Interventional Radiology - Original Article

Long-term results of liver thermal ablation in patients with hepatocellular carcinoma and colorectal cancer liver metastasis regarding spatial features and tumor-specific variables

10.4274/dir.2023.221986

  • Okan Akhan
  • Ahmet Gürkan Erdemir
  • Sevilay Karahan
  • Emre Ünal
  • Türkmen Turan Çiftçi
  • Devrim Akıncı
  • Şuayib Yalçın

Received Date: 29.10.2022 Accepted Date: 08.02.2023 Diagn Interv Radiol 2024;30(3):183-192 PMID: 36994643

PURPOSE

Colorectal cancer liver metastasis (CRLM) and hepatocellular carcinoma (HCC) are widely treated using microwave and radiofrequency ablation. Local tumor progression (LTP) may develop depending on the shortest vascular distance and large lesion diameter. This study aims to explore the effect of these spatial features and to investigate the correlation between tumor-specific variables and LTP.

METHODS

This is a retrospective study covering the period between January 2007 and January 2019. One hundred twenty-five patients (CRLM: HCC: 64:61) with 262 lesions (CRLM: HCC: 142:120) were enrolled. The correlation between LTP and the variables was analyzed using the chi-square test, Fischer’s exact test, or the Fisher-Freeman-Halton test where applicable. The local progression-free survival (Loc-PFS) was analyzed using the Kaplan–Meier method. Univariable and multivariable Cox regression analyses were performed to identify prognostic factors.

RESULTS

Significant correlations were observed for LTP in both CRLM and HCC at a lesion diameter of 30–50 mm (P = 0.019 and P < 0.001, respectively) and SVD of ≤3 mm (P < 0.001 for both). No correlation was found between the ablation type and LTP (CRLM: P = 0.141; HCC: P = 0.771). There was no relationship between residue and the ablation type, but a strong correlation with tumor size was observed (P = 0.127 and P < 0.001, respectively). In CRLM, LTP was associated with mutant K-ras and concomitant lung metastasis (P < 0.001 and P = 0.003, respectively). In HCC, a similar correlation was found for Child–Pugh B, serum alpha-fetoprotein (AFP) level of >10 ng/mL, predisposing factors, and moderate histopathological differentiation (P < 0.001, P = 0.008, P = 0.027, and P < 0.001, respectively). In CRLM, SVD of ≤3 mm proved to be the variable with the greatest negative effect on Loc-PFS (P = 0.007), followed by concomitant lung metastasis (P = 0.027). In HCC, a serum AFP level of >10 ng/mL proved to be the variable with the greatest negative effect on Loc-PFS (P = 0.045).

CONCLUSION

In addition to the lesions’ spatial features, tumor-specific variables may also have an impact on LTP.

Keywords: Ablation techniques (D055011), colorectal neoplasm (D015179), hepatocellular carcinoma (D006528), local tumor progression (D009364), survival analysis (D016019)

Main points

• Morphometric features of the lesion, such as larger diameters or shorter vascular prox­imity, were an effective factor in local tumor progression.

• Colorectal cancer liver metastasis, concom­itant lung metastasis, and a host-specific variable had the greatest impact on local progression-free survival after short vascu­lar proximity.

• In hepatocellular carcinoma, a serum al­pha-fetoprotein level of >10 ng/mL proved to be the variable with the greatest negative effect on local progression-free survival.


Colorectal cancer liver metastasis (CRLM) and hepatocellular carcinoma (HCC) are widely treated using local ablation, which provides an increased survival outcome.1,2,3 Both radiofrequency ablation (RFA) and microwave ablation (MWA), the most commonly used ablation techniques, cause necrosis through elevated temperatures, but they have different physical parameters that offer advantages in certain situations.4,5

After local ablation procedures, local tumor progression (LTP) may be confronted out of favor.6,7 This situation is more frequent in large tumors that exceed the perimeter of the ablation zone and in tumors with a blood vessel in close proximity (heat-sink effect).8,9 Of these two variables, tumors with a blood vessel in close proximity are considered the highest risk factor for the development of LTP.5,10,11 Given the extensive literature addressing the development of LTP, it is possible that several non-spatial variables also influence this development in malignancies with different pathogenesis, such as mutant K-ras oncogene in colorectal carcinoma (CRC) and alpha-fetoprotein (AFP) levels before ablation in HCC.

This study has three main objectives. First, to investigate the correlation between common variables (tumor diameter, shortest vascular distance, and the ablation type) and LTP with local progression-free survival (Loc-PFS), which corresponds to the period without LTP. Second, to investigate the association with K-ras mutation, primary tumor location (left- or right-sided), and concomitant lung metastases in CRLM as tumor-specific variables. Third, to investigate the association between the Child–Pugh score, histopathological differentiation grade, serum AFP level, and predisposing factors for chronic liver disease in HCC as tumor-specific variables.


Methods


Study design

This study is a retrospective analysis of liver lesions that received RFA or MWA between January 2007 and January 2019 due to CRLM or HCC. The Hacettepe University Faculty of Medicine Ethics Committee of the Institute approved this study (GO-18/429).

The decision for each thermal ablation was made by the multidisciplinary institutional tumor board, and informed consent was obtained from all enrolled patients.

Imaging-guided ablation therapies were defined according to publications developed by the “International Working Group on Image-Guided Tumor Ablation” and “Results of the SIO and DATECAN Initiative”.12,13


Inclusion criteria

The inclusion criteria for both CRLM and HCC lesions were as follows:

1. Maximum of five liver lesions for each patient with CRC and a maximum of three lesions for each patient with HCC,

2. Maximum diameter of 5 cm for each lesion,

3. Curative intent (the ablation of all liver lesions in the same session),

4. Presence of magnetic resonance imaging (MRI) images within 2 months before the ablation,

5. Presence of follow-up MRI or computed tomography (CT) imaging at 1, 3, 6, and 12 months and semiannually after the first year.

Furthermore, the presence of lung metastasis was not an exclusion criterion.

Seventy-one patients with CRC and 67 patients with HCC who underwent ablation with “curative intent” were identified. However, due to insufficient follow-up, seven patients with CRC (9.85%) and six patients with HCC (8.95%) were excluded. Finally, the remaining 64 patients with CRC (142 lesions) and 61 patients with HCC (120 lesions) were enrolled in this study.

For further information please see the flowchart (Figure 1).


Ablation procedure and follow-up

All procedures were done with ultrasound guidance. StarBurst® (AngioDynamics®) electrodes were used for RFA, while Acculis®/Solero® (AngioDynamics®) antennas were used for MWA. All procedures were performed according to the manufacturer’s instructions, with an ablation margin of at least 5–10 mm.14

After the ablation, contrast-enhanced CT and MRI were performed within the first month. Patients with no residual disease were accepted as “complete ablation”. Tumor development during follow-up in patients with complete ablation was classified as LTP. Loc-PFS was calculated for each lesion, starting with ablation until the development of LTP or patient death, and it was censored at the last follow-up date. The censor date for Loc-PFS estimation was February 2, 2020.


Data collection

All data were reviewed and collected with the consensus of two radiologists (A.G.E. and O.A.) at two different time points to ensure external and internal validity in both patient selection and data collection.

The segmental distribution, LTP development, thermal ablation type (MWA or RFA), and shortest vascular distance of each lesion were recorded. The measurement of the shortest vascular distance was performed on volumetric dynamic T1W slices from the patient’s last MRI before ablation. The shortest perpendicular distance to the vessel with a width of ≥3 mm was estimated through multiplanar reformation images (Figure 2). The longest axial and craniocaudal diameter of each lesion was also recorded.

The complications, ablation type and technique (percutaneous or intraoperative), and the segment of the relevant lesion were recorded.

The presence of K-ras mutation, the site of primary disease (right or left colon), and concomitant lung metastasis were considered CRC-specific variables. In the HCC group, predisposing factors (non-alcoholic steatohepatitis and hepatitis B or C virus), degree of histopathologic differentiation, AFP level, and Child–Pugh score within one month before ablation were recorded.

Histopathologic diagnoses were available in both the CRC and HCC groups that participated in this study. However, in patients with CRC with multiple liver metastases, only one of the lesions was biopsied. In addition, K-ras mutations were analyzed using DNA derived from formalin-fixed paraffin-embedded tumors obtained from primary sites in the colon. The presence of microsatellite instability was also analyzed.


Statistical analysis

The data were processed using the IBM-SPSS® Statistics 24.0, StataCorp LCC-STATA® 14 software, and R® version 4.0.3. Categorical variables were reported as frequencies and percentages, and continuous variables were reported as means and standard deviations.

Categorical variables were evaluated using the chi-square test or the Fisher-Freeman-Halton test where applicable. For all tests, a two-tailed P value of less than 0.05 was considered statistically significant.

The Kaplan–Meier method was used for estimates of Loc-PFS, and the log-rank test was used to compare survival groups. Cox regression models were used to assess the effects of confounding factors on overall survival. Variables with a P value of <0.20 in the univariable analyses were analyzed in multivariable Cox regression models to explore prognostic factors of overall survival. The results are reported with hazard ratios and 95% confidence intervals.15


Results


Background data and complications

The detailed baseline characteristics of 262 lesions and 125 patients are shown with all aspects in Table 1.

The complications of the included patients were biliary obstruction, abscess, and costochondritis (Figures 3, 4, 5). Seventeen lesions (out of 262 lesions, 6.48%) were complicated: eight of them were percutaneous (out of 198 lesions, 4.04%) and nine of them were intraoperative (out of 64 lesions, 14.06%). A significant correlation was found between intraoperative ablation and the occurrence of complications: when all 17 complications were included and when only the abscess [three percutaneous (3/198 = 1.51%) and six intra-operative (6/64 = 9.37%) lesions] were included (P < 0.005). Six of the nine abscesses had a history of hepaticojejunostomy (two lesions) and endoscopic sphincterotomy (four lesions) due to gallstones. All lesions (n = 7, 2.67%) that developed biliary dilatation were in the central segments (segments 1, 4b, and 5) (P < 0.001). A transient costochondritis complication was observed in only one patient with a subcapsular localized lesion in segment eight.

Residue occurred in three CRLM (MWA: RFA: 0:3) and four (MWA: RFA: 1:3) HCC lesions, and six of seven residual lesions (out of 262 lesions) were observed after RFA (P = 0.127). Moreover, all six of them had a diameter of 30–50 mm (P < 0.001). For all lesions with residual occurrences, that were reablated with complete ablation, were included in the cohort from the time of complete ablation.

More detailed information on complications and residue is shown in Table 2.


Correlations of common variables with local tumor progression development and local progression-free survival

Regarding the ablation type (MWA or RFA), no statistically significant difference was found for LTP development and Loc-PFS in the CRC group (P = 0.141 and P = 0.161, respectively). In the HCC group, no correlation was found between the development of LTP and Loc-PFS considering the ablation type (P = 0.771 and P = 0.699, respectively).

The development rate of LTP in CRLM was statistically significant in those with a lesion diameter of 30–50 mm (P = 0.019). Loc-PFS also decreased in this group but failed to reach a statistically significant result (P = 0.085). In HCC lesions with a lesion diameter of 30–50 mm, a statistically significant correlation was observed between both LTP development and Loc-PFS (P < 0.001 for both).

The shortest vascular distance of ≤3 mm in both HCC and CRLM was statistically associated with both LTP development (P < 0.001 for each group) and decreased Loc-PFS (P < 0.001 and P = 0.014, respectively).

More detailed information on the common variables of both groups can be found in Tables 3 and 4.


Multivariable analysis and correlations of colorectal carcinoma-specific variables with local tumor progression and local progression-free survival

In the Cox regression analysis for the CRC-specific variable model (Supplementary Table 1), the P value was 0.0006.

Mutated K-ras oncogene was found to be statistically correlated with both LTP development and decreased Loc-PFS (P < 0.001 and P = 0.021, respectively). Similar results were observed with the existence of concomitant lung metastasis for both LTP development and decreased Loc-PFS (P = 0.003 and P = 0.044, respectively). Although LTP development and decreased Loc-PFS were more associated with right-sided CRLM, no statistically significant results were observed (P = 0.064 and P = 0.358, respectively).

Although microsatellite instability was also analyzed in all patients, it was not detected in any of them.

In the multivariable analysis for CRLMs, the shortest vascular distance of ≤3 mm was found to be the variable with the largest negative effect on Loc-PFS (P = 0.007), followed by concomitant lung metastasis (P = 0.027).

More detailed information on CRC-specific variables and multivariable analysis can be found in Table 3 and Supplementary Table 1.


Multivariable analysis and correlations of hepatocellular carcinoma-specific variables with local tumor progression and local progression-free survival

In the Cox regression analysis for the HCC-specific variable model (Supplementary Table 1), the P value was 0.0002.

Child–Pugh B, a serum AFP level of >10 ng/mL, and moderate histopathological differentiation showed a highly significant statistical correlation with both LTP development and decreased Loc-PFS (P < 0.001, P = 0.008, and P < 0.001, respectively).

Poor histopathologic differentiation was not observed in the entire HCC cohort.

The LTP development rate in lesions with HBV was statistically significant (P = 0.027). However, although Loc-PFS decreased in this group, no statistically significant results were obtained (P = 0.210).

In the multivariable analysis for HCC lesions, a serum AFP level of >10 ng/mL was found to be the variable with the largest negative effect on Loc-PFS (P = 0.045).

More detailed information on HCC-specific variables and multivariable analysis can be found in Table 4 and Supplementary Table 1.


Discussion

This study’s results showed that in both CRLM and HCC lesions, albeit with lower LTP rates in the lesions treated with MWA, there was no significant difference between RFA and MWA. In contrast, two other common variables in both lesion groups were statistically associated with LTP: the shortest vascular distance of ≤3 mm and a lesion diameter of 30–50 mm. Important results were obtained from the observations for tumor-specific variables. A significant correlation of CRC-specific variables was observed with mutant K-ras and concomitant lung metastasis, while the same was observed for HCC-specific variables with Child–Pugh B, a serum AFP level of >10 ng/mL, HBV, and moderate histopathological differentiation.

Extensive meta-analyses have shown that the most important difference in the clinical outcome between MWA and RFA is the size of the larger liver lesion treated, with RFA having some possible disadvantages over LTP.9,16,17 In this study, although residues were seen more frequently in tumors treated with RFA, all of these lesions were 30–50 mm in diameter. Although no statistically significant results were obtained in this cohort, LTP was more frequent, and Loc-PFS was shorter in patients treated with RFA. In accordance with this study, numerous articles have been published in the literature showing the association of tumor size and shortest vascular distance with LTP.7,11,18,19,20

Recent retrospective studies have shown a strong correlation between the K-ras mutation, which is one of the CRC-specific variants, and LTP.21,22 In the study by Jiang et al.23, which is one of the most recent studies conducted in this context, similar results were obtained, but they only included lesions with RFA. The second tumor-specific variable studied in CRLM lesions was the primary origin of the tumor. There are few studies in the literature that address primary origins. Zhou et al.24, who studied patients with MWA, and Gu et al.25, who studied patients with RFA, conducted their studies considering “patient-based” survival and observed better outcomes in patients with left-sided primary origin. In this “lesion-based” study, where more variables were considered, LTP was observed more frequently in CRLMs originating from the right colon. However, the study failed to achieve significant results. Concomitant lung metastasis, another CRC-specific variable, is one of the most important variables affecting survival and LTP.26,27 In the study by Shady et al.26, which only included patients with RFA, the presence of lung metastases was targeted as one of the most important prognostic factors. In this study, which included more comprehensive variables, the presence of lung metastases had an impact on LTP and Loc-PFS, and it proved to be more important than the primary origin of metastases and K-ras mutations. This suggests that concomitant lung metastases may be an important overall indicator of aggressive neoplastic behavior. Moreover, in the multiples analysis, it was found to be the second most important factor in lowering Loc-PFS after the shortest vascular distance of ≤3 mm.

There are numerous articles in the literature that include HCC-specific variants. One of the largest prospective studies that included patients with RFA and MWA, by Chong et al.28 and Vietti Violi et al.29, examined predisposing factors, the Child–Pugh score, and AFP levels, but they were not included in the statistical analysis. In another study comparing RFA with liver resection, in which 109 patients were treated with RFA, these three HCC-specific variables were included, and no effect of these three variables on “disease-free survival” was reported.30 In a study examining 48 lesions with RFA, in addition to these three HCC-specific variables, the degree of histopathological differentiation was also included, of which only a high AFP level before ablation was correlated with “intrahepatic distant recurrence”.31 There are other studies that correlate with higher AFP levels.32,33 In this study, a correlation was found between all these four variables and LTP. In addition, the multiples analysis revealed that the AFP level was the most important variable affecting the poor Loc-PFS outcome. This result is valuable in that it indicates that a host factor such as the AFP level is an important poor prognostic factor that outperforms even a tumor-based variable such as the shortest vascular distance of ≤3 mm or large tumor size.

Complications were also investigated in this study as ancillary findings. Previous retrospective studies have shown that there was no difference in safety between ablation types.8,34,35 A significant association was found between the occurrence of complications and intraoperative ablation, either when only the abscess or all complications were included. This could be due to a more invasive procedure and greater surface area of the peritoneum. There was also a strong correlation between the dilation of the bile duct and the ablation of the central segments (segments 1, 4b, and 5). It is understandable that the ablation of zones closer to the portal hilum may lead to this biliary obstruction. Costochondritis was observed in only one patient with a subcapsular lesion. The ablation of a subcapsular lesion in close proximity to the costochondral arcus may have caused this inflammation. This is the first time such a case was reported with the corresponding images.

This study has some limitations. First, it is a single-center, retrospective study. However, it represents the results of a large tertiary oncology center with a long-established thermal ablation protocol. Second, although an inspection was carried out, no patients with CRC with microsatellite instability and no patients with HCC with poor histopathological morphology were detected. Third, there were only three predisposing factors for chronic liver disease, and no other chronic liver diseases were included. However, this study provides a suitable basis for future thermal ablation studies to include more tumor-specific variables.

In conclusion, large tumor size and the shortest vascular distance of ≤3 mm are important factors with effects on LTP. However, host variables such as concomitant lung metastasis in patients with CRC and high pre-ablation AFP levels in patients with HCC may be important indicators of poor prognosis. Prospective randomized studies with tumor-specific variables and spatial characteristics are needed to explain the exact effects.


Conflict of interest disclosure

The authors declared no conflicts of interest.

  1. Van Cutsem E, Cervantes A, Adam R, et al. ESMO consensus guidelines for the management of patients with metastatic colorectal cancer. Ann Oncol. 2016;27(8):1386-422.
  2. Liver EAFTSOT. EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69(1):182-236.
  3. Minagawa M, Makuuchi M, Torzilli G, et al. Extension of the frontiers of surgical indications in the treatment of liver metastases from colorectal cancer: long-term results. Ann Surg. 2000;231(4):487-499.
  4. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation: standardization of terminology and reporting criteria--a 10-year update. J Vasc Interv Radiol. 2014;25:1691-1705.
  5. Kim C. Understanding the nuances of microwave ablation for more accurate post-treatment assessment. Future Oncol. 2018;14:1755-1764.
  6. Chinnaratha MA, Chuang MyA, Fraser RJ, Woodman RJ, Wigg AJ. Percutaneous thermal ablation for primary hepatocellular carcinoma: a systematic review and meta‐analysis. J Gastroenterol Hepatol. 2016;31(2):294-301.
  7. Shady W, Petre EN, Do KG, et al. Percutaneous microwave versus radiofrequency ablation of colorectal liver metastases: ablation with clear margins (A0) provides the best local tumor control. J Vasc Interv Radiol. 2018;29(2):268-275.
  8. Poulou LS, Botsa E, Thanou I, Ziakas PD, Thanos L. Percutaneous microwave ablation vs radiofrequency ablation in the treatment of hepatocellular carcinoma. World J Hepatol. 2015;7(8):1054-1063.
  9. Facciorusso A, Di Maso M, Muscatiello N. Microwave ablation versus radiofrequency ablation for the treatment of hepatocellular carcinoma: a systematic review and meta-analysis. Int J Hyperthermia. 2016;32(3):339-344.
  10. Pathak S, Jones R, Tang JM, et al. Ablative therapies for colorectal liver metastases: a systematic review. Colorectal Disease. 2011;13(9):252-265.
  11. Berber E, Siperstein A. Local recurrence after laparoscopic radiofrequency ablation of liver tumors: an analysis of 1032 tumors. Ann Surg Oncol. 2008;15(10):2757-2764.
  12. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation: standardization of terminology and reporting criteria -- a 10-year update. Radiology. 2014;273(1):241-260.
  13. Puijk RS, Ahmed M, Adam A, et al. Consensus guidelines for the definition of time-to-event end points in image-guided tumor ablation: Results of the SIO and DATECAN initiative. Radiology. 2021;301(3):533-540.
  14. Crocetti L, De Baere T, Lencioni R. Quality improvement guidelines for radiofrequency ablation of liver tumours. Cardiovasc Interv Radiol. 2010;33(1):11-17.
  15. Hosmer Jr DW, Lemeshow S, Sturdivant RX. Applied logistic regression, 3rd ed. John Wiley & Sons; 2013.
  16. Glassberg MB, Ghosh S, Clymer JW, et al. Microwave ablation compared with radiofrequency ablation for treatment of hepatocellular carcinoma and liver metastases: a systematic review and meta-analysis. OncoTargets Ther. 2019;12:6407-6438.
  17. Luo W, Zhang Y, He G, et al. Effects of radiofrequency ablation versus other ablating techniques on hepatocellular carcinomas: a systematic review and meta-analysis. World J Surg Oncol. 2017;15(1):126.
  18. Veltri A, Gazzera C, Calandri M, et al. Percutaneous treatment of Hepatocellular carcinoma exceeding 3 cm: combined therapy or microwave ablation? Preliminary results. Radiol Med. 2015;120(12):1177-1183.
  19. Di Vece F, Tombesi P, Ermili F, Maraldi C, Sartori S. Coagulation areas produced by cool-tip radiofrequency ablation and microwave ablation using a device to decrease back-heating effects: a prospective pilot study. Cardiovasc Intervent Radiol. 2014;37(3):723-729.
  20. Geyik S, Akhan O, Abbasoglu O, et al. Radiofrequency ablation of unresectable hepatic tumors. Diagn Interv Radiol. 2006;12(4):195-200.
  21. Calandri M, Odisio BC. Tailoring ablation strategies for colorectal liver metastases based upon rat sarcoma viral oncogene mutation status. Chin Clin Oncol. 2019;8(5):51.
  22. Rhaiem R, Rached L, Tashkandi A, Bouché O, Kianmanesh R. Implications of RAS Mutations on Oncological Outcomes of Surgical Resection and Thermal Ablation Techniques in the Treatment of Colorectal Liver Metastases. Cancers. 2022;14(3):816.
  23. Jiang BB, Yan K, Zhang ZY, et al. The value of KRAS gene status in predicting local tumor progression of colorectal liver metastases following radiofrequency ablation. Int J Hyperthermia. 2019;36(1):211-219.
  24. Zhou F, Yu X, Liang P, et al. Does primary tumor location impact the prognosis of colorectal liver metastases patients after microwave ablation?-Lessons from 10 years’ experience. Oncotarget. 2017;8:100791-100800.
  25. Gu Y, Huang Z, Gu H, Gao F, et al. Does the site of the primary affect outcomes when ablating colorectal liver metastases with radiofrequency ablation? Cardiovasc Intervent Radiol. 2018;41(6):912-919.
  26. Shady W, Petre EN, Gonen M, et al. Percutaneous radiofrequency ablation of colorectal cancer liver metastases: factors affecting outcomes -- a 10-year experience at a single center. Radiology. 2016;278:601-611.
  27. Merkel S, Weber K, Croner R, et al. Distant metastases in colorectal carcinoma: a proposal for a new M1 subclassification. Eur J Surg Oncol. 2016;42:1337-1342.
  28. Chong CCN, Lee KF, Cheung SYS, et al. Prospective double-blinded randomized controlled trial of Microwave versus RadioFrequency Ablation for hepatocellular carcinoma (McRFA trial). HPB (Oxford). 2020;22(8):1121-1127.
  29. Vietti Violi N, Duran R, Guiu B, et al. Efficacy of microwave ablation versus radiofrequency ablation for the treatment of hepatocellular carcinoma in patients with chronic liver disease: a randomised controlled phase 2 trial. Lancet Gastroenterol Hepatol. 2018;3(5):317-325.
  30. Ng K, Chok K, Chan A, et al. Randomized clinical trial of hepatic resection versus radiofrequency ablation for early-stage hepatocellular carcinoma. J Br Surg. 2017;104(13):1775-1784.
  31. Zytoon AA, Ishii H, Murakami K, et al. Recurrence-free survival after radiofrequency ablation of hepatocellular carcinoma. A registry report of the impact of risk factors on outcome. Japan J Clin Oncol. 2007;37(9):658-672.
  32. Izumi N, Asahina Y, Noguchi O, Uchihara M, et al. Risk factors for distant recurrence of hepatocellular carcinoma in the liver after complete coagulation by microwave or radiofrequency ablation. Cancer. 2001;91(5):949-956.
  33. Harrison LE, Koneru B, Baramipour P, et al. Locoregional recurrences are frequent after radiofrequency ablation for hepatocellular carcinoma. J Am Coll Surg. 2003;197(5):759-764.
  34. Lahat E, Eshkenazy R, Zendel A, et al. Complications after percutaneous ablation of liver tumors: a systematic review. Hepatobiliary Surg Nutr. 2014;3(5):317-323.
  35. Hauser K, Matthes J. Medical students’ medication communication skills regarding drug prescription - qualitative analysis of simulated physician-patient consultations. Eur J Clin Pharmacol. 2017;73(4):429-435.