Can We Add the History of the Nonoperative Therapy of Varices to Other Success Chapters of Modern Medicine?

Short abstract

Watch an interview with the author

Abbreviations

AASLD

American Association for the Study of Liver Diseases

HV

hepatic vein

ISMN

isosorbide mononitrate

NSBB

nonselective beta‐blocker

NTG

Nitroglycerin

SMA

superior mesenteric artery

TIPS

transjugular intrahepatic portosystemic shunt

Mina Niazi

Juan G. Abraldes

Variceal bleeding, one of the dreaded complications of portal hypertension, can be a frightening experience for patient and physician alike, and could mean the end of a patient’s life. Overall survival has increased in the last three decades because of improved treatments. Despite this, the 30‐day mortality rate in the setting of acute bleeding from varices remains as high as 20%.1 It is for this reason that some clinicians may argue that we have not reached the biblical millennium to add the treatment of variceal bleeding to other success chapters of modern medical science, such as tetanus, typhoid fever, poliomyelitis, smallpox, and even hepatitis C—to bring us closer to home.

It is virtually impossible to narrate the history of medical treatment of esophagogastric varices without mentioning the name of the famous surgeon and twice Chair of Surgery, the formidable Charles Gardner Child III, who, parenthetically, insisted on the use of the Roman numeral III (and not the Arabic/Latin 3rd) for the suffix that documented his lineage.2 In the 1950s, management of variceal bleeding was in the hands of the surgeons, which meant hours in the operating room to create various types of portosystemic shunts.3 These unique and inventive surgical approaches were indeed effective in reducing rebleeding in the acute setting, but accelerated the progression of liver parenchymal dysfunction, increased the frequency and severity of encephalopathy, and most importantly, failed to improve overall survival. With experience, surgeons learned that “sicker patients,” later referred to as those in Child‐Turcotte‐Pugh class C, not only had higher intraoperative mortality but dismal overall outcome despite the creation of portosystemic shunts. It was not until 1964, when Charles Child and Jeremiah Turcotte in a chapter in Child’s textbook of surgery4 proposed a classification system (based on the description 3 years earlier by Wantz and Payne5) that was later modified by Pugh and colleagues6 in 1973, and is today referred as the Child‐Turcotte‐Pugh classification.2, 4, 6 Sad to report, but in keeping with the popular and influential aphorism of the Spanish‐American philosopher Jorge Agustín Nicolás Ruiz de Santayana v Borrás (better known as George Santayana), who warned that “those who do not remember the past are condemned to repeat it,”¹ Turcotte’s contribution is repeatedly forgotten and his name omitted from the eponym that is carelessly referred to as “Child‐Pugh” and sometimes even “Child’s Pugh”, as if Pugh were somehow a device with which cirrhosis is assessed. In those bygone days, medical therapy offered only temporizing measures, which comprised gastric ice lavage, balloon tamponade, and in privileged centers, infusion of pituitary extract that was later identified as vasopressin.7

The breakthrough that underpins the modern era of variceal therapy was a better understanding of the pathophysiology of portal hypertension. This time, the investigators ultimately got their theories correct. In the early 1950s, clinicians made the astute observation that patients with cirrhosis had clinical findings similar to that of patients with a hyperdynamic circulation, as seen in beriberi and aortic insufficiency, namely, “warm extremities, a wide pulse pressure, and capillary pulsations in the nail beds,”8 as well as tell‐tale vascular spiders on the skin and elsewhere. This proved to be the case when Henry Kowalski and Walter Abelmann8 demonstrated increased cardiac output and decreased peripheral vascular resistance in patients affected by alcohol‐related liver disease. At that time, the accepted theory for portal hypertension was “backward flow,” meaning a low‐flow state within the splanchnic circulation, which was further supported by low flow at the hepatic hilum. Hence it was postulated that increased intrahepatic resistance was the main cause for the increase in portal pressure.9 This notion followed logically from seminal experimental histopathological and perfusion observations of increased intrahepatic resistance in human cirrhotic liver,10 published in 1928 by a young New Zealand fellow at Mayo Clinic in Rochester, Minnesota, Archibald Hector McIndoe. McIndoe’s explanation for portal hypertension made more sense than the delusional “forward flow” hypothesis of Guido Banti that an enlarged spleen associated with hemolysis and anemia (splenomegalic anemia or Banti’s disease) in patients with cirrhosis was the cause and not the result of the liver disease (i.e., Banti’s syndrome11). However, the backward flow hypothesis was misleading because it did not take into consideration that some, even considerable, portal flow is diverted through the collateral circulation.

In 1972, Roberto Groszmann (Fig. 1),12 through his work with cardiologist Jay Cohn on splanchnic hemodynamics (Fig. 2),13 was able to debunk the dogma of “backward flow” and reintroduce the concept of “forward flow” in a new and verifiable form that set the stage for the use of vasoconstrictors in patients with portal hypertension.

Fig 1.

Young Roberto Groszmann (second right) during his 1968‐1971 days with clean‐shaven Jay Cohn (first left). Reproduced with permission from Hepatology.12 Copyright 2010, American Association for the Study of Liver Diseases.

Fig 2.

Using an indicator dilution technique demonstrated that in patients with alcoholic liver disease the mean transit time and appearance time of ¹³¹I‐albumin from the SMA to the HV was shorter than in normal controls. Based on these findings, the authors suggested for the first time that there was high flow in the mesenteric bed in patients with alcohol‐related liver disease. Reproduced with permission from Gastroenterology.13 Copyright 1972, American Gastroenterological Association.

Later on, Groszmann developed rodent models that allowed a detailed characterization of the pathophysiology of the splanchnic circulation in portal hypertension14, 15 and the sequence of events leading to the “arterial vasodilation hypothesis”: first, vasodilation; second, increased neurohumoral systems activity due to an ineffective circulating systemic volume; and lastly, plasma volume expansion ultimately causing increased cardiac output.8, 9, 16, 17

Sir Keith Peters, former Regius Professor of Physic at the University of Cambridge and head of the School of Clinical Medicine, beautifully illustrated the intricate interplay between basic research and clinical science, and highlighted the contrast between 21st and 20th century medicine,18 as exemplified by the lifework of Roberto Groszmann19:

Today’s catch phrase is from “bench to bedside” and university medical centers are struggling with the promotion of translational research. But, in the relatively recent past, the debt of science to medicine has in fact grown. . . . [There are examples] where astute observations of . . . small groups of patients have been the starting point of voyages of scientific discovery—from, as it were, the bedside to the bench . . . but in doing . . . so, illustrate[s] the synergy between clinical and basic biomedical research and that now provides medical scientists with investigative tools of hitherto unimagined power.18

—Sir Keith Peters, F.R.C.P.

The figure of Didier Lebrec then entered the scene (Fig. 3).

Fig 3.

Young Didier Lebrec (encircled) during his fellowship days in 1973‐1975 with mustachioed Jay Cohn and, now wiser, back in Paris (inset). Courtesy Dr. Didier Lebrec.

During his postdoctoral training alongside Jay Cohn, between the years of 1973 and 1975, Lebrec learned about the hemodynamic interplay between the portal and systemic circulations, which eventually led him to propose beta‐adrenergic blockade as a treatment for portal hypertension.20 In 1981, his landmark trial that was published in the New England Journal of Medicine provided the first evidence that medical treatment using propranolol reduces portal pressure and was effective in reducing the rebleeding risk in cirrhosis,21 thereby making beta‐blocker therapy the first effective medical treatment for the prevention of first variceal bleeding and later rebleeding. Later, meta‐analyses of existing randomized controlled trials suggested that propranolol is a safe and effective means of reducing risk for rebleeding by about 40% and mortality related to variceal bleeding by about a fourth.22 Lebrec proposed that, when giving propranolol, the dose should be increased until a 20% to 25% decrease in heart rate is achieved. When directly asked why, he explains that whereas in cardiology the preferred dose of propranolol is the dose that reduces the heart rate to 50 to 60 beats/min, that would have been difficult for patients with cirrhosis because they have different heart rates than patients without cirrhosis. Later on, confirming the major impact of beta‐blockers in patients after variceal bleeding, Jaime Bosch in Barcelona showed that in patients receiving beta‐blockers who achieved portal pressure reductions of more than 20% or to less than 12 mm Hg, the rates of rebleeding from varices were as low as those achieved by portosystemic shunts.23 Somatostatin24 and later its synthetic analogue octreotide were also shown to have the capacity to decrease splanchnic blood flow, but their short half‐lives limited their use to in‐hospital management of episodes of acute bleeding as alternative agents to vasopressin.

Another avenue for the development of treatment for varices and portal hypertension can be traced to 1972, when pathologist Prithi Bhathal25 in Melbourne discovered contractile cells within the liver. Akin to other scar tissue, the cirrhotic liver could contract and, therefore, relax. Next, Bhatal, working with a different Grossman, demonstrated that nitrates—as nitric oxide donors—could decrease the hepatic resistance to blood flow and thus portal pressure.26 Subsequently, results of human studies from Jaime Bosch’s group in Barcelona showed that a greater reduction in portal pressure could be achieved by combining beta‐blockers and vasodilators, such as nitrates or the alpha‐blocker prazosin; beta‐blockade results in reduced portal blood flow, and vasodilators reduce hepatic vascular resistance.27, 28 Prazosin never entered clinical practice, but the concept of alpha‐blockade set the stage for the use of carvedilol, a mixed alpha₁/beta‐adrenoreceptor blocking agent that could achieve the effect of combining propranolol with prazosin, as first proposed by Peter Hayes in Edinburgh.29 Nitrates fell out of use because of poor tolerance and difficulties with administration, but the concept of providing nitric oxide to the liver led, two decades later, to repurpose statins, which are up‐regulators of nitric oxide synthase, for portal hypertension.30

When beta‐blockers were introduced into clinical practice, surgery had been the only recourse to reduce rebleeding, because endoscopic therapies were in their infancy. Sclerotherapy was equal, if not superior, to portacaval or splenorenal shunts in controlling bleeding but was limited by many complications and up to 2% mortality. The beginnings of sclerotherapy were terrifying. Patients would get prophylactic sclerotherapy followed by compression with a Sengstaken‐Blakemore balloon for a few hours, to allow the sclerosant to act. Fortunately for the patients, this noxious approach fell from grace as variceal ligation therapy emerged and was shown to have fewer procedure‐related complications and lower mortality.31

In 1986, Greg Van Stiegmann’s description of endoscopic variceal band ligation (EBL or EVL) portrays a war zone with the analogy of the varix as the target on a bloody battlefield32: “the banding cylinder surrounds the intended target, drawing it into the chamber just before the loaded device fires to strangulate the varix at the base”33 (Fig. 4).

Fig 4.

(A) Spurting, (B) sclerotherapy, and (C) banding. Reproduced with permission from New England Journal of Medicine.33 Copyright 1994, Massachusetts Medical Society.

The banding device that Greg Van Stiegmann invented34 is only an attachment to the flexible endoscope, and therefore we must briefly review and give due credit to its predecessors and the inventors of this equipment that we are privileged to use on a daily basis. In 1950, George Crile Jr,35 one of the four physicians who founded the Cleveland Clinic on returning from service in World War I, advocated alternative surgical procedures for bleeding varices and described transthoracic ligation of esophageal varices in seven patients. Drawing an analogy with the management of hemorrhoids, as did Van Stiegmann nearly 40 years later,34 Crile suggested that direct obliteration of the varices should result in prolonged or even permanent control of bleeding. He did not claim priority because Boerema36 had not only performed such a procedure in 1949 but also had injected sclerosing agents between the ligated varices. At that time, critics of transthoracic variceal ligation and sclerotherapy saw these manipulations rather as “Band‐Aid therapy”—almost as invasive as surgical shunts and limited by numerous complications. An endoscopic approach to band ligation was not a feasible option because this remarkable gadget was not even at an embryonic stage of development. Indeed, Van Stiegmann’s experience had been with variceal therapy via the rigid esophagoscope and general anesthesia, during a brief training period with John Terblanche in Cape Town, South Africa.37

The era of flexible endoscopic therapy of varices was ushered in in 1957, when Basil Hirschowitz38 engineered the first fiberoptic gastroscope38 and elegantly surpassed the generations of gastrocameras39 that had been used worldwide since their invention by Dr. T. Uji40 at Tokyo University, Japan, (working with Sugiura and Fukami, engineers of the Olympus Company), in the 1950s, a half century since the first intragastric photographs were published in 1898.41

Another major milestone in the field of hepatology occurred when a visionary mentor and student saw the potential of turning a diagnostic tool into a therapeutic technique, as was the case in the development of transjugular intrahepatic portosystemic shunts (TIPSs) in 1968.42 Josef Rösch was a diagnostic radiologist in the Central Military Hospital in Prague, who became well acquainted with the portal circulation after performing more than 1,000 splenoportograms.43 After a year of fellowship under the supervision of Charles Dotter at the University of Oregon Medical School, his career evolved from a diagnostic radiologist to an interventionalist. Whether it was because his mentor’s conviction and/or advice never left Rösch’s imagination, or serendipitously, for, as Pasteur remarked,44 “in the fields of observation, chance favors only the mind which is prepared…” he discovered transjugular portography42, 43 while working as a visiting professor in the University of California, Los Angeles (Fig. 5). Although the concept of TIPSs was sound, it was not yet possible to introduce supporting material within the liver to keep the TIPS track patent until the late 1980s, when the insertion of expandable metallic stents became a clinically feasible therapeutic reality, thanks to the Freiburg group, as ably reviewed by Martin Rössle.45 Lining of the bare metal device with polytetrafluoroethylene, curiously termed “covered stents,” greatly increases the durability of the procedure by avoiding thrombosis and stenosis.46

Fig 5.

TIPS was first created by Josef Rösch in late 1968 using Teflon tubing, which formed the basis of the TIPSs we use today.42, 43 Reproduced with permission from Journal of Vascular and Interventional Radiology.42 Copyright 2015, Society of Interventional Radiology.

It takes a brave soul to have the audacity to puncture the liver for the first time or inject glue into living tissue. In this context, Lunderquist et al.47 first described the use of cyanoacrylate (“crazy”) glue in the intralesional treatment of esophageal and gastric varices, administered via the percutaneous transhepatic route. Nowadays, the endoscopic approach to gluing is the treatment choice for acute gastric variceal bleeding.48

Up to the late 1970s, the treatment of variceal bleeding was in the hands of the surgical team. However, in the last four decades, we have witnessed enormous advances in the management of patients with cirrhosis with bleeding. These were systematized in successive consensus meetings in Baveno, Italy, originally launched in 1990 under the leadership of Roberto de Franchis, the late and much‐missed Andy Burroughs, and Jaime Bosch, at which dedicated delegates suffered on the shore of the magnificent Lake Maggiore. The endurance of Lebrec’s landmark findings that showed that oral treatment with nonselective beta‐blockade decreases portal pressure and bleeding risk is an example of our progress in the care of varices in patients with cirrhosis; such beta‐blockade, or some modification thereof, remains today the bedrock of standard treatment in the prophylaxis of variceal bleeding, both for its first occurrence and secondarily in the prevention of rebleeding. Endoscopic ligation is used as an alternative or in conjunction with beta‐blockade for the secondary prophylaxis of variceal bleeding and to staunch the acutely bleeding varix. Somatostatin analogues, such as octreotide or preferably terlipressin (where it is available), that replaced vasopressin are used in combination with banding in acute bleeding. Yet vasopressin, after falling out of favor, made a reappearance in 2016 in the American Association for the Study of Liver Diseases (AASLD) guidelines for treatment of varices49 (Table 1). Like the American cavalry, TIPS rides to the rescue when bleeding is deemed to be refractory to endoscopic therapy, and it can even serve preemptively in high‐risk patients with advanced cirrhosis.50 The care of varices in portal hypertension is far from canonized as new interventional radiology techniques are introduced, such as balloon‐occluded retrograde transvenous obliteration for the treatment of gastric varices and many others.51, 52, 53 There are ongoing clinical trials in most of the treatments for portal hypertension and varices around the world, which can be found by searching The International Trials Registry Platform of the World Health Organization (https://www.who.int/ictrp/en/); ClinicalTrials.gov (https://clinicaltrials.gov/ct2/home) of the US National Library of Medicine, which is a database of privately and publicly funded clinical studies conducted around the world; and the European Clinical Trials Register (https://www.clinicaltrialsregister.eu).

Table 1.

Pharmacological Treatments Recommended Since 1990 by Baveno Conferences and AASLD Guidelines

Baveno I 1990	Baveno II	Baveno III	Baveno IV	Baveno V /AASLD 2007	Baveno VI 2015	AASLD 2016
Propranolol	Propranolol	Propranolol	Propranolol	Propranolol	Propranolol	Propranolol
Nadolol	Nadolol	Nadolol	Nadolol	Nadolol	Nadolol	Nadolol
	ISMN
				NSBBs+ISMN
				Carvedilol	Carvedilol	Carvedilol
Vasopressin+NTG	Vasopressin+NTG	Vasopressin+NTG				Vasopressin+NTG
Terlipressin	Terlipressin	Terlipressin	Terlipressin	Terlipressin	Terlipressin	Terlipressin
Somatostatin	Somatostatin	Somatostatin	Somatostatin	Somatostatin	Somatostatin	Somatostatin
		Octreotide	Octreotide	Octreotide	Octreotide
			Vapreotide	Vapreotide

It remains to be seen what consensus and identification of unmet needs the next Baveno Conference VII has in store for us (Fig. 6), not to mention the camaraderie and gastronomic delights.

Fig 6.

Banner of Baveno VII consensus workshop (2020). Courtesy Baveno VII Consensus Workshop.

Series Editor’s Postscript

In keeping with the multiethnic, multicultural makeup of AASLD, we now welcome two additional authors from Canada. However, this description belies their true origins. The señor, i.e., senior, author has pilgrimed from his Iberian birthplace of Santiago de Compostela, at the end of El Camino, in his native Galicia, by way of Barcelona in Catalunya to Edmonton, Alberta, in North America. In contrast, his coauthor’s treks that she describes as “a rollercoaster” have taken her to Saskatoon in Saskatchewan from her native Kabul in Afghanistan, via Moscow in Russia, Edmonton, Alberta, to London (that is still) in the United Kingdom, and back to Saskatoon—phew.

Their sweeping historical account of the evolution of nonsurgical therapy of varices in portal hypertension is simultaneously a scientific treatise and lyrical prose. Not bad for non‐native anglophones. Their essay leads us though the evolution of these therapies that departed from the bewildering array of surgical diversions of blood from the portal to the systemic circulations, once an understanding developed of the pathophysiology of portal hypertension. Thus, the current contribution serves as a link between the ongoing elucidation of the pathogenesis of portal hypertension, so ably described by Jaime Bosch, the pictorial history of the aforementioned surgery reported by Michael Henderson—both elsewhere in this series—and the mini‐invasive techniques of endoscopic variceal obliteration and transvenous decompression of portal hypertension, the latter by ever‐resourceful interventional radiologists who will join us in the future with their own essays, and even more acronyms.