TIPS is a cost effective alternative to surgical shunt as a rescue therapy for prevention of recurrent bleeding from esophageal varices☆
Article Outline
Portosystemic shunt surgery has played a major role in the evolution of management of portal hypertension. Prophylactic shunts were abandoned in the early 1970s after 4 RCTs showed increased encephalopathy and mortality in operated patients, although bleeding was virtually abolished. The increase in mortality and portal systemic encephalopathy was attributed to the progressive liver function deterioration, induced by the blood diversion from the portal to the systemic circulation [1].
Selective variceal decompression by the distal splenorenal shunt (DSRS) was ideated in the late 1960s to maintain a sufficient portal hepatopetal flow, and therefore preserving liver function, while reducing the risk of variceal bleeding. However, since RCTs and meta-analyses showed no advantage of DSRS over total shunt and sclerotherapy [2], it was confined to the role of rescue therapy for failure of endoscopic therapy for the prevention of rebleeding in patients with relatively well preserved liver function.
More recently the trans-jugular intra-hepatic portal systemic shunt (TIPS) has been added to the therapeutic options for portal hypertension and it is currently widely used as a rescue therapy after failure of combined medical and endoscopic therapy for the prevention of recurrent variceal bleeding. In fact, since two meta-analyses showed that TIPS is effective in reducing the risk of variceal rebleeding, but also increases the risk of portal hepatic encephalopathy and does not improve survival, it is not recommended as a first-line treatment for the prevention of recurrent variceal bleeding [3]. Besides the lack of clinical advantage, bare stent TIPS has also the disadvantage of requiring a strict surveillance program with ultrasonography and repeated reinterventions for preserving its patency, with a marked increase of the overall cost over that of the initial procedure. However, the rates of stent dysfunction and reinterventions have been significantly reduced by the use of covered stents.
Current AASLD guidelines for the prevention of variceal hemorrhage in cirrhosis recommend that TIPS should be considered in patients who are Child A or B who experience recurrent variceal hemorrhage despite combination pharmacological and endoscopic therapy. In centers where the expertise is available, surgical shunt can be considered in Child A patients [3]. A similar position was reached in the last Baveno consensus conference [4].
Rebleeding rate under combined endoscopic and pharmacological therapy ranges from 15% to 25% within 1 to 2 years of follow-up in published RCTs, suggesting that rescue therapy with TIPS or DSRS may be required in a sizable proportion of patients [5], [6], [7].
The question whether either treatment should be preferred to the other has been assessed in a recently reported large, multicenter RCT including 140 Child-Pugh class A and B patients with cirrhosis surviving a variceal bleeding episode [8]. After a mean follow-up of 46 months, no significant difference in recurrent bleeding was found between bare stent TIPS (10.5%) and DSRS (5.5%), although reintervention was significantly more frequent for TIPS (82%) than for DSRS (11%). Five-year survival was 61% and 62%, respectively, and encephalopathy was 50% in both groups.
In this issue of the Journal, a cost effectiveness analysis of TIPS compared with DSRS from the same RCT is reported [9]. Treatment costs for DSRS were those assigned to the relevant Diagnostic Related Group (DRG). Since no specific DRG for TIPS is available, the relative difference of cost for TIPS and DSRS was derived from the Cleveland Clinic’s cost accounting database and used to calculate a similar difference in DRG for TIPS vs DSRS. Quality of life was assessed by the SF-36 form completed by all patients at randomization and yearly thereafter. Utilities, derived from this form, were used to calculate quality adjusted life years (QALY).
Median initial costs for TIPS and DSRS were, respectively, $ 21,607 and $ 28,734; final costs were, respectively, $ 70,527 and $ 48,796 for patients still alive at the end of the 5-year follow-up; corresponding figures for patients dead were $ 74,267 and $ 54,975. Excess costs for TIPS compared with DSRS were accumulated for hospitalizations and for outpatient visits during the whole 5-year study period. In a probabilistic sensitivity analysis by Monte Carlo simulation (n
=1.000), ICER was $ 61,737/QALY for TIPS, and a sensitivity analysis showed that if covered rather bare stents were used, the cost effectiveness of TIPS would have increased only slightly. Acceptability analysis showed that probability for TIPS being cost effective would be approximately 0.50 with an ICER threshold of $ 50,000 and 0.65 with a threshold of $100,000. The Authors’ conclusion is that TIPS is as effective as DSRS in preventing variceal rebleeding and may be more cost effective.
The study is unique in that it is based on observed data and not on assumptions, usual for cost effectiveness analyses. This applies both to the incidence of events and to utilities which were those directly perceived from the patients and prospectively collected throughout the study. In fact one major pitfall of cost-effectiveness analyses is that the assumptions underlying statistical models may be not satisfactorily validated, and utilities are rough approximations from other studies often not strictly relevant to the specific setting of interest. These limitations may bring the developed model far away from the real world. As an example, a cost effectiveness analysis of strategies for the prevention of first variceal bleeding in cirrhosis supported beta-blocker therapy independent of the presence of esophageal varices [10], but a subsequent large multicenter trial failed to validate this strategy [11].
The main limitation of the study by Dr. Boyer and colleagues, however, is that TIPS was performed using bare stent [8], since it may be well expected that the use of covered stent (nowadays the standard of use) would much improve the cost effectiveness of TIPS over DSRS. In fact, since the only difference between the two treatments was the number of reinterventions for maintaining the shunt patent, it is conceivable that overcoming this problem would result not only in a marked reduction of the overall treatment cost, but also in an appreciable improvement of the quality of life, thus providing a base for a sustained reduction in the incremental cost effectiveness ratio.
Current experience with TIPS using covered stents strongly supports this view. In a RCT of polytetrafluoroethylene covered (n=39) vs bare (n=41) stent [12], [13], after a mean follow-up of 486 days (range 1–832) 6 vs 18 patients developed stent dysfunction requiring 9 (23%) and 31 (75%) reinterventions, respectively (relative difference 69%); 2-year cumulative rate of primary patency was 76% vs 36% (relative difference 52%). Similar findings have recently been reported from consecutive series with non-parallel controls [14], [15].
The experience at the Mediterranean Institute for Liver Transplantation and Advanced Specialized Therapies (ISMETT, Palermo), with 106 patients who underwent bare stent and 111 covered stent TIPS, further confirms these results (Luca A, unpublished data). After a mean follow-up of 20 months, stent dysfunction (defined as a portal pressure gradient >12mmHg) occurred in 58/106 (55%) with the bare stent and 10/111 (9%) with the covered stent (Fig. 1; p<0.00001). Reinterventions were 104 for the bare stents and 14 for the covered stents (relative difference per patient/year 77%).
Fig. 1.
Cumulative proportion of patients free of TIPS dysfunction according to the type of stent received. Numbers below the abscissa are patients still free of TIPS dysfunction per each observation period. Patients were treated at ISMETT, Palermo (Luca A, unpublished). [This figure appears in colour on the web.]
Therefore, available data consistently show that compared with bare stents, covered stents reduce the per patient dysfunction rate by 52–77%, and the reintervention rate by 69–77% in the first two to three years after TIPS insertion.
It is hard to estimate how much such a reduction of shunt dysfunction and reinterventions may affect the cost effectiveness analysis of TIPS compared with DSRS. Clearly such an analysis should account not only for the reduction of the crude costs for TIPS revision with dilatation or restenting but also for the reduction of hospitalizations and outpatient visits including Doppler ultrasonography. Moreover, the reduction of costs related to the reduction of major clinical events saved by the shunt patency, such as bleeding, ascites and possibly also encephalopathy, should be accounted for. These events were in fact significantly reduced by the covered stent as compared to the bare stent in the only reported RCT [12], [13]: it is also of note that a trend towards a reduction in mortality has been reported [12], [13], [14].
Another major issue for the need of reassessment of cost effectiveness analysis of the covered stent is quality of life. It is conceivable that patients with such a marked reduction of the risk of stent dysfunction and correspondingly of reinterventions would perceive a better quality of life. This would result in an increase in the denominator of the ICER which would be consequently reduced.
Unfortunately, Dr Boyer and colleagues, who must be complimented for their excellent work, did not report details on the sensitivity analysis they performed according to different shunt dysfunction rates, in an attempt to simulate the analysis with covered stents. The difference they reported in per patient re-stent cost when compared to the base case scenario does not seem to account for the factors above mentioned. However, even assuming an overall cost reduction as low as 5% for the covered stent and a utility increment of 0.05, the 5-year estimated ICER reported in table 5 of their article, would change from $ 61,737 to $ 31,527 per QALY (Table 1), well-below the standard acceptability threshold of $50,000. This example is quite a conservative one, based on the actual cost of a single TIPS reintervention with or without covered stent insertion of $9,808 and $5,062, respectively, accounted at ISMETT, Palermo. Clearly, for higher cost reductions or utility increments, ICER would be even more favourable.
Table 1. Effect of a total cost reduction of 5% and utility increase of 0.05 hypothesized for the covered stent on a cost effectiveness analysis of TIPS compared to DSRS
| Strategy | Cost | Incremental cost | Effect QALY | Incremental effect QALY | Cost/effect | ICER |
|---|---|---|---|---|---|---|
| DSRS | $ 66,685 | 2.441 | $ 27,319 | |||
| TIPS bare | $ 84,033 | $ 17,348 | 2.722 | 0.281 | $ 30,872 | $ 61,737 |
| TIPS covered | $ 79,832 | $ 13,147 | 2.858 | 0.417 | $ 27,932 | $ 31,527 |
As outlined by Dr. Boyer and colleagues, a factor that may hamper generalizability of their analysis is the difference of costs assigned to the specific DRGs 191 and 192 (Pancreas, Liver and Shunt procedures with or without complications, respectively) in USA ($ 27,725 and $9,970, respectively) compared to other countries (in Italy $ 19,051 and $11,781, respectively). The use of actual costs from a local cost accounting database to derive a relative difference between TIPS and DSRS to calculate a similar difference in DRG costs may further reduce the generalizability of the results, because actual costs may be a further source of difference. As an example, in two studies based on bare stents [16], [17], overall costs for TIPS placement varied from $ 9560 to $23,789, and at ISMETT, Palermo, the actual cost accounted for covered stent TIPS is currently $12,211 including: covered stent (a mean of 1.1 stent per patient) $ 4746, medical supply $ 3168, staff (radiologist, nurse, technician, anesthesiologist, and aide) $1119, floor admission (mean length of stay 2.5 days) $1143, amortization expense $ 2035.
However, even with the limitations of generalizability and of the use of bare stent, the study of Dr. Boyer and colleagues provides thorough information on the comparison of TIPS with DSRS and a solid rationale supporting the use of TIPS for clinical practice. The characteristics of long-term patency and reduced need for reintervention of the covered stent are a logical premise for reinforcing this conclusion which should be validated by prospective studies of the cost effectiveness of this type of stent.
A final consideration is that we probably have to admit that the time of shunt surgery for portal hypertension is over. However, we guess the medical community involved in portal hypertension will continue to refer to the many lessons learned from the shunt surgery experience for a long time to come.
References
- Conn HO. The rationale evaluation and management of portal hypertension. In: Shaffner F, Sherlock S, Leevy CM, editors. The Liver and its disease. Intercontinental, 1974. p. 289–306.
- Distal splenorenal shunt versus endoscopic sclerotherapy in the prevention of variceal bleeding. A meta-analysis of 4 randomized clinical trials. J Hepatol. 1992;16:338–345
- . AASLD practice guidelines. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology. 2007;46:922–938
- . Evolving consensus in portal hypertension. Report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol. 2005;43:167–176
- Variceal ligation plus nadolol compared with ligation for prophylaxis of variceal rebleeding. A multicenter trial. Hepatology. 2005;41:572–578
- Endoscopic variceal rebleeding: a prospective, randomized trial. Hepatology. 2000;32:461–465
- Spanish cooperative variceal rebleeding study group. Multicenter RCT comparing drug therapy vs the combination of drug therapy+endoscopic band ligation in the prevention of rebleeding in patients with cirrhosis. Hepatology 2006;44:202A.
- Distal splenorenal shunt versus trans-jugular intrahepatic portal systemic shunt for variceal bleeding: a randomized trial. Gastroenterology. 2006;130:1643–1651
- Cost of preventing variceal rebleeding with trans-jugular intrahepatic portal systemic shunt and distal splenorenal shunt. J Hepatol. 2008;48:407–414
- . Endoscopic screening for esophageal varices: is it ever cost-effective?. Hepatology. 2003;37:366–377
- Beta-blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med. 2005;353:2254–2261
- Improved clinical outcome using polytetrafluoroethylene-coated stents for TIPS: results of a randomized study. Gastroenterology. 2004;126:469–475
- Patency of stent covered with polytetrafluoroethylene in patients treated by trans-jugular intrahepatic portosystemic stents: long-term results of a multicenter study. Liver Int. 2007;27:743–747
- Improved clinical outcome with trans-jugular intrahepatic portosystemic stent-shunt utilizing polytetrafluoroethylene-covered stents. Eur J Gastroenterol Hepatol. 2006;18:225–232
- The use of a polytetrafluoroethylene-covered stent graft for trans-jugular intrahepatic portosystemic shunt (TIPS): long-term follow-up of 100 patients. Acta Radiol. 2006;47:660–666
- . Cost-effectiveness analysis of trans-jugular intrahepatic portosystemic shunt (TIPS) versus endoscopic therapy for the prevention of recurrent esophageal variceal bleeding. Hepatology. 2000;31:358–363
- Trans-jugular intrahepatic portosystemic shunting versus paracentesis plus albumin for refractory ascites in cirrhosis. Gastroenterology. 2002;123:1839–1847
☆ The authors declare that they do not have anything to disclose regarding funding from industries or conflict of interest with respect to this manuscript.
PII: S0168-8278(07)00687-3
doi:10.1016/j.jhep.2007.12.009
© 2007 European Association for the Study of the Liver. Published by Elsevier Inc. All rights reserved.
