Introduction
Hepatocellular carcinoma (HCC) is the sixth most common malignancy worldwide and causes about 1.25 million deaths each year [
1,
2]. Surgical intervention (e.g., resection and transplantation) is normally the first therapeutic option for solid tumors, however, HCC is complicated because of coexisting chronic liver disease and poor hepatic reserve in a large proportion of patients, or the lack of suitable donors. This uniqueness of HCC substantially limits the use of surgical intervention, but has also increased the development of percutaneous local ablative therapy (PLAT). PLAT includes radiofrequency ablation (RFA), percutaneous ethanol injection, microwave ablation, and cryoablation. With the constant improvement in the past few decades, there is now a broad consensus that PLAT, which is predominantly presented by RFA, is a promising first-line treatment for early-stage HCC, with or without surgical prospects [
1–
3].
Despite being introduced for curative treatment of HCC over a decade ago and even longer in cancer treatment [
4,
5], cryoablation is somehow still largely limited in China and a few overseas centers. Its low level of popularity has made it a relative “silent player” in HCC treatment, compared to RFA, and there is limited evidence about its usefulness in HCC. This has led to the situation in which most of the published studies have been of early and small-volume trials, which has made it difficult to judge its therapeutic value [
6,
7].
Working in a tertiary referral center, our team started percutaneous argon-helium cryoablation for the treatment of HCC in the early 2000s. Generally, cryoablation is performed in two groups of patients who are precluded from surgical resection. The first group is, as suggested by the current guidelines, patients with small HCC according to the Milan criteria, with a single nodule≤5 cm in diameter or up to three nodules≤3 cm in diameter [
8]. Cryoablation had been expected to provide first-line curative treatment in these patients. The second group is patients with advanced or recurrent HCC but who still have well-preserved liver function and no extrahepatic metastases. Cryoablation, usually as a part of the multidisciplinary treatment, provides palliative therapy for these patients, to reduce the tumor load and prolong life expectancy to the best extent.
In the current study, we summarized our experience of cryoablation in 1595 treated patients and tried to give a comprehensive profile about the effectiveness, safety and long-term outcome of cryoablation for HCC.
Materials and methods
Patients
The diagnosis of HCC was according to the criteria issued by the American Association for the Study of Liver Diseases or from cytohistological evidence [
9]. The inclusion criteria were: (1) unresectable tumor because of the number and size of HCCs; (2) tumor location and/or underlying illnesses that contraindicated surgery; (3) reluctant to undergo hepatic resection or liver transplantation; (4) recurrence after resection; (5) tumor with partial portal vein thrombosis (PVT) or vena cava invasion without distant metastasis and the patient wished to destroy the tumor to reduce tumor load, even though they would not normally be considered for local ablative therapy; (6) cirrhosis had been diagnosed based on a long history of chronic liver disease, clinical findings, imaging findings or endoscopic findings [
10], and Child-Pugh class A or B with life expectancy of at least 12 weeks, with a serum bilirubin level<51.3 μmol/L; and (7) Eastern Cooperative Oncology Group Performance Status (ECOG PS)≤2. The exclusion criteria were: (1) BCLC staging D; (2) ECOG PS>2; (3) Child-Pugh C; (4) tumor thrombosis at the main branch of the portal vein, and the size of the thrombosis; and (5) obvious cirrhotic ascites.
From January 2003 to December 2013, 3046 patients with HCC or metastatic liver tumor were consecutively referred to our center. After review by a multidisciplinary team, 1595 of them underwent cryoablation alone or in combination with other treatments; 1074 patients received single or multiple rounds of the relevant treatment, including 186 who underwent surgical resection, 35 who received liver transplantation, 192 RFA, 311 transarterial chemoembolization (TACE), 55 sorafenib alone, 179 RFA plus TACE, 116 TACE plus sorafenib, and 377 best supportive care [151 were Child-Pugh class C; 135 had a life expectancy<12 weeks and 91 were Eastern Cooperative Oncology Group (ECOG) performance status (PS)≥3]. The study met the requirements of the Declaration of Helsinki, and was approved by the Research Ethics Committee of the hospital. Written informed consent was obtained from all patients before treatment.
All patients underwent complete history taking, physical examination, essential laboratory tests, and imaging studies, including ultrasound (US), triphasic dynamic computed tomography (CT), and/or magnetic resonance imaging (MRI). The Barcelona Clinic Liver Cancer (BCLC) staging classification [
11] was applied to identify each patient’s tumor stage, and there were 99 patients in stage 0, 592 in stage A, 557 in stage B, and 347 in stage C.
Patients with a single HCC nodule≤5 cm or up to three nodules≤3 cm in diameter were intended to receive curative therapy after one or multiple cryoablations. For tumors that exceeded the criteria, cryoablation was given as part of the palliative treatment, which was usually followed by TACE and/or oral administration of multikinase inhibitor sorafenib after cryoablation. For TACE, superselective catheterization was performed directly on the artery supplying the tumor, wherever possible, to avoid hepatic failure and severe adverse events.
Cryoablation procedure
Cryoablation was performed with an argon-helium gas-based EndoCare system (EndoCare, Irvine, CA, USA) as previously described [
12,
13]. Various cryoprobes (2 or 3 mm in diameter), were applied, solely or multiply, according to the size and number of the nodules to be ablated. Generally, the formed freezing area covers both the whole tumor and at least 5–10 mm of the paraneoplastic hepatic tissue. After sonographic determination and local anesthesia, cryoprobes were inserted into the tumor under US or CT guidance through the most favorable percutaneous approach. The cryoprobe was advanced until the distal margin of the targeted lesion was reached. Tumors were then ablated under a dual freeze-thaw cycle comprising 20 min freezing, 10 min thawing, and a further 15 min freezing. After removal of the probes, all tracts were packed with Surgicel (Johnson & Johnson, Arlington, TX, USA) through the sheath introducer to control bleeding, and the sheath introducer was removed.
For patients with tumor>5 cm, it was probably that the tumor would not be ablated by one cryoablation session, therefore, an overlapping protocol was applied. Tumors were gradually ablated by multiple cryoablations that were performed every two weeks until elimination. A typical cryoablation procedure was shown in Fig. 1.
Follow-up
Treatment efficacy was assessed by MRI or triphasic CT scan 1 week after each cryoablation session. Complete response (CR) was defined as no CT enhancement in the entire tumor site. Tumors that failed to show a CR were defined as incomplete response (IR). For patients intended to receive curative treatment, the IR tumors were iteratively ablated up to three times, and tumors not eliminated after three cryoablation sessions were regarded as treatment failure (TF), and TACE and/or other treatments were considered. Serum α-fetoprotein (AFP) was also measured 2 weeks after ablation if the baseline level was abnormal and was rechecked every month thereafter.
Ectopic anti-tumor effect (EAE) is a unique postoperative phenomenon of cryoablation [
14], through which, residual tumors, not only the local ones, but the distant ones without ablation, manifest significant necrosis and shrunken. The release of tumor-specific antigens in the initial cryoablation was considered to subsequently evoke robust and long-term immunity against tumor cells in residual or distant untreated lesions. Generally, the EAE was assessed 8 weeks after cyroablation by MRI or spiral CT scan.
A unique complication of cryoablation, referred to as “cryoshock” elsewhere [
15], was particularly monitored in our study. Cryoshock was defined as the presence of multiorgan failure (at least 2), and/or severe coagulopathy and disseminated intravascular coagulation (DIC) after cryoablation.
All patients were followed up and received serial monitoring of serum AFP, chest radiographs, and CT or MRI scans every 3 months in the first year and twice a year thereafter to detect tumor recurrence. Local recurrence (LR) was defined as tumor recurrence within or at the periphery of the ablated lesion. Distant intrahepatic recurrence (DIR) was defined as new intrahepatic lesions that appeared in regions at least 1 cm away from the original ablated area. Extrahepatic recurrence (ER) referred to any tumor recurrence outside the liver. Patients who developed LR with or without DIR or ER were considered to have local tumor progression and patients who developed DIR or ER, or both, were considered to have distant recurrence.
Statistical analysis
Quantitative variables are expressed as medians or means and standard deviations, and qualitative variables are expressed as absolute frequencies or percentages. A P value<0.05 was considered statistically significant. Data analyses were performed using the IBM-SPSS Statistical Package version 20.0 (SPSS Inc, Chicago, IL, USA). Curves of OS rate were calculated by the Kaplan-Meier method.
Results
Patients and procedure efficacy
A total of 2313 tumors (mean size, 3.8±2.5 cm; range, 1.2–15 cm) in 1595 patients (1317 male, 278 female, mean age 53.1±10.2 years) were treated with cryoablation. Among them, 1069 patients had one tumor, 334 had two, and 192 had three. A total of 2958 cryoablations were performed, with one, two, three, four, five, six, seven, and eight procedures being performed in 774 (33.4%), 376 (23.6%), 373 (23.4%), 54 (3.4%), 14 (0.9%), two (0.1%), one (0.05%), and one (0.05%) patients, respectively. Triphasic CT or MRI scans showed that 1893 lesions in 1294 patients with a mean diameter of 3.4±2.2 cm were completely ablated, while 420 lesions with a mean diameter of 7.6±1.4 cm in 301 patients were not completely ablated. Thus, the CR rate was 81.2% (1893/2313), 99.4% (780/784), 94.4% (1622/1719), and 45.6% (271/594) in all tumors, tumors<3 cm, tumors<5 cm, and tumors>5 cm, respectively. The median follow-up was 33.4 months (range, 3.2–120 months). Fifty-two (3.2%) patients were lost to follow-up and they were censored at the date of their last visit. The baseline characteristics of all patients are summarized in Table 1.
The incidence of EAEs in all of the 347 patients with BCLC stage C was monitored, and there were 22 (6.3%) patients with EAEs. The mean OS of patients with post-cryoablation EAEs was significantly longer than that of patients without EAEs (14.5 months, range: 10.3–23.0 months vs. 9.5 months, range: 4.2–11.5 months, P = 0.0032). A representative case of EAE after cryoablation is shown in Fig. 2.
Mortality and complications
There was no treatment-related mortality. Cryoablation-related complications were assessed according to the patients’ complaints, US, and laboratory examinations (Table 2). Major complications were observed after 80 (3.4%) cryoablations; and minor complications after 330 (14.3%). None of the patients manifested cryoshock; nevertheless, several milder, nonspecific symptoms (e.g., chills, fever, tachycardia, tachypnea and temporary renal damage,), which we prefer to term as cryoreaction in distinction to cryoshock, were observed after 25 (1.08%) cryoablation sessions. Cryoreaction is also a major complication that usually needs prolonged in-hospital time, but caused no death. All the 25 patients recovered within 5 days with conservative treatment.
Consistent with our previous studies [
12,
13], postoperative pain, postoperative fever and transient elevations of alanine aminotransferase were the most common complications of cryoablation, which were usually self-limited and recovered within 2 weeks without treatment. Severe liver dysfunction occurred in 12 patients, and four of them had ascites. After treatment of 4–7 weeks, 10 patients recovered and two died of liver failure. All the 12 patients had advanced stage HCC with Child-Pugh class B liver function, and most of their tumors were not completely ablated. Other major complications included hepatic hemorrhage in nine patients, which were all diagnosed within 3–6 h after cryoablation. Tumor location was a critical factor that was associated with hepatic hemorrhage; two of the nine patients had a tumor in the caudate lobe of the liver, and the other seven had tumors near the liver capsule without sufficient surrounding liver parenchyma. Subcapsular hemorrhage distant from the ablation zone was detected in one patient. Fortunately, his liver capsule was intact, and after emergency TACE, the patient was fully recovered. Bowel injury and intestinal fistulas developed in one patient who received surgical repair, with a good outcome. Seven patients developed hemothorax that required drainage. Liver and pleural abscesses were observed in four and three patients, respectively, and all seven received drainage.
Follow-up and survival
Patients whose tumors were>5 cm, or were incompletely ablated, routinely received postoperative TACE (up to four times), and if they were BCLC stage C, oral administration of sorafenib. At the end of December 2013, 1134 of 1595 patients had died. As for the cause of death, 771 (68.1%) were due to HCC progression; the remaining 363 were unrelated to HCC, including esophagogastric variceal bleeding (n = 94, 25.9%), refractory ascites-induced renal failure (n = 69, 19.1%), liver failure (n = 80, 22.3%), and liver-unrelated diseases (n = 120, 33.1%). The median time from first cryoablation to death was 30.1 months (mean, 32.4±16. 6 months). The overall 1-, 3-, 5- and 10-year survival rates were 85.7%, 44.6%, 25.7% and 9.2%, respectively, with a mean survival of 36.9 months and median survival of 31.8 months (Fig. 3).
Tumor recurrence
HCC recurrence occurred in 937 patients who achieve CR from cryoablation. Among them, intrahepatic recurrence occurred in 787 (83.9%) patients, intrahepatic plus extrahepatic recurrence occurred in 122 (13.0%), and extrahepatic recurrence alone occurred in 28 (2.9%). Among 787 patients with intrahepatic recurrence, 306 (38.8%) developed local recurrence, whereas the other 481 (61.2%) developed distant intrahepatic recurrence. For the 150 patients who developed extrahepatic recurrence, 72 (48%) had pulmonary metastasis, 57 (38%) had retroperitoneal or celiac lymph node metastasis, 12 (8%) had bone metastasis, and nine (6%) had adrenal metastasis. The median duration from treatment to first local HCC recurrence was 18.4 months (mean, 21.1±20.4 months). The 1-, 3-, 5- and 10-year rates of local recurrence were 14.8%, 23.5%, 32.7% and 39.3%, respectively. The 1-, 3-, 5- and 10-year rates of non-local HCC progression were 17.4%, 28.1%, 34.2% and 42.3%, respectively. The median duration from cryoablation to first recurrence was 16.1 months (mean, 19.3±22.3 months).
Discussion
Cryoablation induces the death of tumor cells through intracellular ice crystal formation, followed by one or two cycles of solute-solvent switch, and tumor cells develop dehydration and rupture [
16]. Early cryoablation techniques applied liquid nitrogen and large cryoprobes which can cause damage to the liver and complications, thus, their utility in HCC is controversial and their use is largely restricted to laparoscopic or open surgery. However, with the significant advances in cryotherapy technology made in recent years, minimally invasive percutaneous approaches and a new generation of argon-helium cryo-equipment with thinner probes have been introduced clinically [
16], although data about the efficacy and safety of percutaneous cryoablation in HCC are limited. Nevertheless, cumulative evidence suggests that cryoablation is a feasible, safe, and effective treatment for HCC. In a pioneering work by Orlacchio and colleagues, the safety and efficacy of US- or CT-guided percutaneous cryoablation using 17-G cryoprobes were demonstrated in four patients with small HCCs. The CR rate was 100%, and all patients survived>6 months without severe complications [
17]. In another series, Shimizu
et al. assessed MRI-guided percutaneous cryoablation for HCC in 15 patients who had no more than three HCCs with a maximum size<5 cm, or more tumors but maximum size<3 cm. The CR rate was 80.8% and 1- and 3-year OS rates were 93.8% and 79.3%, respectively [
18]. Chen
et al. reported the outcomes of cryoablation in 40 patients with 76 HCCs. The 1- and 3-year OS rates were 81.4% and 60.3%, respectively. Complications occurred in 12.1% of the 40 patients with no procedure-related mortality [
19]. As demonstrated by our data, cryoablation is an effective local ablative therapy for HCC; the CR rate was 81.2% (1893/2313), 99.4% (780/784), 94.4% (1622/1719), and 45.6% (271/594) in all tumors, tumors<3 cm, tumors<5 cm, and tumors>5 cm, respectively. The rate was even better than that of RFA, which had a CR rate of 50%–80% in HCCs of 3–5 cm [
20], and could be as low as 25% in tumors>5 cm [
21]. Although RFA is more widely used than cryoablation in the treatment of HCC, cryoablation offers several potential advantages versus RFA. First, multiple cryoprobes can be used simultaneously to generate a large ice ball. Second, the size and shape of the cryotherapy generated ice ball can be readily visualized using intra-procedural computed tomography (CT), magnetic resonance (MR), or ultrasound (US) imaging. Third, in contrast to heat-based ablation, percutaneous cryoablation is a relatively painless procedure, and can therefore be performed under local anesthesia, rather than general anesthesia, which is required for RFA (Table 3). We therefore suggest cryoablation as first-line treatment for patients with HCC<3 cm, and for tumors of 3–5 cm, because>90% of them could be completely ablated, cryoablation can also be considered as first-line or adjuvant treatment. For tumors>5 cm, cryoablation is also helpful, and might be superior than other ablative therapies, in reducing tumor load and bridging patients to curative treatments or improving their quality of life.
Percutaneous argon-helium cryotherapy for HCC is minimally invasive. However, its safety has long been a concern and a major obstacle to the broader application of cryoablation in HCC [
3,
7]. For the lack of heat coagulation effect, cryoablation was pointed out to have a higher risk of post-cryoablation hemorrhage [
7]. Indeed, life-threatening hemorrhage was infrequently encountered in our early series [
13]. The most common cause of bleeding after cryoablation is needle tract related, and other possible risks include hypervascularity in the ablated tumor, and coagulopathy and thrombocytopenia resulting from underlying cirrhosis. However, by routinely using the needle tract hemostatic technology and the vascular sheath tamponade with stanching ghatpot, the incidence and severity of needle tract bleeding has been well controlled in our center. There were only nine cases of bleeding in our cohort, and compared to RFA, there was no significant risk of bleeding. It is worthy of mention that tumors under the liver capsule were more likely to cause post-cryoablation bleeding. In these cases, we suggest reaching the target tumor by passing through sufficient liver parenchyma, but not by direct puncture. Besides, intensive monitoring of vital signs should be carried out in the first 24 h after cryoablation to ensure timely diagnosis of bleeding and subsequent emergency interventions. Although liver rupture and bleeding are rare in percutaneous cryoablation, tumors located on the surface of the liver still carry a high possibility of liver rupture and should be avoided for cryoablation.
Cryoshock is a rare but fatal cryoablation-related complication that may present with multi-organ failure, severe coagulopathy, and disseminated intravascular coagulation. The incidence of cryoshock, according to a postal survey, was about 1% in patients who received cryosurgery [
15], however, the number in percutaneous cryoablation remains to be determined. We assumed that in the era of percutaneous cryoablation, cryoshock may be not be as critical as it used to be, not only because we failed to observe cryoshock after 2958 cryoablation sessions, but also in our entire treatment history of>2000 patients, cryoshock happened only once. The patient had lung cancer that was surgically resected, and cryoablation was carried out after a new metastatic lesion was found in the liver. Two days later, she developed thrombocytopenia, severe hemolysis, renal and respiratory failure, followed by death [
30].
EAE is an attractive feature of cryoablation. The mechanism of EAE is largely unknown but we have found reduced peripheral regulatory T cells [
31] and significantly suppressed PD-1/PD-L1 after cryoablation [
32]. Compared to other PLATs, cryoablation maximally reserves and releases the cellular components of dead tumor cells, which perfectly mimics
in situ vaccination and may give a boost to the immune system in HCC patients [
16]. This point is of importance when considering that most of the patients have chronic hepatitis with long-term immunosuppression. The current study, to the best of our knowledge, profiled the incidence of EAE in patients received cryoablation for the first time. According to our data, about 7% of patients gained extra benefit from cryoablation. We believe the number is expandable, in combination with novel adjuvants or immunotherapies in the future.
Tumor seeding is also an infrequent and nonspecific complication that may develop in all the ablative treatments using a percutaneous approach [
28]. According to the current study and one of our previous series [
29], the incidence of tumor seeding was 0.76%–0.9%, thus, tumor seeding could be a rare complication of cryoablation. The formation of tumor seeding was associated with direct puncture of the hepatic subcapsular tumor and could be reduced by applying a vascular sheath system. Besides, other complications, including liver abscess and neighboring organ damage are all rare and preventable in cryoablation [
12,
13].
In conclusion, the present study clearly indicated the clinical benefits and safety of cryoablation in treating HCC, including advanced stage disease. There is currently no evidence linking cryoablation with more severe complications compared to other ablative therapies. With careful patient selection, comprehensive preparation and accumulated experience, cryoablation holds strong potential to be an essential player in the local ablative treatment of HCC.
Higher Education Press and Springer-Verlag Berlin Heidelberg
PDF
(1938KB)
