A history of the treatment of primary liver cancer

INTRODUCTION

It took many centuries, perhaps a dozen or more, for the 3-millenium-old practice of Hepatoscopy [also known as Haruspicy and Hepatomancy], ie, an inspection of the liver for the purpose of divining the intentions of the gods and predicting the future^1–3 (Figure 1), to be transformed into Hepatology, namely the study of liver structure, function, and disease. Presumably, the Haruspex—a priest who practiced haruspicy—then retrained as a hepatologist.

FIGURE 1.

Adjuncts to the practice of hepatoscopy. (A) Clay liver models found in 1935–36 by the French archeologist André Parrot, at Mari, Syria, dated to the 19th and 18th centuries BCE. (B) Inscribed Babylonian clay model of sheep liver ~1600 BCE, showing peg holes for recording the findings seen on inspection of the sacrificed animal’s liver. Courtesy—the Trustees of the British Museum, London. (C) Etruscan life-size bronze model of sheep’s liver ~100 BCE (left) with explanatory inscriptions (right), found in Piacenza. Courtesy—Musei Civici di Palazzo, Florence.

Of course, hepatologists of yore were hampered in their ability to make accurate anatomical and pathological diagnoses, by the limited tools at their disposal, which were initially only visual examination, olfaction, palpation, dissection, and later inspection with a hand lens, and ultimately microscopy enhanced by tissue staining. Little wonder that there was agonizingly slow reliable confirmation of the occurrence of primary cancer of the liver. Along the way, investigators were also stymied by many obstacles to the elucidation of the nature of primary liver malignancies, including the difficult distinction between benign inflammatory nodular entities, like tuberculosis, and all forms of cirrhosis. Indeed, the evocative term tubercle was often applied to any small protuberant rounded lesion distinct from its surroundings. Even when hepatic histopathology was available, there were competing microscopic classifications, and a corresponding lexicon to name them, which was a Tower of Babel of semantic confusion. This perplexity was only made worse when a pathologist as eminent as Rudolf Virchow (Figure 2A) declared confidently that “organs commonly affected by metastases are rarely the site of primary neoplasia.”⁵

FIGURE 2.

(A) Rudolf Ludwig Karl Virchow by Sir Leslie Ward, chromolithograph, published in Vanity Fair, May 25, 1893, NPG D44648 © National Portrait Gallery, London https://creativecommons.org/licenses/by-nc-nd/3.0/. (B) Aretaeus of Cappadocia. Source: Wikipedia. (C) Friedrich Theodor von Frerichs, Lithography by P. Rohrbach, 1859, from a photograph by G. Schauer. Source: Wikipedia. (D) Galen, Reprinted from Cheselden W: Frontispiece. Osteographia, or the Anatomy of the Bones. London 1733.⁴ (E) Paulus Aeginata, reprinted from Chriba M, Levell NJ. The bridge from west to east: Paulus Aeginata (625–690 AD).

A comprehensive history of the discovery of HCC—the preferred term nowadays for primary liver cancer that also includes cholangiocarcinoma and other less frequent tumors—has recently been consummately reviewed in this series by Yamini Natarajan, MD, and renowned digestive disease epidemiologist Hashem B El-Serag, MD, MPH.⁶ Other less memorable terms for malignant tumors of hepatocytes included cancer trabeculaire, adenocarcinoma, carcinoma hepatocellulaire, malignant adenoma, and hepatoma that was applied to benign, malignant, and even intermediate lesions.

Aside from the challenge of devising therapy for malignancies in general, it is evident that knowledge of the etiology and pathogenesis of HCC, determination of its stage, awareness of demographics, and acknowledgment of both related and independent comorbidities (such as diabetes mellitus, chronic viral hepatitis, and the number and severity of complications of cirrhosis), are adjuvant keys to successful treatment. Many of these variables are used in the various published HCC treatment decision trees, including those generated (Figure 3A)⁷ and regularly updated (Figure 3B)^8,9 by the celebrated Barcelona Clinic Liver Cancer (BCLC) group—discussed further below.

FIGURE 3.

(A) First BCLC staging classification and treatment schedule. Four stages are considered. Stage A: Patients with early HCC are suitable for radical therapies (resection, liver transplantation, or percutaneous treatments); stage B: patients with intermediate HCC may benefit from chemoembolization; stage C: patients with advanced HCC may receive new agents in the setting of Randomized Controlled Trials; stage D: patients with end-stage disease will receive symptomatic treatment. Reprinted with permission from Llovet et al.⁷ (B) BCLC staging and treatment strategy in 2022. The prognosis is established in accordance with the 5 stages that are linked to first-line treatment recommendations. No longer divided according to curative or palliative, but stressing that recommended treatments improve survival. The expected outcome is expressed as the median survival of each tumor stage according to the available scientific evidence. Individualized clinical decision-making, according to the data available on November 15, 2021, is defined by teams responsible for integrating all available data with the individual patient’s medical profile. Liver function is evaluated beyond the conventional Child-Turcotte-Pugh staging. Reprinted with permission from Reig et al.⁸ Abbreviations: AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; BSC, best supportive care; CLT, cadaveric liver transplantation; ECOG-PS, Eastern Cooperative Oncology Group-performance status; LDLT, living donor liver transplantation; LT, liver transplantation; MELD, Model for End-Stage Liver Disease; PEI, percutaneous ethanol injection; PS, performance status; PST, performance status test; RF, risk factor; TAE, transarterial embolization; TACE, transarterial chemoembolization.

According to Natarajan and El-Serag,⁶ the Egyptian authors of the Ebers papyrus¹⁰ (1500 BCE, Figure 4A), as well as the originators of the equally venerable Edwin Smith surgical papyrus (3000 BCE),¹³ suspected the existence of liver cancer even though its differentiation from inflammatory swellings and chronic ulcerations was not achievable then. Naturally, it was in the writings (Figure 4B)¹¹ of Hippocrates (460–375 BCE), a contemporary of Socrates (469–399 BCE), and Plato (427–347 BCE) that the term Karkinoma was coined (derived from the Greek word karkinos for crab, καρκίνος)¹², which was latinized centuries later by the encyclopedist Galen to carcinoma, representing the malignant form of carcinos, a tumor. Confusingly, in the Hippocratic vernacular carcinoma apertus denoted malignant ulcers, while carcinoma occlusus referred to deep-seated or occult tumors. A less colorful term Phuma, which refers to any concentrated swelling in the body, is sometimes encountered but only rarely. As we have seen, technically Cancer meant a nonhealing malignant ulcer, similar to the 17th-century English words for certain ulcers, like canker and chancre. In contrast, scirrhus, a hard malignant tumor, was different from carcinos and carcinoma,¹⁴ in fact, commonly, it would have probably been a carcinoma occlusus.

FIGURE 4.

(A) The Edwin Smith Papyrus.¹⁰ (B) To the writings of Hippocrates, printed in Venice in 1588.¹¹ (C) The frontispiece to the collected writings of Galen, printed in Venice in 1625.¹² (D) Title page of Pauli Aeginetae Medici Opera, Apud Gulielmum Rovillium 1589. Source: Wellcome Library, London. (E) The title page from the second volume of Friedrich Theodor von Frerichs’s Clinical Treatise on Diseases of the Liver, in an English translation by Charles Murchison, MD, FRCP, published by the New Sydenham Society in 1861.

But why a crab? It is said that growing tumors allegedly reminded Hippocrates of a moving crab, resulting as we have seen in carcinos for tumor, and carcinoma for a malignant tumor. It was either Galen (second CE, Figure 2D ) in his writings (Figure 4C) or seventh century Paulus of Ægina (Figure 2E)¹⁵ who observed (Figure 4D) “…it appears at length with turgid veins shooting out from it, so as to resemble the figure of a crab; or as others say, because like a crab, where once it has got, it is scarcely possible to drive it away.” According to the Theory of Atra Bile favored by Galen and later promoted by Paulus Æginata, an excess of black bile (atra bilis, melancholos, or choler niger), particularly thick black bile, caused severe ulcerated and incurable cancer^16,17 unlike thin bile that causes nonulcerated curable cancer. And before Galen’s teacher, Aglancon, thought that when the liver was too warm it generated atrabilious blood, ie, blood that was decomposed and not normal, and which overwhelmed the spleen.¹⁵ Cancer aside, melancholia, one of the disorders of the humors (ie, dyscrasias, from Ancient Greek dyskrasis “δυσκρασισ,” meaning “a bad mixture,” is also due to black bile, which comes from the spleen. The popular but erroneous lay notion (as mentioned in the nonspecialist account of cancer “The emperor of all maladies: a biography of cancer” by Siddhartha Mukherjee, published in 2010 by Charles Scribner and Sons, New York) and the professional idea that the Hippocratics thought that black bile causes cancer, may have arisen by innuendo from Paul of Aegina’s reading of Galen. Ovid’s poetic description,¹⁸ quoted by the great English lexicographer Dr. Samuel Johnson, is perhaps the most graphically startling, “As when a cancer on the body feeds, And gradual death from limb to limb proceeds; So does the chillness to each vital part, Spread by degrees, and creeps into the heart” (Figure 5). Incidentally, repeated failure to confirm the existence of black bile once anatomical dissection was permitted during the Renaissance proved fatal to the Atra Bile Theory of cancer pathogenesis.

FIGURE 5.

(A) Sixty-five-year-old man with otherwise uncomplicated alcohol-associated cirrhosis underwent echocardiography during the investigation of recent onset breathlessness. (A) The transthoracic (left) and transesophageal (right) echocardiograms showed a mass in the right atrium, which seemed to be arising from the inferior vena cava. (B) A combined abdominal and chest CT scan confirmed the presence of an intracardiac mass originating from a HCC in the liver that had invaded the vena cava and, as Ovid observed,¹⁸ “creeps into the heart.” (C) A biopsy of the cardiac mass confirmed the presence of HCC (inset square) that had developed within the cirrhotic liver. Images courtesy—Adrian Reuben, series editor, from his personal collection.

Arguably, the first clinically recognizable description of liver cancer was reported in the second-century CE¹⁹ by the Alexandrian physician Aretæus of Cappadocia (Figure 2B) (or Capadocia [Turkish: Kapadokya], the land of fairy chimneys, is a historical region in Central Anatolia, ie, Turkey. It is largely in the provinces of Nevşehir, Kayseri, Aksaray, Kirşehir, Sivas, and Niğde). Aretæus, whose exact dates are unknown even though he was a contemporary of Galen, described patients with livers swollen by inflammation, which transformed to a hardened state, eventually becoming scirrhus, ie, cancerous. This malignant transition then heralded early death because of a deterioration in liver function, which we now appreciate as decompensation of cirrhosis or “acute-on-chronic liver failure” (ACLF).²⁰ This dismal portrait of the disease persisted during the centuries, as related in the 19th-century German Textbook of Liver and Biliary Diseases (Figure 4E) by Friedrich Theodor von Frerichs (Figure 2C). However, von Frerichs did acknowledge that some patients may present with indolent disease that would seem to preclude malignancy, despite vascular invasion and extrahepatic spread at necropsy.

Primary liver cancer was considered almost nonexistent in the West so that when Charles Berman, a physician working in South Africa published, over an interval approaching a quarter of a century, a large series of monographs wholly devoted to Primary Carcinoma of the Liver,^21–28 principally among the Bantu, these publications were seen as amounting to a unique collection of all the available knowledge. Berman provided information covering the descriptions published by physicians spanning the years from Aretæus to Morgagni, while extensively describing clinical scenarios and showing pathological findings. All the clinical cases were based on his personal experience with Bantu patients who worked in the Witwatersrand gold mines and were seen at the Johannesburg Non-European Hospital. Michael Kew (1939–2021) MB BCh, FCP(SA), MD, PhD, DSc, FRCP, (Figure 6A) single-mindedly pursued the study of all aspects of liver cancer in South Africa. Among the trove of his clinical and translational scientific achievements—upward of 500 publications (https://scholar.google.co.za/citations?user=ONGrV0QAAAAJ&hl=en accessed 2/14/2023)—he is perhaps most noted for his key role in establishing the clinical and molecular association between HBV infection and HCC, in a series of epidemiological, clinical, histological, virological, and molecular studies.^29,30 Indeed, he was involved in the seminal description of p53 mutations in HCC,³¹ thereby pioneering the field of molecular profiling. Later on, this was pursued by Snorri Thorgeirsson, Jessica Zucman Rossi, Josep Llovet, Xin We Wang, and JC Nault among others, but it is still in need of robust validation for use in clinical decision-making. Mike Kew remained throughout a proud citizen of South Africa, which included his appointment as Personal Physician to the President (Figure 6B), but he also exported knowledge in the field through his mentees Morris Sherman (Figure 7A), who migrated to Toronto, Canada, where he developed a career in liver cancer and was a leader of major international clinical practice guidelines for liver cancer; Geoff Dusheiko³³ and Adrian Di Bisceglie,³² who relocated to London, UK, and St Louis, MO, respectively (Figure 7B, C).

FIGURE 6.

Michael Charles Kew (1939–2021). (A) Consummate clinician, hepatologist, and hepatology investigator, reprinted from Thomson et al.²⁹ https://creativecommons.org/licenses/by-nc/4.0/. (B) Personal Physician (1996–2010) to Nelson Mandela, President of South Africa, reprinted with permission from Kassianides et al.³⁰

FIGURE 7.

(A) Morris Sherman MB BCh, PhD, FRCP(C), reproduced from the inaugural Toronto Liver Cancer Symposium, at which tribute was paid to the “truly phenomenal” career of Dr. Morris Sherman. Credit: UHC News. (B) Geoffrey M. Dusheiko MB BCh, FCP(SA), FRCP, FRCP (Edin), courtesy Geoffrey M. Dusheiko. (C) Adrian Di Bisceglie MB BCh, FAASLD, reprinted with permission from Hoofnagle.³²

It goes without saying that end-stage clinical status and an absence of effective treatment were common all over the world until it was feasible to diagnose the tumor at a stage before the patient was close to death. Diagnosis by physical examination, sometimes supplemented by obtaining a skinny needle biopsy, was not followed by any effective therapy. Indeed, the biopsy results could be received when the patients had already died, as tissue processing could take its time and, in some instances, may be performed at a distant location. Pathological guidelines for diagnosis and grading were established by Edmonson and Steiner,³⁴ while modern criteria were prepared by a group of international pathologists led by Masamichi Kojiro³⁵ (Figure 8A). Angiography was a conventional diagnostic tool until the development of ultrasonography, CT, and MRI. These latter noninvasive visualization techniques displaced angiography and scintigraphy for diagnosis and staging, and allowed diagnoses to be brought forward to an asymptomatic phase and early stage so that curative treatment resulting in long-term disease-free survival became feasible. The discovery of α-fetoprotein by Gari I Abelev (1928–2013)^36,39 in 1963 in Moscow in the former Soviet Union (Figure 8B), provided a serum tumor marker for HCC⁴⁰ that offered the potential for screening protocols using regular measurements, as was done in China⁴¹ and Alaska,⁴² where the high prevalence of HBV infection in the population was linked to a high liver cancer incidence. Indeed, the high prevalence of liver cancer in Asia and sub-Saharan Africa (Figure 9)⁴³ together with the low prevalence reported in Europe and the Americas (likely due to underdiagnosis) corroborated the erroneous idea that HCC was not a disease that should occasion any attention in the West. Vaccination against HBV translates into a sharp decrease in HCC incidence in children and adults as shown in studies in Taiwan led by Din-Shin Chen^44,45 and the same impact is expected by the Gambia project established by Ruggero Montesano.⁴⁶

FIGURE 8.

(A) Professor Masamichi Kojiro with the author, from the author’s personal collection (Kurume, 2005). Dressed up as required after many years of friendly relationships with Japanese leaders, which started in 1990. (B) Gari Izrailevich Abelev. Reprinted with permission from Engelhardt et al.³⁶ (C and D) Kunio Okuda and his pastimes, reprinted with permission from Reuben et al.³⁷ (E) Claude Couinaud working with his collection of “liver casts” at the school of medicine in Paris 1988. Reprinted with permission from Felekouras et al.³⁸ (F) The somewhat flamboyant gravestone of Claude Couinard’s grave, which illustrates his lifelong and seemingly posthumous devotion to the segmental anatomy of the liver. Sad to report that when Claude Couinard died in 2008, he was alone, away from his family. In his honor, anatomists from Tunisia, including Jacques Belghiti, a native of Morocco (who is seen in this photograph) organized the creation and subsequent installation in 2018 of the gravestone that celebrates Couinard’s contribution to our knowledge of liver surgical anatomy, and particularly its segmental arrangement. Photograph courtesy of Jacques Belghiti.

FIGURE 9.

The most prevalent causes of HCC and their percent contributions in causing HCC in different countries in northern Africa and the Middle East. The countries in (1) North Africa were: Egypt, Tunisia, Morocco, Algeria, and Somalia, and in (2) the Middle East were: Iran, Lebanon, Turkey, Yemen, Saudi Arabia, and Pakistan. Reprinted from Alavian, et al.⁴³ https://creativecommons.org/licenses/by-nc/4.0/.

Trials in China explored the potential of early detection through ultrasound and α-fetoprotein determination.⁴¹ The wide use of ultrasound in Asia, more intensely in Japan⁴⁷ and Taiwan, allowed HCC detection at an early stage. Such capability was also shown in Europe by studies from Massimo Colombo in Milano⁴⁸ and Luigi Bolondi in Bologna.⁴⁹ The ability to diagnose liver cancer at an early stage emphasized the need to stratify the patients according to stage and prognosis, in order to discern if any intervention might be of benefit, compared to the natural history of the disease (ie, no intervention). The 1975 study, by Olweny et al,⁵⁰ testing the effect of Adriamycin on the survival of patients with advanced disease provided suggestively positive results and thereby gave credence to the Kampala staging system⁵¹ (Table 1), based on a cohort with a median survival of less than 2 months. Kampala staging was later replaced by several other schemes with slightly more accurate clinical profiling. Cochrane et al⁵² developed a western scoring system (Table 2) that included 7 variables aimed at identifying those patients among a cohort of 18 individuals who would survive for more than 14 weeks and thus be considered for systemic treatment. The Okuda staging system⁵³ (Table 3) was proposed later by the Japanese Liver Cancer Study Group and was widely used when the treatment of liver cancer gained more interest. Kunio Okuda (Figure 8C and D) was a worldwide key opinion leader based in Chiba (Japan), whose major research and educational achievements facilitated international progress in the field.³⁷ Some additional scoring or stratifying proposals were put forward but ultimately the emergence of the BCLC model^7–9 (Figure 4A) became the backbone of clinical decision-making that was endorsed by several scientific association practice guidelines, research consortia, and reviews. The BCLC model has been regularly updated to include the advances in knowledge with the necessary background evidence; the last update was released in 2022⁸ (Figure 4B).

TABLE 1.

Kampala staging classification stratifying patients with liver cancer into 3 categories according to expected survival

Stage Ia	No ascites, weight loss, or portal hypertension (ie, no abdominal collaterals), bilirubin<2.0 mg/dL
Stage IIa	Ascites and/or moderate (25%) weight loss, no portal hypertension, bilirubin <2.0 mg/dL
Stage IIIa	Severe weight loss, portal hypertension, bilirubin >2 mg/dL

^aIA, IIA, and IIIA signify appropriate stages with associated distant metastatic disease.

Reprinted with permission from Primack et al.⁵¹

TABLE 2.

The Cochrane colleagues scoring system to stratify patients according to prognosis

Variable	Parameter	Score	Parameter	Score
Age (y)	<50	10.0	>50	2.5
Duration of symptoms (mo)	>3	5.8	<3	3.3
Weight loss (%)	<15	6.7	>15	3.3
Ascites	Absent	7.3	Present	1.4
Cirrhosis	Absent	10.0	Present	2.5
AST	<2x normal	6.3	≥2x normal	4.0
Bilirubin	<2x normal	6.0	≥2x normal	0.0

Note: Interpretation: Patients with a score >35 were classified as Grade A and those with a score ≤35 as Grade B, predicting survivals longer or shorter than 14 weeks, respectively.

Abbreviation: AST, aspartate transaminase.

Reprinted with permission from Cochrane et al.⁵²

TABLE 3.

The Okuda staging system for HCC

	Points
Variables	0	1
Tumor size	<50% of liver volume	>50% of liver volume
Ascites	Absent	Present
Albumin (g/dL)	>3	<3
Bilirubin (mg/dL)	<3	>3

Note: Interpretation: The sum of points according to the profile of four variables is used to stratify patients into 3 stages, namely stage I: 0 points, stage II: 1–2 points, stage III: 3–4 points, which are that would associate with clearly distinct life expectancies.

Reprinted with permission from Okuda et al.⁵³

The following sections summarize the development of each of the effective treatment options while highlighting the landmark contributions that paved the way for our current knowledge. Ancient practitioners, especially the Greco-Roman physicians, formulated an imposing spectrum of therapies, which included topical applications—albeit caustic ones, like calcined verdigris (copper sulfate), salts of lead and sulfur, and so-called “Egyptian ointment,” ie, arsenic paste, and less noxious concoctions like boiled cabbage and a salted mixture of honey and egg white or fig. Hippocrates wrote that tumors that are not cured by medicine are cured by iron (the knife), those that are not cured by iron are cured by fire (cautery), and those that are not cured by fire are incurable.¹¹ It was even realized that therapy was most likely to be successful at the precursor or first stage of malignancy, known as cacoethes^14,15 from the Latin cacoēthes, which was borrowed in turn from the Greek, kakóēthes, ie, κακοήθης—its many meanings include malignant or literally, of “bad character.”¹⁵

SURGICAL TREATMENT

Tumor resection was once the mainstay and the sole available treatment that could offer a long-term cure. However, the desperate situation of patients presenting with advanced disease prompted a less restrictive selection of candidates. This resulted in major morbidity and postoperative mortality due to liver failure. Understaging because of suboptimal accuracy of imaging techniques was common and the presence of advanced liver disease with severe portal hypertension and even clinical decompensation (ascites and overt jaundice) resulted in very poor outcomes, and also in high recurrence rates in those who would survive the early follow-up period. Hippocrates also cautioned against operating on carcinoma occlusus because such surgery was likely to make the patients worse and hasten their demise.¹⁵ Although Galen recommended leaving a wide margin of healthy tissue when removing cancer, he believed, notwithstanding, that the best surgeon was the one who only operated as a last resort. Perhaps Galen’s affirmation of the Roman prejudice against surgery was underpinned to some extent by his experiences as a surgeon to the gladiators. In the setting of surgery for advanced cancer, Pliny the Elder (23–79 CE), in his Materia Medica (a largely plant-based pharmacopeia), recommended various potions of herbs and other remedies for internal use before and after operations, the most highly praised being a brew of boiled ash of sea crabs, egg white, honey, and powdered feces of falcons.^12,16 After the fall of the Western Roman Empire, the baton of medicine passed to physicians of The Islamic Civilization, as deftly related in the forthcoming essay in this series, on the liver in Islam by Rifaat Safadi.⁵⁴ Arabic and Persian surgeons were similarly wary of operating on cancer, a reluctance that persisted for many centuries.¹²

Results of several studies clearly established the risk of postoperative morbidity and death related to impaired liver function due to underlying cirrhosis,⁵⁵ and the prognostic relevance of so-called clinically significant portal hypertension, ie, a portal pressure of >10 mm Hg⁵⁵ (Figure 10A, B) usually determined by the Hepatic Venous Pressure Gradient method.^55–60 At the same time, it was repeatedly shown that resection of multifocal disease or when vascular invasion was already present was associated with a very high risk of recurrence.⁶¹ All these observations stimulated the progressive refinement of candidate selection and emphasized the need to avoid major removal of liver tissue in order to spare as much nontumoral liver as possible. Incorporation of the hepatic anatomy according to segments as defined by the Parisian anatomist and surgeon Claude Couinaud (1922–2008),³⁸ from Neuilly-sur-Seine (Figure 8E), gave the rationale for anatomical resection⁶² (Figure 8F).

FIGURE 10.

The outcomes of hepatic resection for HCC, in patients with cirrhosis stratified according to portal pressure determined by HVPG measurement. (A) Results in 18 patients who had uneventful perioperative outcomes, including 5 individuals with HVPG ≥10 mm Hg, and in 11 patients who all had HVPG ≥10 mm Hg and all decompensated. There were no deaths in this 29-patient cohort. Reprinted with permission from Bruix et al.⁵⁶ (B) Long-term postoperative survival (%) in 77 patients, stratified by HVPG and serum bilirubin. Five-year cumulative survivals were (1) 74% in 35 patients with HVPG <10 mm Hg, (2) 50% in 15 patients with HVPG ≥10 mm Hg and normal serum bilirubin, (3) only 25% in 27 patients with HVPG ≥10 mm Hg and bilirubin ≥1 mg/dL. Reprinted with permission from Llovet et al.⁵⁷

Successful surgical achievement was impressive in countries such as China under the leadership of the Shanghai University group led by Zhao-You Tang⁶³ and the Shanghai East Hepatobiliary Surgery Hospital led by Meng-Chao Wu.⁶⁴ Major contributions such as using intraoperative ultrasonography to guide segmentectomy and a very clinically oriented and still useful decision tree (Figure 11) to decide the magnitude of resection based on liver function^65,66 were made by the innovative Masatoshi Makuuchi (Figure 12A) who also developed intraoperative ultrasound to guide resection. Currently, postoperative HCC resection deaths should remain below 1%, the risk of major operative bleeding should be <10%, and 5-year survival rates could be as high as 70%, even though successful tumor resection may still be hampered long-term because of tumor recurrence. No effective intervention existed to reduce this recurrence risk until the positive results achieved by the combination of atezolizumab and bevacizumab.⁷¹ So major hope is now invested in further progress with the new immune-oncology agents. Laparoscopic⁷² and robotic⁷³ resections allow less invasive interventions with less morbidity and shorter hospital stays.

FIGURE 11.

Flowchart to select patients for surgical resection and extent of liver resection according to the presence of ascites and its response to treatment, serum bilirubin (mg/dL), and ICG retention (%) at 15 minutes as proposed by Professor Masatoshi Makuuchi.^65,66 Abbreviation: ICG, indocyanine green.

FIGURE 12.

(A) Professor Masatoshi Makuuchi, Reprinted from Xu et al.⁶⁷ https://creativecommons.org/licenses/by-nc-nd/4.0/. (B) Professor Leslie H Blumgart (1931–2022). Reprinted with permission from Bhattacharya et al.⁶⁸ (C) Professor Henri Bismuth (b 1934). Reprinted with permission from He.⁶⁹ (D) Professor Jacques Belghiti; recipient of the 2015 International Hepato-Pancreato-Biliary Association Distinguished Service Medal, among a host of other awards. Courtesy—Jacques Belghiti. (E) Professor Thomas E. Starzl (1926–2017). Reprinted with permission from Rakela et al.⁷⁰ (F) Professor Sir Roy Y. Calne (1930–2024), on the occasion of the naming of the specialist transplant unit at Addenbrooke’s Hospital as “The Roy Calne Transplant Unit,” in 2021. Picture courtesy of Cambridge University Hospitals NHS Foundation Trust. (G) A somewhat menacing Mafia-style photograph of the author with Professors Michel Beaugrand, Olivier Seror, and Jean-Charles Nault, from the author’s personal collection (Saint Victor des Oules, 2019). At a variation of a multispeciality tumor board dinner at Michel Beaugrand’s place, there was ample time to debate the past, present, and future of liver cancer research, and offers that could not be refused!!. (H) The author with the members of the Founding Governing Board of the International Liver Cancer Association (ILCA, Milan, 2009), from the author’s personal collection. Front row from left to right: Masamichi Kojiro (Japan), Nelson Fausto (USA), and Jordi Bruix (Spain); Back row from left to right: Ricardo Lencioni (Italy), Pierre A Clavien (Switzerland), Vincenzo Mazzaferro (Italy), Snorri Thorgeirsson (USA), Josep Llovet (Spain and USA), Peter Galle (Germany), Sheng-Long Ye (China), and Kwang-Hyub Han (South Korea). Overdue appropriate gender balance was initiated immediately afterward, culminating in the appointment of Jessica Zucman-Rossi (France) as President, in 2022.

While several of the giants of liver surgery feature in this account of the care of patients with HCC,⁷⁴ it would be remiss to omit mention of Les Blumgart and Henri Bismuth. Professor Leslie Harold Blumgart BDS, MB BCh, MD, (1931–2022 and Figure 12B) migrated peripatetically from his upbringing and dental education in Johannesburg, South Africa, to New York City, via Sheffield, Nottingham, and London in England, Cardiff in Wales, Glasgow in Scotland, and Bern in Switzerland. He pioneered many advances in abdominal surgery, with a particular focus on patients with metastatic cancer, and biliary and pancreatic diseases. Professor Henri Bismuth, MD (Figure 12C), born in 1934 in Tunisia, exerted a major influence on hepatobiliary surgery in Europe and indeed worldwide from his Centre Hépato-Biliaire at the Hôpital Paul-Brousse in Villejuif—for which he was recognized with a slew of honorary degrees worldwide. At Paul-Brousse, Bismuth was the first to perform liver transplantation (LT) in France, and there he also devised strategies for augmenting the hepatic parenchyma to better withstand resection. In this surgical pantheon, we must include a second French surgeon with North African roots, namely Professor Jacques Belghiti (Figure 12D) from Beaujon, who exercised a major world leadership role, and, along with Masatoshi Makuuchi⁶⁷ (Figure 12A), helped to bring East and West together, negating Rudyard Kipling’s assertion, in his “Ballard of East and West,” that “ …never the twain shall meet.” Among other awards that Jacques received was the “Legion d’Honeur.”⁷⁵

A transformative development in surgery for liver cancer was brought about by the transition of LT from an experimental procedure into routine practice. Groundbreaking surgical innovation and the pivotal development and introduction of safe effective immunosuppressives were achieved in the United States by Tom Starzl MD, PhD,⁷⁰ (Figure 12E) and by Sir Roy Yorke Calne FRCP, FRCS, FRS, in the United Kingdom (who died in January 2024, aged 93) (Figure 12F).^76,77 All the available experience then suggested that if patients could only be diagnosed and transplanted at an early stage, the outcomes could be highly competitive with other therapies.⁷⁸ Vincenzo Mazzaferro et al⁷⁹ published their seminal study in 1996 and the resultant “Milan criteria” for the selection of patients with HCC for LT, became the established clinical guidelines to allow survival beyond 80% at 5 years with a recurrence rate of approximately 15%. This favorable result turned out to be driven mainly by the absence of microvascular invasion detected at explant pathology. Several proposals to expand the HCC limits for LT have been raised while still maintaining a relative increase in size. In addition, downstaging proposals to bring patients within Milan criteria (the condition used by the US United Network for Organ Sharing (UNOS) to define optimal candidacy) have increased the potential for LT for patients who would otherwise be excluded by adhering to strict Milan limits. Results of studies initiated by Francis Yao et al in San Francisco, CA,⁸⁰ have consolidated a more permissive approach, but extensive prospective validation and the establishment of robust imaging criteria to be applied at the time of listing, as well as rules for removing patients from the LT waitlist, are keenly awaited.⁸¹

A major limitation to an optimal outcome after LT historically was the presence of underlying infection with HBV or HCV. Didier Samuel et al⁸² demonstrated that successful control of HBV replication could prevent its negative impact on the new organ. This therapy meant that it was now no longer necessary to exclude patients infected with HBV from listing for LT, thereby allowing this significant subset of patients to benefit from LT. HCV infection was also a major condition that resulted in organ loss and the need for repeat LT. Effective antiviral agents have resolved this contraindication⁸³; currently postoperative viral hepatitis infection of the donor organ is no longer a major concern, and even HCV-positive donor organs are being used without adverse outcomes.⁸⁴ Shortage of LT donors is still a major limitation in the West, and this has prompted the development of living donor liver transplantation programs that do provide similar outcomes to cadaveric donation,⁸⁵ but this adds the consideration of balancing the benefit for the recipient with the risk to the donor. The feasibility and success of living donor liver transplantation is the rule in most of the Far East, as is well-illustrated by results achieved in Turkey,⁸⁶ Canada,⁸⁷ USA,⁸⁸ and Japan.⁸⁹ The Kyoto group, led by Koichi Tanaka, has played a leading role in implementing living donor liver transplantation for HCC.⁹⁰

Overall, it is amazing to realize how much the field of LT, which Tom Starzl likened to “. a desperate appliance…” (paraphrasing Claudius in Shakespeare’s Hamlet), has advanced to a really effective, well-considered, evidence-based approach to HCC, and one that results in high survival rates and good quality of life.

ABLATION

Tumor ablation through the injection of alcohol, acetic acid, or boiling saline, or via heating or freezing by radiofrequency, microwave, laser, cryotherapy, or radiotherapy provides potentially curative treatment for those patients who would otherwise not be fit for conventional surgery. Studies using tumoral ethanol injection under ultrasound guidance were pioneered in the late 1980s, by Masaaki Ebara⁹¹ and Shuichiro Shiina⁹² in Japan and Tito Livraghi⁹³ in Italy, and thereafter in Barcelona by Concepció Bru.⁹⁴ Early reluctance to accept the efficacy of this technique was short-lived, and ethanol injection became an established option. Multihole needles to facilitate tumor infiltration and the implantation of radiodense seeds to facilitate locating the previous site of the tumors made invisible radiologically by ablation, were developed in Taiwan by Jin-Chuan Sheu.⁹⁵ As tumor ablation became established, ethanol use was replaced by radiofrequency destruction of small tumors,^96,97 giving the same degree of necrosis in just one session, compared to several treatments. The no-touch approach extensively evaluated by Oliver Seror avoids puncturing the tumor, consequently lowering the risk of tumor seeding.⁹⁸ Thereafter, ablation by microwave was shown to achieve more extensive necrosis than by radiofrequency.⁹⁹ Nowadays, several devices, with different limitations and advantages, are available.¹⁰⁰ Results of cohort studies and a randomized trial in Japan¹⁰¹ have shown that ablation through radiofrequency or microwave may offer the same survival as surgical resection; hence, the decision in favor of one or another option relies on detailed profiling of the clinical status and tumor characteristics of each patient.⁸ The location of the lesion is likely to influence the probability of complications and efficacy of ablation, as is also true for surgical resection. High-intensity focused ultrasound, cryoablation, laser therapy, and radiation are still relatively under-evaluated with respect to their specific roles.

A byproduct of the success of HCC ablative therapy has been the need to refine the criteria that define a response. Ascertaining tumor size may not be informative after successful ablation and thus, according to conventional oncological criteria such as RECIST (Response Evaluation Criteria in Solid Tumors), the response to treatment would artificially be considered negative. The demonstration that necrosis is paralleled by the absence of contrast uptake at dynamic imaging allows the inclusion of this variable in the evaluation of a response as was already apparent in the early days of ablation but was formally endorsed in the 2000 European Association for the Study of the Liver Monothematic Conference.¹⁰² The latter decision became a turning point in the preparation of guidelines and clinical recommendations, by scientific associations. The determination of responses based on contrast uptake would thereafter be known as the “European Association for the Study of the Liver criteria,” which have become the mainstay for locoregional treatment evaluation and ultimately were incorporated into the mRECIST proposal that proposes to rigorously and numerically stratify the magnitude of responses.¹⁰³ Notwithstanding, responses to ablation are simply considered either complete or failed, judged by the visibility of residual tumor.

TRANSARTERIAL PROCEDURES

Hemoperitoneum was evidence of tumor rupture¹⁰⁴ and such a medical emergency could be effectively treated by hepatic artery embolization.¹⁰⁵ Since surgical hepatic dearterialization or hepatic artery embolization was known to induce major tumor necrosis as shown by the results from patients so treated,^106,107 therapeutic arterial obstruction became a promising approach. In this context, an inflatable instrument was devised to be placed around the hepatic artery, to permit the induction of repeated ischemic insults.¹⁰⁸ However, selective hepatic artery access via peripheral artery catheterization allowing the injection of gelatin sponges or spheres, or placement of coils, was less invasive and yet caused major necrosis—ultimately to be termed succinctly as embolization or bland embolization.¹⁰⁹ Kenichi Takayasu in Japan then introduced the intra-arterial delivery of a chemotherapy agent emulsified in lipiodol,¹¹⁰ which would be defined by the tongue-twister “chemolipiodolization.” When combined with arterial obstruction caused by the injection of embolic materials, it was labeled with the more easily pronounced term “chemoembolization.”¹¹¹ A survival benefit accrued following chemoembolization but not after bland embolization, as was shown in randomized controlled trials and a cumulative meta-analysis.¹¹² As with surgical resection, discussed above, optimal results depend on strict candidate selection,^8,100 as exemplified by the decision tree for resection candidates in Figure 11.^65,66 Desirable long-term survival is achieved in patients without cancer-related symptoms and without hepatic decompensation, in whom accurate staging has excluded vascular invasion and/or extrahepatic spread. Chemoembolization and bland embolization achieve extensive necrosis that is not captured by the conventional RECIST model.¹¹³ Thus, the response assessment must include dynamic imaging appearances. Slow-release chemotherapy spheres have been developed, which are effective yet have reduced systemic toxicity;^114,115 combining the use of these new devices with systemic agents, however, has not provided survival benefits.

Transarterial radioembolization, also termed selective intra-arterial radiation therapy, employs the selective injection of plastic or glass beads charged with ß-emitting isotopes to deliver radiation to the tumor that induces significant necrosis. Riad Salem in Chicago has successfully developed a program using this approach,^116,117 which may compete in outcomes with ablation and chemoembolization. However, the failure of phase 3 trials of transarterial radioembolization/selective intra-arterial radiation therapy versus or in combination with systemic therapy, highlights the need to identify the subset of patients for whom this new approach could be the first choice to achieve a survival benefit. External beam radiotherapy allows accurate targeting of the tumor and highly selective intervention while avoiding damage to the surrounding nontumor liver and nearby structures, as caused by conventional radiotherapy.¹¹⁸ As with transarterial radioembolization /selective intra-arterial radiation therapy, the ideal target population for whom external beam radiation could be the first treatment option has not yet been defined, and results of prospective trials are eagerly awaited.¹¹⁸

SYSTEMIC THERAPY

Systemic therapy was ineffective¹¹⁹ until the performance of the successful SHARP trial and its landmark publication.¹²⁰ Before such advancement the evolution of the modality of systemic therapy languished in the doldrums (The maritime doldrums referred to a belt of calms and light winds north of the Equator, in which sailing ships in the 18th century were often becalmed. The word is possibly derived from Middle English “dold,” the past participle of dollen, to make or become dull-witted or inactive), suffering from unacceptable rates of severe adverse events. The encouraging results of Adriamycin published by Olweny et al⁵⁰ in Uganda explain the incorporation of this drug as a conventional treatment approach for HCC. Even in the absence of any evidence of survival benefit^121,122 Adriamycin was seen by some as a “de facto” first-line option.^121,122 Several other agents had raised positive expectations but unfortunately, none of the proposed drugs, either as a single agent or in combination, showed any benefit.¹²² Pilot trials testing novel agents were run by several investigators, but when randomized trials were performed there was a key difference between medical oncologists and hepatologists. In medical oncology–based trials the control arm was active drug administration (usually Adriamycin), whereas the control arm in hepatology-based trials was placebo and/or “best supportive care.” Michel Beaugrand, in France (Figure 12G), led the Groupe d’Etude et de Traitement du Carcinome Hépatocellulaire, which was instrumental in testing different screening intervals,¹²³ chemoembolization,¹²⁴ and antiandrogenic drugs.¹²⁵ Together, we tested a vitamin D analog and obtained negative results,¹²⁶ while at the BCLC we were also testing bland embolization,¹²⁷ chemoembolization,¹²⁸ interferon,¹²⁹ and tamoxifen¹³⁰ All these therapeutic trials turned out to be negative, with the exception of chemoembolization.¹²⁸ Notwithstanding, the aforementioned trials did establish the basis for the current trial design in terms of target population, stratification prior to randomization, and by setting survival as the main endpoint, thereby avoiding faulty endpoint surrogates, such as progression-free survival.¹³¹

The design of the SHARP trial that tested sorafenib against a placebo in a double-blind fashion, exemplifies what we learned but, notwithstanding, we had to defend the need to avoid active treatment in the control arm. One of the arguments we used was that the potential toxicity and adverse events due to conventional chemotherapy could impair the outcome of the patients and in this manner result in a “false-positive” result. Two relevant additional points were introduced. Since RECIST would classify the appearance of ascites or pleural effusion as progression, and the appearance of new intrahepatic nodules could simply result from better recognition of regenerative nodules, we produced an innovative radiology charter, specific for HCC, so as to avoid erroneously recording “progression.”¹³² It is worth mentioning that in this charter a new nodule >10 mm displaying arterial uptake at dynamic imaging, would be classified as progression, while the later mRECIST would also require washout in the venous delayed phase.¹⁰³ In practice, mRECIST delays recognition of progression (and its potential treatment), but in the final analysis, this is actually depicted by the time to progression, which is in fact registered at the time of first suspicion. Furthermore, we envisioned that a treatment benefit could accrue despite progression and that treatment could be maintained beyond progression. Indeed, the co-primary end point was “time to symptomatic progression,” thereby indicating that some marginal progression would not actually mean that there had been treatment failure nor imply a significant impairment of prognosis as it is not associated with any deterioration in liver function. The importance of the pattern of progression in predicting a better or worse prognosis was confirmed later on by Reig et al¹³³ at the BCLC. Indeed, current trials of immunotherapy reveal that a large proportion of patients persist on treatment beyond progression, because of perceived clinical benefit.

The SHARP trial turned out positive in 2007, and as such it represented a watershed in the research of systemic therapy for HCC. It was an era when so-called targeted therapies were portrayed as if they were acting mechanistically and/or highly specifically, on well-defined targets of the relevant pathways involved in tumor progression. In reality, they were dirty molecules, so to speak in the jargon of the field, that acted both on and off target and such dirtiness certainly played a role in the outcomes. Accordingly, several previous agents were tested as first-line systemic therapy against HCC over the next 10 years but, unfortunately, they failed despite promising phase 1 or phase 2 data.¹³⁴ In contrast, the performance of trials of second-line agents, ie, regorafenib,¹³⁵ cabozantinib,¹³⁶ and ramucirumab,¹³⁷ were motivated by the positive results with sorafenib. Trial designs were progressively refined, which increased understanding of the natural history of the disease and how to stratify patients prior to randomization in order to have well-balanced treatment arms with respect to prognosis. Lenvatinib was proven noninferior to sorafenib in 2018¹³⁸ as first-line therapy, but otherwise, the paucity of further major advances triggered pessimism somewhat, about the possible success of further investigations along the same lines.¹³⁹

Fortunately, the therapeutic armamentarium changed dramatically with the development of active immune-oncology agents that are able to prevent tumor cells from evading immune surveillance. The use of antibodies against tumor proteins, such as α-fetoprotein or ferritin, to deliver radioiodine to the tumor was the first immune-based strategy,¹⁴⁰ although the potential of immunotherapy was optimistically envisioned already in 1975 by pioneers, such as Steven Rosenberg at the University of Texas, MD, Anderson Cancer Center in Houston, TX.¹⁴¹ A breakthrough materialized when it became possible to activate the Programmed Cell Death Pathway, by blocking either the receptor programmed cell death protein-1 or its ligand programmed cell death ligand-1.¹⁴² The aforementioned responsible agents had changed the treatment landscape of other cancer types, but it took longer to prove their efficacy in treating HCC. Single-agent regimens, using nivolumab¹⁴³ and pembrolizumab,¹⁴⁴ had antitumoral activity, but a survival benefit was only first seen when atezolizumab was tested in combination with the anti-angiogenic bevacizumab¹⁴⁵ or when both cytotoxic T lymphocyte–associated antigen 4 and programmed cell death ligand-1 were blocked (tremelimumab and durvalumab, respectively).¹⁴⁶

The advances that have transformed the field of primary liver cancer treatment, also encouraged the emergence of physicians who are best identified by the neo-epithet Hepato-Oncologist, which implies having the combined knowledge of hepatology and liver cancer expertise. The need to bring separate fields of expertise into a common forum triggered the genesis of the International Liver Cancer Association (ILCA, Figure 12H), which provides a focused forum for those deeply involved clinicians and researchers. This teaming up to advanced knowledge and quality care fosters the creation of health care teams dedicated to HCC, which comprise hepato-oncologists, hepatobiliary and liver cancer surgeons, physicians with professional expertise in diagnostic and interventional imaging—formerly known as radiologists—as well as separate specialists involved in cancer management, advanced practice nurses, social workers, and appropriate pharmacists, in medical centers where patients with HCC are managed. When validated molecular profiling guided therapy becomes available, tumor boards will have to incorporate such expertise. Facilities to provide outpatient care, facilitate drug infusions, and supervise safety monitoring should be set up and also equipped with home delivery of treatment and planning for end-of-life care.

SERIES EDITOR’S POSTSCRIPT

The landscape of treatments for HCC is a veritable minefield of competing surgical and other ablative manoeuvers and a burgeoning range of systemic therapies. But who better to guide us through this maze of therapeutic options than Professor Jordi Bruix of the BCLC that he co-founded and the internationally acclaimed Liver Unit at the Hospital Clinic of Barcelona, at the University of Barcelona, Spain!

Jordi is a proud Catalan, who was born, raised, and educated in Barcelona, the capital of the autonomous Spanish community of Catalonia where, during his upbringing, respect and appreciation of his large Catalan family roots, culture, and language were inculcated, including support of the Barcelona football team in which his grandfather had been a first-line player. Catalonia (Catalunya in Catalan, Cataluña in Spanish, and Catalunha in Occitan), which is designated as a nationality by its Statute of Autonomy, lies on the northeast of the Iberian Peninsula, between the Pyrenees, France, and Andorra to the north, the Spanish autonomous communities of Aragon to the west and Valencia to the south, and the Mediterranean Sea to the East. The history of Catalonia, like that of Spain itself, has been tumultuous throughout the centuries, both before and after the latter’s occupation as Hispania, a province of the Roman Empire.

Jordi was born during the loathsome rule (1939–1975) of the dictator Generalisimo Francisco Paulino Hermenegildo Teódulo Franco Bahamonde and his fascist government, when among other indignations was the banning of the Catalan language and culture. Indeed, Jordi was only able to be officially known by his Catalan given name in the 1980s, following Franco’s death in 1975. Although Catalan is the official language of Barcelona, including the Hospital and Liver Unit, the BCLC, and the rest of the University of Barcelona, it is not always plain sailing.

Pathology was Jordi’s first choice for a profession, but after he diverted to Internal Medicine and Digestive Diseases and even escaped the gravitational pull in the Liver Unit of Portal Hypertension, he established an interest in HCC. Jordi attracted collaborators to craft a team that ultimately became the BCLC in 1986, which has been a magnet for trainees, researchers, collaborators, and visiting faculty from home and abroad, ever since. Under the aegis of the BCLC and with the watchwords of teamwork and honesty, investigations were performed into HCC epidemiology, diagnosis, prognosis, therapy, and trial design and analysis—these led to the establishment of the BCLC Prognosis and Treatment model^6,7 that has been endorsed at a worldwide level as providing an evidence-based approach to patients with primary liver cancer. The 2022 update of the prognosis and treatment model for HCC⁸ was introduced at the 35th BCLC anniversary celebrations in November 2021.

Professor Bruix has been awarded the International Hans Popper Award, the Golden Medal Award of the Canadian Association for the Study of the Liver, the Recognition Award of the European Association for the Study of the Liver, the Nelson Fausto Award of the International Liver Cancer Association, and the Distinguished Achievement Award of the American Association for the Study of the Liver, which he (right-hand side in the figure below) received in the company of the series editor (middle), who received the Distinguished Service Award, and future essayist in this series, Professor K. Rajender Reddy (left-hand side), who received the Distinguished Clinician Educator/Mentor Award, in Boston MA in November 2019. Readers are encouraged to access forthcoming related essays, in this History of Hepatology series, by (i) Riad Salem and colleagues entitled Interventional Radiology and the Liver Vasculature: A Personal Perspective, and (ii) Deepak Hariharan and Abhirup Banerjee, entitled History of Liver Surgery, which deal with radioembolization (as cited in references 116 and 117), and non-HCC tumor localization in the liver and its extirpation, respectively.