Journal Logo

Article

Inhibition of Neointimal Proliferation After Coronary Angioplasty by Low-Molecular-Weight Heparin (Clivarine) and Polyethyleneglycol-Hirudin

Buchwald, Arnd; Hammerschmidt, Stefan; Stevens, Jens; Göring, Jens; Nebendahl, Klaus*; Unterberg, Christina

Author Information
Journal of Cardiovascular Pharmacology: October 1996 - Volume 28 - Issue 4 - p 481-487
  • Free

Abstract

The mechanisms resulting in the migration and proliferation of medial smooth muscle cells (SMC) after coronary angioplasty, ultimately resulting in significant luminal renarrowing, are incompletely understood. Heparin inhibits SMC proliferation in culture (1,2). Administration of high doses of a low-molecular-weight heparin (LMWH) has been shown to reduce neointimal proliferation in a porcine coronary and rabbit femoral angioplasty model (3,4).

Furthermore, platelets aggregating acutely at the site of angioplasty have been suggested as a major source of growth stimulatory factors, such as platelet-derived growth factor (PDGF) (5). Recently, we and other researchers showed a correlation between the amount of platelets deposited acutely after coronary angioplasty and the subsequent degree of late neointimal proliferation. Reduction of platelet deposition, measured either directly by radioactive labeling or indirectly by cyclic flow variations, resulted in less neointima formation (6,7). However, the doses of LMWH and hirudin required to achieve in vivo SMC inhibition or thrombus reduction are even higher than those associated with a significant bleeding risk in the clinical situation. We investigated the relative effects of a novel, polyethyl-eneglycol (PEG)-coupled hirudin (LU 57291) with a long plasma half-life (t½), resulting in smooth, continuous anticoagulation and avoiding high peak concentrations associated with administration of conventional hirudin and the LMWH clivarine three times daily. Clivarine has a molecular weight between 2.5 and 10 kD of 86% of its components, and <1.5% are >10 kD. The reported anti-factor-Xa activity is 124 IU/mg (8). Clivarine has been reported to be effective in reducing experimental neointima formation at doses that are apparently safe in the clinical setting (9,10). These individual compounds were compared with a clinical standard anticoagulation during coronary angioplasty with a single bolus of unfractionated heparin with regard to their effectiveness in reducing neointimal proliferation after coronary stent angioplasty.

METHODS

Animal model

Minipigs of the Göttingen strain weighing 25-45 kg and aged 12-24 months were sedated by an intramuscular injection of azaperone (10 mg/kg) and intravenous piritramide (7 mg). Anesthesia was induced after orotracheal intubation and ventilation by a mixture of halothane (0.8-1.5%), nitrous oxide (69%), and oxygen (30%) (Sulla 19 respirator, Dräger, Germany). Blood gases were regularly monitored, and ventilation was adjusted to maintain blood gases in the physiologic range.

The right carotid artery was surgically exposed, and an 8F coronary guide catheter was advanced into the ascending aorta. Tantalum-stents (Wiktor-stents, Medtronic, Düsseldorf, Germany), premounted on a balloon with a nominal inflated diameter of 3.0 mm, were placed in coronary artery segments of slightly smaller diameter (0.3-0.5 mm less) under fluoroscopic guidance. Stents were inflated twice at 8-10 atm for 30 s. Balloons were then deflated and withdrawn. Thereafter, the carotid artery was ligated, and the animals were returned to their cages after recovery from anesthesia. After 4 weeks, angiography was repeated to confirm patency of the stented artery. The hearts were excised and immediately perfused with phosphate-buffered saline at a pressure of 100 mm Hg, followed by perfusion-fixation with buffered 4% formalin (1,000 ml).

Anticoagulation

All 16 animals received a single bolus of 250 mg acetyl salicylic acid before insertion of the arterial catheter. From the day after operation, they received 100 mg acetyl salicylic acid daily in tablet form.

Minipigs were randomly assigned to either one of four treatments (n = 4 per group). In the control group (9 arteries), a bolus of 100 U/kg of unfractionated heparin was given immediately before angioplasty, with no further anticoagulation. A second group (11 arteries) received a bolus of 1 mg/kg i.v. PEG-hirudin before angioplasty, followed by a continuous infusion of 0.3 mg/kg for 2 h. Thereafter, a dose of 5 mg/kg was given subcutaneously, and then once daily for 3 days. Animals in the third group (10 arteries) received the same treatment, but the subcutaneous therapy was continued until day 14, a time after which reendothelialization was observed to be complete on electron microscopy examination in earlier experiments. In the fourth group (11 arteries), 150 IU/kg clivarine was administered as a bolus before angioplasty, followed by a continuous intravenous infusion of 10 IU/kg/h for 24 h with an osmotic pump. The pump was filled with 0.9% saline, and the outlet was connected to an elastic tubing cut to a predetermined length to contain exactly the entire dose of clivarine to be administered in 24 h, as assessed in previous in vitro studies. From the day after angioplasty, animals in this group received 75 IU/kg twice daily subcutaneously for 4 weeks. PEG-hirudin and clivarine were supplied by Knoll AG, Ludwigshafen, Germany.

Histology

At the end of each experiment, stent angioplasty segments were excised exactly at the proximal and distal borders of the stent. The segments thus obtained had a comparable length of 15-17 mm. All stents were further processed as previously described (20). After dehydration in graded alcohol concentrations, they were embedded in methylmethacrylate. Six sections (≈4 μm thick) per stent were prepared without prior removal of the stent wire, and Elastica-van Gieson staining was performed. This procedure allowed microscopic analysis of damage limited to the intima or extending into the media. Morphometric analysis of the free luminal area, total vessel area (lumen plus intima plus media), and neointimal area was performed in six slices per stent. Injury was evaluated according to a score proposed by Schwartz (21). Each stent strut per slice was assigned a numerical value: 1 = internal elastic lamina intact; 2 = internal elastic lamina lacerated, media compressed but not lacerated; 3 = internal elastic lamina and media lacerated, external lamina intace; and 4 = large laceration of the media, stent strut in the adventitia. Mean injury scores were calculated as the quotient of the sum of weights for each wire and the number of wires present. Another indicator of the degree of injury to the vessel wall is the ratio of balloon to artery (B/A). The higher this value, the more injury should be imposed on the artery. This value was calculated by the nominal inflated balloon diameter and the vessel area in the segment proximal to the stent, as assessed during postmortem histologic analysis.

Laboratory testing

Blood samples for hemoglobin were taken before angioplasty, 24 h later, and 4 weeks after angioplasty. Activated partial thromboblastin time (aPTT: control and PEG-hirudin groups) was measured before anticoagulant drugs were administered, 1 and 24 h later, and before each subcutaneous dose in the first 3 postoperative days. Anti-factor-Xa activity in group 4 animals was measured before the first clivarine dose, at the end of the 24-h infusion, and before each subcutaneous injection for 3 days.

Statistics

Laboratory variables at different timepoints in a group were compared by the Wilcoxon test. Values for hemoglobin, hematocrit, injury score, neointimal thickness, neointimal area, and minimal luminal diameter from different groups were compared by analysis of variance (ANOVA). Neointimal areas grouped for injury scores ≷2 in the PEG-hirudin and clivarine groups were compared with those in the control group by Fisher's exact test: p < 0.05 was considered statistically significant.

RESULTS

All animals survived throughout the 4-week experimental period. Animals in the control group, given a bolus of unfractionated heparin only, showed only minor bleeding from the neck wound, but this blood loss was stronger and lasted as long as 24 h in the PEG-hirudin and the clivarine groups. Although no other bleeding sources were observed, this loss resulted in a decrease in the hemoglobin concentration of 15-26 g/L in 2 animals in each of the PEG-hirudin groups and the clivarine groups. No change was noted in hemoglobin and hematocrit values in the control group that received a heparin bolus only. Values are shown in Table 1.

aPTT values in control and PEG-hirudin-treated groups are shown in Table 2. All animals showed a several-fold increase of the aPTT by 1 h after angioplasty that normalized in control animals after 6 h but remained prolonged to more than twice control value in both PEG-hirudin groups during the subcutaneous treatment phase. In clivarine-treated animals, anti-factor-Xa activity increased from 0.027 ± 0,002 to 1.07 ± 0,63 IU/ml plasma after the 24-h intravenous infusion and was 0.19 ± 0.11 IU/ml plasma during the subcutaneous treatment phase 12 h after an injection (Table 2).

Histologic analysis

Morphometric analysis could not be performed on all six slices in all treated arteries because of side branch takeoffs covered by the stent, making assessment of vessel and neointimal area in these cases meaningless. However, at least four slices per stent were analyzed. Although the mean values of injury scores tended to be higher in the control group, these differences were not significant between treatment groups (Table 3).

The B/A ratio was comparable between the control and both PEG-hirudin groups. However, in contrast to the injury score observed, the B/A ratio in the clivarine group was lower (Table 3).

Maximal neointimal thickness was greatest in the control group and significantly less in both PEG-hirudin groups and the clivarine groups (Fig. 1). Similarly, neointimal area was significantly lower in all three treatment groups as compared with the control group (Fig. 2). To assess a possible effect of the slightly higher overall injury score in the control group, slices were further grouped according to the observed score in those with a score ≥2.0 (deep injury into the media, lamina elastica interna lacerated by most stent struts) and those with a score <2 (superficial injury, lamina elastica interna largely intact). The neointimal area was greater in all four groups in the slices with deep arterial injury, irrespective of the treatment given. However, although the area covered by the neointima increased from 28 to 52% in the control group, this increase in neointimal proliferation was smaller in all three treatment groups (Fig. 3). In accordance with these findings of reduced neointimal proliferation in the treatment groups, the minimal area of the free vessel lumen was least in the control group. The high value in the clivarine group reflects the relative larger vessel segments, in which the angioplasty was performed, as indicated by the low B/A ratio with a constant balloon diameter of 3.0 mm used (Table 3).

DISCUSSION

Restenosis of initially successfully dilated coronary stenoses is the major limiting factor for the long-term success of percutaneous transluminal coronary angioplasty, resulting in symptomatic recurrence in 6 months in 30-40% of patients. The underlying mechanisms include acute elastic recoil of dilated vessel segments and late neointimal proliferation of SMC invading from the media into the neointima. This proliferation is most likely induced and controlled by growth factors such as PDGF and others. Several growth factors do stimulate SMC growth in vitro, and antagonism of different growth factors in animal models inhibits neointimal proliferation. However, the exact nature of the mechanisms resulting in this neointimal proliferation are not yet clear (11-13). Until very recently, no pharmacologic treatment or new angioplasty technique has succeeded in reducing the restenosis rate in patients, despite promising results with various substances in different angioplasty models (14).

In 1994, two studies showed a reduced recurrence rate after coronary stent implantation by ≈30% as compared with balloon angioplasty alone (14,15), supporting the hypothesis of elastic recoil as a major contributor to restenosis; however, ≈20% of all patients still had a symptomatic recurrence of their disease. Any vessel trauma is followed by an activation of the coagulation cascade, resulting in the formation of a mural thrombus (17,18). The amount of thrombus formed can contribute to restenosis either by an organization and replacement by invading and proliferating medial SMC, by the release of factors stimulating migration and proliferation of SMC from the media (19) or, most likely, by both. Therefore, the amount of acute thrombus formation could be shown to determine the degree of late neointimal proliferation (6,7).

Our results show that anticoagulation with either the direct thrombin inhibitor hirudin, coupled to PEG to prolong plasma t½, or to the LMWH clivarine significantly reduced neointima formation in a minipig coronary stent-angioplasty model to a similar degree, as compared with the standard clinical single bolus of unfractionated heparin.

Neointimal proliferation in this model depends on the degree of vessel wall injury, increasing with deep injury into the vessel media. Although the inhibitory effects of both strategies, PEG-hirudin and clivarine, is significant for all degrees of injury from mild to severe, it is more pronounced in severely injured arterial segments, as is reflected by the free luminal area of the most greatly narrowed segments, which was at least twice the size of the control group in all three treatment groups.

PEG-hirudin

Hirudin, a direct thrombin inhibitor, has been shown to inhibit platelet and fibrin deposition after deep arterial injury in carotid and coronary angioplasty more effectively than heparin. This effect is dose dependent. Hirudin is rapidly cleared from plasma, with a t½ of 1 h largely due to renal excretion (20), and thus requires continuous intravenous infusion or frequent subcutaneous dosing to maintain the aPTT in the required range of twice the normal value models (21,22). Coupling to PEG results in a markedly diminished elimination (23,24) and consequently in a maintained aPTT prolongation with a single daily subcutaneous dose. As shown for standard hirudin, continuation of PEG-hirudin treatment for >3 days until reendothelialization is complete, and thus thrombogenicity of the foreign material is no longer present, provides no additional reduction of the proliferative response. These results are in accordance with those of a clinical study in which an antibody against the platelet IIb/IIIa integrin, involved in the final common pathway of platelet aggregation (25), was used. In that study, although the factor was assessed only clinically, treatment for 12 h reduced restenosis after 6 months, underscoring the role of platelet aggregation in inducing late neointimal proliferation.

Clivarine

The inhibitory effect of heparin on SMC proliferation has long been established. In the setting of arterial angioplasty, heparin must be given for several weeks to produce an effect, since even administration for 48 h did not affect restenosis rate in a clinical study (26). Due to the low bioavailability of unfractionated heparin given subcutaneously, its extended administration is impractical; this led to the preparation of LMW fragments of heparin, which have a much better bioavailability and a retained inhibitory effect on SMC proliferation. Moreover, because LMWHs have a low anti-factor-II activity and exert their anticoagulatory effect predominantly by inhibiting factor Xa, they cause fewer bleeding complications and do not require frequent laboratory monitoring (2,27,28).

Different LMWHs were shown to inhibit neointimal proliferation after angioplasty (3,10,29). We previously showed a 50% reduction in vessel wall area after 4-week treatment with dalteparine in this coronary stent-angioplasty model. However, the effect was achieved at doses of 200-300 anti-Xa U/kg body weight, resulting in plasma anti-Xa levels of ≈1.0 IU/ml (3). These doses are twice as high as is considered clinically safe in terms of bleeding complications (30). Therefore, the reduction in neointimal proliferation observed with the LMWH enoxaparin in an animal model (29) could not be confirmed in a controlled clinical study in which a lower dose was administered only once a day (31). In contrast, clivarine was shown experimentally to reduce restenosis in the rabbit carotid injury model and to be clinically safe in a coronary angioplasty restenosis trial in much lower doses of an anti-factor-Xa basis (75 IU/kg twice daily) than were used in the present study (9,10). Despite a much lower plasma anti-Xa activity in the present study as compared with our earlier study (3), we noted a comparable inhibitory effect on neointimal proliferation. The reason for this discrepancy between anti-Xa levels achieved and inhibition of neointimal proliferation observed is unclear, but could be due to a different effect on SMC. However, because the mechanism by which heparins inhibit SMC proliferation is not clear (32) and because we neither performed dose-response studies nor a direct comparison of these two LMWHs in a single study, this may well be not a true difference.

We showed that both the direct thrombin inhibitor PEG-hirudin, probably by reducing platelet aggregation, and the LMWH clivarine, probably by inhibiting SMC proliferation, reduce neointimal proliferation to a similar degree after coronary stent angioplasty in a minipig model. These results are in accord with the hypothesis of platelet aggregation as a major determinant for restenosis through stimulation of SMC proliferation and suggest the possibility of combining two effective pharmacologic strategies to reduce restenosis rate after coronary angioplasty. Such a combination might prove useful in light of results of a recent clinical study showing that recombinant desulftohirudin, administered for 12 h after coronary angioplasty, does not reduce restenosis rate as compared with standard heparin (33). Further studies are necessary to discover doses of these two substances, which are both effective and safe. At the doses used in this study, bleeding remains a major concern in light of the decrease in hemoglobin values observed in several animals in the drug-treated groups. Fig. 5

Acknowledgment: This work was supported by a research grant from Knoll AG, Ludwigshafen, Germany. We thank Petra Sejdija and Christina Schäfer for expert technical assistance.

FIG. 1.
FIG. 1.:
Maximal neointimal thickness. Four to six slices per stented segment in each artery were analyzed. Columns represent the means ± SD of the greatest values observed in 9, 11, 10, and 11 arteries treated. Neointimal thickness in all treatment groups was significantly lower than that in the control group (*p < 0.05), whereas differences between treatment groups were not significant.
FIG. 2.
FIG. 2.:
Neointimal area, expressed as percent of total vessel area, was significantly lower in all treatment groups as compared with the control group (*p < 0.02). Columns represent means ± SD of all slices analyzed in the four groups; the number of slices analyzed per group is indicated at the bottom of the columns.
FIG. 3.
FIG. 3.:
Neointimal area in slices grouped for injury score ≷2. Columns represent the means ± SD of the number of slices indicated (n), that showed a score of ≥2 (hatched columns) or <2 (open columns). In all four groups, values for injury scores >2 as compared with those with a score <2 were significantly higher (p < 0.05). Differences between polyethyleneglycol hirudin (PEGH) and clivarine groups were not significant, whereas these treatment groups showed significantly lower values as compared with the control group (*p < 0.01).
FIG. 4. A and B:
FIG. 4. A and B::
Histologic sections through stent-angioplasty segments (Elastica-van Giesson staining) in animals treated with polyethyleneglycol (PEG)-hirudin for 3 days. The amount of neointimal proliferation was very low around stent struts that did not penetrate through the elastic internal lamina [original magnification ×20 (A) ×300 (B)].
FIG. 5. A and B:
FIG. 5. A and B::
With penetration of a strut through the elastic lamina interna into the media, proliferation was much more intense as compared with that shown in Fig. 4A and B. These findings were consistently observed in all groups and suggest that, irrespective of the pharmacologic treatment given, the extent of vessel wall injury primarily determines the amount of subsequent neointimal proliferation.

REFERENCES

1. Clowes AW, Karnovsky MJ. Suppression of heparin of smooth muscle cell proliferation in injured arteries. Nature 1977;265:625-6.
2. Karnovsky MJ, Wright TC Jr, Castellot JJ, Choay J, Lormeau JC, Petitou M. Heparin, heparan sulfate, smooth muscle cells, and atherosclerosis. Ann NY Acad Sci 1989;556:268-81.
3. Buchwald AB, Unterberg C, Nebendahl K, Gröne HJ, Wiegand V. Low-molecular weight heparin reduces neointimal proliferation after coronary stent implantation in hypercholesterolemic minipigs. Circulation 1992;86:531-7.
4. Sarembock IJ, Gertz SD, Gimple LW, Owen RM, Powers ER, Roberts WC. Effectiveness of recombinant desulphatohirudin in reducing restenosis after balloon angioplasty of atherosclerotic femoral arteries in rabbits. Circulation 1991;84:232-43.
5. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 1993;362:801-9.
6. Willerson JT, Yao SK, McNatt J, et al. Frequency and severity of cyclic flow alternations and platelet aggregation predict the severity of neointimal proliferation following experimental coronary stenosis and endothelial injury. Proc Natl Acad Sci USA 1991;88:10624-8.
7. Unterberg C, Sandrock D, Nebendahl K, Buchwald AB. Reduced acute thrombus formation results in decreased neointimal proliferation after coronary angioplasty. J Am Coll Cardiol 1995;26:1747-54.
8. Jeske W, Blazejlowski BS, Walenga JM, Hoppenstaedt D, Ahsan A, Fareed J. Biochemical and pharmacologic profile of low molecular weight heparin (LU 47311, Clivarine). Semin Thromb Hemost 1993;19(suppl 1):229-40.
9. Schmid KM, Preisack M, Voelker W, Sujatta M, Karsch KR. First clinical experience with low molecular weight heparin LU 47311 (Reviparin) for prevention of restenosis after precutaneous transluminal coronary angioplasty. Semin Thromb Hemostas 1993;19(suppl 1):155-9.
10. Hanke H, Oberhoff M, Hanke S, et al. Inhibition of cellular proliferation after experimental balloon angioplasty by low molecular weight heparin. Circulation 1992;85:1548-55.
11. Ferns GA, Raines EW, Sprugel KH, Motani AS, Reidy MA, Ross R. Inhibition of neointimal smooth muscle accumulation after angioplasty by an antibody to PDGF. Science 1991;253:1129-32.
12. Santoian EC, Schneider JE, Gravanis MB, et al. Angiopeptin inhibits intimal hyperplasia after angioplasty in porcine coronary arteries. Circulation 1993;88:11-4.
13. Lindner V, Reidy MA. Proliferation of smooth muscle cells after vascular injury is inhibited by an antibody against basic fibroblast growth factor. Proc Natl Acad Sci USA 1991;88:3739-43.
14. Herrmann JPR, Hermans WRM, Vos J, Serruys PW. Pharmacologic approaches to the prevention of restenosis following angioplasty. Drugs 1993;46:18-52.
15. Serruys PW, De Jaegere P, Kiemeneji F, et al., for the Benestent Study Group. A comparison of balloon-expandable-plant implantation with balloon angioplasty in patients with coronary artery disease. N Engl J Med 1994;331:489-95.
16. Fishman DL, Leon MB, Baim DS, et al., for the Stent Restenosis Study Investigators. A randomized comparison of coronary stent placement and balloon angioplasty in the treatment of coronary artery disease. N Engl J Med 1994;331:496-501.
17. Waller BF. Coronary luminal shape and the arc of the disease-free wall: morphologic observations and clinical relevance. J Am Coll Cardiol 1985;6:1100-1.
18. Waller BF, Pinkerton CA, Orr CM, Slack JD, van Tassel JW, Peters T. Restenosis 1 to 24 months after clinically successful coronary balloon angioplasty: a necropsy study of 20 patients. J Am Coll Cardiol 1991;17:58B-70B.
19. Liu MW, Roubin GS, King SB. Restenosis after coronary angioplasty: potential biologic determinants and role of intimal hyperplasia. Circulation 1989;79:1374-87.
20. Bichler J, Siebeck M, Fichtl B, Fritz H. Pharmacokinetics, effect on clotting tests and assessment of the immunogenic potential of hirudin after a single subcutaneous or intravenous bolus administration in man. Haemostasis 1991;21(suppl 1):137-41.
21. Heras M, Chesebro JH, Webster MWI, et al. Hirudin, heparin, and placebo during deep arterial injury in the pig. Circulation 1990;82:1476-84.
22. Buchwald AB, Sandrock D, Unterberg C, et al. Platelet and fibrin deposition on coronary stents in minipigs: effect of hirudin versus heparin. J Am Coll Cardiol 1993;21:249-54.
23. Rübsamen K, Hornberger W, Laux V, Schwarz M, Schweden J. Antithrombotic efficacy of the polyethyleneglycol-coupled hirudin LU 57291 in experimentally induced venous and arterial thrombosis [Abstract]. Thromb Hemost 1993;69:2717.
24. Esslinger HU, Dübbers K, Müller-Peltzer H. First data on safety and anticoagulant activity after single i.v. and s.c. PEG-hirudin administration in healthy volunteers [Abstract]. Thromb Hemost 1995;73:2109.
25. Topol EJ, Califf RM, Weisman HF, et al. Randomised trial of coronary intervention with antibody against platelet IIb/IIIa integrin for reduction of clinical restenosis: results at six months. Lancet 1994;343:881-6.
26. Ellis GS, Roubin GS, Wilentz J, Douglas JS, King SB. Effect of 18- to 24-hour heparin administration for prevention of restenosis after uncomplicated coronary angioplasty. Am Heart J 1989;117:777-82.
27. Hirsh J, Levine MN. Low molecular weight heparin. Blood 1992;79:1-17.
28. Lockner D, Bratt G, Tornebohm E, Aberg W. Pharmacokinetics of intravenously and subcutaneously administered fragmin in healthy volunteers. Haemostasis 1986;16(suppl 2):8-10.
29. Currier JW, Pow TK, Haudenschild CC, Minihan AC, Faxon DP. Low molecular weight heparin (Enoxaparin) reduces restenosis after iliac angioplasty in the hypercholesterolemic rabbit. J Am Coll Cardiol 1991;17:118B-25B.
30. Faxon DP, Spiro TE, Minor S, et al. Low molecular weight heparin in prevention of restenosis after angioplasty. Circulation 1994;90:908-14.
31. Wallentin LC, Swahn E. FRISC: low molecular weight heparin (fragmin) during instability in coronary artery disease. Eur Heart J 1995;16(suppl):A2544.
32. Castellot JJ Jr, Cochran DL, Karnovsky MJ. Effect of heparin on vascular smooth muscle cells: I. Cell metabolism. J Cell Physiol 1985;124:21-8.
33. Serruys PW, Herrman JPR, Simon R, et al. A comparison of hirudin with heparin in the prevention of restenosis after coronary angioplasty. N Engl J Med 1995;333:757-63.
Keywords:

Clivarine; Polyethyleneglycol-Hirudin; Coronary stents; Neointima

© Lippincott-Raven Publishers