1429

similar enough to be recognised by the same

polyclonal antibodies, but antibodies recognising onlyone protein have also been produced.9 9The cDNA encoding the larger isofonn of brain

GAD has been cloned from several non-human

species.10-14 The sequences for these genes are highlyconserved and code for a protein of predictedmolecular weight 66-67 K. One group15 has cloned alarger 80 K protein (also enzymatically active whenexpressed as a fusion protein). There is now evidencefor at least two genes for GAD isoformsl6 on different

chromosomes; moreover, the isofonns expressed inislet and brain of both rat and man seem to bethe same,17,18 as shown in two letters in this issue(pp 1468-69).So what is the connection between IDDM and stiff

man syndrome? Stiff man syndrome tends to affectmore females than males, despite its name, and ischaracterised by a progressive symmetrical increase inaxial muscle tone, painfully exacerbated by sensorystimuli such as loud noises or emotion. About 60% of

patients have serum or cerebrospinal fluid antibodiesto GAD,3 and GABA-ergic agonist therapy forms themainstay of treatment. Some 25% of anti-GAD-positive patients have IDDM.3There are many issues still to be settled. Why

should the antibody responses in IDDM and stiff mansyndrome differ in specificity?2 Stiff man syndromemay be associated with IDDM and other organ-specific autoimmune diseases but, unlike IDDM, it israre, affects mainly females, and starts later in life. Arethese characteristics related to the fact that neurons,unlike pancreatic p-cells, are hidden behind the

immunological veil of the blood-brain barrier and arepoor expressors of major histocompatibility genecomplex molecules? Whether IDDM and stiff mansyndrome result from autoimmune reaction to GAD(and, if so, how) or whether the antibodies are merelymarkers of the process is unclear. Thus, the cellularbasis and pathogenetic mechanisms of these diseasesremain to be elucidated; meanwhile, it is possible thatassays for antibodies to GAD may facilitate the

diagnosis of a rare disease or predict a more commondisease that wreaks havoc with many organs.

1. Editorial. The 64 K question in diabetes. Lancet 1990; 336: 597-98.2. Baekkeskov S, Aanstoot H-J, Christgau S, et al. Identification of the 64K

autoantigen in insulin-dependent diabetes as the GABA-synthesizingenzyme glutamic acid decarboxylase. Nature 1990; 347: 151-56.

3. Solimena M, Folli F, Aparasi R, Pozza G, De Camilli P. Autoantibodiesto GABAeric neurons and pancreatic &bgr;-cells in stiff-man syndrome.N Engl J Med 1990; 322: 1555-60.

4. Okada Y, Taniguchi H, Shimada C. High concentration of GABA andhigh glutamate decarboxylase activity in rat pancreatic islets and humaninsulinoma. Science 1976; 194: 620-22.

5. Erdo SL, Joo F, Wolff JR. Immunohistochemical localization of

glutamate decraboxylase in the rat oviduct and ovary: further evidencefor non-neural GABA systems. Cell Tissue Res 1990; 255: 431-34.

6. Persson H, Pelto-Huikko M, Metsis M, et al. Expression of theneurotransmitter synthesizing glutamic acid decarboxylase in malegerm cells. Mol Cell Biol 1990; 10: 4701-11.

7. Erlander MG, Tobin AJ. The structural and functional heterogeneity ofglutamic acid decarboxylase: a review. Neurochem Res 1991; 16:

215-26.

8. Kaufman DL, Houser CR, Tobin AJ. Two forms of the &ggr;-aminobutyricacid synthetic enzyme glutamic acid decarboxylase have distinctintraneuronal distributions and cofactor interactions. J Neurochem1991; 56: 720-23.

9. Chang Y-C, Gottlieb DI. Characterization of the proteins purified withmonoclonal antibodies to glutamic acid decarboxylase. J Neurosci 1988;8: 2123-30.

10. Kobayashi Y, Kaufman DL, Tobin AJ. Glutamic acid decarboxylasecDNA: nucleotide sequence encoding an enzymatically active fusionprotein. J Neurosci 1987; 7: 2766-72.

11. Wyborski RJ, Bond RW, Gottlieb DI. Characterization of a cDNAcoding for rat glutamic acid decarboxylase. Mol Brain Res 1990; 8:193-98.

12. Jackson FR, Newby LM, Kulkarni SJ. Drosophila GABAergic systems:sequence and expression of glutamic acid decarboxylase. J Neurochem1990; 54: 1068-78.

13. Julien JF, Samama P, Mallet J. Rat brain glutamic acid decarboxylasesequence deducted from a cloned cDNA. J Neurochem 1990; 54:703-05.

14. Katarova Z, Szabo G, Mugnaini E, Greenspan RJ. Molecularidentification of the 62 kD form of glutamic acid decarboxylase fromthe mouse. Eur J Neurosci 1990; 2: 190-202.

15. Huang W-M, Reed-Fourquet L, Wu E, Wu J-Y. Molecular cloning andamino acid sequence of brain L-glutamate decarboxylase. Proc NatlAcad Sci 1990; 87: 8491-95.

16. Erlander MG, Tillakaratne NJ, Feldblum S, Patel N, Tobin AJ. Twogenes encode distinct glutamate decarboxylases. Neuron 1990; 7:

91-100.17. Karlsen AE, Hagopian WA, Grubin CE, et al. Cloning and primary

structure of a human islet isoform of glutamic acid decarboxylase fromchromosome 10. Proc Natl Acad Sci USA 1991; 88: 8337-43.

18. Michelsen BK, Petersen JS, Boel E, Moldrup A, Dryberg T,Madsen OD. Cloning, characterization and autoimmune recognition ofrat islet glutamic acid decarboxylase in insulin-dependent diabetesmellitus. Proc Natl Acad Sci USA 1991; 88: 8754-58.

Managing liver trauma conservativelyA liver tear due to blunt or penetrating injury

damages blood vessels (hepatic artery, portal vein,hepatic vein) and bileducts. Surgical dogma andcommonsense suggest a need to operate to preventcontinued haemorrhage or infected bile collections.Thus, a radical policy of early laparotomy andextensive resection became widely adopted. Now,reports on blunt liver trauma from Cambridge1 andSydney2 emphasise a new policy of surgical restraintand non-operative management. The Cambridgepaper highlights perihepatic packing; both researchgroups show that severe liver injuries displayed bycomputed tomographic scanning (CT) can be treatedconservatively.

Early mortality is caused by massive haemorrhageor associated non-hepatic injuries. The most

important late complication is sepsis due to a bile leakor damage to intra-abdominal organs, particularly thecolon. The principles of management are staunchingof haemorrhage, removal of dead or devitalised livertissue, and oversewing or repair of damaged bloodvessels and bileducts. Comparison of results of

published series is difficult because different complexclassifications have been used and the nature of theliver injuries varies: for instance, penetrating injuries,especially stab and low-velocity missile wounds, arecommon in the USA and have a better prognosis thanblunt injuries. The Houston group class liver injuriesas either simple or complex.3 Retrohepatic vena cavaland hepatic vein tears should probably be regarded asan additional subcategory.4

1430

A deliberate non-operative approach for patientswith diagnosed liver injuries was first introduced

successfully and later widely adopted by paediatric. surgeons;5,6 general surgeons have been slow to followin treating adults.’ Patients who are stable on arrivalor who respond rapidly to fluid resuscitation shouldbe considered for conservative management with CT

monitoring, even if they have penetrating wounds.3,4Indications for laparotomy during conservative

management are evidence of continued bleeding orsigns of peritonitis, which suggest associated bowelinjury. Blood in the peritoneal cavity alone, shown onCT scan or by peritoneal lavage, is not a reason tooperate,4 nor is evidence of severe liver injury on theCT scan.2 However, a laparotomy to exclude otherintra-abdominal injuries is usually associated withlittle morbidity and offers reassurance in

management, especially in patients with associatedhead or spinal injuries. 2

Liver packing for complex liver injuries withbleeding has been practised since the turn of thecentury7 but the procedure fell into disrepute after the1939-45 war, probably because of incorrect use-packs were placed into the liver wound, potentiatingbleeding, and left there for longer than 48 hours,resulting in sepsis. The technique was reintroducedby the Cambridge group, who showed that

perihepatic packing to close the liver injury and stopbleeding allowed the patient to be transferred to aspecialised unit for definitive treatment.8 Their latestpaper’ confirms the value of perihepatic packing insevere liver injuries. If the surgeon encounters

massive bleeding on opening the abdomen, temporarypacking will allow time for resuscitation. Immediateperihepatic packing also allows time to deal with othermajor injuries. Since receiving surgeons in mostdistrict general hospitals have very little experiencewith major liver trauma, they need to decide whetherto call for expert help or to close the abdomen andtransfer the patient. Some surgeons now favour theuse of definitive perihepatic packing in severe liverinjuries, including injuries to the vena cava and

hepatic veins. 1,4 The abdomen is closed and the patienttransferred to an intensive care unit for vigorousresuscitation, including correction of coagulationabnormalities. Within 48 hours the patient should bereturned to the operating theatre and the packs gentlyremoved. Often, no further procedure is required.4Packing for more than 48 hours is associated with

sepsis. Dry laparotomy packs with radio-opaquemarkers are preferred; usually about six packs arerequired. They are placed around the liver andbetween the liver and the diaphragm and chest wall toclose the liver wound. Patients who continue to bleedwith packs in situ may have an intrahepatic arterialinjury which can be treated by radiologicalembolisation. Surgical hepatic artery ligation has beenlargely abandoned. If an angiogram does not reveal anarterial source of bleeding the abdomen should be

reopened because the packs have probably beenincorrectly placed into the liver wound.Most simple liver wounds will not be bleeding; they

should be left alone and probably do not even requiredrainage. Inflow (or porta hepatis) occlusion by meansof a non-crushing vascular clamp, first suggested byPringle in 1908/ has been used for elective liverresections to reduce blood loss9,10 but has only latelybeen introduced in the management of complex livertrauma.3 Extensive anatomical liver resections are notwarranted unless the whole of a liver lobe has beendevitalised. Resectional debridement should berestricted to dead or devitalised liver tissue. Damagedblood vessels and bileducts should be ligated. Severebleeding from the depths of a liver wound requireshepatotomy, as suggested by Pachter et all Thefinger fracture technique is used to open the liverlaceration further, after the porta hepatis occlusion,and the bleeding vessels are ligated. This procedurecan be adapted for juxtahepatic vena caval and hepaticvein injuries. 12 The alternative is total hepaticisolation-porta hepatis occlusion is combined withclamping the vena cava above and below the liver; theupper inferior vena cava is approached by a sternalsplit or via the pericardial surface of the diaphragmfrom within the abdomen.4 The aorta should not be

clamped. Hepatic vascular isolation time should bekept as short as possible. The intracaval shunt

technique has been abandoned.Conservative management of hepatic damage and

surgical restraint have improved the survival of

patients with serious liver damage.1-3 We hope that thedays when young surgical trainees carried out theirfirst major liver resections in patients with severeinjuries have gone forever.

1. Watson CJE, Caine RY, Padhani AR, Dixon AK. Surgical restraint in themanagement of liver trauma. Br J Surg 1991; 78: 1071-75.

2. Hollands MJ, Little JM. Non-operative management of blunt liverinjuries. Br J Surg 1991; 78: 968-72.

3. Feliciano DV, Jordan GL, Bitondo CG, Mattox KL, Burch JM, CrusePA. Management of 1000 consecutive cases of hepatic trauma(1979-1984). Ann Surg 1986; 204: 438-45.

4. Terblanche J, Krige JEJ. Injuries to the liver and bile ducts. In:

Williamson RCN, Cooper MJ, eds. Emergency abdominal surgery.London: Churchill Livingstone, 1990: 21-35.

5. Oldham KT, Guice KS, Ryckman F, Kaufman RA, Martin LW,Noseworthy J. Blunt liver injury in childhood: evolution oftherapy andcurrent perspective. Surgery 1986; 100: 542-49.

6. Cywes S, Rode H, Miller AJW. Blunt liver trauma in children:

nonoperative management. J Pediatr Surg 1985; 20: 14-18.7. Pringle JH. Notes on the arrest of hepatic hemorrhage due to trauma. Ann

Surg 1908; 48: 541-49.8. Calne RY, McMaster P, Pentlow BD. The treatment of major liver

trauma by primary packing with transfer of the patient for definitivetreatment. Br J Surg 1979; 66: 338-39.

9. Huguet C, Nordinger B, Black P, Conrad J. Tolerance of the human liverto prolonged normothermic ischemia: a biological study of 20 patientssubmitted to extensive hepatectomy. Arch Surg 1978; 113: 1448-51.

10. Terblanche J, Krige JEJ, Bornman PC. Simplified hepatic resection withthe use of prolonged vascular inflow occlusion. Arch Surg 1991; 126:298-301.

11. Pachter HL, Spencer FC, Hofstetter SR, Coppa GF. Experience with thefinger fracture technique to achieve intra-hepatic hemostasis in 75patients with severe injuries of the liver. Ann Surg 1983; 197: 771-78.

12. Pachter HL, Spencer FC, Hofstetter SR, Liang HC, Coppa GF. Themanagement of juxtahepatic venous injuries without an atriocavalshunt: preliminary clinical observations. Surgery 1986; 99: 569-75.