Introduction
Hepatocellular carcinoma (HCC) is a major cause of global morbidity and mortality [
1], with an rising incidence in the Western world due to an increasing prevalence of steatohepatitis [
2]. Staged according to the BCLC classification, only very early (single tumour, < 2 cm) and early HCC (up to three tumours, < 3 cm each) are generally considered curable [
3]. Curative treatment options include thermal ablation (TA, which includes microwave ablation and radiofrequency ablation), hepatic resection (HR), and orthotopic liver transplantation (OLTx) [
4]. The latter is generally considered to yield the best recurrence-free survival [
5], but is limited by a scarcity in available donor organs, and considerable procedural morbidity and mortality [
6]
. Conversely, liver-sparing treatment modalities are plagued by higher rates of tumour recurrence [
3]. Typically, initial therapy is with either HR or TA, after which some patients may eventually undergo OLTx [
7].
Controversy still shrouds the optimal treatment strategy for early and very early HCC [
8]. It has been shown that carefully performed TA can result in excellent locoregional control of primary HCC [
9]. Data on outcomes after repeat TA for recurrence following succesful primary TA, however, is sparse. Even more controversial is the timing of OLTx after successful TA in patients with preserved liver function [
8]. Should TA be considered ‘only’ as a stop-gap measure until curative treatment is performed in the form of OLTx, or can TA itself, in selected cases, be considered a definitive therapy, provided that strict follow-up imaging is performed, in order to detect possible recurrences at an early stage? We aim to contribute to this debate by presenting our data on outcomes after primary and repeat TA, with and without subsequent OLTx, in a large tertiary referral centre.
Patients and methods
The present study focuses on outcomes after primary and repeat TA for HCC, with and without subsequent OLTx. In our centre, all cases of HCC are discussed in a multidisciplinary team. From 2000 onward, percutaneous TA was performed in a joint effort by specialised interventional radiologists and hepatobiliary surgeons. All patients who underwent TA, HR, or OLTx for HCC were prospectively entered into a database. In this study, we analysed patients who underwent primary TA between January 2000 and December 2019. Follow-up was completed up to December 2020. Outcomes studied were recurrence rate after primary TA and after its repeat interventions, the occurrence of untreatable recurrence, OS and DSS after primary TA and repeat interventions, and complications after TA (graded according to Clavien-Dindo [
10]). We reconstructed patient-specific treatment paths, registering treatment modalities, time of recurrence (if any), and follow-up status. In order to compare survival without bias from tumour stage, age or comorbidity, we analysed patients who had been waitlisted for OLTx at any time after primary TA (
“intention-to-perform-OLTx”), and compared them to a cohort of patients matched for age (< 70 years), tumour size and number (within Milan criteria) and comorbidity (ASA < 3; no severe cardiopulmonary comorbidity), who upon revision of charts by a liver transplant surgeon would have been eligible for OLTx from a medical point of view, but who were never waitlisted due to other reasons (
“eligible, non-waitlisted”).
All instances of percutaneous TA were performed under computer tomography (CT) guidance while patients were under general anesthesia, and mechanically ventilated. Image acquisition and needle positioning were performed under controlled apnea during an expiratory breath hold, allowing for constant anatomical relations. We believe this to be critical to our TA workflow, as accurate needle positioning is a prerequisite of curative locoregional treatment [
11]. All patients underwent a contrast-enhanced CT scan one week after TA, to assess completeness of ablation and to obtain a baseline scan for optimal comparison during follow-up. If this scan revealed an inadequate ablation zone, patients underwent completion ablation as soon as possible. This was defined as incomplete ablation, and the completion ablation was not counted as a separate procedure, in accordance with the standardised reporting criteria proposed by Ahmed et al. [
12]. Follow-up consisted of 4-monthly contrast-enhanced CT scans, laboratory workup, including serum alpha-foetoprotein (AFP) and, where indicated, MRI or other imaging studies. We defined recurrence as any lesion arising after successful treatment (as evidenced by satisfactory margins on the CT scan at day 7), either with the radiological characteristics of HCC, or based on histopathological examination. Time of recurrence was defined as the timepoint at which any such lesion was first detected. Where missing, dates and causes of death were registered by contacting patients’ general practitioners. In case of doubt, death was assumed to be due to HCC. Survival was calculated separately from the date of primary intervention, and from the date of any subsequent interventions. This eliminates the time span between interventions, which might erroneously be interpreted as an increase in survival when multiple interventions have been performed (comparable to “lead-time” bias). Recurrence-free survival (RFS) was calculated from the time of intervention to recurrence of HCC; DSS was calculated from the time of intervention to the time of death from HCC.
Statistical analysis was performed with IBM SPSS (version 23). A χ2 test was used for categorical variables. Means or medians were compared with a Student’s T-test, or with a nonparametric test, as appropriate. Hazard ratios for all-cause mortality were calculated with a Cox regression model. Survival was calculated with a Kaplan–Meier model, and groups were compared with logostic regression. The threshold for statistical significance was set at p < 0·05.
Discussion
We found TA to be a reliable, repeatable treatment modality for HCC. After primary TA, almost half of patients never developed clinically apparent recurrence during follow-up. Most recurrences could be treated with curative intent, either with repeat TA or with OLTx. Some patients underwent numerous sessions of TA without developing an incurable recurrence. Even though RFS was shorter after repeat TA than after primary TA, we found no difference in OS. In fact, after matching patients for baseline oncological characteristics, comorbidity, and age, we found that OS after TA without ever being waitlisted for OLTx was comparable to OS after TA as a step-up to OLTx. Our findings imply the existence of a group of patients who do remarkably well with TA alone, some despite developing multiple recurrences. There is almost no data on outcomes after repeated sessions of TA. A literature search on PubMed-listed studies from January 2015 to 2021 (search terms: ((HCC[Title]) OR (hepatocellular carcinoma[Title])) AND ((ablation[Title]) OR (RFA[Title])) AND ((redo[Title]) OR (recurrence[Title]) OR (repeat[Title]) OR (recurrent[Title])), yielded 104 hits. Manual screening revealed two studies which reported on survival after repeat TA following primary TA [
13,
14]. Both were performed in Asian populations, with comparatively young patients (median age 50 [
13], and 60 [
14]), with almost exclusively viral hepatitis-induced HCC. One study reported a number of non-cirrhotic patients above 30% [
13]. Therefore, we believe these studies are not generalisable outside of their highly specific demographic contexts.
We found a median OS after primary TA of 96.0 months, and a 5-year OS of 54·1%. A widely varying 5-year OS after TA has been reported in the literature, mostly around 40–60% [
15‐
17], (outliers in selected populations 30%-80% [
18‐
20]). It must be noted that our patient population is relatively old (median 64·5 years; 25% of patients ≥ 70 years), while most studies had a median patient age around 55 years [
17,
20], with only sporadic studies in a comparable age group [
16,
19]. Furthermore, tumours in our study were relatively large (median 2·7 cm, 25% of tumours ≥ 3·3 cm), whereas many studies excluded tumours > 3 cm [
15,
21], or had such tumours only in small numbers [
20]. Lastly, some authors excluded patients with multiple tumours [
21] or neo-adjuvant therapy [
19].
As causes of death are variable in patients with chronic liver disease, DSS may be a better measure of oncologic results, especially when comparing results across different populations. Unfortunately, few studies report DSS. Our 5-year DSS of 73·8% after TA is high compared to the literature [
22,
23].
Our finding that 13·6% of patients developed incurable recurrence following primary TA (24% of first recurrences) appears to be in contrast to the suggestion by Doyle et al. that 42% of recurrences are beyond Milan criteria, when analysing a series of smaller, unifocal primary HCC’s [
21]. Despite some patients undergoing up to four sessions of repeat TA, only 23·3% of our patients eventually developed a recurrence unamenable to repeat TA or OLTx. These results might be because of the relatively high technical success rate of our TA workflow, and due to our strict follow-up protocol, which allows us to detect recurrences early on.
Over the past two decades, TA has established itself as an effective treatment option for HCC. Little evidence supports its formerly assumed inferiority to HR [
19,
24]. Indeed, our DSS after TA is not inferior to rates reported after HR [
25,
26]. More uncertain is the optimal relationship between TA and OLTx. On the one hand, OLTx leads to the best recurrence-free survival [
5], with some authors proclaiming OLTx to be the only curative treatment [
27], and others suggesting that locoregional treatments for HCC should be thought of as “a bridge to nowhere” [
28]. By removing the preneoplastic cirrhotic liver parenchyma in its entirety, OLTx adresses the driving factor behind HCC recurrence [
25]. On the other hand, OLTx is a major procedure with a lasting impact on patients’ lives [
6], and is restricted by a shortage of donor organs. It is therefore imperative that we put these scarce resources to optimal use.
Our findings support a more nuanced view of the role of TA in early and very early HCC than hitherto proposed by many. Until now, clinicians have focused on identifying patients who have the lowest risk of post-transplant recurrence [
29], but this is only one side of the picture. Ideally, we should reserve OLTx for patients who would have bad outcomes with locoregional therapy alone [
30].
A shortcoming of our study is its observational nature. This brings with it an inherent danger of confounding variables and selection bias, for which it is impossible to fully correct. Furthermore, in some subgroup analyses, smaller numbers may have obscured underlying correlations. The heterogeneity of this patient population makes a sound prospective analysis beyond the first treatment modality extremely difficult.
In conclusion, we found that TA is an effective treatment modality both for primary and recurrent HCC. Although many patients develop recurrences during follow-up, we demonstrated that most can be treated with repeat TA. There is a group of patients with HCC who survive many years with TA, without ever undergoing OLTx. In matched cohorts, TA without waitlisting for OLTx was not inferior to TA with the intention to perform OLTx. Therefore, it is imperative that we find a way to differentiate between those patients who would do well with TA alone, and those who will require OLTx due to a high risk of untreatable tumour recurrence. This way, we can put scarce donor organs to best use. It is beyond any doubt that if a “TA-first” approach is taken, intensive follow-up is crucial.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Added value of this study
- It is the first study reporting on the impact of repeat TA for recurrent HCC after primary TA in a non-Asian population of cirrhotic patients.
- It is the first study comparing results after primary and repeat TA, Showing that survival outcomes are comparable.
- It is the first study reporting outcomes after several sessions of repeat TA. We reconstructed patient-specific treatment paths, demonstrating that even though many patients indeed develop recurrence after primary TA, these can most often be managed with repeat TA. Some patients do well despite multiple recurrences.
- It is, to our knowledge, the first study that compares outcomes after TA with the intention to perform OLTx to a TA-only approach, correcting for inherent bias and confounders (age, comorbidity and tumour stage). In our intention-to-treat analysis, we compared two matched cohorts of patients. The first cohort consisted of patients eligible and waitlisted for OLTx, and the second consisted of patients who would have been eligible for OLTx, but were never waitlisted. This analysis revealed the existence of a group of people who will do well without ever undergoing OLTx.
Implications of all the available evidence
- Repeat TA can be an effective therapy and should be considered for recurrent HCC following primary TA.
- Our findings suggest that in patients with preserved liver function, TA should be considered as an alternative to early OLTx. A “TA-only” approach may be reasonable in certain groups of patients with HCC.
- Future research should focus on finding predicting factors that identify those patients who are likely to do well with TA alone, and those who will need OLTx because of a high risk of disease progression despite adequate initial locoregional control.