The Emory prospective randomized trial: selective versus nonselective shunt to control variceal bleeding. Ten year follow-up.

From 1971 to 1975, 55 patients with variceal bleeding secondary to cirrhosis were entered into a prospective randomized trial comparing distal splenorenal (selective) and H-graft interposition (nonselective) shunt. This 10-year follow-up documents that selective shunt is better (p less than 0.05) in four of the five variables monitored. Control of bleeding: selective shunt prevented variceal bleeding better than interposition shunt due to the higher (0.05 less than p less than 0.1) occlusion rate (30%) of interposition shunt. Selective shunt maintained postoperative portal perfusion better (p less than 0.01) than patent interposition shunt. Seventy-five per cent of selective shunt survivors have portal perfusion at 10 years: no patient with a patent nonselective shunt perfuses the liver. Quantitative liver function was better preserved (p less than 0.01) 10 years after selective shunt than nonselective shunt. Postoperative encephalopathy occurred in fewer (p less than 0.01) selective (27%) than nonselective (75%) shunt patients over the 10 years. Survival: in the randomized population, the improved survival in the selective shunt subgroup did not reach statistical significance. However, improved survival was confirmed in nonalcoholics. Five of eight nonalcoholics operated with selective shunt are alive at 10 years with patent shunts. No nonalcoholic, of seven total, operated with nonselective shunt survived 10 years with a patent shunt. These data show that selective shunt was superior to nonselective shunt. There was less rebleeding and encephalopathy after distal splenorenal shunt; postoperative portal perfusion and hepatic function were maintained.     

Portal diversion for portal hypertension in children. The first ninety patients.

Ninety children with portal hypertension were treated by portal diversion. Fifty-two had cavernous transformation of the portal vein and 38 had an intrahepatic block from various causes. There were 59 central splenorenal shunts, 19 mesocaval, 11 portacaval and one distal splenorenal. In 61 peripheral shunts the veins used for the anastomosis were less than 10 mm in diameter. There was no operative mortality in children with extrahepatic block. One child with cystic fibrosis died postoperatively. Thrombosis of the shunt occurred in five children (5.6 per cent) and was responsible for recurrent bleeding in two. Four children with a thrombosed shunt underwent succesful reoperation and one is awaiting another anastomosis. No late complications occurred in the 52 children with extrahepatic block, while encephalopathy developed in four children with intrahepatic block. These figures confirm our earlier results in the management of portal hypertension in childhood and suggest that portal diversion is the treatment of choice. Several precautions have permitted lowering of the rate of thrombosis whichever shunt is performed. Portal diversion should be indicated following the first episode of hemorrhage in children with extrahepatic block. In patients with intrahepatic block, congenital hepatic fibrosis and cystic fibrosis are good indications as are in general the hepatic diseases with no or mild activity.     

The clinical effectiveness of selective portal shunts

The use of four types of selective portal shunts designed to avoid postshunt encephalopathy has been reviewed. The incidence of hospital mortality and recurrent hemorrhage from varices among 187 reported cases have been comparable to standard portacaval shunts. The incidence of postshunt encephalopathy among 154 survivors with patent shunts followed from two to ninety-six months has averaged 8 per cent but with significant differences between the types of operation. The most effective has been the left gastric vein to vena cava shunt followed by the distal splenorenal shunt, the modified central splenorenal shunt, and the central splenorenal shunt, with success in avoiding encephalopathy in direct proportion to the number and size of postoperative collaterals between the persistently hypertensive portal system and the decompressed splenic system. The development of such collaterals in long-term survivors, especially after the splenorenal shunts, may contribute to an incidence of late encephalopathy approaching that of nonoperative patients.     

Ten Years Portal Hypertensive Surgery at Emory: Results and New Perspectives

Five hundred four Shunt procedures have been done at Emory University Hospitals between 1971 and 1981 to decompress bleeding esophageal varices. This paper reviews how far the experiences of a prospective randomized study (55 patients) of distal splenorenal shunts against total shunts is supported by the nonrandomized experience (449 patients), and outlines our current methods of management dictated by this experience. The overall operative mortality for 348 selective shunts is 4.1% and for 156 nonselective shunts, 14.1%. The five-year survival following Selective shunt is 59%, and following nonselective shunt is 49%: more than half the selective shunt patients are alive, in contrast to the median survival of 44.5 months for patients having nonselective shunts. Following Selective shunt, the survival in nonalcoholic patients is significantly better than the median survival of alcoholic patients of 57 months. Encephalopathy, reported at three years after surgery in the randomized patients was significantly (p < 0.001) lower after selective shunt (12%) compared to nonselective shunt (52%): in the same population at seven years, all patients with patent nonselective shunts have clinical or subclinical encephalopathy, but only 30% of the selective shunt patients have subclinical encephalopathy. Shunt patency, immediately after surgery, is 93% following selective shunt, with only two documented late thromboses: nine of nine patients, at a mean of seven years, retain patency in the randomized study. Shunt occlusion increases with time after interposition nonselective shunts: seven of 13 are occluded at a mean follow-up of seven years in the randomized study. Portal venous perfusion is retained in 93% of patients seven to ten days after selective shunt, but in no patient with a patent nonselective shunt. Late portal perfusion is maintained in nine of the eleven patients in the randomized group studied at a mean of seven years after selective shunt. Restoration of portal perfusion has led to clearing of encephalopathy and improvement in hepatic function in six patients. The following conclusions are made: (1) selective shunts can be done with low operative mortality, and long-term patency with excellent control of bleeding; (2) hepatic portal venous perfusion has been maintained after selective shunt for ten years, and this is vital for preventing encephalopathy and maintaining hepatic function; (3) long-term survival after selective shunt is better than any reported series for nonselective shunt; and (4) selective shunts are the operative procedure of choice for variceal decompression and nonselective shunts should rarely be performed for elective decompression.     

The significance of portal vein thrombosis after distal splenorenal shunt

The aims of this study were to determine the incidence of portal vein thrombosis after the distal splenorenal shunt, to identify any predictive factors, and to assess the clinical significance of this complication. Preoperative and postoperative angiograms and clinical evaluation were reviewed in 124 patients who underwent distal splenorenal shunts. Total and partial portal vein thrombosis were seen on 13 (10.5%) and 22 (17.7%) postoperative angiograms, respectively. The only preoperative variable correlating with development of portal vein thrombosis was portal venous perfusion, which was significantly lower in patients with than in those without portal vein thrombosis. In six of 10 patients with postoperative pancreatitis, portal vein thrombosis developed. The frequency of early postoperative complications was significantly greater in patients with total portal vein thrombosis than in those with partial or no thrombosis. Long-term follow-up has shown no significant effects of portal vein thrombosis on late ascites, encephalopathy, or survival.     

Selectivity of the distal splenorenal shunt.

The distal splenorenal shunt is less likely to provoke encephalopathy than conventional shunting procedures, and it may offer a survival advantage for certain cirrhotic individuals, presumably because of its selective nature. This study suggests that the distal splenorenal shunt, even with exceptional efforts to achieve portomesenteric-gastrosplenic (PM-GS) disconnection, is not nearly as selective as it originally was assumed to be. In 11 patients intraoperative pressure determinations showed a significant decrease in portal pressure after end-to-side distal splenorenal anastomosis and no restoration of portal pressure after PM-GS disconnection. Measurements of flow through the shunt were comparable to those reported for portacaval shunts, and shunt flow was not decreased significantly by PM-GS disconnection. Postoperative angiography showed some PM-GS collateral in 17 of 18 patients, and later angiographic studies showed a tendency for progressive collateral development and consequent loss of hepatopetal portal perfusion. The advantages of the distal splenorenal shunt must accrue from gradual, as opposed to abrupt, portal deprivation, rather than from lasting selectivity.     

Liver transplantation in patients with patent splenorenal shunts

Patent distal splenorenal shunts (Warren shunt) have been reported to cause decreases in the portal perfusion pressure and the total hepatic blood flow. Such hemodynamic alterations could have adverse effects on the transplanted liver. The experience with hepatic replacement in four patients with patent Warren shunts is reported. Operative findings were phlebosclerotic portal veins of small size and diminished portal blood flows. Hepatofugal collateral channels created by the construction of the Warren shunt were eliminated by division of the shunt and splenectomy in three patients and splenectomy alone in the other. All patients recovered; thus the presence of a patent Warren shunt should not be a contraindication for hepatic transplantation.     

Distal splenorenal shunt in children.

The distal splenorenal end-to-side anastomosis (Warren shunt) decompresses esophageal varices while maintaining high portal hypertension and avoiding reduction of portal venous blood inflow to the liver. The Warren shunt was performed in seven consecutive patients with portal hypertension, including post-necrotic cirrhosis, portal thrombosis, and schistosomiasis, all with recurrent esophageal bleeding. Five shunts remained patent and two thrombosed. There was no mortality. If long-term follow-up evaluations indicate its effectiveness in preventing esophageal hemorrhage, the distal selective splenorenal shunt would be the more physiologic and safer procedure in children with portal hypertension.     

Surgical treatment of portal hypertension

A switch to decompressive shunt procedures is mandatory if endoscopic therapy fails to control recurrent variceal hemorrhage. Surgical shunt procedures continue to be safe, highly effective and durable procedures to control variceal bleeding in patients with low operative risk and good liver function (Child A). In cirrhotics, elective operations using portal flow preserving techniques such as a selective distal splenorenal shunt (Warren) or a partial portocaval small diameter interposition shunt (Sarfeh) should be preferred. Rarely, end-to-side portocaval shunt may serve as a salvage procedure if emergent endoscopic treatment or TIPS insertion fail to stop bleeding. Until definitive results from randomized trials are available patients with good prognosis (Child-Pugh A and B) can be regarded as candidates for surgical shunts. For patients with noncirrhotic portal hypertension, in particular with extrahepatic portal vein thrombosis, portosystemic shunt surgery represents the only effective therapy which leads to freedom of recurrent bleeding and repeated endoscopies for many years, and improves hypersplenism without deteriorating liver function or encephalopathy. Gastroesophageal devascularization and other direct variceal ablative procedures should be restricted to treat endoscopic therapy failures without shuntable portal tributaries.     

Flow and pressure characteristics of the portal system before and after splenorenal shunts.

The technique of percutaneous transhepatic portal vein cannulation provides a valuable means for determining portal pressure, direction of blood flow, and visualization of the entire portal system in the nonanesthetized patient. This technique, along with selective celiac arterial, superior mesenteric arterial, and renal venous catheterization, was used in the evaluation of a series of 17 splenorenal venous shunts [eight nonselective and nine selective (modified) distal splenorenal shunts]. As a result of these studies it is concluded that (1) prograde portal flow is maintained in the majority of patients following nonselective or selective (modified) distal splenorenal shunts; (2) bidirectional flow occurs in various branches of the portal system before and after splenorenal shunts; (3) a significant drop in portal pressure occurs following the establishment of either type of shunt; and (4) esophageal varices are decompressed by the trans-splenic route following either type of procedure used in this study.     

Current state of portosystemic shunt surgery

BACKGROUND: A switch to decompressive shunt procedures is mandatory if endoscopic therapy fails to control recurrent variceal hemorrhage. Surgical shunt procedures continue to be safe, highly effective, and durable procedures to treat variceal bleeding in patients with low operative risk and good liver function. DISCUSSION: In cirrhotics, elective operations using portal flow preserving techniques such as a selective distal splenorenal shunt (Warren) and a partial portocaval small diameter interposition shunt (Sarfeh) should be preferred. Rarely, end-to-side portocaval shunt may serve as a salvage procedure if emergency endoscopic treatment or transjugular intrahepatic portosystemic shunt insertion fails to stop bleeding. Until definitive results from randomized trials are available patients with good prognosis (Child-Pugh A and B) should be regarded as candidates for surgical shunts. For patients with noncirrhotic portal hypertension, in particular with extrahepatic portal vein thrombosis, portosystemic shunt surgery represents the only effective therapy which leads to freedom of recurrent bleeding and repeated endoscopies for many years, and improves hypersplenism without deteriorating liver function or encephalopathy. Gastroesophageal devascularization and other direct variceal ablative procedures should be restricted to treat endoscopic therapy failures without shuntable portal tributaries.     

A Randomized, Controlled Trial of the Distal Splenorenal Shunt

In 1971 a prospective, randomized trial was initiated to determine efficacy of the distal splenorenal shunt in the management of cirrhotic patients who had previously bled from esophageal varices. When entry into the trial was terminated in 1976, 26 patients had received the distal splenorenal shunt (selective) and 29 had undergone a nonselective shunting procedure (18 interposition mesorenal, six interposition mesocaval, and five other nonselective shunts). Three operative deaths occurred in each group. Early postoperative angiography revealed preservation of hepatic portal perfusion in 14 of 16 selective patients (88%), but in only one of 20 nonselective patients (5%; p < .001). Quantitative measures of hepatic function (maximal rate of urea synthesis or MRUS and Child's score) were similar to preoperative values in the selective group but were significantly decreased in nonselective patients on the first postoperative evaluation (p < .001 for MRUS; p < .05 for Child's score). Eighty-seven per cent of selective and 81% of nonselective patients have now been followed for three to six years since surgery. Late postoperative evaluation of 29 survivors (12 selective, 17 nonselective) still shows an advantage to the selective group with respect to MRUS, Child's score, and incidence of hepatopetal portal blood flow, but differences are no longer statistically significant. However, if the seven patients with portal flow (five selective; two nonselective) are compared to the 20 with absent portal flow (seven selective; 13 nonselective), the former group has significantly higher values for MRUS (p < .05) and Child's score (p < .025). No patient with continuing portal perfusion has developed encephalopathy as compared to a 45% incidence of this complication in individuals without portal flow (p < .05). No significant differences between selective and nonselective groups have appeared with respect to total cumulative mortality (ten selective; 38%; eight nonselective, 28%), shunt occlusion (two selective, 10%; five nonselective, 18%), or recurrent variceal hemorrhage (one selective, 4%; two nonselective, 8%). Overall, significantly fewer selective patients have developed postoperative encephalopathy (three selective, 12%; 15 nonselective, 52%; p < .001). Therefore, we conclude that the distal splenorenal shunt, especially when its objective of maintaining hepatic portal perfusion is achieved, results in significantly less morbidity than nonselective shunting procedures.     

Selective distal splenorenal shunt versus side-to-side portacaval shunt. Clinical results of a prospective, controlled study

A prospective, controlled study comparing the clinical results of the selective distal splenorenal shunt procedure and the side-to-side portacaval shunt procedure was undertaken in 1980. Ninety-three cirrhotic patients with previous episodes of bleeding from esophageal varices underwent a distal splenorenal shunt procedure (47 patients). The operative mortality rate was 2 percent in both groups. The intraoperative decrease of portal hypertension after the portacaval shunt procedure was higher than after the distal splenorenal shunt procedure (p less than 0.05), and in those with patent shunts, there was a 0 percent incidence of early variceal rebleeding after the portacaval shunt procedure compared with a 9 percent incidence after the distal splenorenal shunt procedure (p less than 0.05). Both shunts, however, had similarly satisfactory results in preventing long-term variceal rebleeding (portacaval shunt 2 percent and distal splenorenal shunt 0 percent). Postoperative ascites was more common after the distal splenorenal shunt procedure (58 percent versus 24 percent; p less than 0.01). Analysis of actuarial survival curves showed no difference between the two procedures. The incidences of long-term episodes of chronic encephalopathy were not statistically different after both procedures. The only three instances of severe encephalopathy occurred in patients with the portacaval shunt (p less than 0.05). The distal splenorenal shunt also seemed to have a less negative effect on postoperative liver function than the portacaval shunt. These data suggest that the selective shunt should be viewed as a first choice strategy in the treatment of portal hypertension.     

Hepatic encephalopathy due to congenital splenorenal shunts: Report of a case

A 76-year-old Japanese woman with hepatic encephalopathy was successfully treated for congenital splenorenal shunts by surgical intervention. The flow volume of the splenorenal shunts was 800 ml/min with a shunt pressure of 9.8 mmHg. The portal pressures before and after the shunt resection were 9 mmHg and 12 mmHg, respectively. The portal flows before and after the shunt resection were 230 ml/min and 470 ml/min, respectively. Therefore, both the hepatic sinusoid and the portal vein might provide good compliance for an increased portal flow volume load after shunt resection.     

Portal pseudoperfusion: an angiographic illusion.

Much confusion regarding the hemodynamics following interposition mesosystemic shunts prevails. Many authorities have claimed that portal venous perfusion continues following interposition mesocaval shunts. In 1971, a prospective, randomized trial comparing the distal splenorenal shunt with a variety of interposition mesosystemic shunts (primarily mesocaval or mesorenal) was begun. Visceral angiography was utilized to assess the early and late postoperative hemodynamic changes following both selective and nonselective shunts. None of the patients with patent interposition shunts retained portal perfusion present preoperatively. Searching for an explanation for this hemodynamic discrepancy, we examined two patients of the randomized trial angiographically. Both patients had excellent portal perfusion preoperatively, yet following interposition shunting (one mesocaval and one splenocaval), neither maintained portal perfusion of the liver. Celiac artery injections produced opacification of the entire splenoportal axis; however, it is shown that such portal venous opacification occurred in a retrograde direction by selective hepatic arterial injections demonstrating hepatofugal portal venous flow. Additionally, two nonrandomized patients received interposition mesorenal shunts and exemplify this phenomenon, entitled "portal pseudoperfusion". The explanation for conflicting literature reports lies in the misinterpretation of venous phase celiac and non-selective SMA arteriography in determining the direction of portal flow. A narrative of preoperative and postoperative angiograms of four patients will clarify the mechanism of "portal pseudoperfusion" and demonstrate that interposition shunts totally siphon portal venous perfusion. Clues to the detection and techniques to avoid this phenomenon will be presented.     

Left renal vein ligation for large splenorenal shunt during liver transplantation

Adequate hepatopetal portal vein blood flow is obligatory to ensure proper liver function after liver transplantation. Large collateral veins as shunts impair portal vein flow and even cause hepatofugal blood flow and portal steal syndrome. In particular, splenorenal shunts in liver transplant recipients can lead to allograft dysfunction and possible allograft loss or hepatic encephalopathy. Restoration of portal flow through left renal vein ligation (LRVL) is a treatment option, which is much easier compared to splenectomy, renoportal anastomosis and shunt closure, but bears the risk of moderate and temporary impairment of renal function. In addition, a patent portal vein is mandatory for LRVL. However, although LRVL has been reported to be an effective, safe and easy method to control portacaval shunts and increase hepatopetal flow in some studies, indications and safety are still not clear. In this review, we summarize existing studies on LRVL during liver transplantation.     

Selective variceal decompression in portal vein thrombosis

Thirty-two patients with congenital portal vein thrombosis have been managed for bleeding gastro-oesophageal varices. Fifteen had splenectomy and/or other therapy before referral: nine were managed by endoscopic sclerosis, four by devascularization and two by total shunt; six rebled. Seventeen had their spleen 'in situ' at referral and were evaluated for selective shunt: thirteen had distal splenorenal shunts (DSRS)--one transiently rebled despite a patent shunt and one had shunt thrombosis; four had no veins suitable for shunt, and were managed by splenectomy and devascularization, with two rebleeds. Detailed study of seven patients before, and 1 year after DSRS, showed a rise in platelet count, maintenance of hepatocyte function, portal perfusion, liver blood flow and liver size. The spleen showed a significant (P less than 0.025) reduction in size with trans-splenic decompression. We conclude that DSRS provides an excellent method for long-term control of bleeding in such patients, without alteration of liver function or haemodynamics. Patients managed by splenectomy and direct ablative procedures have a significantly (P less than 0.05) greater risk of rebleeding than patients receiving DSRS.     

Graft interposition splenocaval shunt for total or selective decompression of portal hypertension

A new operation for selective or total decompression of the portal venous system in cases of intrahepatic portal hypertension is described. It involves interposition of a large-caliber Dacron graft between the splenic vein and the inferior vena cava. The graft-interposition splenocaval shunt is performed readily and quickly, satisfying the variable hemodynamic needs of patients with portal hypertension. It can be either selective (S-SCS) or total (T-SCS) from the beginning, or a T-SCS may be converted subsequently to a S-SCS should surgically induced hepatic decompensation supervene. It is less demanding technically than distal splenorenal shunt (D-SRS). The S-SCS conserves portal venous perfusion of the liver, preserves hepatocellular function and architecture at the preoperative levels, avoids precipitation of postshunt portal-systemic encephalopathy, and decompresses gastric-esophageal varices with prevention of further variceal bleeding even better than D-SRS. One hundred percent graft patency has been obtained, and the surgical results have been superior to those following portacaval shunt in patients with large liver blood flow and relative benignity of the liver disease, be it cirrhosis or noncirrhotic portal fibrosis. In patients with advanced cirrhosis, variceal bleeding, and small liver blood flows, T-SCS would be indicated. Patients of this category obtained inferior surgical results and had operative deaths (16.7%) following S-SCS. The concept of the operation has merits and deserves further evaluation.     

Selective splenocaval shunt. Report of 26 cases and review of the literature

Splenocaval shunting was performed in 26 patients for the treatment of variceal bleeding. The indications for this alternative selective shunt were congenital anomalies of the left renal vein, inadequate outflow from the left renal vein on preoperative venography, or an anatomic relationship of the splenic vein favoring direct splenocaval rather than splenorenal anastomosis. Technical considerations in which splenocaval shunts differ from distal splenorenal shunting relate to exposure of the vena cava. Operative mortality was 7.7% (2/26). Immediate shunt patency was documented in 23 of 24 patients studied, and all 14 shunts studied at 13 to 57 months were patent. Portal perfusion was maintained in 95% (20/21) of the patients when studied at 7 to 10 days after shunting and in 57% (8/14) at late follow-up. This experience with distal splenocaval shunting has reaffirmed its place as an alternative technique to selective distal splenorenal shunts, particularly when the left renal vein is abnormal.     

Hepatofugal portal blood flow in hepatic cirrhosis.

A variety of indirect techniques has been claimed to provide evidence of spontaneous reversal of portal blood flow in hepatic cirrhosis but the existence of the phenomenon has been doubted by some who do not accept the validity of the indirect evidence. There are few reports of the demonstration of hepatofugal portal flow by selective hepatic arteriography, which is the only acceptable technique. We report three patients with histologically confirmed cirrhosis in whom hepatofugal portal blood flow was unequivocally demonstrated by arteriography, in whom no surgical portosystemic shunt had been performed and in whom there was no evidence of the Budd-Chiari Syndrome or hepatoma, situations accepted as associated with reversed portal blood flow. Theoretical considerations suggest that shunt surgery for bleeding esophageal varices should not be ruled out on the grounds of hepatofugal portal flow. However, end-to-side portacaval anastomosis and distal splenorenal shunt might predispose to the early redevelopment of esophageal varices when reversed portal flow is present. Side-to-side portacaval and conventional splenorenal shunts might be preferable in having less effect on hepatic parenchyma perfusion than when orthograde portal flow in the case.