive for S aureus. When compared to angioplasty alone, the presence of a stent was strongly associated with pathologic evidence of inflammation [93% versus 7%]. The quality of inflammation in the stented groups also differed. Ninety percent of the stented arteries in groups 1 and 2 had acute inflammation, compared to only 20% in group 3. The remainder of the stented arteries in group 3 had chronic inflammation or were normal. CONCLUSION: In the swine model, intravascular metallic stents have the potential to become infected. This is associated with acute inflammation of the arterial wall and vessel thrombosis. Further studies evaluating the incidence of stent infections in humans are needed.

Endovascular stent infection with delayed bacterial challenge.

BACKGROUND: Reports of endovascular stent infection have recently been described. The purpose of this study was to determine if intravascular metallic stents in a swine model could become infected following a bacterial challenge given remote from the time of stent placement. METHODS: Balloon expandable metallic stents (Palmaz) were implanted in the iliac arteries of 14 swine. An angioplasty, without stent placement, was also performed in the contralateral iliac artery. An intravenous bacterial challenge with Staphylococcus aureus was given 4 weeks after stent placement. Euthanasia was performed 72 hours after the bacterial challenge. At the time of euthanasia, the iliac artery/stent complex and the contralateral angioplastied iliac artery were harvested and sent for microbiologic and pathologic analysis. RESULTS: Seven of the 14 stent/artery complexes were culture positive for S aureus whereas only one of the 14 angioplastied arteries was positive for S aureus (P = 0.03). On histologic examination, 6 of the 14 stent/artery complexes had evidence of acute inflammatory changes in the arterial wall. This compares with only 1 of 14 angioplastied arteries having evidence of inflammatory infiltrate in the arterial wall (P = 0.07). All 6 of the stent/artery complexes with inflammatory infiltrate were culture positive. CONCLUSION: In the swine model, intravascular metallic stents have the potential to become infected when a bacterial challenge is given 4 weeks after stent placement. Further studies evaluating the incidence of stent infections in humans are needed.

Infectibility of endovascular stents following antibiotic prophylaxis or after arterial wall incorporation.

BACKGROUND: Case reports of endovascular stent infection have been accumulating in the last several years. We sought to determine if prophylactic antibiotics would prevent stent/artery complex infections in a swine model if given before a bacterial challenge at the time of stent placement and 4 weeks following deployment. We also investigated whether arterial wall incorporation protected the stent against infection without antibiotic prophylaxis. METHODS: Balloon expandable Palmaz stents were placed in the iliac arteries of 42 swine. At the same time, angioplasty was performed on the contralateral iliac artery as a control. In group A, prophylactic cefazolin was given to 12 swine at the time of stent deployment followed by an intraaortic bacterial challenge of Staphylococcus aureus. In group B, 10 swine received prophylactic cefazolin followed by intravenous S aureus 4 weeks after iliac stenting and angioplasty. In group C, 3 months following iliac stent placement and angioplasty an intravenous bacterial challenge was administered with S aureus. All swine were euthanized, and the iliac stent/artery complex and the contralateral angioplastied iliac artery were harvested and sent for culture and pathology. Experimental groups were compared with results from our previously published swine infection model using the Fisher's exact test. P values were considered significant if less than 0.05. RESULTS: Group A: Two of the 12 (17%) stent/artery complexes in the antibiotic treatment group had positive cultures. This compares with 7 of 10 (70%) in the control group (P = 0.016). In addition, there was one infection in an angioplastied vessel contralateral to one of the two stent infections. Molecular strain typing verified that the positive cultures were the same strain that was used to challenge the animals. No vessel thrombosis occurred in the stented arteries even in the presence of infection. Group B: One of 10 (10%) stented iliac arteries had a culture positive infection. This compares with 7 of 14 (50%) positive cultures in the control group (P = 0.04). In addition, one angioplastied vessel did have mild S aureus growth on culture. Both positive cultures were verified to be the same as the injected strain by molecular strain typing. There were no thrombosed or occluded vessels. Group C: One of 15 patent stents had growth of S aureus on culture and evidence of acute inflammation on histopathologic examination. The stent infection rate of 1 of 15 (7%) patent stents in this study was significantly less than the infection rates with bacterial challenge at placement (7 of 10, 70%; P < 0.01) and at 1 month postplacement (7 of 14, 50%; P = 0.0142). Five stents occluded without evidence of infectious cause. CONCLUSIONS: The results of this study support a recommendation that antibiotic prophylaxis should be used at the time of arterial stent placement and early after placement at times of anticipated bacteremia, but indefinite prophylaxis may be unnecessary due to arterial wall incorporation of the stent.

Cytomegalovirus infection of gastrointestinal endothelium demonstrated by simultaneous nucleic acid hybridization and immunohistochemistry.

To demonstrate the degree of involvement of endothelial cells in cytomegalovirus (CMV) infection of the gastrointestinal tract we have stained sections from gastrointestinal specimens that showed inclusion bodies on hematoxylin-eosin staining. Factor VIII was first detected using a rabbit anti-factor VIII primary antibody and an alkaline phosphatase-labeled sheep anti-rabbit secondary antibody. The CMV was then visualized with a biotin-labeled CMV probe detected by a streptavidin peroxidase technique with aminoethyl carbazole as the chromogen. Factor VIII staining was a bright blue and CMV a brick red. The specimens included one small-bowel resection and four colonic resections, as well as an esophageal biopsy. The patients' diagnoses included bone marrow transplant recipient, acquired immunodeficiency syndrome, ulcerative colitis, and renal transplant recipient. Cells positive for both CMV and factor VIII ranged from 35% to 60% of positive cells in a representative section, and the relative percentages (mean +/- SE) for cell type for infected cells were: endothelial, 48.9 +/- 4.5; vascular luminal (factor VIII negative), 6.1 +/- 1.7; perivascular (factor VIII negative in vascular wall), 16.2 +/- 3.2; and other cell (non-vascular factor VIII negative), 28.9 +/- 5.1. These findings and clustering of infected cells around the vessels provide evidence that CMV infection of the gastrointestinal tract is primarily vasculitic and related to infection of endothelial cells.

Changes in calcium distribution during therapeutic plasmapheresis.

Observations of normal persons indicated that there was a net shift of serum calcium from the ionized and protein-bound fractions to the anion-bound fraction when these persons received citrate during donor plasmapheresis. To confirm these findings and to extend them to diseased persons, we measured total, ionized, and ultrafilterable Ca concentrations in 11 patients who were undergoing therapeutic plasmapheresis. Anticoagulation was achieved with anticoagulant citrate dextrose solution. Mean concentrations of total, ionized, and protein-bound Ca decreased, whereas the ultrafilterable and anion-bound fractions increased. Our results suggest that protein-bound Ca may be relatively labile and may help to maintain constant levels of ionized Ca. Also, the observed increase in the anion-bound Ca level with citrate infusion suggests a shift from the protein-bound and ionized Ca fractions to this pool.

In situ DNA hybridization for cytomegalovirus in colonoscopic biopsies.

We reviewed colonoscopic biopsies of the lower gastrointestinal tract performed during a two-year period. Those representing neoplasia were excluded. Formalin-fixed paraffin-embedded biopsy specimens from 53 patients were studied by in situ DNA hybridization for cytomegalovirus (CMV) using commercially available biotinylated DNA probes detected by an avidin-biotin peroxidase technique. Nine of the patients were severely immunocompromised: four had acquired immunodeficiency syndrome, three had ulcerative colitis and were receiving high-dose steroid therapy, one was a bone marrow transplant recipient, and one had idiopathic pulmonary fibrosis and was receiving therapy with prednisone and cyclophosphamide. Four of these had evidence of CMV infection by routine histology and DNA hybridization. Three additional immunocompromised patients had evidence of CMV infection by DNA hybridization alone. Forty-four patients had inflammatory conditions or ulcerations of the lower gastrointestinal tract. Six of these had evidence of CMV by DNA hybridization alone. Histologically normal as well as enlarged and cytomegalic cells were probe positive, and the cells were sparse to numerous in number. They were found in the epithelium and/or lamina propria. This technique was demonstrated to be applicable to routinely processed colonic biopsy specimens.

Inverted papilloma of the middle ear.

Inverted papilloma is a benign neoplasm confined almost exclusively to the sinonasal tract. We present the first known case report of an inverted papilloma arising from the middle ear. In this paper we discuss the pathophysiology and review the literature of this interesting clinical entity.

Anomalous course of the median nerve medial to the trochlea and anterior to the medial epicondyle of the humerus.

It is generally assumed that in the cubital region the median nerve is directly medial to the brachial artery and that it lies anterior to the trochlea of the humerus, from which it is separated by the brachialis muscle. Three instances are noted where the median nerve passed medial to the trochlea and anterior to the medial epicondyle, deep to the pronator teres, unilateral in one subject and bilateral in another. Its possible significance in elbow trauma is commented upon.

Gentamicin/heparin interactions: effects on two immunoassays and on protein binding.

Heparin has been reported to have two different effects on gentamicin. One is the interference by heparin with the quantitation of gentamicin; the other is an increase in the bound fraction of gentamicin in the presence of serum. The reports conflict regarding what concentration of heparin causes these effects. The main objective of this study was to determine what effect heparin had on the determination of gentamicin using an enzyme immunoassay (EIA) or a fluorescence polarization immunoassay (FPIA). Heparin concentrations studied simulated the amounts that would be found in plasma when blood samples are collected with an evacuated heparinized tube (less than 30 U/ml), or the greater amounts that might be present when blood is collected from indwelling lines (100-200 U/ml). Heparin had no effect on the quantitation of gentamicin by FPIA. In measurements done by EIA, heparin had no effect at 28 U/ml but did significantly inhibit the reaction by 10-20% when its concentration was 100-200 U/ml. We studied the effects of heparin on the distribution equilibrium of gentamicin by measuring free gentamicin produced in an ultrafiltrate. It was found that the bound fraction increased as the heparin concentration increased. Our results with aqueous and serum-based samples suggest that the increased binding was the result of a direct binding of gentamicin to heparin, as well as a more complex interaction involving heparin, gentamicin, and proteins.

Histochemistry of sulfhydryls in acute myocardial infarction.

To study the changes in sulfhydryl and disulfide distribution in myocardial infarction we applied the fluorescent sulfhydryl reagent, monobromobimane to sections of myocardium from patients dying of infarction of 24 h to 7 days duration. Staining for both sulfhydryls and for disulfide after reduction of slides blocked with N-Ethyl maleimide showed that sulfhydryls were decreased in the infarcted areas. Disulfides were increased in the periphery of infarction but cells undergoing cytolysis showed loss of disulfide staining as well as sulfhydryl staining. The causes and implication of these changes are discussed.

The blood supply of the abdominally-placed vertical sigmoid colon.

We have recorded the blood supply to a high or subsplenic sigmoid loop from branches of the left colic and first sigmoid arteries. An accessory middle colic present in this case crossed the region immediately superior to the apex of the sigmoid loop. In this instance it arose from the dorsal pancreatic which also gave origin to inferior pancreaticoduodenal branches. Since the high sigmoid may be retroperitoneal, as in the case described, it would not be readily apparent; it would require lateral views to separate it from the descending colon radiographically.

When the marginal artery is not marginal.

In the course of dissection one does not normally pay much attention to the anastomosing loops forming the marginal artery. The observation of a significant precolic anastomosing loop on the descending colon of a spare subject came as quite a surprise. The same was noted on a second subject, an obese male with sagging abdomen and, to a lesser degree, on a third. Such an occurrence would have surgical significance, if only because the evaluation of the blood supply to the segment of bowel that was to be resected would be much more difficult. In addition, the blood supply to the distal bowel could be impaired following resection, depending upon the collateral circulation.

A proposed method of dissection of the anterolateral abdominal wall -- applied features.

The dissection described allows the segmental nerve supply of the rectus abdominis to be seen in its continuity, namely: intercostal nerves seven to eleven, the subcostal and occasionally the iliohypogastric nerve as shown in Fig. 1. The advantages of transverse or interneural incisions are obvious. The medial retraction of the lower rectus after incision of the anterior part of the rectus sheath, exposing the femoral ring and pectineal ligament is a suitable approach of femoral hernia repair. The posterior part of the dissection shows the relation of subcostal nerve and the ventral ramus of the first lumbar nerve and/or its iliohypogastric and ilioinguinal branches to the superior lumbar triangle, which is of consequence in posterior approaches to the kidney. A lateral cutaneous branch, which normally comes off the iliohypogastric nerve also crosses the field. Brief mention is made concerning the anatomy of local anesthesia for inguinal hernia repair.

Enhanced endothelial cytopathogenicity induced by a cytomegalovirus strain propagated in endothelial cells.

Endothelial involvement has been implicated in cytomegalovirus (CMV) infection, a source of major complications in immunosuppressed individuals (e.g., those with acquired immune deficiency syndrome [AIDS] and organ transplants). Traditionally, CMV has been grown in fibroblasts; however, propagation in these cells may alter characteristics of the virus. In developing an in vitro model system of CMV/endothelial cell interaction, we have addressed this issue by propagating a clinical isolate, CMV VHL 1, in human umbilical vein endothelial (HUVE) cells by serial cocultivation of heavily infected cultures with fresh HUVE monolayers and have compared its infectious properties with those of the fibroblast-raised strain, CMV AD169. In situ hybridization using a biotinylated DNA probe, as well as immunofluorescent staining for CMV-specific antigen, has confirmed infection of HUVE cells inoculated with either strain of the virus. Infection of HUVE by VHL was accompanied by dramatic cytopathology not observed in AD169-infected cells. Plaque assay of culture supernatants revealed greater virus production in VHL-infected HUVE as compared with equivalently inoculated fibroblasts. In contrast, AD169 production in inoculated fibroblasts exceeded that in HUVE. These studies demonstrate the suitability of cultured endothelial cells as a substrate for CMV propagation and suggest that a strain of virus thus propagated may offer an accurate model of CMV/endothelial cell interaction in human disease.

Unexpected death in an infant with AIDS: disseminated cytomegalovirus infection with pancarditis.

Cardiac involvement in patients with acquired immunodeficiency syndrome (AIDS) is being reported with increasing frequency, although the factors responsible for the cardiac abnormalities are rarely identified. We report a case of sudden and unexpected death of an infant with AIDS in whom histologic and virologic studies documented generalized infection with cytomegalovirus (CMV), including pancarditis, sialitis, nephritis, colitis, hepatitis, prostatitis, orchitis, myositis, pneumonitis, and meningoencephalitis. CMV was isolated from four of five tissues cultured. Lymphocytic infiltration in the region of the sinoatrial node could have been responsible for the development of a fatal cardiac arrhythmia, and the autopsy failed to reveal any other cause of death in this infant. Children infected with the human immunodeficiency virus (HIV) need to be closely monitored for cardiac complications bearing in mind that opportunistic infections in AIDS patients may cause cardiac involvement that is atypical or that is overshadowed by the primary manifestations of the infection.

Preservation of natural endothelial cytopathogenicity of cytomegalovirus by propagation in endothelial cells.

Cytomegalovirus (CMV) is a source of major complications in immunosuppressed individuals, and endothelial involvement in CMV infection is well documented. Traditionally the virus has been propagated in fibroblasts, however this process may alter CMV's characteristics, thereby limiting the fibroblast model's utility as a research tool. In our efforts to develop a more accurate in vitro model of CMV/endothelial cell interaction, we have propagated a recent isolate (CMV VHL) through multiple passages in human umbilical vein endothelial cells (HUVE) and, collaterally in neonatal human dermal fibroblasts (NHDF). Infection of HUVE inoculated with either sub-strain of the virus was confirmed by CMV-specific in situ hybridization and by immunocytochemical staining for CMV antigens. Whereas infection of HUVE by substrain VHL/E (endothelial-raised) was accompanied by dramatic cytopathology resembling that observed clinically, the endothelial cytopathic potential of VHL/F (fibroblast-raised) was lost by its 20th passage in NHDF. Similarly, the ability of VHL/F to initiate sustained productive infection in HUVE was severely attenuated; plaque assay of culture supernatants and infected cell fractions, as well as virus-specific DNA polymerase assay of cell lysates, demonstrated progressive viral reproductive activity in VHL/E-inoculated HUVE, whereas VHL/F reproduction was barely detectable. Since properties of VHL/F bear strong resemblance to those of the fibroblast-raised AD169, these studies suggest that while the fibroblast adaptation process commonly employed in the propagation of CMV restricts the host range of the virus and attenuates its spectrum of cytopathic potential, endothelial-based propagation preserves the natural endothelial cytopathogenicity of the original isolate.