A cautionary case: osteoporotic femur fracture after robotic-assisted total knee arthroplasty

Osteoporosis is a disease of low bone mass which leads to an increased risk of fracture. The prevalence of osteoporosis is greatest amongst postmenopausal women who also form a large portion of patients with osteoarthritis presenting for total knee arthroplasty. We present a case of a 77-year-old female with knee osteoarthritis and osteoporosis on denosumab who underwent robotic-assisted total knee arthroplasty. Postoperatively, she suffered an atraumatic midshaft femur fracture at the pin site associated with the use of intraoperative robotic technology. This is a unique complication related to the use of robotics which requires additional skeletal fixation versus standard arthroplasty surgical technique. We recommend caution and further study regarding the use of robotics in total knee arthroplasty in the setting of osteoporosis and denosumab use.

Polycystin expression in the kidney and other tissues: complexity, consensus and controversy

PKD1, the major gene mutated in autosomal dominant polycystic kidney disease, was identified in 1994, and fully sequenced in 1995. The protein which it encodes, polycystin-1, is the first member of a new family of proteins, whose functions presently remain unclear. This review seeks to highlight the difficulties researchers studying polycystin-1 have faced and to summarize the current areas of consensus and controversy between different groups, particularly with regard to the expression pattern, subcellular location and biochemical characterization of polycystin-1. Where relevant, more recent data regarding polycystin-2, the protein encoded by PKD2, will also be discussed.

Polycystin-1 expression in PKD1, early-onset PKD1, and TSC2/PKD1 cystic tissue

BACKGROUND

The mutational mechanism responsible for cyst formation in polycystic kidney disease 1 gene (PKD1) remains controversial, with data indicating a two-hit mechanism, but also evidence of polycystin-1 expression in cystic tissue.

METHODS

To investigate this apparent paradox, we analyzed polycystin-1 expression in cystic renal or liver tissue from 10 patients with truncating PKD1 mutations (including one early-onset case) and 2 patients with severe disease associated with contiguous deletions of TSC2 and PKD1, using monoclonal antibodies (mAbs) to both extreme N-(7e12) and C-terminal (PKS-A) regions of the protein. Truncation of the C-terminal epitope from the putative mutant proteins in each case allowed exclusive assessment of the nontruncated protein with PKS-A.

RESULTS

In adult PKD1 tissue, the majority of cysts (approximately 80%) showed polycystin-1 expression, although staining was absent in a variable but significant minority (approximately 20%), in spite of the normal expression of marker proteins. Unlike adult PKD1, however, negative cysts were rarely found in infantile PKD1 or TSC2/PKD1 deletion cases.

CONCLUSIONS

If a two-hit mutational mechanism is operational, these results suggest that the majority of somatic mutations in adult PKD1 are likely to be missense changes. The low level of polycystin-1-negative cysts in the three "early-onset" cases, however, suggests that a somatic PKD1 mutation may not always be required for cyst formation.

Coordinate expression of the autosomal dominant polycystic kidney disease proteins, polycystin-2 and polycystin-1, in normal and cystic tissue

A second gene for autosomal dominant polycystic kidney disease (ADPKD), PKD2, has been recently identified. Using antisera raised to the human PKD2 protein, polycystin-2, we describe for the first time its distribution in human fetal tissues, as well as its expression in adult kidney and polycystic PKD2 tissues. Its expression pattern is correlated with that of the PKD1 protein, polycystin-1. In normal kidney, expression of polycystin-2 strikingly parallels that of polycystin-1, with prominent expression by maturing proximal and distal tubules during development, but with a more pronounced distal pattern in adult life. In nonrenal tissues expression of both polycystin molecules is identical and especially notable in the developing epithelial structures of the pancreas, liver, lung, bowel, brain, reproductive organs, placenta, and thymus. Of interest, nonepithelial cell types such as vascular smooth muscle, skeletal muscle, myocardial cells, and neurons also express both proteins. In PKD2 cystic kidney and liver, we find polycystin-2 expression in the majority of cysts, although a significant minority are negative, a pattern mirrored by the PKD1 protein. The continued expression of polycystin-2 in PKD2 cysts is similar to that seen by polycystin-1 in PKD1 cysts, but contrasts with the reported absence of polycystin-2 expression in the renal cysts of Pkd2+/- mice. These results suggest that if a two-hit mechanism is required for cyst formation in PKD2 there is a high rate of somatic missense mutation. The coordinate presence or loss of both polycystin molecules in the same cysts supports previous experimental evidence that heterotypic interactions may stabilize these proteins.

Characterisation and expression of the PKD-1 protein, polycystin, in renal and extrarenal tissues

Studies of the nature and expression of the PKD-1 gene product, polycystin, have been complicated by duplication of the PKD-1 gene, the low expression of the PKD-1 gene in adult tissues and the lack of antibodies to epitopes in the duplicated region of polycystin. Using five monoclonal and polyclonal antibodies to epitopes encompassing the whole of the polycystin molecule, we have studied the biochemical and cellular expression of polycystin, and sought to identify homologues and alternative splice forms of polycystin. We find that polycystin exists as a 400-500 kDa protein in normal kidney and in a range of renal epithelial cell lines by immunoblotting, using antibodies to four different epitopes. No evidence of translated products from the genes (HG) homologous to PKD-1 nor any major splice forms of polycystin was found. In renal cells, polycystin could be detected as a cell surface protein but significant intracellular concentrations were also found by cellular fractionation and immunofluorescence. In normal and PKD-1 fetal tissues, immunoreactive polycystin was detected in many different cell types outside the kidney including vascular smooth muscle cells, endothelium, pancreatic, biliary and respiratory ductal epithelia, thyroidal epithelium, endocardium, myocardium, oocytes and Leydig cells. In the developing mouse kidney, polycystin expression is seen in all nephron segments but expression becomes restricted to mature distal tubules and collecting ducts in adult life. These observations clarify the nature of the PKD-1 protein, provide a molecular basis for understanding the systemic nature of PKD-1 and may explain the known phenotypic difference in the nature of renal cysts between 'early-onset' cases and typical adult-onset disease.

Regional variations in endothelin-1 and its receptor subtypes in human coronary vasculature: pathophysiological implications in coronary disease

Endothelin-1 is a potent vasoconstrictor peptide and mitogen for vascular smooth muscle cells. Increased plasma or tissue levels of endothelin-1 have been described after myocardial infarction and in atherosclerosis, suggesting that this peptide may play a pathophysiological role in various coronary syndromes. Here, we have studied regional variations in ET-1 and its receptors in control and atherosclerotic human coronary vasculature using standard immunohistochemistry and in vitro autoradiography. ET-1 immunoreactivity was associated with luminal endothelial cells and smooth muscle cells at regions of atherosclerosis. ET(A) receptors were present on smooth muscle cells of coronary arteries and on cardiac myocytes. Medial ET(B) receptor binding at the proximal region of coronary arteries was weak, but increased significantly towards distal regions of this vessel (p<0.005 in control and p<0.0005 in ischaemic heart disease). Microvascular endothelial cells in the adventitia of coronary arteries, myocardial microvessels and the endocardial endothelium expressed the ET(B) receptor exclusively. The receptor variations revealed in this study provide supporting evidence that ET-1 is associated with (1) vascular smooth muscle and endothelial cell proliferation, including areas of intimal hyperplasia and regions of neovascularization (2) increased ET-1-induced reactivity of distal portions of the human coronary artery, (3) ET-1-mediated constriction of myocardial microvessels. These results provide new insights into different potential roles for this peptide in healthy and diseased human coronary vasculature.

Identification of mutations in the duplicated region of the polycystic kidney disease 1 gene (PKD1) by a novel approach

Mutation screening of the major autosomal dominant polycystic kidney disease gene (PKD1) has been complicated by the large transcript size (> 14 kb) and by reiteration of the genomic area encoding 75% of the protein on the same chromosome (the HG loci). The sequence similarity between the PKD1 and HG regions has precluded specific analysis of the duplicated region of PKD1, and consequently all previously described mutations map to the unique 3' region of PKD1. We have now developed a novel anchored reverse-transcription-PCR (RT-PCR) approach to specifically amplify duplicated regions of PKD1, employing one primer situated within the single-copy region and one within the reiterated area. This strategy has been incorporated in a mutation screen of 100 patients for more than half of the PKD1 exons (exons 22-46; 37% of the coding region), including 11 (exons 22-32) within the duplicated gene region, by use of the protein-truncation test (PTT). Sixty of these patients also were screened for missense changes, by use of the nonisotopic RNase cleavage assay (NIRCA), in exons 23-36. Eleven mutations have been identified, six within the duplicated region, and these consist of three stop mutations, three frameshifting deletions of a single nucleotide, two splicing defects, and three possible missense changes. Each mutation was detected in just one family (although one has been described elsewhere); no mutation hot spot was identified. The nature and distribution of mutations, plus the lack of a clear phenotype/genotype correlation, suggest that they may inactivate the molecule. RT-PCR/PTT proved to be a rapid and efficient method to detect PKD1 mutations (differentiating pathogenic changes from polymorphisms), and we recommend this procedure as a firstpass mutation screen in this disorder.

A stable, nonsense mutation associated with a case of infantile onset polycystic kidney disease 1 (PKD1)

Autosomal dominant polycystic kidney disease (ADPKD) is the most common single gene disorder resulting in renal failure. It is generally an adult onset disease, but rarely, cases of severe childhood polycystic disease arise in ADPKD families. The clear clinical anticipation in these pedigrees has led to the suggestion that the mutation may be an unstable trinucleotide repeat. We have now identified a nonsense mutation, Tyr3818Stop, in one such family (P117) within the major ADPKD gene, polycystic kidney disease 1 (PKD1). The mutation is shown to be a de novo change in the father, and of grandpaternal origin. PKD1 manifests as typical adult onset disease in the father, but is seen as severe disease, detected as enlarged polycystic kidneys in utero, in one of a pair of dizygotic twins; the other twin has the mutation but no evidence of cysts, consistent with an adult onset disease course. The finding of the same stable mutation associated with very different disease severity in this family indicates that phenotypic variation in PKD1 is not due to a dynamic mutation. It seems most likely that a small number of modifying factors may radically affect the course of disease in PKD1; identification of such factors will have important prognostic implications in this disorder.

Polycystin, the polycystic kidney disease 1 protein, is expressed by epithelial cells in fetal, adult, and polycystic kidney

Polycystic kidney disease 1 (PKD1) is the major locus of the common genetic disorder autosomal dominant polycystic kidney disease. We have studied PKD1 mRNA, with an RNase protection assay, and found widespread expression in adult tissue, with high levels in brain and moderate signal in kidney. Expression of the PKD1 protein, polycystin, was assessed in kidney using monoclonal antibodies to a recombinant protein containing the C terminus of the molecule. In fetal and adult kidney, staining is restricted to epithelial cells. Expression in the developing nephron is most prominent in mature tubules, with lesser staining in Bowman's capsule and the proximal ureteric bud. In the nephrogenic zone, detectable signal was observed in comma- and S-shaped bodies as well as the distal branches of the ureteric bud. By contrast, uninduced mesenchyme and glomerular tufts showed no staining. In later fetal (>20 weeks) and adult kidney, strong staining persists in cortical tubules with moderate staining detected in the loops of Henle and collecting ducts. These results suggest that polycystin's major role is in the maintenance of renal epithelial differentiation and organization from early fetal life. Interestingly, polycystin expression, monitored at the mRNA level and by immunohistochemistry, appears higher in cystic epithelia, indicating that the disease does not result from complete loss of the protein.

Candidate 56 and 58 kDa protein(s) responsible for mediating the renal defects in oncogenic hypophosphatemic osteomalacia

The effects of tumor-conditioned media (TCM) derived from cultured cells from an oncogenic hypophosphatemic osteomalacia (OHO) tumor on transformed human kidney cells were investigated. Dose-dependent cell detachment and aggregation occurred in kidney cells cultured in serum-free medium supplemented with TCM, but not in skin fibroblast controls, or in kidney cells cultured in the presence of serum. Kidney cells exposed to TCM in the presence of serum (0.5%) had reduced Na(+)-dependent phosphate cotransport (36%, p < 0.04) and increased 1alpha-hydroxylase activity (48%, p < 0.05). In contrast, TCM had no significant effect on Na(+)-dependent alpha-methyl-glucose transport. To investigate these effects further, serum from an OHO patient, before and after tumor resection, was used to raise polyclonal antiserum to tumor-derived products (preoperative and postoperative antiserum, respectively). Changes in Na(+)-dependent phosphate cotransport and vitamin D metabolism induced by TCM were prevented by the addition of preoperative but not postoperative antisera. Furthermore, Western analysis revealed the presence of two proteins (56-58 kDa) in TCM media screened with preoperative antisera. These proteins were not detected by postoperative antisera and were absent in skin fibroblast control media. Direct inhibition of Na(+)-dependent phosphate cotransport by phosphonoformic acid did not affect 1,25-dihydroxy vitamin D(3) synthesis. These studies provide support for a circulating component affecting phosphate handling and vitamin D metabolism in OHO.

Screening the 3' region of the polycystic kidney disease 1 (PKD1) gene reveals six novel mutations

Recently, the gene for the most common form of autosomal dominant polycystic kidney disease (ADPKD), PKD1 (polycystic kidney disease 1), has been fully characterized and shown to encode an integral membrane protein, polycystin, involved in cell-cell and/or cell-matrix interactions. Study of the PKD1 gene has been complicated because most of the gene lies in a genomic region reiterated several times elsewhere on the same chromosome, and consequently only seven mutations have been described so far. Here we report a systematic screen covering approximately 80% of the approximately 2.75 kb of translated transcript that is encoded by single-copy DNA. We have identified and characterized six novel mutations that, together with the previously described changes, amount to a detection rate of 10%-15% in the population studied. The newly described mutations are two deletions, an insertion of a T-nucleotide causing a frame shift, two single-base-pair substitutions resulting in premature stop codons, and a G-->C transversion that may be a missense mutation. These results have important implications for genetic diagnosis of PKD1 because they indicate that the majority of mutations lie within the duplicated area, which is difficult to study. The regions of polycystin removed in each mutation so far described are assessed for their functional significance; an area disrupted by two new small in-frame changes is highlighted. PKD1 mutations are contrasted with those in the PKD1/TSC2 contiguous-gene syndrome, and the likely mutational mechanism in PKD1 is considered.

An endothelin-1 mediated autocrine growth loop involved in human renal tubular regeneration

Renal tubules have the capacity to regenerate following injury. We have investigated the possibility that tubular-derived endothelins, acting as autocrine growth factors, may be involved in this response in human kidney. ET-1 immunoreactivity was demonstrated by immunohistochemical staining in proximal tubules, distal cortical tubules and medullary collecting ducts of human kidney. In cultured human renal proximal tubular cells, RNAase protection assays demonstrated the expression of ET-1 and ET-2 mRNA's, and radioimmunoassay, following separation of conditioned medium by reverse phase HPLC, showed immunoreactive material which co-eluted with ET-1 and ET-2. Competition binding studies revealed the presence of at least two types of endothelin receptor: one with high and one with low affinity for ET-3 relative to ET-1. Analysis of cellular RNA by RT-PCR demonstrated expression of mRNA's for both ETA and ETB receptor subtypes. Combined blockade of ETA and ETB receptors (by PD-145065) but not that of ETA receptors alone (by BQ-123) blocked the mitogenic effect of exogenous or endogenous ET-1 and also profoundly suppressed endogenous ET-1 synthesis. By contrast, incubation with the ETB receptor agonist, BQ-3020, stimulated endogenous ET-1 synthesis. Exposure of the cells to hypoxia (1% O2 for 16 to 24 hr) resulted in specific up-regulation of ET-1 but not ET-2 gene expression. These findings reveal the existence of a hypoxia-inducible, autocrine growth system in human proximal tubular cells, which is mediated by ET-1 through the ETB receptor, and which could function in vivo as an autoregenerative system for restoring tubular integrity after injury. The widespread distribution of ET-1 peptide in different tubular segment suggests that ET-1 mediated tubular regeneration may also occur in other nephron segments.

Human high density lipoproteins stimulate endothelin-1 release by cultured human renal proximal tubular cells

The vasoconstrictive and mitogenic actions of endothelins have been implicated in the pathogenesis of progressive renal disease. In the present study, we have assessed whether plasma high density lipoproteins (HDL), the major filtered urinary lipoprotein in nephrotic states, can influence endothelin-1 (ET-1) production by cultured human renal proximal tubular cells. Human HDL was found to stimulate ET-1 secretion up to fourfold in a dose- and time-dependent manner; the effect was greater in subconfluent cultures than in confluent ones. There was little difference between the stimulatory effect of HDL2 and the major HDL subclass, HDL3. Preincubation of the cells with albumin did not abolish the HDL effect, while partially- or fully-delipidated HDL3 largely reproduced the effect of whole HDL3. These findings suggest that stimulation of ET-1 secretion was not simply due to protein or lipid repletion of the cells. Rather, the effect was mediated by HDL apolipoproteins, although binding to the HDL receptors involved in cellular cholesterol homeostasis was not required as tetranitromethane-modified HDL3 was an equally effective agonist of ET-1 release. Apolipoprotein (apo) A-I was indirectly implicated in the process since modified HDL3 in which apoA-II largely replaced apoA-I was less potent than HDL3. A one hour exposure of the cells to HDL3 was sufficient to activate ET-1 production for the following 12 hours, although maximum activation required six hours.(ABSTRACT TRUNCATED AT 250 WORDS)

Tubular lipidosis: epiphenomenon or pathogenetic lesion in human renal disease?

Tubular lipidosis is a commonly observed histological lesion in proteinuric renal diseases. We have studied the interaction between native and modified human lipoproteins and human renal proximal tubular cells to investigate whether lipoproteins could be injurious to tubular cells in culture. Human renal proximal tubular cells were cultured and characterized by established methods. Preliminary studies showed that these cells could take up and degrade normal human lipoproteins by high affinity (HDL) and low affinity (LDL) pathways. In subconfluent culture, native lipoproteins, that is, LDL, HDL2 and HDL3, had markedly different effects on cell growth as measured by 3H-thymidine uptake and total cell protein as compared to modified lipoproteins such as minimally modified and oxidized LDL. In addition, we found that renal tubular cells could oxidized native LDL in the presence of copper largely by a superoxide-mediated mechanism. Finally, cellular accumulation of lipid was demonstrated in vitro by incubating cultured cells with varying lipoprotein concentrations for up to 48 hours. Notably, cell detachment was observed only with high concentrations of modified LDL especially with minimally modified LDL. We speculate that uptake and oxidation of filtered LDL by tubular cells may lead to tubular injury in nephrotic states.

Tubular-derived growth factors and cytokines in the pathogenesis of tubulointerstitial fibrosis: implications for human renal disease progression

Detailed histomorphometric analysis of human biopsy tissue over the last 30 years has convincingly demonstrated that preservation of the tubulointerstitial compartment of the kidney is the major determinant of renal outcome in a variety of human renal diseases. Nevertheless, the pathophysiology of tubulointerstitial disease remains obscure. In particular, the primary role of tubular injury has not been explored adequately. There is now accumulating evidence that apart from their many transport functions, tubular cells also secrete an array of cytokines, including chemotactic cytokines, polypeptide growth factors, and vasoactive peptides. Three paracrine growth systems acting at different levels in the kidney are described as examples of potential interactions between tubular and interstitial cells in health and disease. We hypothesize that while glomerular injury may precede tubular injury, it is tubular injury that sets into motion the irreversible process of tubulointerstitial fibrosis characteristic of progressive human renal disease, leading to secondary loss of glomerular function.

Mechanisms of tubulo-interstitial injury in progressive renal diseases

A vast amount of evidence, based upon human renal biopsy material, indicates that the presence of tubular atrophy and interstitial fibrosis is a better indicator of outcome of renal function than is the extent of glomerular sclerosis. The pathophysiological basis for this surprising fact has not been adequately addressed. In this review we point out that the systemic hypertension which accompanies most forms of chronic renal disease could impact adversely upon the vasodilated interstitial vascular compartment which, together with a component of primary capillary injury related to the disease process, could cause progressive obliteration of particular capillaries. This would initiate a process of chronic tubular ischaemia ultimately leading to tubular atrophy. Since tubular cells have been shown to produce an array of cytokines and growth factors which modulate fibroblast proliferation, extracellular matrix production and chemo-attracts for infiltrating cells, it is further proposed that it is the tubular injury which initiates the deleterious cascade of events. Tubular injury may be aggravated by the filtration of potentially 'noxious' molecules through the diseased glomerulus and by infiltrating cells. As the vascular bed into which glomerular blood flow empties is progressively obliterated, glomerular function declines and renal failure advances in relation to the degree of tubulo-interstitial fibrosis.

Effect of cyclosporin A on endothelin synthesis by cultured human renal cortical epithelial cells

We report here for the first time that human renal proximal tubular cells secrete endothelin, clear evidence of de-novo endothelin synthesis by these cells and the effect of cyclosporin A (CsA) on endothelin synthesis both in short-term (24 h) and medium-term (5-day) culture. Human renal cortical epithelial cells were cultured and shown to possess proximal tubular characteristics. These cells produced endothelin in culture in a time-dependent manner, as measured by radioimmunoassay (291.6 +/- 51.4 pg/well/24 h). Furthermore, endothelin production by these cells was significantly decreased by up to 80% by cycloheximide (1051.8 +/- 54.9 pg/mg cell protein/24 h versus 253.2 +/- 12.6 pg/mg cell protein/24 h), showing that these cells actively synthesize endothelin. In short-term culture (24 h), CsA significantly inhibited endothelin synthesis at a medium concentration of 10,000 micrograms/l. No change in endothelin synthesis was seen at lower CsA concentrations. In contrast, over a 5-day period, a non-significant increase in endothelin synthesis was observed at CsA concentrations of 2000 micrograms/l (152.5 +/- 20.4%); however, cell growth was significantly decreased at this concentration (71.33 +/- 6.39%). Using a newly developed two-site immunoradiometric assay specific for endothelin-1 (ET-1), we demonstrate that ET-1 is the major endothelin isoform produced by human renal proximal tubular cells.

Sinus arrest and asystole due to severe lithium intoxication

We report the occurrence of sinus arrest and asystole due to severe lithium intoxication in a patient with preexisting conduction tissue disease. Lithium should be used with extreme caution and frequent monitoring in patients with conduction tissue disease.

The pathogenesis of the ovarian hyperstimulation syndrome (OHS): a possible role for ovarian renin

Two patients with severe ovarian hyperstimulation syndrome are described. Increased plasma concentrations of immunoradiometrically determined total renin are shown, together with greatly increased plasma levels of active renin and aldosterone. These very high values for total renin, renin activity and aldosterone were not suppressed when extracellular compartments were greatly expanded; the values subsequently declined to normal levels, despite the use of diuretics. This suggested that the renin was of non-renal origin since its production was apparently unaffected by influences which control juxtaglomerular secretion. The high concentrations of the renin-angiotensin-aldosterone system suggest that it contributes to the genesis of the ovarian hyperstimulation syndrome.

Atrial natriuretic peptide release responds to atrial stretch and not to atrial pressure: observations during pericardiocentesis in a young woman

The effect of pericardiocentesis on the plasma concentration of atrial natriuretic peptide was studied in a patient with cardiac tamponade. Plasma atrial natriuretic peptide concentrations rose sharply from 42 pmol l-1 to a maximum of 150 pmol l-1 with a corresponding fall in mean right atrial pressure from 20 mmHg to 6 mmHg. Our results suggest that atrial 'stretch' (or transmural pressure) rather than atrial pressure per se, is the primary stimulus for atrial natriuretic peptide release in man.

Effects of noradrenaline infusion on platelet catecholamine levels and platelet aggregation

Noradrenaline infusions were administered to 10 normal subjects through stepped doses and continued at 60 ng/kg per min for 2 h. Plasma noradrenaline rose from 1.7 +/- 0.3 to 7.7 +/- 0.7 pmol/ml and platelet noradrenaline rose from 2.1 +/- 0.2 to 2.6 +/- 0.2 pmol/mg protein. There was no change in plasma or platelet adrenaline. Platelet aggregation studies using ADP, adrenaline, collagen and thrombin as aggregants showed no overall change during the course of the infusion. Blood was sampled from a heated hand vein (hot box at 60 degrees C) to test the degree of arterialization. Plasma noradrenaline and blood pO2 and pCO2 showed intermediate levels at this sampling site compared with venous and true arterial values. Changes in platelet noradrenaline content can occur over 2 h when plasma levels are considerably increased by noradrenaline infusion. No change in platelet sensitivity to aggregation was observed. The heated hand vein did not provide true arterial levels of noradrenaline.

Hypersensitivity vasculitis complicating intravenous streptokinase therapy in acute myocardial infarction

A 52-year-old man developed a hypersensitivity vasculitic rash on his legs nine days after receiving intravenous streptokinase therapy for acute myocardial infarction. The histological and immunological features and the differential diagnosis of this unusual complication of streptokinase therapy are reviewed.