New measurements of high-momentum nucleons and short-range structures in nuclei.

We present new measurements of electron scattering from high-momentum nucleons in nuclei. These data allow an improved determination of the strength of two-nucleon correlations for several nuclei, including light nuclei where clustering effects can, for the first time, be examined. The data also include the kinematic region where three-nucleon correlations are expected to dominate.

Scaling of the F2 structure function in nuclei and quark distributions at x>1.

We present new data on electron scattering from a range of nuclei taken in Hall C at Jefferson Lab. For heavy nuclei, we observe a rapid falloff in the cross section for x>1, which is sensitive to short-range contributions to the nuclear wave function, and in deep inelastic scattering corresponds to probing extremely high momentum quarks. This result agrees with higher energy muon scattering measurements, but is in sharp contrast to neutrino scattering measurements which suggested a dramatic enhancement in the distribution of the "superfast" quarks probed at x>1. The falloff at x>1 is noticeably stronger in 2H and 3He, but nearly identical for all heavier nuclei.

Probing quark-gluon interactions with transverse polarized scattering.

We have extracted QCD matrix elements from our data on doubly polarized inelastic scattering of electrons on nuclei. We find the higher twist matrix element d˜2, which arises strictly from quark-gluon interactions, to be unambiguously nonzero. The data also reveal an isospin dependence of higher twist effects if we assume that the Burkhardt-Cottingham sum rule is valid. The fundamental Bjorken sum rule obtained from the a0 matrix element is satisfied at our low momentum transfer.

Electron microscopy of human factor VIII/Von Willebrand glycoprotein: effect of reducing reagents on structure and function.

The structure of native and progressively reduced human factor VIII/von Willebrand factor (FVIII/vWF) was examined by electron microscopy and SDS gel electrophoresis and then correlated with its biological activities. Highly resolved electron micrographs of well-spaced, rotary-shadowed FVIII/vWF molecules showed their structure to consist of a very flexible filament that contains irregularly spaced small nodules. Filaments ranged from 50 to 1,150 nm with a mean length of 478 nm and lacked fixed, large globular domains as seen in fibrinogen and IgM. A population of multimeric FVIII/vWF species ranging in molecular weight from 1 to 5 million daltons and differing in size alternately by one and two subunits was observed on SDS-2% polyacrylamide-0.5% agarose gel electrophoresis. With progressive reduction of disulfide bonds by dithiothreitol (DTT), the electron microscopic size of FVIII/vWF decreased in parallel with increased electrophoretic mobility on SDS-agarose gels; between 0.1 and 0.5 mM DTT its structure changed from predominantly fibrillar species to large nodular forms. A 50% loss of vWF specific activity and FVIII procoagulant activity occurred at 0.4 mM DTT and 1 mM DTT, respectively, corresponding to the reduction of 4 and 12 disulfide bonds of the 62 disulfides per 200,000-dalton subunit. We conclude that reduction of a few critical disulfide bonds results in a major structural change by electron microscopy and a concomitant loss of approximately 50% of the vWF function.

The role of non-need factors in individual GP utilisation analysis and their implications for the pursuance of equity: a cross-country comparison.

A substantial amount of health care resources is allocated within the UK using formulae that relate funding to measures of population need. The aim of this paper is to demonstrate the importance of non-need factors in determining utilisation of services at an individual level and explore the implications inclusion of such factors has in the consideration of equity. In the paper we develop a utility model that accords a role to non-health factors in the determination of service use. A series of functions incorporating non-health factors as explanatory variables in GP utilisation functions are estimated using data from the British Household Panel Survey. The functions are decomposed to ascertain the role of service structure and examine the role of income across the four countries of the UK in explaining utilisation. The implications of our findings for the pursuance of equity in the NHS when individual choice has an explicit role are discussed.

Studies on human antihemophilic factor. Evidence for a covalently linked subunit structure.

When purified antihemophilic factor (Factor VIII) was rechromatographed on 4% agarose in 0.15 M NaCl or 1.0 M NaCl, a single protein peak, containing both procoagulant activity and von Willebrand factor activity, as defined by ristocetin-induced platelet aggregation, was eluted in the void volume. Purified Factor VIII immediately lost about 30% of its procoagulant activity when dissolved in 0.25 M CaCl2, and when rechromatographed on 4% agarose in 0.25 M CaCl2, the protein peak and von Willebrand factor activity remained coincident in the void volume; however, most of the remaining procoagulant activity was eluted after the void volume. The elution position of Factor VIII procoagulant activity from 4% agarose in 0.25 M CaCl2, and hence its apparent molecular weight, varied with the protein concentration applied to the column; at low protein concentrations it was eluted close to the inner volume. Yet on Sephadex G-200 in 0.25 M CaCl2, the protein and procoagulant activity were eluted together in the void volume. These observations suggested that the Factor VIII procoagulant activity was not eluting according to size or shape, but was adsorbing to some extent to the agarose. Isolated activity peak material from the 0.25 M CaCl2 columns contained protein and had a typical ultraviolet spectrum. Even at high concentrations, the protein contained no thrombin, Factors IX, X, or Xa activity, or detectable phospholipid. In addition to Factor VIII procoagulant activity, which could be inactivated by a human antibody to Factor VIII, the activity peak protein also contained von Willebrand factor activity. Like native Factor VIII and the void volume protein, the activity peak contained protein that did not enter a sodium dodecyl sulfate 5% polyacrylamide gel in the absence of reducing reagent. After reduction of disulfide bonds, several subunits ranging from 195,000 to 30,000 daltons were observed. These results indicate that the protein in the shifted Factor VIII procoagulant activity peak is large and that its anomalous elution pattern from 4% agarose in 0.25 M CaCl2 results from interaction with the agarose. The Factor VIII-like properties of the activity peak protein and its electrophoretic pattern on sodium dodecyl sulfate gels suggest that it is a species of Factor VIII modified by proteolytic cleavage. These results allow an interpretation that is different from the recently proposed "carrier protein-small active subunit" hypotheses for the structure-function relationships of the Factor VIII molecule.

Some effects of calcium on the activation of human factor VIII/Von Willebrand factor protein by thrombin.

When Factor VIII/von Willebrand factor (FVIII/vWF) protein is rechromatographed on 4% agarose in 0.25 M CaCl(2), the protein and vWF activity appear in the void volume, but most of the FVIII procoagulant activity elutes later. Recent evidence suggests that the delayed FVIII procoagulant activity is a proteolytically modified form of FVIII/vWF protein that filters anomalously from agarose in 0.25 M CaCl(2). To test whether or not thrombin is the protease involved, the effect of 0.25 M CaCl(2) on FVIII/vWF and its reaction with thrombin was examined. About 30% of the FVIII procoagulant activity was lost immediately when solutions of FVIII/vWF protein were made 0.25 M in CaCl(2). When FVIII in 0.15 M NaCl was activated with 0.04 U thrombin/ml and then made 0.25 M in CaCl(2), the procoagulant activity of a broad range of FVIII/vWF protein concentrations remained activated for at least 6 h. But, in 0.25 M CaCl(2), the increase in FVIII procoagulant activity in response to thrombin was much more gradual and once activated, the procoagulant activity was stabilized by 0.25 M CaCl(2). When thrombin-activated FVIII/vWF protein was filtered on 4% agarose in 0.15 M NaCl, there was considerable inactivation of FVIII procoagulant activity; however, the procoagulant activity that did remain eluted in the void volume. In contrast, when thrombin-activated FVIII/vWF protein was filtered in 0.25 M CaCl(2), the FVIII procoagulant activity eluted well after the void volume and remained activated for 6 h. The procoagulant peak isolated by filtering nonthrombin-activated FVIII/vWF protein on agarose in 0.25 M CaCl(2) was compared to that isolated from thrombin-activated FVIII/vWF protein. Both procoagulant activity peak proteins had about the same specific vWF activity as the corresponding void volume protein. Before reduction, the sodium dodecyl sulfate gel patterns for the two procoagulant activity peak proteins were the same. After reduction, the gel pattern for the nonthrombin-activated procoagulant activity peak protein contained bands of 195,000, 148,000-120,000, 79,000, 61,000, 51,000, and 18,000 daltons whereas the pattern for the reduced thrombin-activated procoagulant activity peak protein always lacked the higher molecular weight bands, but consistently showed the four lower molecular weight bands to be well resolved. Taken together, these results imply that thrombin generates the FVIII procoagulant activity that is stabilized by 0.25 M CaCl(2) and elutes aberrantly from 4% agarose in that solvent.

Demonstration and characterization of specific binding sites for factor VIII/von Willebrand factor on human platelets.

The presence of specific Factor VIII/von Willebrand factor (FVIII/vWF) binding sites on human platelets has been demonstrated by using 125I-FVIII/vWF and washed human platelets. Binding is ristocetin-dependent and increases in proportion to the concentration of ristocetin from 0.2 to 1 mg/ml. Binding of 125I-FVIII/vWF to platelets can be competitively inhibited by unlabeled human or bovine FVIII/vWF, but not by human thrombin, fibrinogen, alpha 2-macroglobulin, equine collagen, or a lectin of Ricinus communis. Scatchard analysis of binding data indicated that the dissociation constant of FVIII/vWF receptors is 0.45--0.5 nM. There are 31,000 binding sites per platelet at 1 mg/ml of ristocetin concentration. The optimal pH range for binding is from 7.0 to 7.5. At a concentration of 2 mM, EGTA inhibits 86% of the binding; however, 20 mM of Ca++, Mg++, or EDTA have little effect. Binding sites for FVIII/vWF were found only on platelets, and no significant binding was detected with human erythrocytes or polymorphonuclear leukocytes.

Electron microsocpy of plasmic fragments of human fibrinogen as related to trinodular structure of the intact molecule.

We have examined rotary shadowed, purified plasmic fragments of human fibrinogen with the electron microscope and have determined the relation of these fragments to the intact fibrinogen molecule. Both intact fibrinogen and its earliest cleavage product, fragment X, are trinodular. The next largest product, fragment Y, consists of two linked nodules. The two terminal products, fragments D and E, are single nodules. From measurements of simultaneously shadowed specimens of these different species, we conclude that the outer nodules of the trinodular fibrinogen molecule are the fragment D-containing regions and the central nodule is the fragment E-containing region.

The effect of fibrin-stabilizing factor on the subunit structure of human fibrin.

The formation of human fibrin from fibrinogen has been examined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate, a method which separates a mixture of proteins on the basis of differences in molecular weight. It has been found that the plasma from a patient with a congenital deficiency of fibrin-stabilizing factor forms clots lacking the cross links among the alpha- and gammachains found in normal, cross-linked human fibrin. The addition of purified fibrin-stabilizing factor or normal plasma to the deficient plasma results in extensive cross-linking of the chains. Thus, the fibrinogen in the fibrin-stabilizing factor deficient plasma appears to be normal and forms fibrin which contains dimeric, cross-linked gamma-chains and polymeric, high molecular weight forms of alpha-chains. By the use of these electrophoretic methods, it has also been possible to develop a highly sensitive method for measuring the content of fibrin-stabilizing factor in plasma. This method depends upon the use of urea-treated fibrinogen, which is completely devoid of fibrin-stabilizing factor, but which forms the usual cross-linked subunits after conversion to fibrin by thrombin in the presence of fibrin-stabilizing factor.

The subunit structure of normal and hemophilic factor VIII.

Human factor VIII from normals and hemophiliacs was partially purified by ethanol and polyethylene glycol precipitations. Final purification was achieved by gel filtration on 2 or 4% agarose or ion exchange chromatography on diethylaminoethyl cellulose. Comparable amounts of highly purified protein were obtained from normal and hemophilic plasma following the agarose chromatography step. Highly purified factor VIII was not dissociated by 6 M guanidine hydrochloride or 1% sodium dodecyl sulfate. However, when reduced by beta-mercaptoethanol and analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis, a single subunit species with an estimated 195,000 molecular weight was found for both normal and hemophilic factor VIII. By sedimentation equilibrium analysis, the normal factor VIII subunit was homogeneous and had an estimated molecular weight of 202,000. The subunit polypeptides from normal or hemophilic factor VIII contained carbohydrate. Each was homogeneous by isoelectric focusing. Immunodiffusion of purified normal and hemophilic factor VIII against rabbit antiserum to purified normal human factor VIII showed a single line of precipitation. Very low concentrations of purified human thrombin initially increased the activity of normal factor VIII about threefold and then progressively destroyed activity by 3 h. Only minimal activation occurred with hemophilic factor VIII. Both the activation and inactivation of normal and hemophilic factor VIII were unaccompanied by detectable changes in subunit molecular weight. These findings may have implications for the definition of the molecular defect in hemophilic factor VIII.

Mechanism of ancrod anticoagulation. A direct proteolytic effect on fibrin.

Fibrin formed in response to ancrod, reptilase, or thrombin was reduced by beta-mercaptoethanol and examined by sodium dodecyl sulfate polyacrylamide gel electrophoresis. It was found that ancrod progressively and totally digested the alpha-chains of fibrin monomers at sites different than plasmin; however, further digestion of fibrin monomers by either reptilase or thrombin was not observed. Highly purified ancrod did not activate fibrin-stabilizing factor (FSF); however, the reptilase preparation used in these experiments, like thrombin, activated FSF and thereby promoted cross-link formation. Fibrin, formed by clotting purified human fibrinogen with ancrod, reptilase, or thrombin for increasing periods of time in the presence of plasminogen, was incubated with urokinase and observed for complete lysis. Fibrin formed by ancrod was strikingly more vulnerable to plasmin digestion than was fibrin formed by reptilase or thrombin. The lysis times for fibrin formed for 2 hr by ancrod, reptilase, or thrombin were 18, 89, and 120 min, respectively. Evidence was also obtained that neither ancrod nor reptilase activated human plasminogen. These results indicate that fibrin formed by ancrod is not cross-linked and has significantly degraded alpha-chains: as expected, ancrod-formed fibrin is markedly susceptible to digestion by plasmin.

Effects of plasmin on von Willebrand factor multimers. Degradation in vitro and stimulation of release in vivo.

von Willebrand factor (vWF), a multimeric protein that mediates platelet adhesion, circulates in association with the procoagulant Factor VIII (FVIII). In previous reports, plasmin was shown in vitro to inactivate FVIII and cleave the vWF subunit extensively, but to cause only a modest decrease in vWF platelet-agglutinating activity. In the present study, the digestion of vWF multimers by plasmin was analyzed by sodium dodecyl sulfate-agarose gel electrophoresis and radioimmunoblotting. In vitro, plasmin degraded the large vWF multimers to smaller forms that could be distinguished from the small multimers present before digestion only by a slightly increased electrophoretic mobility. These plasmin-cleaved "multimers" were composed of disulfide-linked fragments with no intact vWF subunits. Thus, many plasmin cleavages occur within disulfide loops. The slight increase in mobility of plasmin-digested vWF is in part explained by the early cleavage from the multimers of a 34,000-mol wt peptide, which was purified and partially sequenced. The amino-terminal sequence (33 residues) agrees with the previously reported sequence (15 residues) for the amino terminus of the intact vWF subunit. Analysis of plasmin-digested vWF allowed deduction of a model for the native vWF structure, including the approximate location of the interprotomer disulfide bond(s). To determine whether plasmin would digest vWF in vivo, plasmas from 12 patients and 2 normal volunteers who received intravenous streptokinase (SK) were analyzed. Rather than vWF digestion, a two- to threefold rise in vWF antigen and platelet-agglutinating activity occurred within 2 h after a single SK dose, and the increase was greatest among the largest multimers. In contrast, FVIII clotting activity dropped to 10-20% of pre-SK levels. Thus, although plasmin destroys FVIII, a pharmacologically induced fibrinolytic state is associated with significant release of vWF from endothelial cells, platelets, or some other storage pool.

Substructure of human von Willebrand factor.

Using electron microscopy, we have visualized the substructure of human von Willebrand factor (vWf) purified by two different approaches. vWf multimers, which appear as flexible strands varying in length up to 2 micron, consist of dimeric units (protomers) polymerized linearly in an end-to-end fashion through disulfide bonds. Examination of small multimers (e.g., one-mers, two-mers, and three-mers) suggests that each protomer consists of two large globular end domains (22 X 6.5 nm) connected to a small central node (6.4 X 3.4 nm) by two flexible rod domains each approximately 34 nm long and approximately 2 nm in diameter. The protomer is 120 nm in length when fully extended. These same structural features are seen both in vWf molecules that were rapidly purified from fresh plasma by a new two-step procedure and in those purified from lyophilized intermediate-purity Factor VIII/vWf concentrates. The 240,000-mol wt subunit observed by gel electrophoresis upon complete reduction of vWf apparently contains both a rod domain and a globular domain and corresponds to one half of the protomer. Two subunits are disulfide-linked, probably near their carboxyl termini, to form the protomer; disulfide bonds in the amino-terminal globular ends link promoters to form vWf multimers. The vWf multimer strands have at least two morphologically distinct types of ends, which may result from proteolytic cleavage in the globular domains after formation of large linear polymers. In addition to releasing fragments that were similar in size and shape to the repeating protomeric unit, plasmic degradation of either preparation of vWf reduced the size of multimers, but had no detectable effect on the substructure of internal protomers.

Why alpha-antiplasmin must be converted to a derivative form for optimal function.

BACKGROUND: Human alpha(2)-antiplasmin (alpha(2)AP), the primary inhibitor of fibrinolysis, is secreted from the liver into plasma as a 464-residue protein with Met as the N-terminus. An R6W polymorphism has been suggested to affect fibrinolytic rate. Within circulating blood, antiplasmin-cleaving enzyme (APCE) cleaves Met-alpha(2)AP(R6) faster than Met-alpha(2)AP(W6) at the Pro12-Asn13 bond to yield Asn-alpha(2)AP. OBJECTIVES: To compare Met-alpha(2)AP(R6), Met-alpha(2)AP(W6) and Asn-alpha(2)AP for crosslinking with fibrin and the ability to protect fibrin from digestion by plasmin. METHODS AND RESULTS: Asn-alpha(2)AP utilizes Gln2 (Gln14 in Met-alpha(2)AP) to become crosslinked to fibrin approximately twelvefold faster than Met-alpha(2)AP(R6) or Met-alpha(2)AP(W6), and this enhances the resistance of fibrin to plasmin. All three forms of alpha(2)AP inhibit plasmin at identical rates. The N-terminal 12-residue peptide of Met-alpha(2)AP slows crosslinking of Met-alpha(2)AP(R6) or Met-alpha(2)AP(W6) by limiting access of factor XIIIa to Gln14 rather than shifting crosslinking to other Gln residues. Edman sequencing and mass analyses of tryptic peptides from each alpha(2)AP crosslinked with 5-(biotinamido)pentylamine showed Gln14 as the only major crosslinking site. Residues 5-8, GRQL in Met-alpha(2)AP(R6), and residues 1-8, MEPLGWQL in Met-alpha(2)AP(W6), slow fibrin crosslinking. CONCLUSION: Gln14 in both Met-alpha(2)AP(R6) and Met-alpha(2)AP(W6) is sheltered by the N-terminal 12-residue peptide, which, when cleaved, yields Asn-alpha(2)AP, which is rapidly crosslinked to fibrin and maximally protects it from plasmin. The R6 W polymorphism in Met-alpha(2)AP does not affect its crosslinking to fibrin, but it does slow cleavage by APCE and reduces the amount of Asn-alpha(2)AP available for rapid crosslinking to fibrin.

High levels of p53 protein in UV-irradiated normal human skin.

Exposure of normal adult human skin to doses of UV irradiation that induced mild sunburn resulted in the rapid appearance of p53 protein in the epidermis and superficial dermal fibroblasts. Immunohistological analysis with a panel of antibodies established that while p53 staining was not seen in normal skin it appeared within 2 h of UV exposure. The level of p53 immunostaining peaked at 24 h and returned to undetectable levels within 360 h. The induction of proliferating cell nuclear antigen (PCNA) (which is required for both DNA replication and repair) followed a similar spatial and temporal pattern to p53. The UV irradiation did not induce a mitotic response or the replication-associated antigens DNA polymerase alpha or Ki67. The accumulation of high levels of p53 and PCNA in response to UV doses to which many human populations are routinely exposed provides strong support for a model in which normal p53 acts as part of the DNA damage response in vertebrate cells. Such a model is consistent with the profound tumour-suppressor function of the p53 gene, the high rate of p53 mutation in neoplasia and the exceptionally high tumour susceptibility of p53-deficient mice.

Spurious hypertension in the obese patient. Effect of sphygmomanometer cuff size on prevalence of hypertension.

We used standard, large adult, and thigh-size cuffs in random order to take BPs in 470 patients. The prevalences of definite high BP [( HBP]), greater than or equal to 160/95 mm Hg) and borderline HBP [( BHBP ], greater than or equal to 140/90 less than 160/95 mm Hg) were the same with all three cuffs in patients with an arm circumference less than 35 cm, a body mass index less than 34, and a weight of less than 95 kg. The large adult and thigh cuffs did not underestimate the prevalence of HBP in these nonobese patients. The prevalences of HBP and BHBP were twofold greater with the standard cuff than with the large adult or thigh cuffs in obese patients (arm circumference greater than or equal to 35 cm or body mass index greater than or equal to 34 or weight greater than or equal to 95 kg). Routine use of the large adult cuff will provide accurate BP measurement and avoid unneeded evaluation and treatment.

BP control. Improvement in a university medical clinic by use of a physician's associate.

Many hypertensive patients, especially those in outpatient clinics at large teaching hospitals, do not achieve BP control. We incorporated a physician's associate into an existing house staff medical clinic and evaluated whether this improved BP control. In patients with moderate or severe hypertension, BP control was achieved in 56% of patients observed by both the physician's associate and the house staff and in 32% of patients observed solely by house staff. Possible contributing factors were more frequent follow-up, simplification of drug regimens, reduced waiting time, more time spent with the patients, and overall greater satisfaction with the physician's associate. We conclude that the addition of a physician's associate to an outpatient clinic is an effective method for enhancing BP control. This can be achieved without establishing a separate hypertension clinic or depriving house staff of experience in the management of hypertension.

Predictive value of compression ultrasonography for deep vein thrombosis in symptomatic outpatients: clinical implications of the site of vein noncompressibility.

BACKGROUND: Compression ultrasonography has a high negative predictive value for deep vein thrombosis in symptomatic outpatients. Limited data are available on factors influencing positive predictive value. The objective of this study was to evaluate the positive predictive value of compression ultrasonography according to the anatomic site of vein noncompressibility. METHODS: We performed a prospective cohort study of 756 consecutive outpatients with suspected first-episode deep vein thrombosis. Compression ultrasonography was performed at the initial visit: results were abnormal if a noncompressible segment was identified or normal if all segments were fully compressible. Venography was performed in patients with abnormal compression ultrasonography results. Positive predictive value was determined according to the site of noncompressibility: common femoral vein only, popliteal vein only, or both sites. Venography was the reference standard for the presence of deep vein thrombosis. RESULTS: Positive predictive value was 16.7% (95% confidence interval, 0.4%-64.1%) for noncompressibility isolated to the common femoral vein compared with 91.3% (95% confidence interval, 72.0%-98.9%) for the popliteal vein only and 94.4% (95% confidence interval, 72.7%-99.9%) for both sites (P<.001). Of 15 patients with isolated noncompressibility of the common femoral vein, 8 (53%) had pelvic neoplasm or abscess compared with 2 (5%) of 42 with noncompressibility of the popliteal vein only and 6 (13%) of 47 with noncompressibility of both sites (P<.001). CONCLUSIONS: The positive predictive value of noncompressibility isolated to the common femoral vein is too low to be used alone as the diagnostic end point for giving anticoagulant therapy. Noncompressibility isolated to the common femoral vein is a diagnostic marker for pelvic disease.

Melanoma associated with blue nevus and melanoma mimicking cellular blue nevus: a clinicopathologic study of 10 cases on the spectrum of so-called 'malignant blue nevus'.

The term "malignant blue nevus" refers to a rare and heterogeneous group of melanomas that arise in several clinical settings. This includes melanomas arising in association with a common or cellular blue nevus and those arising de novo and resembling cellular blue nevi. We reviewed the clinicopathologic features of 10 cases of malignant blue nevi. Six cases proved to be de novo melanoma mimicking cellular blue nevus, but lacking a clear-cut benign component. Two melanomas arose in association with a common blue nevus, and two with a cellular blue nevus. The patients' (5 males, 5 females) ages ranged from 11 to 77 years (average age, 48.1 years). The head and neck was the most common location (6 of 10 patients), with five scalp tumors. Four tumors were located on the trunk; none was located on the extremities. Tumor size ranged from 0.5 to 2.2 cm (average size, 1.1cm). Most lesions had been present for many years before surgical removal. Pigmented dendritic cells were observed in 9 of 10 cases. The malignant and benign components were easily distinguished in the four cases that arose in association with a common or cellular blue nevus. Abrupt transition between a benign blue nevus and melanoma was readily recognized at scanning magnification as distinctive nodules of epithelioid to spindled cells with a sheet-like growth pattern. In all cases, malignancy was evidenced by increased mitotic rate, necrosis, nuclear atypia, pleomorphism, hyperchromasia, and prominent nucleoli. All 7 patients with follow-up information experienced recurrence (3 patients) or metastasis (4 patients). Three patients died of disease. Malignant blue nevus is a heterogeneous group of melanomas that are highly aggressive and often lethal, with a propensity for metastasis to the lymph nodes and lungs.