Hepatocellular carcinoma in adult thalassemia patients: an expert opinion based on current evidence

According to the 2019 AIOM report, 33.000 patients had HCC diagnosis in Italy between 2013 and 2018 [12]. Among beta-thalassemia patients, HCC prevalence ranged from 2.3% in Greece to 1% in Italy [1, 13]. Another study reported that HCC prevalence in male patients with beta-thalassemia is 6 times higher than in non beta-thalassemia subjects [14].

HCC incidence, standardized by age in non beta-thalassemia patients can be estimated at 10.9/100.000 P/Y for male and 3.1/100.000 P/Y for female [13]. Incidence of 1.02/100.000 P/Y was estimated in the Italian Thalassemia Registry including 5855 patients [1]. Annual incidence of 2% was reported in a cohort of 108 TM or TI under US surveillance (C) [15].

Among patients without beta-thalassemia, HCC occurs more frequently in patients with HCV or HBV hepatitis or alcohol abuse. In Italy, the number of patients with HCC is increasing, although the etiologic factors are changing. Within the ITALICA cohort HCC increased from 624 to 1130 over time [16]. A decline in the percentage of HCV related HCC from 51 to 49%, and of HBV related HCC from 3.3 to 2.1% was reported [17]. Nowadays, HCC related to HCV and HBV infections account for 50% of liver tumors whereas HCC related to non-viral etiology make up for 32.4% [18].

In beta-thalassemia patients, HCV infection with or without iron overload has been the most frequent etiology until recently. In those who received blood transfusions before 1990, HCV infection prevalence was proportional to the number of transfused units. In studies published in 2000, HCV Ab resulted positive in 85% of Italian beta-thalassemia patients [19, 20]. Active HCV infection confirmed by HCV RNA replication has been shown in 50% of HCVAb positive subjects [21]. Higher rates of anti HCV positive patients were reported up to 1992 in Greece [22].

Careful donor screening reduced HCV related transfusion risk to 0.1 per 1.000000 [23, 24]. Therefore, patients with beta-thalassemia, infected before 1990, represent the main group at risk of developing cirrhosis and HCC. As the majority of patients has been treated with IFN ± RBV in the past or with DAA since 2014, the risk of cirrhosis of viral etiology will become less relevant in the future.

In non beta-thalassemia patients the prevalence of HCC in non-cirrhotic population is reported at 9.6%, with an upwards trend especially in patients with NAFLD [16].

In a letter published by Restivo Pantalone, only 1 out of 9 cases had cirrhosis [33]. Of an Italian cohort published in 2014 only 4 out of 60 had cirrhosis [1]. Of the 2 HCC beta-thalassemia patients reported by Makaron, none had cirrhosis [34]. Fragatou published 5 HCC cases of whom 2 had cirrhosis [32].

Seven cases with occurrence of HCC in non-cirrhotic livers have been described: two patients with TI (one of them with HCV-positivity); one patient with TM and HCV-positivity; two patients with TI (one of them with HCV-positivity); two patients with TI and negativity both for HBV and HCV [35]. All these cases presented iron overload and although episodic, confirm the possibility that iron overload alone could induce the development of HCC [1, 30].

Persistently high LIC (Liver Iron Concentration) values as a consequence of differences in monitoring schedules either in TD or in NTD patients may play a role [36].

The recent introduction of direct acting antiviral (DAA) against HCV led to extremely high SVR12 [11, 37]. In non beta-thalassemia patients, after SVR12 the risk of HCC although reduced is not zeroed [38, 39] although not different from that after Interferon. The risk is increased by severity of the underlying liver disease and presence of co-morbidities. Not achieving DAA response was associated with increased HCC risk [40]. Chronic HBV infection is safely controlled by NUCs. In non beta-thalassemia patients, the risk of HCC after Tenofovir alafenamide or entecavir is not zeroed. HCC predictors are cirrhosis or severe fibrosis, age and male gender. Patients with NUCs-induced HBsAg seroclearance, on top of complete viral suppression, have a lower risk of HCC than those only achieving complete viral suppression under prolonged NUCs treatment [41].

In patients with NUCs -induced HBsAg lost, HCC risk is similar to that of patients with spontaneous HBsAg lost and can be estimated around 1% at 5 years [42].

There are no available data in the setting of beta-thalassemia patients where the risk may be increased by the presence of iron overload.

Biannual surveillance is required in at risk patients regardless of SVR12 achievement or NUCs induced HBV DNA control.

In patients with beta-thalassemia the evaluation of LIC by MR Imaging has replaced iron concentration in mg/g of dry liver on liver biopsy. In parallel, formerly invasive diagnosis of liver damage can be now performed non-invasively evaluating liver stiffness by transient elastography (TE). In not beta-thalassemia patients, TE is the most accurate non-invasive assay in excluding cirrhosis (AUROC 0.90) [43]. The AUROC for fibrosis comparable to a histological stage ≥2 is 0.86 [44, 45].

In patients with beta-thalassemia, this method was evaluated in 4 studies. The first is a controlled cohort study including 56 patients who underwent both liver histology and TE. Results showed that TE is able to discriminate between advanced fibrosis and liver cirrhosis and that the presence of iron overload is not a confounder [45].

A second study was “cross sectional” and included 115 patients with TM or TI. However, only in 14 of them a direct comparison between TE and liver histology was possible [46]. In patients with positive HCV RNA and ferritin levels > 1000 ng/ml, TE overestimated liver fibrosis.

The third study based on 15 patients who underwent both liver biopsy and TE, demonstrated that TE is accurate in assessing the presence of significant fibrosis in patients with beta-thalassemia, showing a sensitivity of 60% and a specificity of 89%. However, the highest accuracy was attained in diagnosis of cirrhosis with a sensitivity of 100% and a specificity of 92% [47]. The fourth study on 49 patients without HCV infection who underwent T2*MR and TE showed a significant direct correlation between iron accumulation and liver stiffness [48]. In conclusions, in the absence of an active HCV infection, TE is as accurate in determining advanced fibrosis status as in patients without beta-thalassemia.

In beta-thalassemia patients, surveillance is needed either to determine iron accumulation into the liver, or to early detect possible HCC growth.

Measurement of iron concentration in liver biopsy (LIC), representative of total body iron stores, has been replaced by MRI LIC assessment [49].

The current surveillance program suggested for beta-thalassemia patients [50], is stratified according to NTDT or TDT condition and baseline LIC. Both NTDT and TDT patients must undergo biannual MRI if baseline LIC is > 15 mg/g. If baseline LIC ranges between 5 and 15 mg/g (NTDT), and between 3 and 15 mg/g (TDT), MRI must be performed annually.

When baseline LIC is < 5 mg/g (NTDT) or 3 mg/g (TDT), MRI is advised every 2 years.

Optimal frequency for HCC surveillance is every 6 months (semi-annual). The observed improvement of overall survival in a large Italian cohort [17] is also owed to the wider use of semi-annual surveillance, expanding the proportion of HCC susceptible of curative treatments. Another study in the same cohort [51, 52], had previously demonstrated that semi-annual US surveillance has a positive factual benefit in a long-term perspective, not due to lead-time bias, compared to annual one.

Also in beta-thalassemia patients at risk of HCC, surveillance should be performed by abdominal US and AFP serum determination every 6 moths [33, 53]. As for viral infection monitoring, patients undergoing blood transfusion should be tested for HCV Abs and HBsAg every year.

According to the current guidelines on HCC diagnosis and treatment [68], “early HCC” is defined into two different stages (“very early”, or stage 0, and “early”, or stage A) according to liver function, performance status (ECOG-PS), and tumor burden (Table 1).

Subject to liver function preserved (Child-Pugh score A), and ECOG-PS 0 in both groups, “very early stage” is defined when the tumor burden is represented by a solitary nodule ≤2 cm in diameter, and “early stage” when a solitary nodule is > 2 cm in diameter or up to 3 nodules are all ≤3 cm in diameter.

Though Asian [70], American [71], and European [68] guidelines still recommend ablation only in patients with stage 0 and A who are not candidates for resection, such technique is now widely considered the first option in solitary tumors ≤2 cm in diameter favorably located within the liver. The most used method is US-guided percutaneous radiofrequency ablation, which causes tumor necrosis by thermal damage [72]. The success rate is inversely related with nodule diameter, so resulting in a worse local control for tumors > 3 cm. However, for nodules up to 3 cm, thermal ablation has been demonstrated equivalent to resection in terms of cumulative recurrence rates and overall survival [73] and is burdened by fewer complications. On the other side, resected patients present higher rates of perioperative mortality and major complications, but also lower rates of local tumor progression [74].

Among ablative therapies, percutaneous ethanol injection (PEI) has been substantially replaced by thermal ablation in the treatment of HCC < 2 cm because of its higher rates in cumulative recurrence and local tumor progression, in spite of similar 5-year overall survival [75]. Moreover, the extension of tumoral necrosis produced by PEI is scarcely predictable. Very recently, microwave ablation is achieving a wider role as compared to RF, thanks to its capacity (due to greater usable powers) to obtain similar results in shorter times.

Resection, in stage 0, is preferred to thermal ablation in selected cases, above all when a laparoscopic approach is feasible, or when the tumor site is not reachable by percutaneous ablative techniques.

In stage A, when the tumor is solitary, resection is the first option regardless of tumor size, above all when there is no clinically significant portal hypertension [76]. If surgical resection is not possible, liver transplantation must be considered.

For beta-thalassemia subjects with early HCC, data available in the literature are scanty, heterogeneous and tailored to individual patients’ needs. In an Italian cohort study published in 2014 [1], HCC was already advanced at the time of diagnosis in 21 out of 62 cases (33.9%), so that only palliative treatment was possible. US surveillance is essential in detecting small lesions susceptible of curative treatment [17].

There are no sufficient data to establish which is the best therapeutic approach in early HCC among thermoablation, liver resection, and liver transplantation. There are no data suggesting that the management of early HCC should be different in patients with or without beta-thalassemia [76]. A multidisciplinary approach involving hematologists, hepatologists, radiologists and oncologists is essential in identifying a tailored treatment [77].

In latest years, only a few studies have evaluated the surgical treatment of HCC in patients with beta-thalassemia [77, 78].

Beta-thalassemia patients with HCC present relevant co-morbidities: hypogonadism (55%), myocardiopathy (52%), hypothyroidism (42%), osteoporosis (31%), diabetes (31%) and chronic renal failure (4%) [79]. This needs to be taken into account when planning surgical resection.

In the largest available study [1] the same therapies currently adopted in non thalassemia patients, either loco-regional, or resective, or chemotherapeutic, or palliative were adopted for beta-thalassemia patients. Such approach, even when aggressive as liver resection, led to a survival improvement in beta-thalassemia patients with HCC (median of 11.5 months) as compared to 2004 (3.5 months median time from diagnosis to death) [1].

Both Maakaron [34], who described a recurrent HCC treated by liver lobectomy after percutaneous thermal ablation, and Moukhadder [35] stated that both surgical and loco-regional ablative therapies should be considered for the treatment of HCC in beta-thalassemia patients. They appear to be safe and effective options [1, 16, 77].

The adoption of a more invasive treatment, albeit potentially radical as surgical resection, is also supported by the scarce availability of relevant alternatives. Obviously, an extensive evaluation of potential co-morbidities is strongly suggested in the preoperative workup.

Even though there might be no need for specific cardiologic evaluations, it is advised to perform, in patients undergoing surgery, an extremely accurate workout for pulmonary hypertension (PH) that seems more common in adult patients with TI not or inefficiently transfused. Indeed, it is well shown how precapillary PH may be a devastating complication in beta-thalassemia [80]. An echocardiogram and an echo-stress test may be suggested in all patients in order to exclude subclinical heart failure (HF) [81].

Pulmonary atelectasia has been frequently described in the setting of beta-thalassemia patients undergoing splenectomy. A preoperative spirometry may be indicated in patients with other risk factors for postoperative ventilatory dysfunction [80].

Present evidence on the best surgical approach between laparotomy and laparoscopy in the setting of HCC are currently limited. However, studies have been published in the particular setting of splenectomy in beta-thalassemia patients [81, 82]. A single randomized study comparing laparotomic vs. laparoscopic splenectomy showed an increased incidence of intraoperative and postoperative bleedings in the laparoscopic group [83]. The results, however, did not reach statistically significant difference [83].

On the other hand, in the HCC general population, laparoscopic approach has been found to be significantly beneficial in terms of postoperative morbidity, mortality, length of hospital stay and blood transfusion requirement when compared to the open approach [81]. On these bases, we suggest that HCC surgical approach in beta-thalassemia patients should be adopted in tertiary surgical centers with high volume of both open and laparoscopic procedures where the laparoscopic choice can be considered in every case and adopted safely.

As far as the postoperative course is concerned, it has to be underlined that thrombotic risk is increased in beta-thalassemia patients. Few recent evidence report a higher risk of thromboembolic events in particular during the early postoperative course [84, 85]. However, no clear evidence supporting thromboembolic prophylaxis different from non thalassemia patients is currently available.

Even though studies specifically conducted in beta-thalassemia patients undergoing liver resection or transplantation for HCC are not available to date, close postoperative surveillance and aggressive coagulation prophylaxis should be adopted in these patients.

Liver Transplantation (LT) is now considered the treatment of choice in patients with HCC. In the past, mainly due to cardiac co-morbidities [77], LT has long been refused to beta-thalassemia patients with end-stage liver disease or HCC. Beta-thalassemia is not longer considered a contraindication to liver transplant anymore provided that severe PH and subclinical HF are excluded [86]. Thanks to better chelation therapy and iron overload monitoring [1] leading to improved survival, both liver failure and HCC have become a more commonly reported complication, with significant adverse impact on prognosis and mortality [1].

In the series of nine beta-thalassemia patients with HCCs reported by Restivo Pantalone, four patients died during follow-up due to decompensated cirrhosis. The single patient undergone LT (after thermal ablation and TACE) survived 69 months, compared to a median survival of 25 months for four not transplanted patients. The authors concluded that beta-thalassemia should not be considered a contraindication for either treating HCC or for LT [33].

Mancuso described two additional patients undergoing successful LT with satisfactory post-transplantation outcomes, without severe complications after 6 months and 2 years of follow-up, respectively [77].

Borgna-Pignatti described three patients undergoing LT, two of whom died shortly afterwards for reasons unrelated to beta-thalassemia [1]. Therefore, after excluding significant co-morbidities as heart dysfunction and PH, beta-thalassemia should no longer be considered a contraindication to liver transplantation which is, as in non beta-thalassemia patients, a potential therapeutic option conferring better survival [85, 86].

As reported, only six patients underwent LT for HCC, but the outcomes were affected by beta-thalassemia complications [33].

As, HCC and liver failure can develop in young beta-thalassemia patients: in this context, the relevant survival benefit offered by LT compared to the standard alternative treatments (transplant benefit) has to be considered.

Prevalence of HLA antibodies is relevantly increased in TM. Even though a high panel reactive antibody percentage is not a contraindication to transplant, the possibility of antibody mediated events has to be carefully considered in postoperative course. On these bases, a regular evaluation of circulating Donor Specific Anti HLA antibodies (DSA) has to be adopted in beta-thalassemia patients undergoing LT [87].

Current indications for combined heart–liver transplantation (CHLT) include end-stage heart and liver disease of varying etiology, in particular familial amyloid polyneuropathy and heart failure with associated cardiac cirrhosis. Liver iron overload leads progressively to chronic liver damage, and some patients may develop liver cirrhosis associated to heart failure.

To date, only one combined heart and liver transplantation has been reported in the setting of beta-thalassemia [88]. The patient, with a homozygous beta-thalassemia, was diagnosed when he was 2 years old and started blood transfusions when he was four. At age 11 he underwent splenectomy in order to maintain adequate level of hemoglobin. He started deferoxamine therapy but with low compliance. At age 17 he began to show signs and symptoms of cardiac failure due to heavy iron deposition, confirmed with MRI. The left ventricular ejection fraction was 21%. Despite patient’s improved compliance to the therapy, sign of liver failure started to rise: prolonged prothrombin time, low albumin level, increase in liver enzymes. Eventually, he developed ascites and a liver biopsy showed heavy iron loading and portal fibrosis with cirrhosis. At age 26 he underwent combined cardiac and liver transplantation. Eighteen months after transplantation liver showed only focal iron deposition and mild rejection. At a 2-year follow-up, the patient’s hepatic and cardiac functions were normal. The authors concluded that they would consider combined organ transplantation for patients with iron loading and severe cardiac dysfunction associated with histological diagnosis of cirrhosis. This may be the only option for patients with end stage iron -related cardiac and liver disease [88].

Despite the absence of specific studies evaluating TACE in the treatment of HCC in this specific patient’s population, important data can be derived from numerous studies demonstrating efficacy and safety of this procedure in HCC treatment of non beta-thalassemia patients.

According to the Barcelona clinic Liver Cancer (BCLC) system, the most commonly used staging system for treatment and prognosis of HCC in Western countries [89], TACE is the only recommended treatment strategy in intermediate stage (BCLC B). However, not all patients benefit from TACE in the same way [90] and such heterogeneity has prompted authors to identify prognostic parameters and different scores enabling patients’ stratification.

Bolondi and coworkers [91] proposed a sub-classification which identified four sub-stages (B1–B4) of intermediate HCC, incorporating the new concept of joint consideration of the tumor burden according to the “beyond Milan” and the “within up-to-7” criteria together with the Child-Pugh score and PS.

For B1 patients, considering their conserved liver function and limited tumor burden, TACE is suggested as first option. For B2 patients [92], 90Y-radioembolization (TARE) or sorafenib could be considered in cases expected not to respond to, to be poor responders or to have contraindications to conventional TACE. For the substage B3, the option of new treatments tested within clinical trials represents the only therapeutic alternative to sorafenib or TACE. In substage B4 patients who met the up-to-7 criteria, liver transplantation is the best option. In patients without alternative therapeutical options, symptomatic treatment to avoid unnecessary suffering from liver damage can be offered [93].

New recent evidence suggest the possibility to successfully and safely combine TACE with sorafenib in non beta-thalassemia patients [94].

Trans-arterial radioembolization (TARE), consisting of selective internal radiation therapy (SIRT), is a well-recognized therapy in a number of guidelines on clinical management of non-resectable HCC [95]. In patients in intermediate stages and history of prior TACE failure or tumoral macrovascular invasion in the absence of extra-hepatic spread, according to ESMO (European Society of Medical Oncology) guidelines, TARE may compete with sorafenib [96].

From a technical point of view, TARE is a catheter-based interventional procedure allowing the emission of β-radiations at therapeutic levels into the tumor through its feeding arteries. Such ultra-selective approach, as for TACE, is aimed to minimizing the effect on healthy liver parenchyma close to the tumor. The local brachytherapy performed by TARE, conversely to TACE, doesn’t result in tumor ischemia due to microvascular embolization [97]. Devices for radioembolization are commercially available in form of implantable glass (Therasphere®) or biocompatible resin-based (SIR-Spheres®) radioactive (Yttrium⁹⁰ - Y⁹⁰) spheres [97].

TARE is safe [98], also compared to conventional Sorafenib treatment, and no particular attention should be paid in beta-thalassemia patients. Usually, after the procedure, oral amoxicillin/clavulanic acid (375 mg 3 times per day) and pantoprazole (40 mg per day) are administered for 3 days [99]. Discharge from the hospital is decided according to individual patients’ clinical status. Clinical and laboratory follow-up is advisable [99].

In recent years, the radial approach has been introduced. This technique guarantees a safe vascular access, especially if compared with the brachial one, and allows patients to spend a more comfortable post-operative period considering the possibility of walking immediately after the procedure. Thanks to a greater efficacy of the hemostasis means, this approach is also advisable in patients with coagulation alterations [99].

Treatment options for HCC in beta-thalassemia patients are largely based on data extrapolated from general population due to the low incidence of HCC in beta-thalassemia. Beside the issue of limited number of beta-thalassemia patients with HCC treated with TKI, the best treatment choice remains controversial. Even more than in other cancers, it is essential that the most appropriate treatment is administered at the right time, considering that suspending therapies might be harmful to the patient. Furthermore, depriving patients of a possible therapeutic alternative available today exposes them to worsening of the liver function that could make an effective alternative treatment no longer feasible.

For these reasons, it is essential that since first diagnosis, beta-thalassemia patients with HCC are taken over by a multidisciplinary group, following them throughout their therapeutic path [107]. The value of a similar approach is increased in the presence of co-morbidities, possible drug interactions (especially hormonal), or metabolic diseases [110]. A number of beta-thalassemia patients has undergone splenectomy and requires medium and long-term prophylaxis protocol with all the scheduled vaccination calls, before starting cancer treatment. Particularly important are C-conjugated or tetravalent anti-meningococcal vaccination and anti-pneumococcal vaccination.

As in non beta-thalassemia patients, a clear therapeutic hierarchy should be adopted in real life in beta-thalassemia patients as well. It will explore, at first, indication and feasibility of potential radical options as liver resection, ablation or liver transplantation. If radical approaches are not indicated, embolo-therapies represent the second choice followed by first and second line systemic therapies to be adopted according to the multidisciplinary team decision.