The inflammatory effects of crystalline cholesterol monohydrate in the guinea pig gallbladder in vivo

The inflammatory effects of crystalline cholesterol monohydrate in the guinea pig gallbladder in vivo Jay B. Prystowsky, MD, and Robert V. Rege, MD, Chicago, Ill.

Background. The etiologic role of crystalline material in inflammatory arthritis is well established. The role of crystals in cholecystitis is unclear. We hypothesized that crystalline cholesterol monohydrate stimulates guinea pig gallbladder inflammation in vivo. Methods. Crystalline cholesterol monohydrate, lipopolysaccharide (LPS), lysolecithin, polystyrene latex spheres (noninflammatory particles), and saline were instilled into guinea pig gallbladders for 24 to 72 hours after cystic duct ligation. Water transport across gallbladder mucosa was measured. Gallbladder tissue was analyzed for mucus layer thickness, myeloperoxidase, prostaglandin E2 (PGE2), prostaglandin F-1α (PGF-1α), and interleukin-1. Luminal fluid was also examined for PGE2 and PGF-1α. Values for each test were compared with saline controls by using Student’s t test (p < 0.05). Results. Crystalline cholesterol, LPS, and lysolecithin caused significant reduction in mucus layer thickness, reversed water absorption to secretion across the gallbladder mucosa, caused significant increases in myeloperoxidase and interleukin-1 in gallbladder tissue, and caused significant increases in PGE2 and PGF-1α in luminal fluid. These effects were generally dose- but not time-dependent. Polystyrene latex particles caused no difference in outcomes compared with saline controls. Conclusions. Crystalline cholesterol monohydrate has dose-dependent inflammatory effects in the guinea pig gallbladder in vivo that are not simply due to mechanical irritation of the gallbladder wall by crystalline particles. Crystals in the gallbladder may have an etiologic role in cholecystitis. (Surgery 1998;123:258-63.) From the Department of Surgery, Northwestern University Medical School, Chicago, Ill.

THE TERM CRYSTAL DEPOSITION DISEASE has been used to describe the condition wherein (1) metabolic factors lead to supersaturated body fluid, (2) pronucleating factors promote crystal formation, (3) crystals induce tissue inflammation, and (4) a chemically defined crystal is specific for each disease.1 The paradigm of crystal deposition disease is gout in that monosodium urate crystals induce acute or chronic articular inflammation.2 In previous studies we observed that anhydrous crystalline cholesterol (1) is phagocytized in vitro by human polymorphonuclear leukocytes with release of oxidative enzymes involved in inflammation and (2) stimulates human mononuclear cells in vitro to release interleukin-1β.3-5 Also, in preliminary experiments we found that anhydrous crystalline cholesterol may cause inflammation in the guinea pig gallbladder in vivo.6 On the basis of Accepted for publication Aug. 5, 1997. Reprint requests: Jay B. Prystowsky, MD, Department of Surgery, Northwestern University Medical School, 300 E. Superior St., Tarry Building 11-759, T231, Chicago, IL 60611. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0 11/56/85939

258 SURGERY

the results of these studies, we speculate that cholecystitis may be a crystal deposition disease in which biliary crystals induce inflammation in the gallbladder wall. In this investigation we hypothesized that crystalline cholesterol monohydrate (crys-XOL; the crystalline form of cholesterol that occurs in human gallbladders) has inflammatory effects in the guinea pig gallbladder in vivo and these effects are dose- and time-dependent. MATERIAL AND METHODS All experiments were performed with the approval of the Animal Care and Use Committee and the Institutional Review Board of Northwestern University. Unless otherwise specified, all agents and solutions were obtained from Sigma Co., St. Louis, Mo. Crys-XOL was prepared by recrystallization of anhydrous cholesterol (reagent grade) from 10% aqueous acetone.7 Lipopolysaccharide (LPS; proinflammatory stimulant) and lysolecithin (LLC; phospholipid with cytolytic properties that causes cholecystitis in some circumstances) were used as positive controls. Polystyrene latex parti-

Surgery Volume 123, Number 3 cles (4 to 5 µm diameter spheres, chemically inert; IDC Spheres, Portland, Ore.) and saline alone were used as negative controls. LPS, LLC, and polystyrene latex particles were used as reagent preparations. Guinea pigs were anesthetized with an intraperitoneal injection of ketamine 60 mg/kg and xylazine 8 mg/kg. Guinea pigs then underwent a midline laparotomy under sterile conditions. After aspiration of gallbladder bile, the cystic duct was ligated, with care taken to preserve the cystic artery. A cholecystostomy was fashioned by using PE-50 tubing. Saline, crys-XOL (0.2 to 2.0 mg/ml as a crystalline suspension), LPS (10 to 100 µg/ml), LLC (2 to 20 mg/ml), and polystyrene latex particles (2.0 mg/ml in suspension) were instilled into gallbladders for 24 and 72 hours (n = 5 in each group). At the conclusion of each time point, animals were anesthetized and gallbladders were rapidly excised. Animals were killed before awakening from anesthesia. Water transport was determined by using the nonabsorbable tracer, 14C-polyethylene glycol-4000 (14C-PEG; DuPont, Boston, Mass.).8,9 A known volume and concentration of the tracer were placed into the gallbladder lumen at the beginning of the experiment. At the conclusion of the experiment, an aliquot of luminal fluid was obtained and counted for 14C-PEG. The final gallbladder volume was determined as Vf = ([PEG]f/[PEG]i ) · Vt, where Vi, Vf, PEGi, and PEGf are initial and final volumes and PEG concentrations, respectively. Absorption or exsorption of fluid was calculated by Vf – Vt. Mucus layer thickness was measured by using dark-field microscopy.10 A strip of gallbladder tissue was placed on a millipore filter, mucosal side up. The strip of gallbladder was cut by using two razor blades mounted 1 mm apart, and the strip adherent to the filter was mounted on a microscope slide between two pieces of cork. Measurement at 15 different sites along the strip of gallbladder was performed by using dark-field microscopy and a calibrated eyepiece, and average thickness was calculated. Myeloperoxidase (enzyme released by inflammatory cells) was measured from gallbladder wall.9,11 Gallbladder tissue was weighed and then homogenized in 2 ml of cold 0.5% hexadecyltrimethyl ammonium bromide in 50 mmol/L phosphate buffer, pH 6.0. The homogenate was sonicated for 20 seconds, freeze-thawed three times, and centrifuged at 1200 rpm for 15 minutes. Supernatant was assayed for myeloperoxidase activity by mixing 40 µl with 460 µl of 1.6 mmol/L 3,3,5,5 tetramethyl benzidine, 0.3 mmol/L hydro-

Prystowsky and Rege 259

gen peroxide, 8% N,N-dimethylformamide, and 80 mmol/L sodium phosphate buffer. The reaction solution was incubated for 3 minutes at 37° C and then immersed in an ice bath. After 10 minutes, 1.75 ml of 200 mmol/L sodium acetate buffer, pH 3.0, was added and absorbance was read at 655 nm within 30 minutes. One unit of myeloperoxidase activity was defined as that degrading 1 µmol/min of peroxide at 37° C. Prostaglandins (PGE2 and PGF-1α) were measured from gallbladder tissue and luminal fluid by using radioimmunoassay (Amersham, Corp., Arlington Heights, IL).12,13 A section of gallbladder tissue was weighed and placed in 3 ml of 5 mmol/L MgCl2 and homogenized with a polytron homogenizer for 1 minute. Centrifugation for 1 hour at 1000 g was performed, and supernatant was centrifuged at 100,000 g for 1 hour at 4° C. Supernatant (1 ml) was added to 3 ml of petroleum ether in a polypropylene tube and vortexed for 30 seconds to remove neutral lipids. The ether layer was removed and added to 3 ml of ethyl acetate buffer (ethyl acetate:isopropanol:0.2N HCl; 3:3:1 v/v/v). After vortexing for 15 seconds twice, 2 ml of ethyl acetate and 3 ml of distilled water were added. After mixing, the layers were separated by centrifugation. The organic phase was transferred to a polypropylene tube and dried. The residue was resuspended in appropriate radioimmunoassay buffer and radioimmunoassay was performed. Interleukin-1 (IL-1) was measured by using a murine thymocyte technique in which IL-1 is known to stimulate the uptake of thymidine by proliferating murine thymocytes.14 Guinea pig gallbladder wall was weighed, minced, and incubated in RPMI medium with 10% fetal calf serum overnight at 37° C.15 Cells and tissue were discarded, and supernatants were frozen until use. The thymus glands of balb/C mice were removed, minced, and then washed in Hank’s balanced salt solution. The gallbladder tissue supernatant (0.1 ml) was placed into a 96-well microtiter plate with 106 mouse thymocytes and then incubated for 48 hours with 5% CO2 at 37° C. Tritiated thymidine (0.5 µCi; ICN Pharmaceuticals, Inc., Irvine, Calif.) was added to each well and incubated again with 5% CO2 at 37° C for 8 hours. Cultures were harvested and counts were measured in a scintillation counter. Recombinant human IL-1α (R & D Systems, Minneapolis, Minn.) was used to construct a standard curve for the bioassay, and results are reported as picograms (IL-1α equivalents) per milligram gallbladder tissue. Values for each outcome were compared with saline controls by using Student’s t test (p < 0.05).

260 Prystowsky and Rege

Surgery March 1998

Table I. Crystal-induced guinea pig cholecystitis in vivo Myeloperoxidase (units/mg/min) Saline Crys-XOL 0.2 mg/ml 1.0 mg/ml 2.0 mg/ml LPS 10 µg/ml 50 µg/ml 100 µg/ml LLC 2 mg/ml 10 mg/ml 20 mg/ml Latex particles 2 mg/ml

Water transport (ml/24 hr)

24 hr

72 hr

24 hr

72 hr

172 ± 51

201 ± 59

0.32 ± 0.18

0.45 ± 0.18

264 ± 29* 505 ± 86* 802 ± 167*

334 ± 53*† 567 ± 131* 952 ± 190*

–0.04 ± 0.12* –0.33 ± 0.07* –0.42 ± 0.07*

–0.16 ± 0.1* –0.41 ± 0.04* –0.75 ± 0.25*†

313 ± 57* 489 ± 59* 806 ± 175*

340 ± 88* 555 ± 99* 1099 ± 274*

–0.17 ± 0.12* –0.38 ± 0.08* –0.7 ± 0.32*

–0.24 ± 0.14* –0.64 ± 0.28* –0.93 ± 0.15*

303 ± 37* 553 ± 48* 851 ± 15*

319 ± 40* 583 ± 155* 914 ± 303*

–0.11 ± 0.05* –0.33 ± 0.02* –0.46 ± 0.2*

–0.13 ± 0.11* –0.31 ± 0.06* –0.72 ± 0.19*

206 ± 40

198 ± 38

0.40 ± 0.15

0.35 ± 0.16

Values are mean ± standard deviation. For water transport, positive values indicate absorption of fluid across gallbladder mucosa, whereas negative values indicate exsorption of fluid into gallbladder lumen. *Significantly different than value for saline control using Student’s t test (p < 0.05). †Significantly

different than value at 24 hours using Student’s t test (p < 0.05).

Values for each concentration of each agent were compared at 24 and 72 hours by using Student’s t test (p < 0.05) to assess the effect of increasing time on inflammation. Values for each agent were compared by using ANOVA to assess the effect of concentration on inflammation, and correlation coefficients were calculated at both 24 and 72 hours to determine dose-dependency of inflammation for each agent. RESULTS Crys-XOL at all concentrations (0.2 to 2.0 mg/ml) caused significant increases in gallbladder tissue myeloperoxidase compared with saline controls at both 24 and 72 hours (Table I). The increase in myeloperoxidase activity was dose dependent at both 24 and 72 hours (r > 0.98) but was not time dependent because there was no significant difference between 24 and 72 hours. LPS and LLC also caused significant increases in gallbladder tissue myeloperoxidase at both 24 and 72 hours compared with saline controls. Again, the increases in myeloperoxidase caused by LPS and LLC were dose dependent (r > 0.98) but not time dependent. Crys-XOL reversed net water absorption to net secretion over the range of concentrations studied at both 24 and 72 hours (Table I). There was no significant difference in water secretion between 24 and 72 hours, but we did observe a dose-dependent effect (r > 0.98) for both time points. Likewise, LPS

and LLC reversed water absorption to secretion over the range of concentrations studied at both 24 and 72 hours. LPS and LLC effects on water secretion were dose dependent (r > 0.99) but not timedependent. Crys-XOL (1.0 and 2.0 mg/ml) caused significant reduction in mucus layer thickness compared with saline controls at 24 hours (Table I). There was a significant decrease in mucus layer thickness from 24 to 72 hours, and the effects at each time point were inversely and linearly related to crystal concentration (r > -0.99). Similarly, LPS (50 and 100 µg/ml) and LLC (10 and 20 mg/ml) caused significant diminution in mucus layer thickness at 24 hours compared with saline. The effects of LPS and LLC were greater at 72 hours compared with 24 hours, and their effects were inversely related to concentration (r > -0.91). Crys-XOL (1.0 and 2.0 mg/ml) stimulated significant increases in IL-1 release in gallbladder tissue at both 24 and 72 hours (Table I). Its effects were dose dependent (r > 0.98) but not time dependent. Similarly, LPS (50 and 100 µg/ml) and LLC (10 and mg/ml) stimulated IL-1 release in gallbladder tissue at both 24 and 72 hours. The LPS-stimulated release of IL-1 was dose dependent (r > 0.96) but not time dependent. The LLC-stimulated release of IL-1 was neither dose nor time dependent. Crys-XOL stimulated no increases in PGE2 or PGF-1α in gallbladder tissue at either 24 or 72

Prystowsky and Rege 261

Surgery Volume 123, Number 3

Mucus layer thickness (µm) 24 hr 72 hr 126 ± 35

152 ± 57

IL-1 (pg/mg) 24 hr 72 hr 5.5 ± 1.4

5.5 ± 1.4

126 ± 6 87 ± 3* 47 ± 4*

78 ± 8*† 36 ± 3*† 20 ± 2*†

7.5 ± 1.2 10.0 ± 1.5* 14.0 ± 1.4*

5.7 ± 1.2 10.1 ± 3.0* 12.8 ± 1.2*

96 ± 20 46 ± 4* 20 ± 5*

64 ± 9*† 13 ± 2*† 1.8 ± 0.8*†

7.2 ± 2.1 11.4 ± 2.6* 20.8 ± 2.7*

6.5 ± 0.9 9.6 ± 2.5* 21.5 ± 3.5*

124 ± 14 76 ± 2* 47 ± 4*

87 ± 4*† 29 ± 3*† 2.4 ± 0.6*†

7.4 ± 0.9* 9.9 ± 2.4* 9.0 ± 2.4*

7.5 ± 1.7 9.1 ± 1.6* 10.0 ± 1.5*

102 ± 13

109 ± 15

6.3 ± 0.9*

5.9 ± 0.7

hours (Table II). However, crys-XOL (1.0 and 2.0 mg/ml) did stimulate significant release of PGE2 and PGF-1α into gallbladder luminal fluid at 24 hours and caused significant release of PGE2, but not PGF-1α, into luminal fluid at 72 hours. The effect of crys-XOL on prostaglandin release into luminal fluid was dose dependent (r > 0.99) but not time dependent. LPS (50 µg/ml and 100 µg/ml) also caused significant release of PGE2 and PGF-1α into gallbladder luminal fluid at 24 hours and caused significant release of PGE2, but not PGF-1α, into luminal fluid at 72 hours. The effects of LPS were dose dependent (r > 0.98) but not time dependent. LPS did not stimulate increases in PGE2 or PGF-1α in gallbladder tissue at either 24 or 72 hours. LLC caused no release of PGF-1α into gallbladder tissue or luminal fluid at either 24 or 72 hours. LLC at all concentrations (2 to 20 mg/ml) caused significant release of PGE2 into gallbladder tissue and luminal fluid at both 24 and 72 hours. The release of PGE2 in gallbladder tissue was not dose dependent, but PGE2 release into luminal fluid was dose dependent (r > 0.98) at both 24 and 72 hours. Furthermore, a significant decrease in tissue and fluid PGE2 was observed from 24 to 72 hours at LLC concentration of 20 mg/ml. Time-dependent decrease in fluid PGE2 was also observed for LLC at 10 mg/ml but was not noted for tissue PGE2. No time-dependent changes were observed for PGE2 release at LLC concentration of 2 mg/ml.

Polystyrene latex particles (2.0 mg/ml) caused a slight increase in IL-1 release in gallbladder tissue at 24 hours but otherwise demonstrated no significant difference compared with saline with respect to all of the tests measured in this investigation. DISCUSSION During the past 20 years several crystals have been implicated in the pathogenesis of acute and chronic articular syndromes such as gout.2 The urate crystal is the causative agent of human gout. It is a consistent finding in the joint fluid of patients with acute gouty arthritis, and the syndrome can be reproduced by injection of synthetic urate crystals into joints.16-18 Urate crystals stimulate white blood cells and synovial cells in vitro and in vivo to release inflammatory mediators. It is proposed here that the effects of urate crystals on white blood cells and synovial cells are a paradigm for the effects of biliary crystals in the gallbladder. As currently conceived, gallstone formation involves three broad areas: (1) supersaturation of bile with cholesterol and/or calcium salts of bilirubinate, carbonate, phosphate, and fatty acylates, (2) nucleation of crystals from bile, and (3) stasis of gallbladder bile that allows time for growth of crystals into stones.19-21 The inflammatory effects of certain crystals in the joint space led to our hypothesis that cholecystitis may be, in part, a crystal deposition disease. Local metabolic disturbances and pronucleating factors lead to crystal formation in the mucus layer overlying the gallbladder epithelium. These crystals then would have the potential to induce inflammation in the gallbladder wall, if they are capable of stimulating gallbladder epithelial cells or white blood cells to release inflammatory mediators. In this experiment we demonstrated that crysXOL does have a significant inflammatory effect in the guinea pig gallbladder in vivo. Crys-XOL caused significant increases in myeloperoxidase and IL-1 in gallbladder tissue, caused significant increases in PGE2 and PGF-1α in gallbladder luminal fluid, reversed net water absorption to secretion, and caused significant diminution in mucus layer thickness. In general, these effects were dose dependent but not time dependent. LPS and LLC served as positive controls in this experiment, and they caused inflammatory effects in the gallbladder similar to those observed with crys-XOL. Again, the inflammatory effects of LPS and LLC were generally dose dependent but not time dependent. In comparison, the severity of gallbladder inflammation produced by these three agents was LPS > Crys-XOL = LLC.

262 Prystowsky and Rege

Surgery March 1998

Table II. Prostaglandin release in crystal-induced guinea pig cholecysitis in vivo Tissue (pg/mg) 24 hr 72 hr Saline <5 Crys-XOL 0.2 mg/ml <5 1.0 mg/ml <5 2.0 mg/ml <5 LPS 10 µg/ml <5 50 µg/ml <5 100 µg/ml <5 LLC 2 mg/ml 1463 ± 332* 10 mg/ml 1364 ± 538* 20 mg/ml 1272 ± 344*† Latex particles 2 mg/ml <5

PGE2

<5

Fluid (pg/ml) 24 hr 72 hr

PGF-1α Tissue (pg/mg) Fluid (pg/ml) 24 hr 72 hr 24 hr 72 hr

<5

< 15

< 15

<5

< 15

< 15

<5 <5 <5

<5 421 ± 324* 1699 ± 433*

<5 409 ± 504* 1224 ± 453*

< 15 < 15 < 15

< 15 < 15 < 15

< 15 < 15 3382 ± 1210*† < 15 6280 ± 785*† < 15

<5 <5 <5

83 ± 120 739 ± 492* 2540 ± 442*

100 ± 138 995 ± 474* 2482 ± 215*

< 15 < 15 < 15

< 15 < 15 < 15

< 15 < 15 2459 ± 1101*† < 15 6692 ± 670*† < 15

2415 ± 539* 2833 ± 335*† 3280 ± 139*†

1872 ± 235* 2405 ± 205* 2784 ± 333*

< 15 < 15 < 15

< 15 < 15 < 15

< 15 < 15 < 15

< 15 < 15 < 15

< 15

< 15

< 15

<15

1055 ± 393* 923 ± 277* 674 ± 381* <5

<5

<5

Values are mean ± standard deviation. *Significantly different than value for saline control using Student’s t test (p < 0.05). †Significantly different than value at 72 hours using Student’s t test (p < 0.05).

In this experimental model all guinea pigs underwent cystic duct ligation to ensure that crystals remained in the gallbladder lumen. It is probable that cystic duct ligation produces an element of stasis that may contribute to gallbladder inflammation. However, in saline control animals that underwent cystic duct ligation, no inflammation was observed suggesting that cystic duct ligation may be necessary but not sufficient for gallbladder inflammation in the guinea pig. Monosodium urate crystals are well-known as the causative agent of gout, but other types of crystals may also cause inflammatory arthritis.2 Some crystals, such as L-cysteine, have no inflammatory effects in the joint space.22,23 We previously observed differences in the neutrophil phagocytosis in vitro of crystalline cholesterol, bilirubinate, and calcium hydroxyapatite, supporting the concept that crystals have varying degrees of inflammatory potential.4 It has been suggested that negative surface charge of crystals correlates with the inflammatory potential of crystals.22,24 It is not clear what characteristic(s) confers inflammatory potential on crystals, but it is well accepted that crystal-induced arthritis is specific for certain crystals and not simply due to mechanical contact of intraarticular crystalline particles with synovial membrane. Polystyrene latex particles served as a negative control in this experiment in that they are chemically inert. The mechanical effects of these particles did not cause significant changes in the mea-

sured inflammatory markers compared with saline controls. The chemical characteristics and surface properties of polystyrene latex particles differ from crys-XOL. Therefore it is not clear whether the inflammatory effects of crys-XOL are due to its chemical characteristics, surface properties, or both. Nevertheless, the lack of inflammation with polystyrene particles suggests that the inflammatory effects of crys-XOL are not solely due to mechanical contact of luminal particles with the gallbladder wall. In summary, crys-XOL caused inflammation in the guinea pig gallbladder in vivo. The inflammatory effects were dose dependent but not timedependent and were similar to the inflammatory effects of known proinflammatory stimulants, LPS and LLC. The inflammatory effects of crys-XOL do not appear to be solely due to mechanical contact of crystalline particles with the gallbladder wall. These data support the concept that crystals in the gallbladder may stimulate inflammation and may have a specific etiologic role in cholecystitis. Future studies will need to be performed to define the inflammatory potential of other biliary crystals in the gallbladder in vivo. REFERENCES 1. McCarty DJ. Crystal deposition joint disease. Annu Rev Med 1974;25:279-88. 2. Terkeltaub R. Gout: crystal-induced inflammation. In: Gallin JI, Goldstein IM, Snyderman R, editors. Inflammation: basic principles and clinical correlates. New York: Raven Press; 1992. p 977-81.

Prystowsky and Rege 263

Surgery Volume 123, Number 3 3. Prystowsky JB, Huprikar JS, Rege RV. Human polymorphonuclear leukocytes phagocytize crystalline cholesterol in vitro. Hepatology 1992;16:154a. 4. Prystowsky JB, Huprikar JS, Rademaker AW, Rege RV. Human polymorphonuclear leukocyte phagocytosis of crystalline cholesterol, calcium hydroxyapatite, and bilirubin in vitro. Dig Dis Sci 1995;40:412-8. 5. Prystowsky JB, Huprikar JS, Rademaker AW, Rege RV. Crystalline cholesterol, bilirubin, and calcium hydroxyapatite stimulate human peripheral blood mononuclear cells to release interleukin-1 in vitro. Surg Forum 1993;44:183-4. 6. Prystowsky JB, Rege RV. Crystalline cholesterol has inflammatory properties in the guinea pig gallbladder in vivo. Surg Forum 1994;45:146-8. 7. Bogardus JB. Importance of viscosity in the dissolution rate of cholesterol in monooctanoin solutions. J Pharm Sci 1984;73:906-10. 8. Svanvik J, Pellegrini CA, Allen B, Bernhoft R, Way LW. Transport of fluid and biliary lipids in the canine gallbladder in experimental cholecystitis. J Surg Res 1986;41:42531. 9. Fitzgerald S, Deshpande YG, Nguyen HQ, Kaminski DL. The effect of capsaicin on gallbladder fluid absorption. Hepatology 1991;14:660-4. 10. Kerss S, Allen A, Garner A. A simple method from measuring thickness of the mucus gel layer adherent to rat, frog and human gastric mucosa: influence of feeding, prostaglandin, N-acetylcysteine and other agents. Biochem Soc Med Res Soc 1982;63:187-95. 11. Suzuki K, Ota H, Sasagawa S, Sakatani T, Fujikura T. Assay method for myeloperoxidase in human polymorphonuclear leukocytes. Ann Biochem 1983;132:345-52. 12. Dray F, Charbonnel B, Maclouf J. Radioimmunoassay of

13.

14.

15.

16. 17. 18.

19. 20. 21. 22. 23.

24.

prostaglandins F-1α, E1 and E2 in human plasma. Eur J Clin Invest 1975;5:311-8. Song MK, Adham NF. Relationship between zinc and prostaglandin metabolisms in plasma and small intestine of rats. Am J Clin Nutr 1984;41:1201-9. Luger TA, Stadler BM, Luger BM, Mathieson BJ, Mage M, Schmidt JA, et al. Murine epidermal cell-derived thymocyte-activating factor resembles murine interleukin 1. J Immunol 1982;128:2147-52. Pearce W, Sweis I, Yao JS, McCarthy WJ, Koch AE. Interleukin-1β and tumor necrosis factor-α release in normal and diseased human infrarenal aortas. J Vasc Surg 1992;16:784-9. McCarty DJ, Hollander JL. Identification of urate crystals in gouty synovial fluid. Ann Intern Med 1962;54:452-60. Seegmiller JE, Howell RR, Malawista SE. The inflammatory reaction to sodium urate. JAMA 1962;180:125-31. Faires JS, McCarty DJ. Acute arthritis in man and dog after intrasynovial injection of sodium urate crystals. Lancet 1962;2:682-4. Nahrwold DL, Prystowsky JB, Rege RV. Biliary lithotripsy. Curr Probl Surg 1990;27:709-72. Moore E. Biliary calcium and gallstone formation. Hepatology 1990;12:206S-18S. Johnston DE, Kaplan MM. Pathogenesis and treatment of gallstones. N Engl J Med 1993;328:412-21. Mandel NS. The structural basis of crystal-induced membranolysis. Arthritis Rheum 1976;9:439-45. Lietman PS, Frazier PD, Wong VG, Shotton D, Seegmiller JE. Adult cystinosis: a benign disorder. Am J Med 1966;40:511-7. Dieppe P, Hornby J, Swan A, Hutton C, Preece A. Laboratory handling of crystals. Ann Rheum Dis 1983;42:60-3.

Surgery is abstracted and/or indexed in Index Medicus, Science Citation Index, Current Contents/ Clinical Medicine, Current Contents/Life Sciences, Excerpta Medica, and MEDLINE. This Journal has been registered with Copyright Clearance Center, Inc., 222 Rosewood Dr., Danvers, MA 01923. Consent is given for the copying of articles for personal or internal use of specific clients. This consent is given on the condition that the copier pay directly to the Center the per-copy fee stated on the first page of each article for copying beyond that permitted by U.S. Copyright Law. This consent does not extend to other kinds of copying, such as for general distribution, resale, advertising and promotional purposes, or for creating new collective works. All inquiries regarding copyrighted material from this publication other than those that can be handled through Copyright Clearance Center should be directed to Associate Journal Publisher, Permissions, Mosby, Inc., 11830 Westline Industrial Dr., St. Louis, MO 63146; phone (314)453-4129; fax (314)432-1380.