Early hepatic dysfunction in a large animal model of nonhypotensive sepsis

S SULLIVAN, M TROSTER, A LINTON. Early hepatic dysfunction in a large animal model of nonhypotensive sepsis. Can J Gastrocnterol 1991 ;5(6):219223 . epsis was induced in 12 sheep by cecal deva~cularization and perforation . lmravenous crystallo id was administe red to maintain th e pulmonary capillary wedge pressure a t preseptic levels. By 24 h t he re was ,1 significant drop in albumin and a lka line phosphatase with inc reases in aspartatc aminotransferase (AST), laccate dehydrogenase and bilirubin. By 48 h the alkaline phosphatase had returned to presepsis levels, the bili rubin continued co rise, t he AST and lacrn te deh•drogenasc had plmeaued while the a lhumin remained low. T hese biochemical alterations were confirmed by exam ining da ta from 25 sheer used in simi lar epsis experiments. These data revealed that marked elevations in AST were as~ociared with a high er cen tral venous pressure an d worse ren a l impa irment, but there was no relationshi p with any other hemodynamic parameter, P02, pl-! or ~erum lactare. l listologically these biochemical alterat ions were associaLed with extensive m icrovesicular fan y change a l 24 h and centrilohula r necros is at 48 h. Electron. microscopy revea led mi tochondrial degeneration , hyperplas ia of th e smooch endoplasmic reticulum :rnd intrace llular ch o lestasis.

ln the present sr udy the auth1ws clocumenL the early biochem ical and histo logical changes associated with nunhyp111 ens1vc sepsis mduccd by expen mem,11 peritnniris in ,heep.
Control measun.?menrs n( h emndynamic, ren,1 I and hepatic (unction were obrn ined 4 h hcfme induct ion o( sepsis.Under general ant::sthesi:i a mid-linc laparotomy was perfimrn:d.The cccum and ilencecal junction were 1dcnttficLI and al l of the cecum to with in 5 cm nf the ileoceca, valvc was dt::\'ascu lari:ed.A qu an my llf fecal material wa, allowed to rema in in the cccum and the distal cecum wa, lig,iu.:dw1Ch ~2 silk .A local omcntcctomy was rerformed tn prevent locali:arion of° int'ect ion .A 2 cm perforatilln was 1nc1de 1n the cecal lip and fec.11 material wa;, a llowed ro spill into the pe rn111,1l C.l\'i ty.T h e ahdomcn was ckised with a running sutu rc.An imals were ,1lluwed to recover in a mt::tnbol,c cage.Over 1hc ensuing 12 h all animals hcg;1n ro , h ow cli nic.1 1 TABLE 1 evidence of ~Yl:>temic infection as evidcnccd by increasing re:.piratory ratc and devclopmem of fever, lethargy and ,morexia.
PoMoperatively, fluid admin1Mrat inn was guided by the pulmonary capillary wedge pressure.Sufficient 1ntravenou:.c ry:.talloid was infused to 111ainta111 the balloon occluded wedge pressure at presepsis levels.
The pulmonary capi llary wedge pressu re, mean arterial pressure, cemrnl venous pressure, h eart rnte, respiratory rate, thermudilution cardiac output and systemic vascular resistance index wt::re measured prior to surgery a nd repcated daily for th e duration of the experiment.Blood was obtained fro m the artt::rial line for daily measuremems of hlood gases and from the venous line fo r h emoglobin, wh ire blood cell count, album in, creatinine, urea, total bili rubi n , asrartate aminotransferase (AST), and lactate dehydmgenase.Blood was abn drawn for aerobic and anaerobic culture in broth a nd subculture on hlond and chocolate ag,1r.U rin e was coll ected for measurement of creatinine ,md calcubtion nf c reati nine clearancc.
ln six sh eep I ivcr biopsies were per-

D1SCUSS10N
The c1usl' of the hepatic dysfunc-L iLlll that uccurs in ~eptic sta tes is unclear.S tudi es in hum ans are difficult ht:c,1use n f the uim plexn ies of the c l in 1ca l ,it uat ion and the uncenaint y ca u,ed hy the necessary conco mit a nt use nf hlllml t ramf1l'tons and porenual h eparot ox m s such as ant1h1mics ,md parenteral nutrit ion.
X 6700 post mo rte m li ve r histology sh ,1ws c hanges ot venous congest inn , isch em, c necni~is, fo rty c h ange, Kupffe r cel l h ype rplasia and intrahepatic ch o les-tasi~.Occasio nall y c hnle~rasis may be evide nt m the c hola ngiolar level (2 ) .
In t he presen t st ud y th e au thors used a large ani nrnl mode l nf syste mic sepsis whi ch m1m1cs the c ha n ges in ,ept ll: humnns unde rgoing ac ti ve resuscna tio n (2 l ).[1 prnduces a normo tcmive ~ta tc c ha racte ri ::ed hy high card iac mn put, low pe riphe ral resistance a nd the gradua l developme nt of re n al dysfunction.The h epat ic bi och e mier1l change~ con~isted o f an early and plateau ing 111• c rease in A ST and lac tate dchydrogl'• nase with progress ive inc reases in biliruhin.The eml y d rop in albumin 1, no t due to impaired h epatic synthesi,.but rn the r w inc reased vasculc1r endothelial permeahil it y a nd lnss of alhumm fro m the vascular space ( l 0, l l).
While a portio n of the increase 111 biliruhin may represent inc reased bi! ,.
rubin load from h e molys1s this is likely a ~mall co mpone n t as rhe re were 1111 significa nt c h a nges 111 h e moglobi n.It 1, likely tha t the primary explanation for hype rbilirubinc mia i~ d ecreased bile fl ow whic h h a:.heen sh o wn in prev1ou., ex periments to fo llow infusio n ot endo toxin o r bacteria.A t whm level this defect occurs is not c lc,ir, hut expl:n• me ntally e ndom xin hns been shown t,i decrease bile fl ow at the cana licular level poss ibly hy inhibitio n o f l l+K+.
A TPase (22).As th e re was no h ypotensinn or h ypoxemi a a nd n o anc rial h c modyna mic Jiffercnccs be tween the sh eep with or wi thout high AST levels, it is like!} th a t this nec rosis re presents ischem1a 1>ccurring a~ a result of ~inusoidal events, pe rhaps fro m sinus,1ida l congestion a n J collagen d epositinn in the space o f D isse.It is possible rhm h e patocellular n ecrosis cnuld arise from cemrilo bular ven ous congestio n .The sh eep wirh ve ry h igh AST s h ad higher central venous pressures tha n tho~e with norm al AST s; h owever, the diffe re nces were small a nd unli kely, by the mse lves, to he of c linical impm• ta nce.Fina ll y, 'tox ic' even ts are a lso a possibility, particu larly in v iew of the marked mi crovc ic ular fatty c hangt and mitoc ho ndrial degen e rnri on .

TABLE 2
Renal and hepatic parameters at baseline, and 24 and 48 h after induction of sepsis