Dig Dis 1992; 10(suppl 1):84—93
D ivision of Vascular Surgery, D epartm ent of Surgery, Yale U niversity School o f M edicine, New H aven, Conn., USA
Key Words Splenorenal shunt Portal hypertension Hepatic cirrhosis
Distal Splenorenal Shunt Premise, Perspective, Practice
Abstract Despite progress in our understanding and management of patients with portal hypertension, the long-term control of variceal bleeding remains a significant challenge. With further clarification of the underlying pathophysiology and technolog ical advances that have facilitated progress in both diagnosis and treatment, the goal of safe, selective management of patients presenting with variceal hemorrhage is closer to reali zation. While a variety of non-operative therapies have been advocated, shunt surgery remains the most reliable and dura ble method of controlling the portal hypertension and the bleeding. More than 20 years ago, Warren and Zeppa intro duced the concept of selective shunting to prevent recurrent variceal hemorrhage. The distal splenorenal shunt (DSRS) was advocated as an approach that could selectively decom press the esophageal and gastric varices (resulting in effective bleeding control) while maintaining prograde portal flow (pre sumably leading to a lower incidence of post-shunt encepha lopathy and hepatic failure). While the hemodynamic basis for the DSRS remains valid, its selectivity is neither uniform nor durable and this shunt is neither applicable nor effective in all patients bleeding from varices. It remains, however, appro priate and safe therapy in selected cirrhotic patients with vari ceal hemorrhage. With careful pretreatment assessment (in the context of the advances that have occurred in both opera tive and anesthetic management), the DSRS retains an impor tant role in the management of patients with variceal bleed ing.
Richard J. Gusberg. MD Division of Vascular Surgery Department of Surgery . Yale University School of Medicine, 333 Cedar Street New Haven. CT 06510 (USA)
01992 S. Kargcr AG. Basel 0257-2753/92/ 0I07-0084S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Richard J. Gusberg
Almost 50 years ago, Whipple [1 ] and Blakemore [2] performed the first portal-sys temic shunt in a cirrhotic patient bleeding from esophageal varices. Despite five decades of advances in anesthetic and operative man agement as well as in our understanding of the pathophysiology of portal hypertension, the search for a safer and more widely applicable method to control variceal bleeding has per sisted. Despite the advent of a variety of non operative alternatives, shunt decompression remains the most definitive and durable ap proach to this problem. While total portal-systemic shunts are ef fective in lowering portal pressure, the resul tant diversion of portal blood flow has been associated with an increased incidence of en cephalopathy and subsequent hepatic failure in some patients. While some of these shunted patients may respond to this decrease in por tal flow with a compensatory increase in he patic arterial flow, such compensation is not predictable and may not be sustained. Fur thermore. not only do total shunts result in a decrease in the flow of oxygenated blood to the liver, but they also deprive the liver of important hepatotrophic portal substances. In 1967, Warren et al. [3] proposed the dis tal splenorenal shunt (DSRS) as an alternative that would theoretically selectively decom press the varices (and provide long-term bleeding control comparable to the control associated with total shunts) and yet preserve prograde portal flow (presumably lessening the likelihood of post-shunt encephalopathy and hepatic failure). Experience with this shunt has now spanned several decades and continents [3-12], Have its objectives been met? Which patients are most likely to benefit from it? Where does it fit in the context of other operative and non-operative options available to treat patients bleeding from
esophageal varices? Though many questions remain, several conclusions can be drawn.
The Distal Splenorenal Shunt Technique and Bleeding Control The DSRS appears to effectively decom press the varices and provides long-term con trol of bleeding in an average of 85% and in as many as 97% of patients [ 13], There is a risk of both early and late re-bleeding. This post shunt bleeding, presumably associated with inadequate decompression, is more likely to occur early, within the first post-shunt month [13. 14]. These early failures are thought to be secondary to: (a) technical errors leading to shunt thrombosis; (b) systemic venous hyper tension (inferior vena cava or renal vein) pre cluding a safe or significant drop in variceal pressure, even with a patent shunt, or (c) de layed variceal decompression due to pro longed development of collateral channels be tween the high-pressure varices and the lowpressure shunt. While thrombosis has been reported in 3-14% of patients undergoing DSRS, late shunt thrombosis has been docu mented in only about 2% of patients followed by serial angiography [13], Clearly, effective and durable decompression requires thor ough. preoperative anatomic and hemody namic assessment as well as technical preci sion in performance of the shunt. Pre-shunt evaluation should provide the relevant hemo dynamic or manometric details in addition to anatomic information regarding the caliber and patency of the veins being considered for shunting. This is particularly important since, with the widespread use of non-invasive vas cular diagnostic techniques, thrombus in the splanchnic venous circulation has been de tected with increasing frequency and may in fluence the performance and outcome of the shunt surgery [15]. The ‘pre-shunt package’
85
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Introduction
86
Gusberg
plishcd, some sort of post-shunt evaluation should be undertaken to document: (a) shunt patency, and (b) prograde portal flow.
DSRS - Outcomes, Problems The accumulated experience since 1967 has confirmed that the DSRS is feasible and effective. In several prospective randomized trials, it has been shown to be comparably as effective as total shunts in the control of variceal bleeding [13. 16-18] and appears to be associated with a lower incidence of encepha lopathy [ 13. 18], This difference in post-shunt encephalopathy has not. however, been uni formly confirmed, particularly in alcoholics. In fact, there has been a wide variability in the reported incidence of post-DSRS encephalop athy [19. 20], In this regard, the maintenance of both shunt selectivity and prograde portal How appear to be of central importance in minimizing the risk of post-shunt encephalop athy [21]. In series reported in which the sple norenal anastomosis has been done without portal-azygous disconnection, the incidence of encephalopathy has been consistently high [22]. While most series show no substantial survival advantage conferred by the DSRS. two non-randomized trials have suggested that in non-alcoholic cirrhotics, the long-term survival following DSRS is significantly better than that following total shunts [23-25] among alcoholic patients, the shunt-related survival differences appear insignificant. It appears, therefore, that the DSRS is both a feasible and effective alternative to the total shunt in patients documented to be bleeding from varices. The late re-bleeding rate is low, the incidence of encephalopathy is low. and the long-term survival (particularly among non-alcoholic cirrhotics) may be better. De spite this positive experience, several ques tions remain: Is the hemodynamic basis for
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
has generally included: (a) an inferior vena cava (I VC) gram with measurement of he patic vein (wedged and free) and IVC (supraand infrahcpatic) pressures, and (b) venous phase visceral angiography to assess the splen ic, superior mesenteric, and portal veins. Newer non-invasive modalities. Doppler ul trasound and magnetic resonance imaging may soon supplant some or all of these pre viously routine angiographic studies. In the technical performance of the DSRS. several points merit emphasis: (1) adequate exposure via a generous midline or left-sided abdominal incision; (2) central division of the splenic vein with adequate distal mobiliza tion to facilitate an unkinked, tension-free splenorenal anastomosis; care should be taken to fully mobilize the splenic vein from its pan creatic bed. freeing a length from the portal vein junction to the splenic hilum with partic ular attention taken during the circumferen tial mobilization to securely and gently con trol the multiple, high-pressure pancreaticosplenic tributaries: (3) the anastomosis itself should be accomplished using atraumatic vas cular clamps, an elliptical venotomy in the cephalo-anterior surface of the left renal vein, and fine vascular sutures using either the run ning or the interrupted technique depending upon the splenic vein size; (4) a compulsive attempt at portal-azvgous disconnection which should include: (a) division of splenic vein tributaries to as close to the splenic hilum as is technically feasible: (b) identifica tion and secure ligation of the coronary vein (recognizing its variable anatomy and its sometimes dual nature with drainage into the portal and/or splenic vein; (c) ligation of the inferior mesenteric vein; (d) ligation of the right gastric vein just superior to the pylorus, and (e) ligation of potential portal-systemic collateral connections from the pylorus to the short gastric veins. In an effort to demonstrate that these technical goals have been accom-
bed at the time of DSRS might ameliorate the development of this potential siphon between the portal and splenic systems and result in more durable preservation of portal flow (and perhaps improved survival) in the alcoholic patients [11, 27]. These observations have been corroborated recently in a study of a large group of cirrhotic patients undergoing DSRS. with and without SPD. In long-term follow-up. post-shunt encephalopathy was more likely to occur in those patients without SPD who had evidence of a loss of prograde portal flow [28]. SPD does, however, increase the technical complexity of an already de manding operation. Support for its wide spread application must await prospective, controlled confirmation of its effectiveness. Clearly, however, if the hemodynamic prem DSRS - Hemodynamic Basis and ise of the DSRS is to be clinically valid, the Concequences variceal decompression must be effective and While effective variceal decompression the portal flow maintained. Assuming that the DSRS has been done and maintenance of prograde portal flow can appropriately and completely, several post be anticipated early on following a DSRS with portal-azygous disconnection, the degree of operative factors might adversely influence prograde portal flow apears to decline with either the decompression or the perfusion. time [21], As post-shunt portal-systemic col Anastomotic stenosis may occur, leading to laterals develop between the high-pressure recurrent hemorrhage secondary to variceal portal-mesenteric compartment and the low- hypertension. Shunt stenosis can be docu pressure. shunt-decompressed esophago-gas- mented by percutaneous, transcaval catheter tric-splenic portion of the circulation an in ization. Balloon dilatation can be effective creasing percentage of portal flow is diverted (and perhaps durable) and is considered the away from the liver via portal-systemic collat treatment of choice [29]. The incidence of late erals. The rapidity and extent of this loss of post-shunt portal vein thrombosis is low (less portal flow appears to be related to both the than 5%) but non-occlusive thrombosis in the nature and progression of the liver disease as portal vein has been reported in up to 20% of well as the completeness of the portal-azygous patients [30]. Thrombosis in this setting could or spleno-pancreatic disconnection (SPD) be related to hypercoagulability or portal ve achieved at the time of the shunt. While as nous stasis associated with either progressive many as 90% of non-alcoholic cirrhotics cirrhosis (and increased outflow resistance) or maintain significant prograde portal flow, progressive diversion of flow via high-resis only 25-50% of alcoholic patients have well- tance portosplenic collaterals (and decreased maintained portal flow following a DSRS inflow). There is no reliable, reported experi [26], There is some evidence that full mobili ence with either thrombectomy or thrombo zation of the splenic vein out of the pancreatic lytic therapy in this setting.
87
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
the shunt valid? Does it provide predictable selective variceal decompression while main taining prograde portal How? And does it do this in a durable fashion? What are its limita tions or contraindications in the management of bleeding varices? Does it have a rational role in the acute, emergent management of variceal hemorrhage? Can the long-term out come be enhanced either by technical altera tions in performance of the shunt or better patient selection? What is the proper role for DSRS in the context of other shunt and non shunt options available to treat variceal bleed ing? And how do these treatments impact on the subsequent performance of the others?
When in the course of cirrhosis and portal hypertension should a shunt be undertaken? Shunts are indicated only after an endoscopically documented episode of variceal hemor rhage. Since the majority (approximately twothirds) of patients with varices never bleed from them [31], prophylactic shunts (under taken prior to any documented bleeding) have been considered contraindicated. In a random ized. controlled trial comparing total shunts to expectant medical therapy in cirrhotics who had never bled from varices, the outcome (in cluding survival) was worse in the shunted patients [32], Though prophylactic surgery has had no supportable role in the management of portal hypertension, it is unclear whether the experience with prophylactic total shunts can be generalized to include DSRS. In recent reports from China in patients with non-alco holic cirrhosis. 39% of the shunts performed were done prophylacticallv [33] and the subse quent incidence of both bleeding and encepha lopathy was low with a long-term survival that was significantly belter than that associated with medical therapy [ 12]. Unless or until bet ter selection and prediction criteria can be developed, this prophylactic approach cannot be recommended widely. Most of the reported experience with the DSRS has been in the elective setting - in patients who are stable hemodynamically and not bleeding. In patients requiring an emer gency shunt for active hemorrhage, total shunts have been thought to provide more direct, immediate, and predictable decom pression. Because of the somewhat circuitous decompression route from the high-pressure varices to the low-pressure splenorenal anas tomosis, the post-DSRS time to safe variceal decompression has been thought to be neither immediate nor predictable. Until more reli
88
Gusherg
able methods are developed to safely and repeatedly measure variceal pressure (and clearly define the natural history of postDSRS decompression), the emergent applica tion of the DSRS should be undertaken onlyin a cautious and highly selective fashion. Are there other settings in which the DSRS is relatively or absolutely contraindicated? Massive splenomegaly with hypcrspicnism. pancreatitis with pseudocyst, portal vein thrombosis, previous splenectomy, and chil dren with variceal hemorrhage have all been thought to pose particular problems. Particularly if approached through a mid line abdominal incision, the technical perfor mance of a DSRS should not be adversely affected by the size of the spleen. Further more. hypersplenism - commonly encoun tered in cirrhotic, portal-hypertensive pa tients - is rarely of inherent clinical signifi cance. In this setting, splenectomy is virtually never indicated for the hypersplenism alone, and the hypersplenism can be expected to improve following portal and/or splenic de compression [34], A history of pancreatitis and its complica tions does not represent an absolute contrain dication to DSRS but should raise the index of suspicion for splenic vein thrombosis (SVT) (precluding a DSRS). or particular and perhaps prohibitive technical difficulty asso ciated with peripancreatic fibrosis or inflam mation and splenic vein phlebitis. Under such circumstances, clear preoperative visualiza tion of the splenic vein (by either venousphase angiography. magnetic resonance imag ing or Doppler ultrasound) is mandatory. Furthermore, endoscopic sclerotherapy has been associated with SVT [15]. In patients being evaluated for DSRS following sclero therapy failure, particular attention should be directed at ruling out SVT. The role for thrombectomy or thrombolytic therapy in this setting has not been defined.
The Distal Splenorenal Shun!
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
DSRS - When Is It Indicated and When Is It Not?
DSRS - and the Alternatives Since Whipple [1] and Blakemore [2] re ported their initial experience with portacaval shunting in patients with portal hypertension and variceal bleeding, a variety of operative and non-operative approaches have become available. With advances in anesthesia, peri operative monitoring, and operative tech niques. surgery has become more broadly ap plicable and safer. With the further advances in immunobiology, even transplantation has become a viable option in some patients with variceal hemorrhage associated with endstage liver disease. The réintroduction and now wide experience with sclerotherapy as
well as the emerging role of (3-blockade in the treatment of some cirrhotics with variceal hemorrhage have broadened both the treat ment options and the spectrum of patients who can be treated. Since its introduction more than two de cades ago, the DSRS has been widely applied throughout the world. Significant experience has accumulated in North America. South America, Europe, Africa, and Asia. As we approach the next century, almost 50 years since the first reported decompressing shunt for variceal bleeding, this collective experi ence provides a useful perspective of the role of the DSRS in the management of the patient with portal hypertension and variceal bleed ing. While its hemodynamic premise has been largely validated, the DSRS provides variceal decompression that is effective but not as selective as originally hoped. Prograde portal flow is maintained initially, but may be lost with time. Operative alterations and patient selection may lead to more prolonged preser vation of portal flow (and survival). The DSRS provides effective and long-term bleed ing control, a low incidence of post-shunt encephalopathy, but no uniform or predicta ble survival advantage. In non-alcoholic pa tients. however, this selective shunt has been associated with improved survival. The DSRS has shown to be feasible in a wide variety of settings (elective and emer gent. in adults and children, in cirrhosis and extrahepatic portal hypertension). With the advances in operative technique, anesthesia, monitoring, and post-operative management, shunt surgery in general (and the DSRS in particular) can be recommended with an in creasing confidence in its anticipated safety. In a variety of series, the operative mortaliy, re-bleeding rate, and incidence of encephalop athy have been acceptably low (see table 1). In view of this encouraging and accumulat ing experience, what is its proper role in the
89
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Portal vein thrombosis is a recognized extrahepatic cause of portal hypertension and variceal bleeding. In the management of bleeding in this setting, DSRS has been shown to be safe and effective [35]. Is DSRS feasible in patients who have pre viously undergone splenectomy? If the splenic vein is patent and the variccal-short gastric vein-splenic vein pathway remains intact, variceal decompression by DSRS is feasible and should be effective [36], Variceal bleeding in children poses an un usual but special challenge. While many dif ferent shunt types have been advocated, large central total shunts have generally been con sidered preferable, presumably because the larger, higher flow shunt should be associated with better patency. Maksoud and Mies [37], in 1982. reported on a group of 21 children, ages 4-12, with variceal bleeding treated by DSRS; the long-term patency rate and bleed ing control were excellent, and no patients developed encephalopathy. While shunt deci sions in children need to be individualized, at least the DSRS appears to be feasible in this setting.
Table 1. Distal spleno-rcnal shunt results - selected series Series
Follow-up years
Patients
Mean years
O perative m ortality
Re bleeding
PSE
D eaths
Fischer. 19 8 1 [5] Millikan, 1985 [13] Langer. 1985 [18] Rikkers, 1987 [4] Grace. 1988 [ 16] Gusberg, 1992 (this study)
5 11 5.5 2 3.5 3.5
23 26 38 23 43 19
48 49 49 NA 54 68
1 (5) 3(12) 5(13) 1 (5) 4 (9 ) 0 (0 )
1 (4) 2 (8 ) 1 (3) 5(19) 7 (18) 0 (0 )
1 (4) 7(27) 8 (24) 4 (1 6 ) 20(51) 4 (21)
4 (1 5 ) 15(58) 20(56) 10(45) 20 (4 6 ) 5 (26)
context of other available operative and non operative options? The DSRS provides bleeding control that is equivalent to that achieved with total shunts and is associated with a lower inci dence of encephalopathy and. at least in non alcoholics, a perhaps prolonged survival. Its effectiveness in providing immediate decom pression in the setting of acute hemorrhage remains to be proven. Whether the smalldiameter portacaval H graft, introduced by Sarfeh et al. [38]. can comparably achieve the dual objective of variceal decompression and preservation of portal flow (and do so in a durable fashion) requires a broader and lon ger experience. The operation is technically less demanding, may provide more imme diate variceal decompression, and the early data on its patency and hemodynamic effects are encouraging. Liver transplantation is the most definitive and curative therapeutic ap proach to these bleeding, portal hypertensive cirrhotics. While it must be applied with great caution and selectivity, it is clearly the only approach that is non-palliative and that can cure both the cirrhosis and the portal hyper tension. In selected end-stage cirrhotic pa tients bleeding from varices, it is the treat ment of choice. Furthermore, the prospect of future candidacy for liver transplantation
90
Gusberg
should influence the choice of therapy in pa tients with still well-preserved liver function who are bleeding from varices. The DSRS, requiring no dissection of the portal struc tures, should have no adverse technical or hemodynamic impact on the subsequent per formance of a transplant and may be the best pre-transplant elective shunt option in these patients [39], While P-blockade has been associated with few short- or long-term risks, its long-term effectiveness in controlling variceal hemor rhage is uncertain and unpredictable. Lebrec et al. [40. 41] have demonstrated that oral propranolol can produce a decrease in both portal pressure and the rate of recurrent vari ceal bleeding in some patients. Neither Bur roughs et al. [42] nor Villeneuve et al. [43] were able to confirm the effectiveness of pro pranolol as compared to placebo in control ling re-bleeding in unselected advanced cir rhotics with documented variceal hemor rhage. In recommending pharmacologic ther apy for variceal bleeding, the challenge re mains one of selection: many patients will re bleed despite propranolol treatment: as many as 30% of patients will have no propranololinduced drop in portal pressure [44], and the compensatory response to hemorrhage resus citation has been inhibited in some P-blocked
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Figures shown in parentheses are percentages.
Table 2 M anagem ent of Variceal Hemorrhage
Acute sclerotherapy
Vasopressin - NTG
Child's A non-alcoholic cirrhosis
Child’s B and/or unstable and/or persistent alcoholic
Selective shunt
Sclerotherapy or selective shunt or ß-blockade
Child's C
I
[49, 50], Multiple sessions are required for complete obliteration, gastric varices or por tal hypertensive gastropathy cannot be effec tively treated by sclerosis, and about half of the patients treated bleed again. Furthermore, sclerotherapy is not without complications and some of the complications (chronic ul cers, extravariceal splanchnic venous throm bosis) might limit future treatment options [15]. While sclerotherapy appears to be an appropriate and widely applicable option in the acute management of patients bleeding from esophageal varices - with initial control in as many as 80-90% of patients [51 ]. its role as definitive therapy in the long-term control of variceal bleeding remains uncertain. It would not appear to be appropriate therapy aimed at providing chronic bleeding control in all portal-hypertensive cirrhotic patients.
91
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
patients presenting with variceal hemorrhage [43]. At the present time, problems with com pliance. tolerance, and effectiveness limit the uniform applicability of p-blockade to pre vent recurrent variceal bleeding in cirrhotic patients. While it can be effective and useful, it cannot be considered a better alternative to DSRS in all such patients. While sclerotherapy was first reported in the treatment of esophageal varices, more than 50 years ago [44], it has been used widely in the primary treatment of variceal hemor rhage only in the past decade. Sclerotherapy has been advocated in the emergency, initial, and chronic management of variceal hemor rhage. While several controlled trials have documented a decrease in re-bleeding rates in patients undergoing sclerotherapy [46-48]. the control is neither certain nor durable and no survival advantage has been documented
Transplant or sclerotherapy or ß-blockade
Conclusion The safe and appropriate management of variceal hemorrhage is a persistent challenge. Biological and technological advances over the past five decades have increased both the variety and safety of the interventions avail able for cirrhotic patients with variceal hem orrhage. While many unanswered questions remain, current practice is closer to the prom ise of matching a particular therapy to a par ticular patient (table 2). Shunt surgery re mains an effective and durable approach to the prevention of recurrent bleeding. Ad vances in anesthesia, monitoring, and opera tive technique have made such surgery safer and more widely applicable. While neither the clinical nor the hemodynamic consequence of the DSRS are all that its developers might have hoped, it is a feasible and effective treat-
nient in many cirrhotics with variceal hemor rhage. In good-risk, stable, non-alcoholic pa tients, it is probably the treatment of choice. In end-stage non-alcoholic patients, evalua tion for transplantation appears warranted. In the remaining patients, the choice among shunt surgery, p-blockade. sclerotherapy or transplantation should be dictated by the pa tient’s overall status, the nature of the varices and site of bleeding, the type and severity of the liver disease, the activity or chronicity of the bleeding, and the patient’s ability to ac cept and tolerate the various alternatives. Ul timately, the optimal choice will be deter mined by a better understanding of the hemo dynamic changes associated with portal hy pertension and its treatment as well as the development of more reliable predictors of the treatment outcomes.
1 Whipple AO: The problem of portal hypertension in relation to the hepato-splenopathies. Ann Surg 1945: 122:449-475. 2 Blakemore AH: Portacaval shunt for portal hypertension. Follow-up results in cases or cirrhosis of the liv er. JAMA 1951:145:1335-1339. 3 Warren WD. Zeppa R. Forman JS: Selective transplenic decompression of gastroesophageal varices by distal splenorenal shunt. Ann Surg 1967; 166:437. 4 Rikkers LF. Rudman D. Galantbos JT. et al: A randomized controlled trial of the distal splenorenal shunt. Ann Surg 1978;188:271-282. 5 Fischer JE. Bower Rl l. Atamian S. et al: Comparison of distal and prox imal splenorenal shunts. A random ized prospective trial. Ann Surg 1981:194:531-542.
92
6 Rikkers LF. Soper NJ, Cormier RA: Selective operative approach for variceal hemorrhage. Am J Surg 1984:147:89-90. 7 Orozco H. Juarez F. Gantillan P. et al: Ten years of selective shunts for hemorrhagic portal hypertension. Surgery 1988:103:27-31. 8 Salam A. Ezzat F. Abu-Elmag I: Se lective shunt in schistosomiasis in Egypt. Am J Surg 1990:160:90-92. 9 Pezzuoli G, Spina GD, Santambrogio R. et al: The distal spleno renal shunt: An update on experi ence with 106 cases. Int Surg 1987; 92:144-148. 10 Myburgh JA: Selective shunts: The Johannesburg experience. Am J Surg 1990:160:67-74. 11 Inokuchi K. Beppu K. Kayanagi N, et al: Exclusion of non-isolated splenic vein in distal splenorenal shunt for prevention of portal malcirculation. Ann Surg 1984:200: 711-717.
12 Jin. G: Current status of the distal splenorenal shunt in China. Am J Surg 1990:160:93-97. 13 Millikan WJ. Warren WD. Hender son JM. et al: The Emory prospec tive randomized trial: Selective ver sus non-selective shunt to control varciceal bleeding. Ten year follow ing-up. Ann Surg 1985:201:712— 722. 14 Eckhauser FF.. Pomerante RA. Knol JA. et al: Early variceal re-bleeding shunt. Arch Surg 1986:121:547552. 15 Leach SD. Meyer GH. Gusberg RJ: Endoscopic sclerotherapy: A risk factor for splanchnic venous throm bosis. J Vase Surg 1989:10:9-13. 16 Grace ND. Conn HO. Resnick Rll. et al: Distal splenorenal vs portosys temic shunts after hemorrhage from varices: A randomized controlled trial. Hepatology 1988:8:1475— 1481.
Gusberg
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
References
29 Henderson JM. Khishen MA. Milli kan WJ. el al: Management of steno sis of distal splenorenal shunt by bal loon dilation. Surg Gynecol Obstet 1983:157:43-48. 30 Henderson JM. Millikan W J.Chipponi J. ct al: The incidence and nat ural history of thrombus in the por tal vein following distal splenorenal shunt. Ann Surg 1982:196:1-7. 3 1 Burroughs AK. Heygere F. McIntyre N: Pitfalls in studies of prophylactic therapy for variceal bleeding in cir rhotics. Hepatology 1986:6:1407— 1413. 32 Conn HO. Lindcnmuth WW. May CT, et al: Prophylactic portacaval anastomosis. A tale of two studies. Medicine (Baltimore) 1972:51:2740. 33 Yantin H: Operations for portal hy pertension. Clin Arch Surg 1985: 120:1197-1199. 34 Hutson DG. Zeppa R. Levi JU: The effect of distal splenorenal shunt on hvpersplenism. Ann Surg 1977; 185: 605-682. 35 Warren W'D. Henderson JM. Milli kan WJ. et al: Management of vari ceal bleeding in patient with non cirrhotic portal vein thrombosis. Ann Surg 1988:207:623-634. 36 Warren WD. Millikan WJ, Hender son JM. et al: Selective variceal de compression after splenectomy or splenic vein thrombosis. Ann Surg 1984:199:694-702. 37 Maksoud JG. Mies S: Distal sple norenal shunt (DSRS) in children. Ann Surg 1982:195:401-405. 38 Sarfeh 1.1. Rypins EB. Mason GR: A systematic appraisal of portacaval 11-graft diameters: Clinical and he modynamic perspectives. Ann Surg 1986:204:356. 39 Wood RP. Shaw B\V. Rikkers L: Liver transplantation for variceal hemorrhage. Surg Clin North Am 1990:70:449-461. 40 Lebrcc D. Nouel O. Berneau J, et al: Propranolol in the prevention of re current variceal hemorrhage in cir rhotic patients. Lancet 1981 :ii: 180— 182.
4 1 Lebrec D. Nouel O. Berneau J, et al: Propranolol in the prevention of re current variceal hemorrhage in cir rhotic patients. Lancet 1981 :i:920— 921. 42 Burroughs AK. Jenkins W'J. Sher lock S, et al: Controlled trial of pro pranolol for the prevention of recur rent variceal hemorrhage in patients with cirrhosis. N Engl J Med 1983: 309:1539-1542. 43 Viileneuve JP. Pauier-Layrargues G. Infante-Richard C. et al: Short term effects of propranolol on portal venous pressure. Hepatology 1986: 6:101-106. 44 Garcia-Tsao G, Grace ND. Groszmann RJ. ct al: Short-term effects of propranolol on portal venous pres sure. Hepatology 1986:6:10 1—106. 45 Crafoord C. Frcukner P: New surgi cal treatment of varicose veins of the esophagus. Acta Otolaryngol 1939: 27:422-427. 46 The Copenhagen esophageal varices and sclerotherapy project: Sclero therapy after first variceal hemor rhage in cirrhosis: A randomized multi-center trial. N Engl J Med 1984:311:1594-1597. 47 Korula J. Balart LA, Radran G .et al: A prospective, randomized trial of chronic esophageal variceal sclero therapy. Hepatology 1985:5:584— 589. 48 Wcstaby D. Hayes PC. Grimson AES. et al: Controlled clinical trial o f injection sclerotherapy for active variceal bleeding. Hepatology 1989; 9:274-277. 49 Terblanche J. Bornman PC. Kahn D. et al: Failure of repeated injec tion sclerotherapy to improve long term survival after esophageal vari ceal bleeding. A five-year prospec tive controlled clinical trial. Lancet 1982:ii: 1328—1331. 50 Wcstaby D. Williams R: Status of sclerotherapy for variceal bleeding in 1990. Am J Surg 1990; 160:32— 36. 51 Terblanche J, Yakoob HI. Bornman PC, el al: Acute bleeding varices. A five-year prospective evaluation of tamponade and sclerotherapy. Ann Surg 1981:194:521-524.
93
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
17 Harley HAJ. Morgan T. Redeker AG. et al: Results of a randomized trial of end-lo-side portacaval shunt and distal splenorenal shunt in alco holic liver disease and variceal bleeding. Gastroenterology 1986:91: 802-809. 18 I.angcr B. Taylor BR. Mackenzie DR. et al: Further report of a pro spective randomized trial compar ing distal splenorenal shunt with end-to-side portacaval shunt. Gas troenterology 1985:88:424-429. 19 Rikkers LF: Is the distal splenorenal shunt better? Hepatology 1988:8: 1705-1707. 20 Henderson JM: Variceal bleeding: Which shunt? Gastroenterology 1986:91:1021-1023. 21 Maillard JN. Flamant YM. HavJM. et al: Selectivity of the distal sple norenal shunt. Surgery 1979:86: 663-671. 22 Vang J . Simcrt G. Hansson J A. el al: Results of a modified distal spleno renal shunt for portal hypertension. Ann Surg 1977:185:224-228. 23 W'arren WD. Millikan WJ. Hender son JM. et al: Ten years of portal hypertensive surgery at Emory: Re sults and new perspectives. Ann Surg 1982:195:530-542. 24 Zeppa R. Hensley GT. Levi JU. et al: The comparative survival of alco holics versus nonalcohol ics after d istal splenorenal shunt. Ann Surg 1978:187:510. 25 Zeppa R. Lee PA. Hutson DG. et al: Portal hypertension. A fifteen-vear perspective. Am J Surg 1988:155: 6-9. 26 Henderson JM. Millikan WJ: Wright-Bacon L. et al: Hemody namic differences between alcoholic and non-alcoholic cirrhotics follow ing distal splenorenal shunt: Effect on survival? Ann Surg 1982:198: 325-334. 27 Henderson JM. Warren WD. Milli kan WJ, et al: Distal splenorenal shunt with splenopancreatic discon nection: A four year assessment. Ann Surg 1989:210:332-341. 28 MalTei-Faccioli A. Gerunda GE, Neri D. et al: Selective variceal de compression and its role relative to other therapies. Am J Surg 1990: 160:60-66.
D ivision of Vascular Surgery, D epartm ent of Surgery, Yale U niversity School o f M edicine, New H aven, Conn., USA
Key Words Splenorenal shunt Portal hypertension Hepatic cirrhosis
Distal Splenorenal Shunt Premise, Perspective, Practice
Abstract Despite progress in our understanding and management of patients with portal hypertension, the long-term control of variceal bleeding remains a significant challenge. With further clarification of the underlying pathophysiology and technolog ical advances that have facilitated progress in both diagnosis and treatment, the goal of safe, selective management of patients presenting with variceal hemorrhage is closer to reali zation. While a variety of non-operative therapies have been advocated, shunt surgery remains the most reliable and dura ble method of controlling the portal hypertension and the bleeding. More than 20 years ago, Warren and Zeppa intro duced the concept of selective shunting to prevent recurrent variceal hemorrhage. The distal splenorenal shunt (DSRS) was advocated as an approach that could selectively decom press the esophageal and gastric varices (resulting in effective bleeding control) while maintaining prograde portal flow (pre sumably leading to a lower incidence of post-shunt encepha lopathy and hepatic failure). While the hemodynamic basis for the DSRS remains valid, its selectivity is neither uniform nor durable and this shunt is neither applicable nor effective in all patients bleeding from varices. It remains, however, appro priate and safe therapy in selected cirrhotic patients with vari ceal hemorrhage. With careful pretreatment assessment (in the context of the advances that have occurred in both opera tive and anesthetic management), the DSRS retains an impor tant role in the management of patients with variceal bleed ing.
Richard J. Gusberg. MD Division of Vascular Surgery Department of Surgery . Yale University School of Medicine, 333 Cedar Street New Haven. CT 06510 (USA)
01992 S. Kargcr AG. Basel 0257-2753/92/ 0I07-0084S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Richard J. Gusberg
Almost 50 years ago, Whipple [1 ] and Blakemore [2] performed the first portal-sys temic shunt in a cirrhotic patient bleeding from esophageal varices. Despite five decades of advances in anesthetic and operative man agement as well as in our understanding of the pathophysiology of portal hypertension, the search for a safer and more widely applicable method to control variceal bleeding has per sisted. Despite the advent of a variety of non operative alternatives, shunt decompression remains the most definitive and durable ap proach to this problem. While total portal-systemic shunts are ef fective in lowering portal pressure, the resul tant diversion of portal blood flow has been associated with an increased incidence of en cephalopathy and subsequent hepatic failure in some patients. While some of these shunted patients may respond to this decrease in por tal flow with a compensatory increase in he patic arterial flow, such compensation is not predictable and may not be sustained. Fur thermore. not only do total shunts result in a decrease in the flow of oxygenated blood to the liver, but they also deprive the liver of important hepatotrophic portal substances. In 1967, Warren et al. [3] proposed the dis tal splenorenal shunt (DSRS) as an alternative that would theoretically selectively decom press the varices (and provide long-term bleeding control comparable to the control associated with total shunts) and yet preserve prograde portal flow (presumably lessening the likelihood of post-shunt encephalopathy and hepatic failure). Experience with this shunt has now spanned several decades and continents [3-12], Have its objectives been met? Which patients are most likely to benefit from it? Where does it fit in the context of other operative and non-operative options available to treat patients bleeding from
esophageal varices? Though many questions remain, several conclusions can be drawn.
The Distal Splenorenal Shunt Technique and Bleeding Control The DSRS appears to effectively decom press the varices and provides long-term con trol of bleeding in an average of 85% and in as many as 97% of patients [ 13], There is a risk of both early and late re-bleeding. This post shunt bleeding, presumably associated with inadequate decompression, is more likely to occur early, within the first post-shunt month [13. 14]. These early failures are thought to be secondary to: (a) technical errors leading to shunt thrombosis; (b) systemic venous hyper tension (inferior vena cava or renal vein) pre cluding a safe or significant drop in variceal pressure, even with a patent shunt, or (c) de layed variceal decompression due to pro longed development of collateral channels be tween the high-pressure varices and the lowpressure shunt. While thrombosis has been reported in 3-14% of patients undergoing DSRS, late shunt thrombosis has been docu mented in only about 2% of patients followed by serial angiography [13], Clearly, effective and durable decompression requires thor ough. preoperative anatomic and hemody namic assessment as well as technical preci sion in performance of the shunt. Pre-shunt evaluation should provide the relevant hemo dynamic or manometric details in addition to anatomic information regarding the caliber and patency of the veins being considered for shunting. This is particularly important since, with the widespread use of non-invasive vas cular diagnostic techniques, thrombus in the splanchnic venous circulation has been de tected with increasing frequency and may in fluence the performance and outcome of the shunt surgery [15]. The ‘pre-shunt package’
85
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Introduction
86
Gusberg
plishcd, some sort of post-shunt evaluation should be undertaken to document: (a) shunt patency, and (b) prograde portal flow.
DSRS - Outcomes, Problems The accumulated experience since 1967 has confirmed that the DSRS is feasible and effective. In several prospective randomized trials, it has been shown to be comparably as effective as total shunts in the control of variceal bleeding [13. 16-18] and appears to be associated with a lower incidence of encepha lopathy [ 13. 18], This difference in post-shunt encephalopathy has not. however, been uni formly confirmed, particularly in alcoholics. In fact, there has been a wide variability in the reported incidence of post-DSRS encephalop athy [19. 20], In this regard, the maintenance of both shunt selectivity and prograde portal How appear to be of central importance in minimizing the risk of post-shunt encephalop athy [21]. In series reported in which the sple norenal anastomosis has been done without portal-azygous disconnection, the incidence of encephalopathy has been consistently high [22]. While most series show no substantial survival advantage conferred by the DSRS. two non-randomized trials have suggested that in non-alcoholic cirrhotics, the long-term survival following DSRS is significantly better than that following total shunts [23-25] among alcoholic patients, the shunt-related survival differences appear insignificant. It appears, therefore, that the DSRS is both a feasible and effective alternative to the total shunt in patients documented to be bleeding from varices. The late re-bleeding rate is low, the incidence of encephalopathy is low. and the long-term survival (particularly among non-alcoholic cirrhotics) may be better. De spite this positive experience, several ques tions remain: Is the hemodynamic basis for
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
has generally included: (a) an inferior vena cava (I VC) gram with measurement of he patic vein (wedged and free) and IVC (supraand infrahcpatic) pressures, and (b) venous phase visceral angiography to assess the splen ic, superior mesenteric, and portal veins. Newer non-invasive modalities. Doppler ul trasound and magnetic resonance imaging may soon supplant some or all of these pre viously routine angiographic studies. In the technical performance of the DSRS. several points merit emphasis: (1) adequate exposure via a generous midline or left-sided abdominal incision; (2) central division of the splenic vein with adequate distal mobiliza tion to facilitate an unkinked, tension-free splenorenal anastomosis; care should be taken to fully mobilize the splenic vein from its pan creatic bed. freeing a length from the portal vein junction to the splenic hilum with partic ular attention taken during the circumferen tial mobilization to securely and gently con trol the multiple, high-pressure pancreaticosplenic tributaries: (3) the anastomosis itself should be accomplished using atraumatic vas cular clamps, an elliptical venotomy in the cephalo-anterior surface of the left renal vein, and fine vascular sutures using either the run ning or the interrupted technique depending upon the splenic vein size; (4) a compulsive attempt at portal-azvgous disconnection which should include: (a) division of splenic vein tributaries to as close to the splenic hilum as is technically feasible: (b) identifica tion and secure ligation of the coronary vein (recognizing its variable anatomy and its sometimes dual nature with drainage into the portal and/or splenic vein; (c) ligation of the inferior mesenteric vein; (d) ligation of the right gastric vein just superior to the pylorus, and (e) ligation of potential portal-systemic collateral connections from the pylorus to the short gastric veins. In an effort to demonstrate that these technical goals have been accom-
bed at the time of DSRS might ameliorate the development of this potential siphon between the portal and splenic systems and result in more durable preservation of portal flow (and perhaps improved survival) in the alcoholic patients [11, 27]. These observations have been corroborated recently in a study of a large group of cirrhotic patients undergoing DSRS. with and without SPD. In long-term follow-up. post-shunt encephalopathy was more likely to occur in those patients without SPD who had evidence of a loss of prograde portal flow [28]. SPD does, however, increase the technical complexity of an already de manding operation. Support for its wide spread application must await prospective, controlled confirmation of its effectiveness. Clearly, however, if the hemodynamic prem DSRS - Hemodynamic Basis and ise of the DSRS is to be clinically valid, the Concequences variceal decompression must be effective and While effective variceal decompression the portal flow maintained. Assuming that the DSRS has been done and maintenance of prograde portal flow can appropriately and completely, several post be anticipated early on following a DSRS with portal-azygous disconnection, the degree of operative factors might adversely influence prograde portal flow apears to decline with either the decompression or the perfusion. time [21], As post-shunt portal-systemic col Anastomotic stenosis may occur, leading to laterals develop between the high-pressure recurrent hemorrhage secondary to variceal portal-mesenteric compartment and the low- hypertension. Shunt stenosis can be docu pressure. shunt-decompressed esophago-gas- mented by percutaneous, transcaval catheter tric-splenic portion of the circulation an in ization. Balloon dilatation can be effective creasing percentage of portal flow is diverted (and perhaps durable) and is considered the away from the liver via portal-systemic collat treatment of choice [29]. The incidence of late erals. The rapidity and extent of this loss of post-shunt portal vein thrombosis is low (less portal flow appears to be related to both the than 5%) but non-occlusive thrombosis in the nature and progression of the liver disease as portal vein has been reported in up to 20% of well as the completeness of the portal-azygous patients [30]. Thrombosis in this setting could or spleno-pancreatic disconnection (SPD) be related to hypercoagulability or portal ve achieved at the time of the shunt. While as nous stasis associated with either progressive many as 90% of non-alcoholic cirrhotics cirrhosis (and increased outflow resistance) or maintain significant prograde portal flow, progressive diversion of flow via high-resis only 25-50% of alcoholic patients have well- tance portosplenic collaterals (and decreased maintained portal flow following a DSRS inflow). There is no reliable, reported experi [26], There is some evidence that full mobili ence with either thrombectomy or thrombo zation of the splenic vein out of the pancreatic lytic therapy in this setting.
87
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
the shunt valid? Does it provide predictable selective variceal decompression while main taining prograde portal How? And does it do this in a durable fashion? What are its limita tions or contraindications in the management of bleeding varices? Does it have a rational role in the acute, emergent management of variceal hemorrhage? Can the long-term out come be enhanced either by technical altera tions in performance of the shunt or better patient selection? What is the proper role for DSRS in the context of other shunt and non shunt options available to treat variceal bleed ing? And how do these treatments impact on the subsequent performance of the others?
When in the course of cirrhosis and portal hypertension should a shunt be undertaken? Shunts are indicated only after an endoscopically documented episode of variceal hemor rhage. Since the majority (approximately twothirds) of patients with varices never bleed from them [31], prophylactic shunts (under taken prior to any documented bleeding) have been considered contraindicated. In a random ized. controlled trial comparing total shunts to expectant medical therapy in cirrhotics who had never bled from varices, the outcome (in cluding survival) was worse in the shunted patients [32], Though prophylactic surgery has had no supportable role in the management of portal hypertension, it is unclear whether the experience with prophylactic total shunts can be generalized to include DSRS. In recent reports from China in patients with non-alco holic cirrhosis. 39% of the shunts performed were done prophylacticallv [33] and the subse quent incidence of both bleeding and encepha lopathy was low with a long-term survival that was significantly belter than that associated with medical therapy [ 12]. Unless or until bet ter selection and prediction criteria can be developed, this prophylactic approach cannot be recommended widely. Most of the reported experience with the DSRS has been in the elective setting - in patients who are stable hemodynamically and not bleeding. In patients requiring an emer gency shunt for active hemorrhage, total shunts have been thought to provide more direct, immediate, and predictable decom pression. Because of the somewhat circuitous decompression route from the high-pressure varices to the low-pressure splenorenal anas tomosis, the post-DSRS time to safe variceal decompression has been thought to be neither immediate nor predictable. Until more reli
88
Gusherg
able methods are developed to safely and repeatedly measure variceal pressure (and clearly define the natural history of postDSRS decompression), the emergent applica tion of the DSRS should be undertaken onlyin a cautious and highly selective fashion. Are there other settings in which the DSRS is relatively or absolutely contraindicated? Massive splenomegaly with hypcrspicnism. pancreatitis with pseudocyst, portal vein thrombosis, previous splenectomy, and chil dren with variceal hemorrhage have all been thought to pose particular problems. Particularly if approached through a mid line abdominal incision, the technical perfor mance of a DSRS should not be adversely affected by the size of the spleen. Further more. hypersplenism - commonly encoun tered in cirrhotic, portal-hypertensive pa tients - is rarely of inherent clinical signifi cance. In this setting, splenectomy is virtually never indicated for the hypersplenism alone, and the hypersplenism can be expected to improve following portal and/or splenic de compression [34], A history of pancreatitis and its complica tions does not represent an absolute contrain dication to DSRS but should raise the index of suspicion for splenic vein thrombosis (SVT) (precluding a DSRS). or particular and perhaps prohibitive technical difficulty asso ciated with peripancreatic fibrosis or inflam mation and splenic vein phlebitis. Under such circumstances, clear preoperative visualiza tion of the splenic vein (by either venousphase angiography. magnetic resonance imag ing or Doppler ultrasound) is mandatory. Furthermore, endoscopic sclerotherapy has been associated with SVT [15]. In patients being evaluated for DSRS following sclero therapy failure, particular attention should be directed at ruling out SVT. The role for thrombectomy or thrombolytic therapy in this setting has not been defined.
The Distal Splenorenal Shun!
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
DSRS - When Is It Indicated and When Is It Not?
DSRS - and the Alternatives Since Whipple [1] and Blakemore [2] re ported their initial experience with portacaval shunting in patients with portal hypertension and variceal bleeding, a variety of operative and non-operative approaches have become available. With advances in anesthesia, peri operative monitoring, and operative tech niques. surgery has become more broadly ap plicable and safer. With the further advances in immunobiology, even transplantation has become a viable option in some patients with variceal hemorrhage associated with endstage liver disease. The réintroduction and now wide experience with sclerotherapy as
well as the emerging role of (3-blockade in the treatment of some cirrhotics with variceal hemorrhage have broadened both the treat ment options and the spectrum of patients who can be treated. Since its introduction more than two de cades ago, the DSRS has been widely applied throughout the world. Significant experience has accumulated in North America. South America, Europe, Africa, and Asia. As we approach the next century, almost 50 years since the first reported decompressing shunt for variceal bleeding, this collective experi ence provides a useful perspective of the role of the DSRS in the management of the patient with portal hypertension and variceal bleed ing. While its hemodynamic premise has been largely validated, the DSRS provides variceal decompression that is effective but not as selective as originally hoped. Prograde portal flow is maintained initially, but may be lost with time. Operative alterations and patient selection may lead to more prolonged preser vation of portal flow (and survival). The DSRS provides effective and long-term bleed ing control, a low incidence of post-shunt encephalopathy, but no uniform or predicta ble survival advantage. In non-alcoholic pa tients. however, this selective shunt has been associated with improved survival. The DSRS has shown to be feasible in a wide variety of settings (elective and emer gent. in adults and children, in cirrhosis and extrahepatic portal hypertension). With the advances in operative technique, anesthesia, monitoring, and post-operative management, shunt surgery in general (and the DSRS in particular) can be recommended with an in creasing confidence in its anticipated safety. In a variety of series, the operative mortaliy, re-bleeding rate, and incidence of encephalop athy have been acceptably low (see table 1). In view of this encouraging and accumulat ing experience, what is its proper role in the
89
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Portal vein thrombosis is a recognized extrahepatic cause of portal hypertension and variceal bleeding. In the management of bleeding in this setting, DSRS has been shown to be safe and effective [35]. Is DSRS feasible in patients who have pre viously undergone splenectomy? If the splenic vein is patent and the variccal-short gastric vein-splenic vein pathway remains intact, variceal decompression by DSRS is feasible and should be effective [36], Variceal bleeding in children poses an un usual but special challenge. While many dif ferent shunt types have been advocated, large central total shunts have generally been con sidered preferable, presumably because the larger, higher flow shunt should be associated with better patency. Maksoud and Mies [37], in 1982. reported on a group of 21 children, ages 4-12, with variceal bleeding treated by DSRS; the long-term patency rate and bleed ing control were excellent, and no patients developed encephalopathy. While shunt deci sions in children need to be individualized, at least the DSRS appears to be feasible in this setting.
Table 1. Distal spleno-rcnal shunt results - selected series Series
Follow-up years
Patients
Mean years
O perative m ortality
Re bleeding
PSE
D eaths
Fischer. 19 8 1 [5] Millikan, 1985 [13] Langer. 1985 [18] Rikkers, 1987 [4] Grace. 1988 [ 16] Gusberg, 1992 (this study)
5 11 5.5 2 3.5 3.5
23 26 38 23 43 19
48 49 49 NA 54 68
1 (5) 3(12) 5(13) 1 (5) 4 (9 ) 0 (0 )
1 (4) 2 (8 ) 1 (3) 5(19) 7 (18) 0 (0 )
1 (4) 7(27) 8 (24) 4 (1 6 ) 20(51) 4 (21)
4 (1 5 ) 15(58) 20(56) 10(45) 20 (4 6 ) 5 (26)
context of other available operative and non operative options? The DSRS provides bleeding control that is equivalent to that achieved with total shunts and is associated with a lower inci dence of encephalopathy and. at least in non alcoholics, a perhaps prolonged survival. Its effectiveness in providing immediate decom pression in the setting of acute hemorrhage remains to be proven. Whether the smalldiameter portacaval H graft, introduced by Sarfeh et al. [38]. can comparably achieve the dual objective of variceal decompression and preservation of portal flow (and do so in a durable fashion) requires a broader and lon ger experience. The operation is technically less demanding, may provide more imme diate variceal decompression, and the early data on its patency and hemodynamic effects are encouraging. Liver transplantation is the most definitive and curative therapeutic ap proach to these bleeding, portal hypertensive cirrhotics. While it must be applied with great caution and selectivity, it is clearly the only approach that is non-palliative and that can cure both the cirrhosis and the portal hyper tension. In selected end-stage cirrhotic pa tients bleeding from varices, it is the treat ment of choice. Furthermore, the prospect of future candidacy for liver transplantation
90
Gusberg
should influence the choice of therapy in pa tients with still well-preserved liver function who are bleeding from varices. The DSRS, requiring no dissection of the portal struc tures, should have no adverse technical or hemodynamic impact on the subsequent per formance of a transplant and may be the best pre-transplant elective shunt option in these patients [39], While P-blockade has been associated with few short- or long-term risks, its long-term effectiveness in controlling variceal hemor rhage is uncertain and unpredictable. Lebrec et al. [40. 41] have demonstrated that oral propranolol can produce a decrease in both portal pressure and the rate of recurrent vari ceal bleeding in some patients. Neither Bur roughs et al. [42] nor Villeneuve et al. [43] were able to confirm the effectiveness of pro pranolol as compared to placebo in control ling re-bleeding in unselected advanced cir rhotics with documented variceal hemor rhage. In recommending pharmacologic ther apy for variceal bleeding, the challenge re mains one of selection: many patients will re bleed despite propranolol treatment: as many as 30% of patients will have no propranololinduced drop in portal pressure [44], and the compensatory response to hemorrhage resus citation has been inhibited in some P-blocked
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
Figures shown in parentheses are percentages.
Table 2 M anagem ent of Variceal Hemorrhage
Acute sclerotherapy
Vasopressin - NTG
Child's A non-alcoholic cirrhosis
Child’s B and/or unstable and/or persistent alcoholic
Selective shunt
Sclerotherapy or selective shunt or ß-blockade
Child's C
I
[49, 50], Multiple sessions are required for complete obliteration, gastric varices or por tal hypertensive gastropathy cannot be effec tively treated by sclerosis, and about half of the patients treated bleed again. Furthermore, sclerotherapy is not without complications and some of the complications (chronic ul cers, extravariceal splanchnic venous throm bosis) might limit future treatment options [15]. While sclerotherapy appears to be an appropriate and widely applicable option in the acute management of patients bleeding from esophageal varices - with initial control in as many as 80-90% of patients [51 ]. its role as definitive therapy in the long-term control of variceal bleeding remains uncertain. It would not appear to be appropriate therapy aimed at providing chronic bleeding control in all portal-hypertensive cirrhotic patients.
91
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
patients presenting with variceal hemorrhage [43]. At the present time, problems with com pliance. tolerance, and effectiveness limit the uniform applicability of p-blockade to pre vent recurrent variceal bleeding in cirrhotic patients. While it can be effective and useful, it cannot be considered a better alternative to DSRS in all such patients. While sclerotherapy was first reported in the treatment of esophageal varices, more than 50 years ago [44], it has been used widely in the primary treatment of variceal hemor rhage only in the past decade. Sclerotherapy has been advocated in the emergency, initial, and chronic management of variceal hemor rhage. While several controlled trials have documented a decrease in re-bleeding rates in patients undergoing sclerotherapy [46-48]. the control is neither certain nor durable and no survival advantage has been documented
Transplant or sclerotherapy or ß-blockade
Conclusion The safe and appropriate management of variceal hemorrhage is a persistent challenge. Biological and technological advances over the past five decades have increased both the variety and safety of the interventions avail able for cirrhotic patients with variceal hem orrhage. While many unanswered questions remain, current practice is closer to the prom ise of matching a particular therapy to a par ticular patient (table 2). Shunt surgery re mains an effective and durable approach to the prevention of recurrent bleeding. Ad vances in anesthesia, monitoring, and opera tive technique have made such surgery safer and more widely applicable. While neither the clinical nor the hemodynamic consequence of the DSRS are all that its developers might have hoped, it is a feasible and effective treat-
nient in many cirrhotics with variceal hemor rhage. In good-risk, stable, non-alcoholic pa tients, it is probably the treatment of choice. In end-stage non-alcoholic patients, evalua tion for transplantation appears warranted. In the remaining patients, the choice among shunt surgery, p-blockade. sclerotherapy or transplantation should be dictated by the pa tient’s overall status, the nature of the varices and site of bleeding, the type and severity of the liver disease, the activity or chronicity of the bleeding, and the patient’s ability to ac cept and tolerate the various alternatives. Ul timately, the optimal choice will be deter mined by a better understanding of the hemo dynamic changes associated with portal hy pertension and its treatment as well as the development of more reliable predictors of the treatment outcomes.
1 Whipple AO: The problem of portal hypertension in relation to the hepato-splenopathies. Ann Surg 1945: 122:449-475. 2 Blakemore AH: Portacaval shunt for portal hypertension. Follow-up results in cases or cirrhosis of the liv er. JAMA 1951:145:1335-1339. 3 Warren WD. Zeppa R. Forman JS: Selective transplenic decompression of gastroesophageal varices by distal splenorenal shunt. Ann Surg 1967; 166:437. 4 Rikkers LF. Rudman D. Galantbos JT. et al: A randomized controlled trial of the distal splenorenal shunt. Ann Surg 1978;188:271-282. 5 Fischer JE. Bower Rl l. Atamian S. et al: Comparison of distal and prox imal splenorenal shunts. A random ized prospective trial. Ann Surg 1981:194:531-542.
92
6 Rikkers LF. Soper NJ, Cormier RA: Selective operative approach for variceal hemorrhage. Am J Surg 1984:147:89-90. 7 Orozco H. Juarez F. Gantillan P. et al: Ten years of selective shunts for hemorrhagic portal hypertension. Surgery 1988:103:27-31. 8 Salam A. Ezzat F. Abu-Elmag I: Se lective shunt in schistosomiasis in Egypt. Am J Surg 1990:160:90-92. 9 Pezzuoli G, Spina GD, Santambrogio R. et al: The distal spleno renal shunt: An update on experi ence with 106 cases. Int Surg 1987; 92:144-148. 10 Myburgh JA: Selective shunts: The Johannesburg experience. Am J Surg 1990:160:67-74. 11 Inokuchi K. Beppu K. Kayanagi N, et al: Exclusion of non-isolated splenic vein in distal splenorenal shunt for prevention of portal malcirculation. Ann Surg 1984:200: 711-717.
12 Jin. G: Current status of the distal splenorenal shunt in China. Am J Surg 1990:160:93-97. 13 Millikan WJ. Warren WD. Hender son JM. et al: The Emory prospec tive randomized trial: Selective ver sus non-selective shunt to control varciceal bleeding. Ten year follow ing-up. Ann Surg 1985:201:712— 722. 14 Eckhauser FF.. Pomerante RA. Knol JA. et al: Early variceal re-bleeding shunt. Arch Surg 1986:121:547552. 15 Leach SD. Meyer GH. Gusberg RJ: Endoscopic sclerotherapy: A risk factor for splanchnic venous throm bosis. J Vase Surg 1989:10:9-13. 16 Grace ND. Conn HO. Resnick Rll. et al: Distal splenorenal vs portosys temic shunts after hemorrhage from varices: A randomized controlled trial. Hepatology 1988:8:1475— 1481.
Gusberg
The Distal Splenorenal Shunt
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
References
29 Henderson JM. Khishen MA. Milli kan WJ. el al: Management of steno sis of distal splenorenal shunt by bal loon dilation. Surg Gynecol Obstet 1983:157:43-48. 30 Henderson JM. Millikan W J.Chipponi J. ct al: The incidence and nat ural history of thrombus in the por tal vein following distal splenorenal shunt. Ann Surg 1982:196:1-7. 3 1 Burroughs AK. Heygere F. McIntyre N: Pitfalls in studies of prophylactic therapy for variceal bleeding in cir rhotics. Hepatology 1986:6:1407— 1413. 32 Conn HO. Lindcnmuth WW. May CT, et al: Prophylactic portacaval anastomosis. A tale of two studies. Medicine (Baltimore) 1972:51:2740. 33 Yantin H: Operations for portal hy pertension. Clin Arch Surg 1985: 120:1197-1199. 34 Hutson DG. Zeppa R. Levi JU: The effect of distal splenorenal shunt on hvpersplenism. Ann Surg 1977; 185: 605-682. 35 Warren W'D. Henderson JM. Milli kan WJ. et al: Management of vari ceal bleeding in patient with non cirrhotic portal vein thrombosis. Ann Surg 1988:207:623-634. 36 Warren WD. Millikan WJ, Hender son JM. et al: Selective variceal de compression after splenectomy or splenic vein thrombosis. Ann Surg 1984:199:694-702. 37 Maksoud JG. Mies S: Distal sple norenal shunt (DSRS) in children. Ann Surg 1982:195:401-405. 38 Sarfeh 1.1. Rypins EB. Mason GR: A systematic appraisal of portacaval 11-graft diameters: Clinical and he modynamic perspectives. Ann Surg 1986:204:356. 39 Wood RP. Shaw B\V. Rikkers L: Liver transplantation for variceal hemorrhage. Surg Clin North Am 1990:70:449-461. 40 Lebrcc D. Nouel O. Berneau J, et al: Propranolol in the prevention of re current variceal hemorrhage in cir rhotic patients. Lancet 1981 :ii: 180— 182.
4 1 Lebrec D. Nouel O. Berneau J, et al: Propranolol in the prevention of re current variceal hemorrhage in cir rhotic patients. Lancet 1981 :i:920— 921. 42 Burroughs AK. Jenkins W'J. Sher lock S, et al: Controlled trial of pro pranolol for the prevention of recur rent variceal hemorrhage in patients with cirrhosis. N Engl J Med 1983: 309:1539-1542. 43 Viileneuve JP. Pauier-Layrargues G. Infante-Richard C. et al: Short term effects of propranolol on portal venous pressure. Hepatology 1986: 6:101-106. 44 Garcia-Tsao G, Grace ND. Groszmann RJ. ct al: Short-term effects of propranolol on portal venous pres sure. Hepatology 1986:6:10 1—106. 45 Crafoord C. Frcukner P: New surgi cal treatment of varicose veins of the esophagus. Acta Otolaryngol 1939: 27:422-427. 46 The Copenhagen esophageal varices and sclerotherapy project: Sclero therapy after first variceal hemor rhage in cirrhosis: A randomized multi-center trial. N Engl J Med 1984:311:1594-1597. 47 Korula J. Balart LA, Radran G .et al: A prospective, randomized trial of chronic esophageal variceal sclero therapy. Hepatology 1985:5:584— 589. 48 Wcstaby D. Hayes PC. Grimson AES. et al: Controlled clinical trial o f injection sclerotherapy for active variceal bleeding. Hepatology 1989; 9:274-277. 49 Terblanche J. Bornman PC. Kahn D. et al: Failure of repeated injec tion sclerotherapy to improve long term survival after esophageal vari ceal bleeding. A five-year prospec tive controlled clinical trial. Lancet 1982:ii: 1328—1331. 50 Wcstaby D. Williams R: Status of sclerotherapy for variceal bleeding in 1990. Am J Surg 1990; 160:32— 36. 51 Terblanche J, Yakoob HI. Bornman PC, el al: Acute bleeding varices. A five-year prospective evaluation of tamponade and sclerotherapy. Ann Surg 1981:194:521-524.
93
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/29/2018 10:45:51 PM
17 Harley HAJ. Morgan T. Redeker AG. et al: Results of a randomized trial of end-lo-side portacaval shunt and distal splenorenal shunt in alco holic liver disease and variceal bleeding. Gastroenterology 1986:91: 802-809. 18 I.angcr B. Taylor BR. Mackenzie DR. et al: Further report of a pro spective randomized trial compar ing distal splenorenal shunt with end-to-side portacaval shunt. Gas troenterology 1985:88:424-429. 19 Rikkers LF: Is the distal splenorenal shunt better? Hepatology 1988:8: 1705-1707. 20 Henderson JM: Variceal bleeding: Which shunt? Gastroenterology 1986:91:1021-1023. 21 Maillard JN. Flamant YM. HavJM. et al: Selectivity of the distal sple norenal shunt. Surgery 1979:86: 663-671. 22 Vang J . Simcrt G. Hansson J A. el al: Results of a modified distal spleno renal shunt for portal hypertension. Ann Surg 1977:185:224-228. 23 W'arren WD. Millikan WJ. Hender son JM. et al: Ten years of portal hypertensive surgery at Emory: Re sults and new perspectives. Ann Surg 1982:195:530-542. 24 Zeppa R. Hensley GT. Levi JU. et al: The comparative survival of alco holics versus nonalcohol ics after d istal splenorenal shunt. Ann Surg 1978:187:510. 25 Zeppa R. Lee PA. Hutson DG. et al: Portal hypertension. A fifteen-vear perspective. Am J Surg 1988:155: 6-9. 26 Henderson JM. Millikan WJ: Wright-Bacon L. et al: Hemody namic differences between alcoholic and non-alcoholic cirrhotics follow ing distal splenorenal shunt: Effect on survival? Ann Surg 1982:198: 325-334. 27 Henderson JM. Warren WD. Milli kan WJ, et al: Distal splenorenal shunt with splenopancreatic discon nection: A four year assessment. Ann Surg 1989:210:332-341. 28 MalTei-Faccioli A. Gerunda GE, Neri D. et al: Selective variceal de compression and its role relative to other therapies. Am J Surg 1990: 160:60-66.
