Percutaneous Arteriovenous Fistula Creation with the WavelinQ 4-French EndoAVF System: A Single-Center Retrospective Analysis of 30 Patients

Clinical question
To assess the safety and efficacy of percutaneous arteriovenous fistula (pAVF) creation with WavelinQ 4F EndoAVF System.

Take away point
The WavelinQ 4F EndoAVF is a safe and effective system to create pAVFs with high maturation rate and long-term patency.

Reference
Kitrou PM, Balta L, Papachristou E, Papasotiriou M, Katsanos K, Theofanis M, Papadoulas S, Anagnostopoulos F, Georgopoulou GA, Goumenos D, Karnabatidis D. Percutaneous Arteriovenous Fistula Creation with the WavelinQ 4-French EndoAVF System: A Single-Center Retrospective Analysis of 30 Patients. J Vasc Interv Radiol. 2022 Jan;33(1):33-40. doi: 10.1016/j.jvir.2021.09.021.

Click here for abstract  

Study design
Single-center retrospective analysis with 30 patients requiring pAVFs for long-term hemodialysis.

Funding Source
The authors of this study receive consultancy fees and lecture honoraria from BD.

Setting
Academic setting. Patras University Hospital, Patras, Greece.

Figure


Figure 2a and b. Catheter alignment and different access options (only parallel approach is shown). Images showing a parallel brachial access (brachial artery and vein) of the catheters aiming to create a percutaneous arteriovenous fistula at the ulnar side (a, b). The images show the distance between the electrode (arrow) and the ceramic backstop before (a) and after (b) activation. The distance between the catheters in image (a) reflects the walls of the vessels and the tissue between the vessels. Following activation and anastomosis creation, the electrode is in contact with the ceramic backstop. In the image (a), the square orientation of the indicators at the front tip of the catheters is shown (arrowhead). 

Figure 3. Final angiogram of a procedure performed with brachial arterial access. Contrast was injected from the brachial artery. The different parts of the newly formed vascular access circuit are pointed out. 1, brachial artery; 2, radial artery; 3, anastomosis; 4, radial vein (with some spasm); 5, perforator vein; 6, cephalic vein; and 7, basilic vein. The ulnar artery (A), main ulnar artery (B), and interosseous artery (C) outside the vascular access circuit are also apparent.

Figure 4. Kaplan-Meier survival curve of the 26 cases that matured and achieved cannulation with subjects at risk.

Summary

Percutaneous arteriovenous fistulas (pAVFs) are a new endovascular technique to create hemodialysis access for patients with chronic kidney failure. Two systems are currently available: the WavelinQ EndoAVF and the Ellipsys Vascular Access System. The WavelinQ creates an anastomosis between the radial and ulnar vessels, while the Ellipsys System uses the perforating vein and proximal radial artery. pAVFs offer an additional technique for long-term hemodialysis access and may have similar outcomes with surgical arteriovenous fistula, however limited data has been published about the safety and efficacy.

Thirty patients met the inclusion criteria of being >18 years old, Stage V CKD, Target A/V diameters >2.5mm with perforating forearm vein of >2mm. Patient with active hypercoagulable state, infection, extensive vascular calcification, and contrast allergy were excluded from the study The study was a single-center, single-arm retrospective study of arteriovenous fistulas created by the 4Fr WavelinQ system.

• 100% (30/30) patients had technical success in pAVF creation.
• In thirty patients that underwent pAVF creation, the cannulation rate was 87% at mid-term follow-up time (mean f/u time 547 +/- 315 days).
• 16 reinterventions (i.e. coil embolization or angioplasty) were necessary to facilitate cannulation without need for open surgical procedures.
• 7 maintenance procedures were performed in 6 of 27 matured pAVFs.
• 22 out of 33 (73%) pAVFs were non-thrombosed and cannulated at the end of the study.
• Relatively safe with adverse event rate (i.e. brachial artery pseudoaneurysm) in 2/30 patients (6.7%).

Commentary

This study is one of the first to address the safety and efficacy of pAVFs. One distinct advantage of pAVFs is that the procedure does not compromise the creation of future surgical AVFs at the same site. While this technique is not risk-free, it appears that it is a safe alternative for providing hemodialysis access to patients with CKD. Other studies with the WavelinQ EndoAVF system have validated the high success rate and efficacy: 

• Endovascular Access System Enhancements study (Berland et al. 2019; 4Fr) was the first to evaluate the current 4Fr system and reported 100% technical success in 32 patients with both ulnar and radial AVFs.
• The FLEX study (Rajan et al. 2015; 6Fr) enrolled 33 pts with a success rate of 97% and one adverse effect (brachial pseudoaneurysm). 
• The Novel Endovascular Access Trial (Lok et al. 2017; 6Fr) with 60 patients demonstrated a high success rate of 98% with only 2% adverse effects in ulnar pAVFs.
• Zemela et al. 2021 (6Fr) with 35 patients reported technical success rate of 100% and one pseudoaneurysm.
• Radosa et al. 2017 (6Fr) with 8 patients reported technical success rate of 100%.

A recent study by Inston et al. compared pAVRs (with both generations of the WavelinQ system) and surgical radiocephalic AVFs and reported comparable patency rates. A study by Shahverdyan demonstrated that WavelinQ and Ellipsys are no different in outcome. A Multicenter European/Canadian discovered that coil embolization at index reduced further need to treat. 

Similar to other studies on novel techniques of percutaneous arteriovenous fistula creation, this study's clinical value remains uncertain and practice adoption of the studied devices/techniques difficult given the small sample size, selection bias, inconsistent follow-up duration, procedural and fistula metrics incompleteness, and lack of prospective control arm(s). The authors rightfully described their findings as early experience, limited in scope, and will require external larger prospective clinical trials for validation.

Post Author
Brian Stephen Wong, MD MPH
PGY-3 Diagnostic Radiology Resident

University of Texas Medical Branch

Edited and formatted by @NingchengLi

Interventional Radiology Resident

Dotter Institute, Oregon Health and Science University

Thrombotic Risk Associated with Inferior Vena Cava Filter Placement in Patients with Heparin-Induced Thrombocytopenia

Clinical question
Are mortality and thrombotic risk increased in patients with heparin-induced thrombocytopenia (HIT) undergoing IVC filter (IVCF) placement compared to patients with either HIT or IVCF alone?

Take away point
IVC filter placement did not significantly increase risk of thromboembolism or death in patients with HIT, suggesting IVCF placement may be a viable treatment option for patients with elevated thromboembolic risk secondary to HIT in whom anticoagulation has failed or is contraindicated.

Reference
Moghbel M, Chen Z, Liu C, Rajan S, Vempaty H, and Wang S. Thrombotic Risk Associated with Inferior Vena Cava Filter Placement in Patients with Heparin-Induced Thrombocytopenia. J Vasc Interv Radiol. 2021; 32:1629-1634. doi.org/10.1016/j.jvir.2021.08.025

Click here for abstract

Study design
Retrospective, single-network cohort study of 26 patients with HIT and IVCF placement vs controls with either HIT or IVCF alone.

Funding Source
No reported funding.

Setting
Private hospital network, Kaiser Permanente Northern California, USA.

Figure

No significant difference in mortality was observed in the HIT+IVCF group versus controls.

Summary

Heparin-induced thrombocytopenia (HIT) is an adverse reaction to heparin resulting in reduced platelet count. Thromboembolic events occur in up to half of patients with HIT. Currently, it is not recommended to routinely place IVC filters (IVCFs) in patients with HIT in whom anticoagulation has failed or is contraindicated due to the paucity of outcomes data in this relatively small population. The authors perform a retrospective cohort study comparing outcomes in patients with HIT receiving IVCFs compared to patients with HIT who did not receive an IVCF and patients receiving an IVCF without a diagnosis of HIT.

Electronic medical records were queried for patients with HIT and/or IVCFs using diagnostic and procedural codes. Patient charts coded with HIT were reviewed manually and excluded if they had been miscoded. The two control groups were unable to be manually reviewed due to the immensity of the sample sizes. Patients in the HIT+IVCF group were stratified into subgroups based on likelihood of HIT (high likelihood = positive serotonin release assay or HIT antibody level >1; intermediate likelihood = HIT antibody level <1) as well as timing of IVCF placement (within 14 days of HIT diagnosis versus >14 days of HIT diagnosis). This 14-day cutoff was chosen based on estimated peak timing of HIT-associated thromboembolic risk.

Primary outcomes were 6- and 12-month mortality and thromboembolic risk, which included new or extended DVT/PE, IVC thrombosis, lower extremity phlegmasia cerulea dolens, and critical limb ischemia. Propensity score matching was performed to control for potential confounders based on demographic data. Comparisons were made using chi-squared and Fisher exact tests.

A total of 4774 patients were analyzed, 26 with HIT+IVCF, 814 with HIT without IVCF , and 3,934 with IVCF without HIT. The most common indication for IVCF placement in the HIT+IVCF group was proximal DVT/PE with contraindication to anticoagulation (n=18, 69.2%).

Mortality rates in the HIT+IVCF group were 26.9% (n=7) at 6 months and 30.8% (n=8) at 12 months compared to 29.5% (n=240) and 34.9% (n=284) in the HIT only group and 33.1% (n=1,288) and 38.2 (n=1,486) in the IVCF only group. No significant difference was noted among groups at either 6 or 12 months (p=0.11 and 0.17, respectively).

Additionally, HIT+IVCF subgroup analyses showed no significant difference in 6- or 12-month mortality between HIT likelihood (p=0.59 and 0.27, respectively) or timing of IVCF placement (p=0.69 and 0.84, respectively). Thromboembolic risk was not significantly different within these two subgroups.

The authors discuss that current guidelines recommend against the routine placement of IVCFs in patients with HIT despite limited comparative data assessing patients with HIT and IVCFs versus either alone. The results of this study suggest that patients with HIT in whom IVCFs are placed are at no higher mortality risk compared to controls nor are they at higher thromboembolic risk compared to previously reported data on patients with either HIT or IVCFs alone.

Commentary

This study compares mortality rates in patients with HIT receiving IVCFs compared to patients with either HIT or IVCF alone as well as evaluates the thromboembolic risk in patients with HIT receiving IVCFs. The authors note that current guidelines are largely based on data describing high rates of thromboembolic events in patients with HIT in whom central venous catheters or IVCFs were placed, but remark that there is limited data assessing the true thromboembolic risk of patients with HIT and IVCF compared to controls. To that end, this study aimed to more accurately characterize the thromboembolic risk and mortality in this population.

The extremely large sample sizes of the control populations, while advantageous for powering the study, did not allow for manual chart review to ensure that diagnostic and procedural codes were accurately assigned in the electronic medical records. In the HIT+IVCF group alone, 9 out of the original 39 cases (23%) were excluded due to miscoding, which raises concern regarding the accuracy of the HIT and IVCF only groups. Additionally, since the control groups were so considerably large, it was not possible for the authors to review all charts to establish thromboembolic risk in each control group. They rightfully discuss that their thromboembolic risk comparison is therefore restricted to previously published data, rather than having an internal comparison. This introduces potential biases between institutional practice patterns, patient populations, and variability over time.

Although generally quite thorough in the evaluation of the HIT+IVCF group, the authors do not mention retrieval rates of IVCFs in either this population or their controls. Given that IVCFs are known to be associated with greater thrombotic risk if not removed, it would have been interesting to see if timing of IVCF removal had any influence on thromboembolic risk or mortality.

Furthermore, although 6- and 12-month follow up periods are relatively long term in relation to duration of thromboembolic risk associated with HIT (which peaks over 14 days), it may also be of benefit to assess longer (3- or 5-year) mortality rates to validate the conclusions over an extended time period. Lastly, the retrospective design has inherent limitations where a prospective study or randomized controlled trial would offer more accurate comparison, although likely not feasible in this relatively small patient population.

The major strength of this study is its large sample size of patients with HIT receiving IVCFs compared to other studies. Their thorough review of laboratory and clinical data to determine likelihood of HIT diagnosis and timing of IVCF placement adds to the comprehensiveness of the morbidity and mortality evaluation. The scope of their discussion also appropriately addresses the limitations of the data and highlights the need for additional research and external validation.

Overall, this study sheds new light on the question of whether IVCF placement should be considered in patients with HIT in which anticoagulation has failed or is contraindicated. For patients with HIT and IVCFs, there was no significant difference in thromboembolic risk compared to rates from other studies nor in overall mortality compared to internal controls. These findings challenge current guidelines that recommend against routine placement of IVCFs in patients with HIT; however, outside validation of these results is essential prior to implementing changes in clinical practice.

Post Author
Catherine (Rin) Panick, MD
Resident Physician, Integrated Interventional Radiology
Dotter Interventional Institute
Oregon Health & Science University
@MdPanick

Edited and formatted by @NingchengLi

Interventional Radiology Resident

Dotter Institute, Oregon Health and Science University

Absolute Ethanol Embolization of Lip Arteriovenous Malformations: Observational Results from 10 Years of Experience

Clinical question
What is the safety and efficacy of lip arteriovenous malformation (AVMs) ethanol embolization via direct puncture or transarterial catheterization technique, with or without follow-up surgical resection?

Take away point
Lip AVM treatment remains variable. Ethanol embolization of lip AVMs in single or multiple sessions was able to achieve complete devascularization in 67% of the patients and complete cosmetic resolution in 53% of patients. When including preoperative embolization followed by surgical resection, 71% achieved complete cosmetic relief and all patients endorsed major to complete improvement of functional impairment.

Reference
Su LX, Li XY, Zhao ZJ, Shao YH, Fan XD, Wen MZ, Yang XT. Absolute Ethanol Embolization of Lip Arteriovenous Malformations: Observational Results from 10 Years of Experience. J Vasc Interv Radiol. 2022 Jan;33(1):42-48.e4. doi: 10.1016/j.jvir.2021.09.004. Epub 2021 Sep 20. PMID: 34547475.

Click here for abstract

Study design
10-year retrospective, observational study of 76 patients

Funding source
(1) Fundamental research program funding of Ninth People’s Hospital affiliated to Shanghai Jiao Tong University School of Medicine (No. JYZZ076)
(2) Clinical Research Program of Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (No. JYLJ201801 and JYLJ201911)
(3) China Postdoctoral Science Foundation (No. 2017M611585)
(4) National Natural Science Foundation of China (No. 81871458)

Setting
Urban academic center at the Shanghai Ninth People’s Hospital which is affiliated with the Shanghai Jiao Tong University School of Medicine

Figures


Figure 1: Intraprocedural digital subtraction angiography (DSA) of a 17-year-old male treated for a pulsating AVM. (a) A focal AVM nidus is opacified with contrast (arrow). (b) Lateral view demonstrates the 1.7 Fr microcatheter at the AVM nidus via superselective embolization from the facial artery. (c) Subsequently, direct 21-guage needle puncture was used to perform an additional embolization of the nidus. (d) The post-procedure lateral view DSA demonstrates 100% AVM devascularization.


Figure 2: Of the 76 patients treated with embolization, most with multiple sessions, the majority went on to have complete cosmetic resolution either solely with ethanol embolization or after ethanol embolization and subsequent surgical resection, and 100% of patients endorses complete or major resolution of cosmetic appearance.

Summary

Lip arteriovenous malformations (AVM) may be present at birth and can enlarge over time secondary to trauma or hormonal changes. Radical surgical resection can result in functional/ cosmetic defects and may risk severe intraoperative hemorrhage. Incomplete resection fails to provide symptom relief. Thus, AVM embolization is increasingly becoming the treatment of choice.

This study utilized ethanol as a permanent embolic agent. Other studies of peripheral AVM embolization have used ethylene vinyl alcohol copolymer and n-butyl cyanoacrylate (n-BCA) glue, however reports show nidus recanalization can occur, and subcutaneous accumulation of the embolic glue can cause infection. Ethanol injection permanently destroys the vascular endothelium by causing protein denaturation and nidus thrombosis. Downsides of ethanol embolization include risks of systemic ethanol toxicity which can include cardiopulmonary arrest and disseminated intravascular coagulation (DIC).

This 10-year retrospective study includes patients with lip AVM Schobinger stage II or higher (growing or symptomatic AVM). Many patients had prior treatments (including surgical resection, sclerotherapy, propranolol, radiofrequency ablation and artery ligation), but only those at least 3 months post prior therapy were included. Other exclusion criteria included contrast or ethanol allergy and major cardiac, respiratory, renal or hepatic impairment.

Lip AVM embolization was performed under general anesthesia, with a Foley catheter in place to monitor for hemoglobinuria secondary to ethanol administration and arterial line for blood pressure monitoring. Pulmonary artery pressure was measured in those expected to have high ethanol volume administered.

Digital subtraction angiography (DSA) was performed via superselective angiography of the external and internal carotid arteries using a 4 or 5 Fr catheter to identify the AVM nidus based on feeding vessels, draining veins, and internal blood flow. Then, a 1.7 Fr microcatheter was advanced into the nidus. If complete nidus access was not feasible by this technique alone, they performed ultrasound-guided direct puncture using a 21-guage butterfly needle. If a draining outflow vein (DOV) was identified, it was coil-embolized prior to ethanol embolization.

Required ethanol volume was estimated based on the amount of contrast injected to fill the AVM nidus. Usually 100% ethanol was used except in patients with a diffuse infiltrating AVM where diluted ethanol was used to reduce risks of systemic toxicity. In all cases, less than 1 cc/kg of ethanol was used during each session, and staged embolization was used in patients to reduce risks of ethanol toxicity. Additionally, manual pressure was applied around the target lesion to reduce non-target embolization.

The 76 patients enrolled underwent a mean of 2.3 procedures (range of 1-7 procedures) and were followed for a mean of about 2.5 years. Patients were asked to self-report cosmetic changes on a scale of 1-5 after treatment, and functional results were evaluated by oral maxillofacial surgeons.

Roughly 67% of patients had 100% AVM devascularization on arteriography. One third of patients had 76-99% devascularization and 26.3% experienced 50-75% devascularization. Symptomatically, 53% of patients had complete cosmetic improvement after embolization alone. A third of patients self-reported major cosmetic improvement, and 44% of them underwent surgery to see further improvement. Those who rated their cosmetic changes as minorly improved or worsened underwent surgery, all of whom felt they had major improvement or complete improvement postoperatively. In total, 71% had complete cosmetic resolution of their AVM and 39% had major cosmetic improvement with embolization alone or followed up by surgical resection.

Immediate post-procedure swelling was seen in almost all patients and resolved within 2 weeks. The most common complication was local skin bullae, which occurred in 9.2% of procedures and spontaneously resolved. There were 3 cases of skin necrosis, one of whom developed serious lip necrosis causing cosmetic deformity that required surgery. Two patients developed transient hemoglobinuria which was treated with IV infusion. No severe complications were reported.

Commentary

The study is limited by its retrospective nature, potential for patient selection bias, relatively small sample size and nonstandardized treatment algorithm which included a mix of embolization, coil embolization of the draining vein when present, and postprocedural surgery. A randomized control trial with a placebo arm as well as comparative groups randomly assigned to surgical or embolization therapy would be useful. Additionally, comparing the efficacy and complications of other embolic materials and combination of coils or plugs would be advantageous for clinical decision-making. Although ethanol is highly effective, rare reports of systemic effects causing DIC and cardiopulmonary arrest may be a considerable drawback for centers that do not perform lip AVM commonly.

Future studies could consider enrolling treatment-naïve patients. In this study, more than half of the patients had prior laser treatment, sclerotherapy, radiofrequency ablation or arterial embolization, performed at least 3 months prior. This may confound the results of ethanol embolization, especially in cases of prior surgical resection or arterial ligation, which can complicate future embolization access. Additionally, the study on average followed patients for 2.5 years, and though ethanol embolization is permanent when injected, AVM recurrence may occur in untreated regions of a diffuse, infiltrating AVM beyond that time frame. Another limitation is that post-procedure cosmetic results were quantified based on a subjective patient-derived scale; however, this appears to have been used in prior studies.

In conclusion, the study authors convey that lip AVM treatment can be challenging due to the nature of the AVM vascular anatomy, however multifactorial treatment options can produce acceptable results in most patients. Treatment options were customized throughout the course of therapy and tailored to patient preferences regarding cosmetic improvement. These options include multiple courses of ethanol embolization via superselective angiography and/or direct puncture technique, post-procedure surgical resection, and draining outflow vein coil embolization. The study reports that most patients had complete or major improvement of cosmetic and functional symptoms and embolization and/or surgical resection. However, a subset of patients was not able to achieve complete resolution of their cosmetic defects despite multiple sessions of embolization and surgical resection, underscoring the complicated treatment course of peripheral AVMs and the importance of accurately conveying to patients the realities of AVM treatment.

Post author
Surbhi B. Trivedi, MD
Diagnostic Radiology PGY-3 Resident
University of Illinois, Chicago
@surbhitrivedi3

Edited and formatted by @NingchengLi

Interventional Radiology Resident

Dotter Institute, Oregon Health and Science University

Clinical Safety and Efficacy of Locoregional Therapy Combined with Adoptive Transfer of Allogeneic γδ T Cells for Advanced Hepatocellular Carcinoma and Intrahepatic Cholangiocarcinoma

Clinical question
Is locoregional therapy plus adoptive transfer of allogeneic gamma delta (γδ) T cells safe and effective in treating patients with hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC)?

Take-away point
The combination of both locoregional ablation and adoptive transfer of allogeneic γδ cells is both safe and effective in patients with HCC and ICC.

Reference
Zhang T, Chen J, Niu L, et al. Clinical safety and efficacy of locoregional therapy combined with adoptive transfer of allogeneic γδ t cells for advanced hepatocellular carcinoma and intrahepatic cholangiocarcinoma. Journal of Vascular and Interventional Radiology. 2022;33(1):19-27.e3.

Click here for abstract

Study design
Randomized Controlled Trial

Funding source
This study was funded by grants from the National Key R&D Program of China (2018YFA0108304), National Natural Science Foundation of China (81771721, 81971505), and Fujian Provincial Health Education Joint Research Project (WKJ2016-2-20).

Setting
Single institution, Fuda Cancer Hospital of Jinan University, Guangzhou, China


Figure 5. Survival curves. Overall survival curves of hepatocellular carcinoma (HCC) patients (a). Distant progression-free survival (PFS) curves of HCC patients (b). Local PFS curves of HCC patients (ablation sites) (c). Overall survival curves of intrahepatic cholangiocarcinoma (ICC) patients (d). Distant PFS curves of ICC patients (e). Local PFS curves of ICC patients (ablation sites) (f).

Summary

Adoptive transfer of T cells has emerged as a novel treatment of malignant neoplasms. A subset of T cells known as γδ T cells has been specifically utilized to recognize and eliminate tumor cells based on tumor-specific antigens. In an allogeneic adoptive transfer strategy, the γδ T cells are provided by a donor, amplified in vitro, and subsequently infused into a patient with a known cancer (as supposed to in vivo activation by zoledronate and IL-2). Few clinical trials have been conducted in patient populations with advanced disease due to the larger tumor burden and associated immunosuppressive environment. It is postulated by the authors that the combination of locoregional ablative therapy in concert with systemic treatment using allogeneic γδ T cells may be an effective treatment. 

Cryoablation and irreversible electroporation (IRE) are locoregional therapies that have proven efficacy and safety, while also preserving tumor cell surface architecture to potentially allow immune system recognition and subsequent response. Thus, the merit behind combining these modalities with systemic immune therapy is evident. Additionally, adoptive transfer has been shown to be safe and effective in several late-stage cancers as a monotherapy as well as in combination with IRE, and infused γδ T cells are known to preferentially migrate to the liver. Thus, a study of the safety and efficacy of a combination of locoregional therapy and adoptive transfer in the treatment of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) was devised.

In this randomized controlled trial, patients between the age of 18-75 years of age with stage III or IV HCC or ICC by 7th AJCC TNM guidelines were included. The study also required patients have multifocal disease that made them unsuitable for surgery or chemotherapy. They could not have tumor invasion of the portal vein, hepatic vein trunk, or secondary branches, and had to have a projected survival of > 3 months. Additional inclusion criteria included Eastern Cooperative Oncology Group (ECOG) performance status (PS) score ≤ 2, normal laboratory parameters, and adequate hepatic and renal function. Patients with coagulation disorders, anemia, severe coronary artery disease, respiratory disease, myelosuppression, acute or chronic infection, or autoimmune disease were excluded. Patients were randomized to either the combined treatment or local treatment groups using a computer-generated random number table. Patients in the combined treatment group underwent either IRE or cryoablation followed by adoptive transfer of allogeneic γδ T cells. Patients in the local treatment group only underwent either IRE or cryoablation. IRE was used for tumors adjacent to large vessels such as the hepatic artery, portal vein, or bile ducts in the liver hilum. Other tumors were treated with cryoablation. Additionally, transarterial chemoembolization was done prior to ablation on tumors > 5 cm in size to reduce tumor size prior to locoregional therapy. In patients undergoing adoptive transfer, donor γδ cells were collected, isolated, and expanded as previously described in the literature on day 1 of treatment. The patients then underwent locoregional therapy on day 9 before receiving donated cells in 3 equal infusions, with one each on days 13, 14, and 15. This was followed by a 2nd donation on day 16 and infusions on days 28, 29, and 30. Additionally, peripheral blood was drawn on day 8 and a month after day 30 to determine CTC levels. Follow-ups included CT or MRI at 1 month and every 3 months after treatment. These were compared to imaging collected 1 week prior to locoregional therapy. Modified Response Evaluation Criteria in Solid Tumors (mRECIST) was used to evaluate treatment response. Technical success was defined as complete ablation while local progression was defined as reappearance of arterial phase enhancement of the ablated area after initial technical success. Distant progression was defined as the appearance of new intrahepatic or extrahepatic tumors or an increase of over 20% in diameter of any unablated tumors. Overall survival was defined as time between locoregional therapy and death. Additional primary endpoints included safety (adverse events), distant progression-free survival (PFS), and local PFS. Secondary endpoints were CTC levels, AFP and CA19-9 (for HCC and ICC respectively), and change in ECOG PS.

Median survival was improved in the HCC group undergoing combination therapy compared to locoregional therapy, with median OS being 13 months for combination and 8 months for locoregional therapy, 1-year OS rates of 57.3% and 7.3% respectively, and 2-year OS rates of 26.7% and 7.3% respectively (p = 0.029). Median distant PFS in combination and locoregional were 8 and 4 months, with 1-year distant PFS rates of 0% and 21.8%, respectively (p = 0.040). Differences in local PFS rates were not significant (p = 0.949, 28.6% in combination group and 14.5% in locoregional group).

For patients with ICC, differences in OS were not significant; median OS was 9 months for combination and 8 months for locoregional therapy, 1-year OS rates of 33.3% and 44.3% respectively and 2-year OS rates of 25% and 8.3% respectively (p = 0.546). Median distant PFS in combination and locoregional groups were 8 and 4 months, with 1-year distant PFS rates of 14.1% and 0%, respectively (p = 0.021). Similar to HCC groups, differences in local PFS rates were not significant (p = 0.394, 16% in combination group and 27.5% in locoregional group). The authors mention that 28.6% of ICC patients and an unspecified number HCC in their respective combination groups did not receive locoregional therapy and dropped out of the study due to progression or withdrawal of consent, prompting a protocol analysis. The results of the protocol analysis were reported as consistent with the intention-to-treat (ITT) analysis.

AFP decreased in all groups, with a more pronounced drop in the combined treatment groups (HCC: p = 0.046, ICC: p = 0.028). Change in CA-19-9 was not significantly different between the experimental and control groups. CTCs decreased in all groups, but was more pronounced in the combined treatment groups (HCC: p = 0.035, ICC: p = 0.003). ECOG PS improvements were also statistically significant in both combined treatment groups (HCC: p = 0.039, ICC: p = 0.007), whereas the locoregional groups did not significantly change after treatment.

No severe adverse events occurred during this study. Only grade 1 and 2 events were reported, with most being grade 1 (least significant). These included fever, nausea, thrombocytopenia, arrhythmia, hemothorax, pneumothorax, pleural effusion, and fatigue. The majority of these were grade 1 and all adverse events were resolved spontaneously or with symptomatic treatment. Most events were associated with locoregional therapy, and there was no significant difference in incidence between the two groups. There was no significant difference in lab markers for hepatic or renal function.

Commentary

This study aimed to demonstrate safety and explore hypothesized efficacy of adoptive transfer of healthy volunteers' (allogeneic) γδ T cells in combination with locoregional therapies in the treatment of advanced HCC and ICC. This was done by employing previously-described techniques and timelines that have been proven safe and modestly effective. Results of this study demonstrated statistically significant if modest improvements in distant PFS for both HCC and ICC patients, but no differences were seen in local PFS. Additionally, OS was improved after combination therapy for HCC patients, but not for ICC patients. It is supposed that the lack of difference in local PFS is due to the primary contribution of locoregional therapy to this metric in all groups. Despite this, PS did improve in both combination groups while not in locoregional groups. It is hard to say, however, if this is clinically significant. The authors also posited that the significantly pronounced decrease in CTCs in both combination groups was likely secondary to systemic effects of γδ T cells and also likely the reason for the differences in distant PFS. Indeed, decrease in CTCs is a previously-demonstrated effect of γδ T cell infusion, and is known to be associated with increased PFS as well as impaired metastasis and recurrence. The decrease in AFP in HCC patients with combination therapy and the respective lack of decrease in CA19-9 in ICC patients, in concert with the less-pronounced survival differences in ICC patients, further suggests that combination therapy may be less effective in treating ICC. ICC is also acknowledged as a morbid diagnosis that progresses rapidly, is refractory to first-line treatments, and is often diagnosed in advanced stages, and thus may require a larger sample to determine true efficacy. 

While this trial was an RCT, it does have some limitations. One is that a higher proportion of patients in the combination groups did not undergo intended therapy. Although the PP and ITT analyses were consistent, this still may affect true therapeutic effect. Additionally, the study was open-label without allocation-concealment, which undermines the randomization design. Finally, the use of both IRE or cryoablation, with their different utilities and mechanisms of action, may have resulted in unknown effects on the data. Future trials will benefit from larger sample size, standardization of locoregional therapy or subgroup analysis of the modalities used, and double-blinding to ensure randomization.

Post Author
Jared Edwards, MD
General Medical Officer
Medical Readiness Division
Naval Surface Forces Pacific, San Diego, CA
@JaredRayEdwards

Edited and formatted by @NingchengLi

Interventional Radiology Resident

Dotter Institute, Oregon Health and Science University