ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 23, No. 1 Copyright 1993, Institute for Clinical Science, Inc. Alterations in von Willebrand Factor Antigen in Premature Infants with Respiratory Distress Syndrome and Chronic Lung Disease* CELESTE H. PATRICK, M.D., JOHN LAZARCHICK, M.D., and BRENDA S. HORN, M.T. (ASCP) Departments of Pediatrics and Pathology/Laboratory Medicine Medical University of South Carolina Charleston, SC 29425 ABSTRACT Elevated levels of von W illebrand Factor Antigen (vwf:ag) may occur in the presence of endothelial injury, a component in the pathology of acute pulmonary insufficiency. The vwf:ag levels were examined in 13 well infants (controls) and 20 infants with respiratory distress syndrome (RDS), nine of whom developed bronchopulmonary dysplasia (BPD). All infants were very low birth weight (730 to 1500 g) and prem ature (25 to 34 weeks estimated gestational age). Plasma samples were obtained at birth and weekly through 28 days of age and frozen at 70 C. The vwf:ag was quantified by the enzyme linked immunosorbent assay (ELISA) m ethod for 138 plasma specimens; in addition, 77 samples were analyzed for multim er pattern by SDS-agarose (1.7 percent) electrophoresis and densitom etric scanning. All groups had elevated mean levels of vwf:ag, compared to adults. Although levels rem ained stable over the four week period, the group of infants with BPD had a significantly high mean level of vw F: Ag at 21 days than those groups without BPD (p < 0.05). Visual examination of vwf m ultim er patterns revealed absence of unusually large vwf multimers and triplet patterns suggestive of increased proteolytic degradation of von W illebrand factor. However, densitometer scanning revealed that samples with higher vwf:ag levels (>200 percent) had increased amounts of moderate to smaller sized multimers, regardless of presence or absence of BPD. It is our conclusion that von W illebrand factor antigen levels are nonspecifically elevated in prem ature infants and that chronic lung disease is associated with even higher plasma values, possibly owing to pulmonary endothelial injury. Send reprint requests to: John Lazarchick, M.D., Department of Pathology/Lab Medicine, Medical University of South Carolina, 171 Ashley Avenue, Charleston, SC 29425. 39 0091-7370/93/0100-0039 $01.20 Institute for Clinical Science, Inc.
40 PATRICK, LAZARCHICK, AND HORN Introduction Factor V III: von W illebrand factor antigen (vw F:Ag) is sy n th esized by megakaryocytes and endothelial cells of arteries and is released into the circulation in resp o n se to various stim uli, including endothelial cell injury.8,9 Low levels of antigen are often associated with a bleeding diathesis, while high concentrations may be noted in certain prim ary throm botic states, including throm botic throm bocytopenic purpura and adult respiratory distress syndrome. Von W illebrand factor circulates as a m ultim er, ranging in size from one to approxim ately 20 X 106 daltons; abnormalities of the structure or proportion of the various m ultimers may also result in bleeding or thrombosis. Adequate levels of normal vwf:ag are essential for platelet adhesion to collagen exposed by damage to the endothelium.11 Severe respiratory failure is characterized by diffuse pulmonary microvascular injury and produces marked increases in Factor V III and von W illebrand factor antigen in adult patients with respiratory failure.4 To explore the alterations of plasma von W illebrand factor antigen in a comparable neonatal population who are at risk for pulmonary endothelial injury, von W illebrand factor antigen levels and multim er patterns in 33 very low birthw eight infants w ere studied. It was hypothesized by us that prem ature newborns with respiratory distress syndrome would have elevated von W illebrand factor antigen levels and alterations in m ultim er patterns. Materials and Methods Blood samples were collected in ethyle n e d ia m in e te tra a c etic acid (EDTA)- anticoagulated tubes from 33 very low birthweight (<1500 gm) infants at birth, on day 3, and w eekly for four weeks. P atients w ere selected as consecutive admissions to the neonatal intensive care unit of the Medical University of South Carolina Children s Hospital Charleston, South Carolina, whose parents provided in fo rm ed co n sen t. S p ecim en s w ere obtained via umbilical vessel catheters or by puncture of peripheral artery or vein. Platelet-poor plasm a was obtained by centrifugation of whole blood at 2500 X g for 20 m inutes at 37 C and stored at -70 C. A total of 136 adequate plasma samples were collected from the infants. A pool of plasma from 20 normal subjects similarly drawn and prepared was stored at 70 C in small aliquots for com parison studies. G roups w ere d iv id e d into control p a tie n ts (13 w ell in fan ts) and sick patients (20 infants with a clinical and x-ray diagnosis of respiratory distress syndrome [RDS]). Upon completion of sample collection, the latter group was further divided into a resolved RDS group (11 infants) and an RDS group who subsequently developed bronchopulm o nary dysplasia (BPD) by chest radiograph findings and 28 day oxygen requirem ent (nine infants). V on W ille b r a n d fa c to r a n tig e n (vw F:A g) assay was p e rfo rm e d by enzym e imm unoassay using the Sandw ich technique.* This technique has been previously described.2 Von W illebrand factor multim er analyses w e re p e rfo rm e d u sin g so d iu m dodecyl sulfate (SDS)-1.7 p ercen t agarose gel electrophoresis.14 Ten jxl of a 1: 10 dilution of test samples were added to each well, and electrophoresis carried out at 20 ma for approximately two hours and then at 3 to 5 ma overnight. M ultimers w ere then transferred from the agarose gel to a nitrocellulose membrane as described by Towbin et al.15 The transblots w ere developed by using a rabbit * ASSERACHROME vwf Diagnostica Stago, Asnieres-sur-seine, France.
VON WILLEBRAND FACTOR ANTIGEN IN PREMATURE INFANTS WITH RDS & BPD 41 antihum an vwf antibody as the primary antibody and peroxidase-conjugated goat a n ti-rab b it IgG as the secondary antibody. The membranes were developed with 4-chloro-l-naphthol as the indicator reagent. M ultimeric patterns were examined visually and by densitom etric scanning using the Beckm an Appraise densito m e te r.t T h e m u ltim e ric p a tte rn obtained on electrophoresis and transblot of the control normal pooled plasma was arbitrarily divided into three equal vertical se g m e n ts d e s ig n a te d as sm all, medium, and large multimers. The multimeric patterns on all test samples were th e n c o m p a re d to th e d is trib u tio n o b ta in e d u sin g th is norm al p o o led plasma as the standard. All Factor VIII: von W illebrand factor antigen results are expressed relative to pooled normal plasm a (100 U per dl) obtained from 20 healthy donors. R esp irato ry d istress was m anaged using conventional ventilation with an effort to keep p C 0 2 betw een 45 and 55 torr, ph g reater than 7.25, and p 0 2 greater than 50 torr. Other medical managem ent included nutritional support, antibiotics w here clinically indicated, and enteral or parenteral diuretics for treatm ent of fluid retention associated w ith bronchopulm onary dysplasia. S ta tistical com parisons w ere p e r form ed using analysis of variance for main effects of tim e and group. Additionally, group tim e interactions were examined. Adjustm ents w ere m ade for m ultiple com parisons using the Bonferroni m ethod. D ifferences were considered statistically significant w hen p < 0.05. Results Com parison of the groups by birthw eight and gestational age is presented t Beckman Instrum ents, Inc., Palo Alto, CA 94303. in table I. There was no statistical difference betw een these groups, although the control patients tended to be slightly heavier and more advanced in gestational age. Mean levels of von W illebrand factor antigen in all groups were higher than adult normal values (upper lim it 150 j, per dl)). Comparison of group m ean values in each time period betw een control infants and those with resolved RDS, revealed no statistical difference over 28 days (table II). At three weeks of age, however, infants who eventually developed bronchopulm onary dysplasia had significantly elevated von W illebrand factor antigen levels versus controls or resolved RDS infants (p <.05, figure 1). Values of vwf:ag in infants with BPD tended to be higher for each sampling interval. Antigen levels did not correlate with the degree of severity of BPD, as m easured by oxygen requirem ents and radiological evidence of fibrosis. Visual inspection of m ultim ers did reveal a qualitative difference in patterns betw een both of these groups and the normal pool plasma (figure 2). No ultra high m olecular w eight (HMW) m ultimers were evident in either the BPD or h y a lin e m em b ran e d ise a s e (H M D ) groups. The HMW multimers were relatively decreased in both of these groups com pared to the pattern seen w ith normal pool plasma. The triplet structure not only had a prom inent central band simi- TABLE I Population Characteristics * Number Weight:grams GA: weeks Controls 13 1274 (±216) 31.0 (±2.2) RDS 11 1128 (±181) 29.0 (± 1.5) RDS with BPD 9 966 (± 191) 28.7 (± 2.7) * (± standard error of the mean). GA = gestational age. RDS = respiratory distress syndrome. BPD = bronchopulmonary dysplasia.
42 PATRICK, LAZARCHICK, AND HORN TABLE II Group Means of von Willebrand Factor Antigen (± SEM) for Each Study Period (% of Standard) Day Controls RDS RDS with BPD 1 265 ± 71 (13) 229 + 131 (7) 272 ± 62 (7) 3 190 ± 87 (4) 209 ± 75(11) 247 ± 179 (9) 7 263 ± 72 (9) 237 ± 93(11) 308 ± 135 (8) 14 262 ± 90 (11) 254 ± 64(10) 332 ± 108 (8) 21 231 ± 67 (12) 230 ± 76 (9) 342 + 98 (8) 28 255 + 70 (11) 228 ± 102 (7) 288 + 147 (7) RDS = Respiratory distress syndrome. BPD = bronchopulmonary dysplasia. Numbers In parentheses are numbers of samples analyzed. lar to the control plasma but also minor bands, especially evident in the lower molecular weight m ultim er triplets. No aberrant or extraneous bands were noted in e ith e r study group. D ensitom etric scanning, how ever, did not show any striking alteration of patterns in those individuals with von W illebrand antigen levels w ithin the normal adult range (figure 3). H igher von W illebrand factor antigen levels, i.e., those over 200 percent activity, were associated with increased amounts of the moderate to smaller size m ultim ers, regardless of the group in which the infant was categorized. Of note is that four patients in the RDS group who developed sepsis and died within the first 28 days of life had very high levels of von W illebrand factor antigen (>250 percent activity), the highest value being 592 percent. Discussion Von W illebrand factor antigen, both quantitative and qualitative, has been evaluated previously in both healthy and ill newborns and adults.3 4,7,10 18 N eonatal levels of von W illebrand factor have been reported to be elevated compared to adults, with the highest values in ill newborns. Preterm infants have levels comparable to terms, and levels in both groups decrease to adult concentrations by six m onths chronologic age.1,12 The aim of this study was to compare a control group of w ell neonates to a group of infants w ith the specific diagno- F ig u r e 1. T im e course of changes in the von W illebrand factor antigen (vwf:ag) in the group of infants with hyaline membrane disease (HHMD) and those who developed bronchopulmonary dysplasia (BPD). Mean and standard deviations are indicated for each group and tim e period. DAY
VON WILLEBRAND FACTOR ANTIGEN IN PREMATURE INFANTS WITH RDS & BPD 1 2 3 4 5 6 7 8 43 9 We if v m * Ijfe t* * m FIGURE 2. Electrophoretic multimeric profile of von Willebrand antigen from normal pooled plasma (NPP, lanes 4 and 5), plasma from patients with bronchopulmonary dysplasia (BPD, lanes 1, 2, and 3) and plasma from patient with hyaline membrane disease (HMD, lanes 6, 7, 8, and 9) following sodium dodecyl sulfate (SDS)-agarose (1.7 percent) gel electrophoresis, nitrocellulose blotting and immunoperoxidase development. sis of acute lung disease, half of whom ultim ately d eveloped the chronic lung disease bronchopulm onary dysplasia. It was c o n firm ed by us th a t p rem a tu re infants have levels of vw F:Ag higher than normal adults. O ur study population included m ore prem ature and extremely low birthw eight infants than previously published surveys;1,12 group m ean val ues for vw F:A g w ere h ig h e r than in these reports, perhaps reflecting extreme immaturity. Although the m echanism has not been elucidated, higher levels could result from increased production of the m olecule by endo th elial cells w ith or w ithout increased proteolysis.6 An alter n a t i v e e x p l a n a t i o n c o u ld in v o k e d e c re a s e d c le a ra n c e or m e ta b o lis m owing to the relatively poor renal func tion in e x tre m e ly p re m a tu re in fan ts, since there are reported to be increased levels in adults with renal failure.12 It w as also fo u n d by th e p r e s e n t authors that infants with respiratory dis tress syndrom e who continued to have pulm onary insufficiency at three weeks of age had even higher levels of vwf:ag versus those with no disease or in whom acute disease had resolved. O ur data are som ew hat consistent with that of Carval h o s study of 100 ad u lt p a tie n ts w ith acute lung injury4 in that our study found a correlation of high von W illebrand fac tor antigen levels in ventilated patients
44 PATRICK, LAZARCHICK, AND HORN Greyvalue NPP: 100% activity 8 16 6 Greyvalue 206 308 408 500 BPD: 348% activity A. h A V F i g u r e 3. D ensitometrie scanning of the von Willebrand antigen multimeric profile following electrophoresis of normal pooled plasm a (NPP, top), plasma from a patient with bronchopulmonary dysplasia (BPD, middle), and plasma from a patient w ith hyaline m e m b r a n e d i s e a s e (HMD, bottom). e 100 Greyvalue.w zee 400 No BPD: 296% activity POINTS
VON WILLEBRAND FACTOR ANTIGEN IN PREMATURE INFANTS WITH RDS & BPD 45 presum ed to be at risk for pulm onary endothelial injury. D espite surfactant deficiency and the potential for ventilator in d u ced barotraum a in these infants, there was not an exaggerated increase in vwf:ag in the prem ature infants with acute pulm onary insufficiency versus w ell controls. This disparity betw een the response of prem ature infants and adults with acute respiratory failure can not be fully explained, but it can be speculated that prem ature infants may be releasing vwf:ag at the maximal rate owing to stress associated with perinatal adaptation. Alternatively, the damaged lung in adults and neonates with acute respiratory failure may release elastase, or o th er p ro teo ly tic enzym es w hich destroy or alter the molecule, although at different rates. The relative decrease of HMW multimers in both study groups and the atypical triplet pattern, especially prominent in the lower molecular weight m ultimer triplets, are consistent w ith increased endothelial cell release of von W illebrand factor with subsequent proteolysis of the HMW multimers. This increase in sm aller to m oderate size m ultim ers is similar to the findings observed by Brenner et al3 in a group of obstetric patients with pregnancy-induced hypertension, a disorder associated w ith m icroangiopathy and en d o th elial cell injury. O ur group did not study the crossed immunoelectrophoresis pattern of vw F:Ag as Carvalho4 did; however, our findings of increased amounts of small and moderate sized multimers when levels are elevated may be analogous to a shift in electrophoretic patterns toward the faster migrating (and therefore smaller) components seen in adult patients with respiratory failure. The so called ultra large vwf multimers (ULvWFM) noted by W einstein et al18 in a study of cord blood samples was not visualized by us. There was no correlation of the degree of chronic lung w ith the levels of von W illebrand factor antigen, which may reflect the heterogeneity of the clinical diagnosis of bronchopulm onary dysplasia. The am ount of endothelial stim ulation or damage necessary to cause the release of von W illebrand factor into the circulation is difficult to quantitate in vivo, even though cultured hum an endothelial cells can be stim ulated to release vwf:ag by treatm ent w ith trypsin and epinephrine.16 Again, enzymes responsible for the metabolism of factor V III are produced in lung tissue; the immature or inflamed lung may alter the m etabolism of circulating antigen, changing either the qualitative or quantitative m easurem ents. The finding that von W illebrand factor antigen levels are extremely high in the presence of fatal sepsis is of interest. The release of tumor necrosis factor (TNF) and interleukin-1 (IL-1) following exposure to bacterial endotoxin is well docum ented.5 13 W hether the rise in vwf:ag during sepsis is a parallel phenom enon related to nonspecific endothelial dam age or is caused by high levels of TNF and/or IL-1 is an area for further investigation. An additional confounding factor with sepsis is the possible alteration in m etabolism and clearance of vw F:Ag in th e h e m o d y n a m ic a lly u n s ta b le infant; changes in renal and/or pulmonary physiology could be responsible for excessive release or decreased excretion of the molecule. Conclusion It is our conclusion that very low birthweight premature infants have elevated levels of von W illebrand factor antigen compared to adults; in the presence of chronic lung injury or sepsis, there is a fu rth e r in c re a se in a n tig e n le v e ls. D etailed examination of qualitative characteristics of the antigen (i.e., multimeric analysis) reveals extrem e elevations (>200 percent of norm al values) to be
46 PATRICK, LAZARCHICK, AND HORN associated with increases in the small to m oderate sized m ultim ers. T he m echanism ^) inducing release in th e circulation is unknown. References 1. An d r e w, M., P a e s, B and MILNER, R.: Development of the human coagulation system in the healthy preterm infant. Blood 72: 1651 1657, 1988. 2. B a r t l e t t, A., D o r m a n d y, K. M., H a w k e y, C. M., St a b l e f o r t h, P., and V o l l e r, A.: Factor V I II related antigen: M easurem ent by enzyme immunoassay. Brit. Med. J. 1:994-996, 1982. 3. B r e n n e r, B., Z w a n g, E., B r o n s h t e i n, M., and S e l ig s o h n, U. Von Willebrand factor multimer patterns in pregnancy induced hypertension. Thromb. Haemost. 62:715 717, 1989. 4. C a r v a l h o, A. C., B e l l m a n, S. M., Sa u l l o, V. J., QUINN, D., and ZAPOL, W. M.: Altered factor VIII in acute respiratory failure. New Engl. J. Med. 307:1113-1119, 1982. 5. DlNARELLA, C. A.: Interleukin-1 and the pathogenesis of the acute phase response. New Engl. J. Med. 311:1413-1418, 1984. 6. H a m il t o n, K. K., F r e t t o, L. S., G r ie r s o n, D. S., and M c Ke e, P. A.: Effects of plasmin on von Willebrand factor multimers. Degradation in vitro and stimulation of release in vivo. J. Clin. Invest. 76:261-270, 1985. 7. F u k u i, H., T a k a s e, T., I k a r i, H., M u r a k a m i, Y., O k u h o, Y., and N a k a m u r a, K.: Factor VIII procoagulant activity, factor VIII related antigen, and von Willebrand factor in newborn cord blood. Brit. J. Haematol. 42:637-646, 1979. 8. Ja f f e, E. A., H o y e r, L. W., and N a c h m a n, R. L.: Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J. Clin. Invest. 52:2757-2764, 1973. 9. J o h n s o n, A. R.: Human pulmonary endothelial cells in culture: Activities of cells from arteries and cells from veins. J. Clin. Invest. 65:841 850, 1980. 10. Ka t z, J. A., M o a k e, J. L., M c P h e r s o n, P. D., W e i n s t e in, M. J., M o is e, K. J., C a r p e n t e r, F. J., and Sa l a, D. J.: Relationship between human developm ent and disappearance of unusually large von Willebrand factor multimers from plasma. Blood 73:1851-1858, 1989. 11. Ke s s l e r, C. M., F l o y d, C. M., Ric k, M. E., Kr i z e k, D. M., L e e, S. L., and G r a l n ic k, H. R.: Collagen-factor VIII/von Willebrand factor protein interaction. Blood 63:2391-2398, 1984. 12. Kr i e s, R., J ü r g e n s, H. Hv., V o s s, H., and G ö BEL, U.: Clinical relevance of factor VIII:C and factor VIII R: Ag determination in newborns. Eur. J. Pediatr. 137:189-194, 1981. 13. N a t a n s o n, C., E ic h e r n o l z, P. W., and D a n n e r, R. L.: Endotoxin and tumor necrosis factor challengers in dogs simulate the cardiovascular profile of human septic shock. J. Exp. Med. 169:823-832, 1989. 14. R u g g e r i, Z. M. and Z im m e r m a n, T. S.: The complex multimeric composition 2 of factor VIII/von Willebrand factor. Blood 57:1140 1143, 1981. 15. T o w b in, H. and GORDON, J.: Immunoblotting and dot immunobinding: Current status and outlook. Immunol. Methods 72:313-400, 1984. 16. W a l l, R. T., C o u n t s, R. B., H a r k e r, L. A., and S t r ik e r, G. E.: Binding and release of factor VIII/von W illebrand s factor by human endothelial cells. Brit. J. Haemotol. 46:287-298, 1980. 17. W a r k e l l, R. P., J r., H u l t o n, M. B., and C o l - l e r, B. S.: Increased factor VIII/von Willebrand factor antigen and von Willebrand factor activity in renal failure. Am. J. Med. 66:226-228, 1979. 18. W e i n s t e in, M. J., B l a n c h a r d, R., M o a k e, J. L., V o s b u r g h, E., a n d M o is e, K.: Fetal and neonatal von Willebrand factor (vwf) is unusually large and similar to vwf in patients with thrombotic thrombocytopenic purpura. Brit. J. Haematol. 72:68 72, 1989.