28 Primary and secondary liver malignancies

SECOND EDITION

The GEC ESTRO Handbook of Brachytherapy

PART II: CLINICAL PRACTICE 28 Primary and secondary liver malignancies Stefanie Corradini, Lukas Nierer, Maya Rottler, Franziska Walter, Konrad Mohnike, Marc Mühlmann, Max Seidensticker, Jens Ricke, Peter Hass

Editors Bradley Pieters Erik Van Limbergen Richard Pötter

Peter Hoskin Dimos Baltas

Primary and secondary liver malignancies

3

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

28 Primary and secondary liver malignancies

Stefanie Corradini, Lukas Nierer, Maya Rottler, Franziska Walter, Konrad Mohnike, Marc Mühlmann, Max Seidensticker, Jens Ricke, Peter Hass

1. Summary 2. Introduction 3. Anatomy 4. Pathology

3 3 3 5 5 6 7 8

9. Treatment planning

9

10. Dose, dose rate and fractionation

10 11 12 14 14 15

11. Monitoring

12. Results

5. Work up

13. Adverse side effects 14. Key messages

6. Indications, contra-indications 7. Tumour and target volumes

15. References

8. Technique

1. SUMMARY

With interstitial HDR liver brachytherapy, excellent local control rates of >90% at 12 months can be achieved for various entities with prescription doses of 15-25 Gy (depending on histology) in a single session. In contrast to stereotactic body radiation therapy (SBRT), liver brachytherapy allows the application of higher doses to the tumour while sparing surrounding organs at risk. Unlike thermal procedures, tumour size is not a limitation with brachytherapy, nor is central location (heat sink effect). Moreover, exposure of adjacent organs at risk can be determined during treatment planning, and dose constraints for OARs can be met to avoid normal tissue toxicities. Therefore, the available evidence suggests that this minimally invasive treatment option is particularly beneficial in the treatment of large liver tumours, multiple lesions, or tumours in central location.

2. INTRODUCTION

placement. In the hands of experienced radiation oncologists or interventional radiologists, toxicity rates and serious procedural complications (major bleeding or infection after the procedure) are very limited [3].

High-dose rate interstitial brachytherapy (HDR brachytherapy) is a treatment option with high local control rates for primary liver malignancies, such as hepatocellular carcinoma (HCC) or cholangiocarcinoma (CCC), and for secondary liver malignancies in oligometastatic disease. HDR brachytherapy is performed using the afterloading technique and high dose rate radioactive sources containing β- or γ-emitting radioisotopes such as iridium-192, which allows short irradiation times of about 10minutes for small lesions to 90 minutes for large lesions [1]. With this technique, very high doses can be applied to liver lesions while optimally sparing healthy liver tissue and adjacent organs at risk (OAR) due to the inherent steep radiation dose gradient. The steep dose gradient is mainly achieved due to the geometric conditions (small source; decrease of radiation fluence by square of distance) and to some minor extent due to attenuation in the tissue. Recent analyses comparing HDR brachytherapy with virtually planned stereotactic body radiotherapy for ablative treatment of liver malignancies showed superiority of HDR brachytherapy in terms of dose coverage of the target volume and liver volume irradiated to 5 Gy [2]. For optimal HDR brachytherapy of the liver, close collaboration between radiation oncologists, medical physicists, and interventional radiologists is required. First, brachytherapy catheters are placed under image guidance (usually computed tomography). Achievable dose coverage of the target depends largely on the catheter

3. ANATOMY

Procedure planning, catheter placement, target delineation and dose optimization require in-depth knowledge of general and patient specific anatomy and anatomic relationships between target lesions and organs at risk. In general, the Couinaud classification divides the liver into 8 segments with further subdivision of segment 4 in 4a and 4b (see figure 1) [5,6]. Liver segments 5-8 are in the right liver lobe, liver segments 2-4 are in the left liver lobe. Each segment is functionally independent having main portal, arterial and biliary branches centrally as well as draining hepatic veins peripherally. The portal veins delineate the horizontal, the liver veins the vertical divisions between segments (segment 1 being the exception). CT and MRI are used in at least three steps to assess the required anatomic information i) before, ii) during and iii) after catheter placement. Preprocedural scans (step i) visualize general and patient specific anatomic features including congenital variations,

Primary and secondary liver malignancies

4

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

Figure 1: Couinaud classification of liver anatomy divides the liver into eight functionally independent segments

TABLE 1 ANATOMIC STRUCTURES AND POTENTIAL RISK ASSESSMENT FROM CATHETER PLACEMENT AND BRACHYTHERAPY TREATMENT Anatomic structure Potential risk of A) catheter placement Costodiaphragmatic recess Pneumothorax, Hemothorax, Compartment barrier disruption Arteries (intercostal, epigas-tric, mammarian, liver) Hemorrhage, arterio-portal Fistulas, arterio-biliary Fistulas, Sepsis Portal vein Hemorrhage, Fistulas Liver veins Air embolism, minor Hemorrhage, Fistulas Biliary tracts Biliary leakage, Abscess, Fistulas (to vessels, to abdominal cavity, to skin) Stomach, bowel, colon Perforation (Chilaiditi syndrome) Intercostal nerves Nerve damage

B) radiation Duodenum

Ulcer, Perforation, Abscess Ulcer, Perforation, Abscess

Stomach

Leakage, Biliary leakage, Duct scaring and pre-stenotic Dilatations, Abscess (increased when papillary Insufficiency)

Biliary tract

Adrenal glands

Radiation induced hypofunctionality Radiation induced hypofunctionality

Kidney

Lung

Radiation pneumonitis

such as accessory lobes, accessory fissures or vessels, as well as acquired changes such as cirrhosis, post-operative and radiogenic alterations. During the catheter planning process, information on target location, distance to radiation sensitive surrounding OARs, anatomic structures in the path of ideal catheter position (e.g. costodiaphragmatic recess, epigastric, intercostal and mammary arteries as well as hepatic arteries, portal-veins and biliary tracts) are evaluated to plan a safe but comprehensive procedure. During the catheter placement, fluoroscopic CT (step ii) is used to account for dynamic changes that might affect liver position or morphometry such as inspiration depth. Furthermore, anticipation of pain,

even under conscious sedation, can lead to subconscious evasion movements or unilateral muscle contractions, which in turn can result inmovement of the liver. Recognition of vascular structures on CT (see figure 2) during catheter placement facilitates the avoidance of arteries and biliary structures as well as the associated risk of bleeding, biliary leakage or fistulas. The radiotherapy simulation CT (step iii) is used to assess distances between the placed catheters, the target lesion andOARs, including duodenum, bowel, stomach, kidneys, adrenal glands, coronary arteries, ribs, spleen and lung (see table 1).

Primary and secondary liver malignancies

5

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

and in Southeast Asia the liver fluke infections [13]. There are significant differences in surgery and locoregional therapy between these tumours. The main approach to curative therapy is complete surgical resection, although this is only possible in 30-50% of patients [14]. Because all bile duct cancers are rare tumours that require complex multidisciplinary therapy, treatment decisions should be made in a specialized multidisciplinary team. Hellman and Weichselbaum first described oligometastatic disease (OMD) in 1995 as a distinct cancer stage between locally confined and systemic metastatic disease [15]. While significant heterogeneity exists in the current OMD definitions in the literature, a recent ESTRO-ASTRO consensus defined OMD as 1–5 metastatic lesions, a controlled primary tumour being optional, but where all metastatic sites must be safely treatable [16]. However, because no biomarker is clinically available to identify patients with true oligometastatic disease, the diagnosis of oligometastatic disease is based solely on imaging findings. A small number of metastases on imaging could represent different clinical scenarios associated with different prognoses and requiring different treatment strategies [17]. The ESTRO-EORTC has made an attempt to classify these different clinical situations of OMD, considering scenarios of induced oligometastatic disease (history of polymetastatic disease) and genuine oligometastatic disease (no history of polymetastatic disease). In addition, oligorecurrence, oligoprogression, and oligopersistence were subdivided, taking into account whether oligometastatic disease was diagnosed during a treatment-free interval or during active systemic therapy and whether or not an oligometastatic lesion had progressed on current imaging. Preprocedural imaging is key in the identification of primary or secondary liver malignancies suitable for brachytherapy. Most commonly, patients with metastatic cancer after previous chemotherapy and/or surgery and either stable or new liver metastasis will be referred for brachytherapy of the liver for consolidation or therapy intensification. MRI of the liver with gadoxetic acid or PET/CT increase sensitivity towards smaller metastasis and thus provide a clearer picture of the total tumour burden in metastatic liver disease [18]. Concerning primary liver malignancies, for instance HCC, MRI with gadoxetic acid and multiphase imaging including late phase imaging allow for tumour burden estimation with high specificity, high accuracy and even the detection of potential precursor lesions, when the European association for the study of the liver (EASL) criteria are applied [19–21]. Routine staging CT, MRI and, depending on availability and suspected tumour type, PET/CT, may be combined to provide information for the key questions before moving forward with brachytherapy: 1. Is there evidence of systemic disease? 2. Are previous malignancies known? 3. Is the nature of the liver lesion proven - visually or histologically and is a biopsy warranted? 4. Are tumour size and lesion number favoring other treatments? 5. Is transplantation an option? The latter questions especially should be discussed in an interdisciplinary tumour board and specific recommendations 5. WORK UP

Figure 2: Liver anatomy and vascular structures. AA: Aorta abdominalis, TC: Truncus coeliacus, AHC: Arteria hepatica communis, AL: Arteria lienalis, AHP: Arteria hepatica propria, AHD: Arteria hepatica dextra, AES: Arteria epigastrica superior, VCI: Vena cava inferior, PV: Vena porta, AG: adrenal gland, CR: Costodiaphragmatic recess, C9-C12: ribs 9-12, L1: lumbar vertebra L1

4. PATHOLOGY

Indications for liver brachytherapy include primary liver malignancies such as HCC or CCC and liver oligometastases from various origins. HCC is the most common primary liver malignancy (>80%) and mostly develops on the background of pre-existing liver cirrhosis or chronic liver disease [7]. It is the sixth most common malignancy and fourthmost common cause of cancer mortality worldwide [8]. The incidence varies geographically with Asia having the highest incidence worldwide (72.5%), followed by Europe (9.8%), Africa (7.7%), North America (5%), Latin America and the Caribbean (4.6%) and Oceania (0.4%) [9]. Risk factors include chronic liver disease leading to cirrhosis, like chronic viral hepatitides (HBV and HCV), heavy alcohol consumption or non-alcoholic fatty liver disease [10]. The treatment of HCC is stage dependent and determined by several variables, including liver function and performance status, in addition to the number and size of lesions. Biliary cancers are a rare and inhomogeneous tumour entity, accounting for less than 3% of all gastrointestinal malignancies [11]. There are significant geographic differences in the overall incidence of biliary cancers. They occur rarely in Europe, Australia and the USA with an incidence of 0.3-3.5/10 5 population. In Asian countries with frequent liver fluke infections, the incidence is much higher [12]. The most common histologic subtype of cholangiocarcinoma in all locations is adenocarcinoma with a biliary differentiation. They are divided according to their location into intrahepatic, perihilar, and distal bile duct cancers. The risk factors differ depending on the location. The main risk factors for intrahepatic carcinoma of the bile ducts correspond to those for hepatocellular carcinoma, i.e. primarily liver cirrhosis and hepatitis B or C infections. For extrahepatic carcinoma of the bile ducts, chronic inflammation of the bile ducts has been identified as a risk factor, especially primary sclerosing cholangitis and other strictures of the bile ducts with biliary cysts and Caroli syndrome,

Primary and secondary liver malignancies

6

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

6. INDICATIONS, CONTRA-INDICATIONS

(European Society for Medical Oncology [ESMO] or European Association for the Study of the Liver [EASL] Guidelines) should be followed. This also includes a critical assessment of patient fitness for surgery and brachytherapy. Frequently used indices to assess the patient’s performance are ECOG status and the Karnofsky index. Furthermore, laboratorymeasurements of GGT, GOT/GPT, albumin, bilirubin, creatinine, INR, partial thromboplastin time (PTT) and platelets should be made and the (creatinine modified) Child Pugh score should be calculated in order to estimate liver function. There is, to the best of our knowledge, no published limit for either ECOGor Child Pugh-Scores for liver malignancy ablation usingHDR brachytherapy, but it should be usedwith caution inCPS class >B8. However, in a smaller cohort of 24 patients with HCC (and Child-Pugh A or B cirrhosis) undergoing either yttrium-90 selective internal radiation therapy (SIRT) or brachytherapy, only mild changes in liver function parameters compared to baseline, with recovery within 6-12 weeks were detected [22]. Nonetheless, critical discussion on a case by case basis should be undertaken, especially when brachytherapy ablation of lesions with >5 cm largest diameter are planned. The goal of brachytherapy should be clarified and documented in the interdisciplinary tumour-board discussion and should be known to the patient (Palliative or curative approach? Down-staging before liver transplantation?). The patient’s blood count must be adequate to perform interventional procedures (e.g. INR 1.5, platelet count >50×10 9 /l, PTT <50 sec) with results not older than one week. Potential anticoagulation should be interrupted or bridged according to contemporary guidelines with pausing of low molecular weight heparin or oral heparin-analogues for at least 48 hours before the intervention, whenever possible (e.g. consensus guidelines of the Society of interventional radiology [SIR] / Cardiovascular and Interventional Radiological Society of Europe [CIRSE]) [23,24]. In contrast, intake of acetylsalicylic acidmay be continued in patients receiving secondary prophylaxis. Medical history assessment should reflect if there is a potential insufficiency of the sphincter of Oddi or presence of a hepatojejunostomy, as this leads to an increased risk of infection, abscess formation and potential sepsis after brachytherapy. Additional risk factors for infection are cirrhosis and diabetes. In these high-risk cases, periprocedural antibiotic prophylaxis is now recommended in the updated standard operating procedure (SOP) as published by SIR and supported by CIRSE [25]. While these recommendations focus primarily on thermal ablations, they may be used as guide for brachytherapy of liver malignancies as well. To date there are no published randomised controlled trials that investigate antibiotic prophylaxis in low-risk patients undergoing hepatic tumour brachytherapy. In these patients “the use of a single agent targeted to skin flora (i.e. cefazolin) may be reasonable”, according to the updated consensus guideline [25]. Preprocedural allergy assessment is needed regarding contrast- agents, local anaesthetics, systemic pain-medications as well as sedatives. Depending on the severity of the known allergic reaction potential modifications to the ablation plan should be discussed (modality switch or avoidance possible? Premedication and anaesthesia stand-by?). General best practice recommendations concerning contrast-agent application as issued by the European society of urogenital radiology (ESUR) should be followed – including for example assessment of thyroid function to exclude or manage hyperthyroidism.

Patients with primary liver malignancies or with oligometastatic disease of the liver can be considered suitable for liver brachytherapy if surgical options or other locally ablative treatments are not feasible or desired by patients. All patients should be discussed in an interdisciplinary tumour board before treatment and deemed amenable for local ablative treatment using brachytherapy. Relative contraindications include liver cirrhosis, Child-Pugh-stage B or higher, more because of a considerably higher risk of severe bleeding and catheter dislocation events than concerns regarding the post-interventional liver function [26,27]. Ascites is a technical contraindication, and even though it may be drained before application, it indicates either advanced liver cirrhosis, peritoneal carcinomatosis or cardiac comorbidity, questioning the value of local ablation with respect to the impact on overall survival in the context of possible adverse events. Serious attention should be paid to oral or s.c. anticoagulation due to the increased risk of severe bleeding events and subsequently increased mortality [26]. A biliodigestive anastomosis appears to increase the risk of post-interventional infection and abscess formation [28]. Hepatocellular carcinoma Optimal treatment of primary liver cancer is highly dependent on a multidisciplinary approach due to the complexity of diagnosis, staging, medical comorbidities (especially underlying cirrhosis), and the myriad treatment options.The contribution and collaboration of the disciplines of diagnostic radiology, pathology, hepatology, transplant surgery, surgical oncology, medical oncology, radiation oncology, and interventional radiology are critical to achieve individualised and evidence-based patient care. Multiple treatment modalities are used for the definitive treatment of primary liver cancer. For HCC, orthotopic liver transplantation (OLT), surgery, and thermal ablation (radiofrequency [RFA] or microwave) are standard treatment modalities with curative intent. Catheter-based therapies (e.g. transarterial embolisation [TAE], transarterial chemoembolisation [TACE], and transarterial radioembolisation [TARE]) are considered acceptable treatment options for locoregional tumour control. According to the Barcelona Clinic Liver Cancer (BCLC) staging system, patients with "very early" and "early" stage HCC benefit especially from local ablation [29]. RFA is the most commonly used ablative technique and, because of its high efficacy in small HCCs (≤3 cm largest diameter) it has been recommended as a first-line curative treatment option for small HCCs [30]. However, despite its high efficacy in small tumours, RFA remains limited by a number of tumour factors such as size (>3 cm), tumour location in close proximity to a large vessel that may result in a heat sink or proximity to organs at risk (e.g. bowel), multiple lesions and vascularization [31]. HDR brachytherapy is an alternative, minimally invasive, radioablative technique. The published data support the use of brachytherapy as a safe and effective treatment strategy in the management of primary and secondary liver tumours (see table 3 and 4). Compared with SBRT, HDR brachytherapy is less affected by respiratory motion because the implanted catheters move with the tumour and liver [4]. In addition, the radiation dose to nearby OARs can be significantly lower with HDR brachytherapy because the dose gradient is generally steeper compared to SBRT [1]. For patients with liver cirrhosis and unresectable early-stage HCC (BCLCA), Orthotopic liver transplantation (OLT) is another option that provides curative treatment of the underlying liver disease in

Primary and secondary liver malignancies

7

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

addition to potentially curative treatment for HCC.The prognosis of patients undergoing OLT due to HCC has greatly improved after the introduction of the Milan criteria [32], which determine eligibility for OLT based on the maximal number and size of HCC lesions. Specifically, a patient is suitable for transplantation if a single tumour is present and does not exceed 5 cm in size (largest diameter), or if up to three nodules with a diameter ≤ 3 cm are present [33]. While patients within theMilan criteria are considered eligible for primary OLT, patients outside theMilan Criteria may be considered for OLT based on an individual evaluation that includes the response to local ablative treatment [34]. A limitation is the shortage of liver grafts resulting in a considerable delay in treatment for many patients while waiting for donor organs. In HCC patients, bridging therapies are frequently applied to avoid tumour progression which could result in delisting. Moreover, it was shown that successful local ablative therapy before liver transplantation is an independent statistically significant factor for long-term tumour-related survival for patients with HCC in cirrhosis [35]. Several local ablative treatments are available as treatment options for either downstaging or bridging before OLT. Most commonly, TACE or RFA are used as bridging- therapies with good results [36]. However, for a subset of patients, these treatment options are not ideal for reasons such as tumour size, localisation or proximity to adjacent structures. In these cases, brachytherapy may be an alternative bridging approach until liver transplantation with good local control rates [2,37–40]. Cholangiocellular carcinoma In patients with intrahepatic cholangiocarcinoma, surgery is the only potentially curative treatment option. However, in patients with unresectable disease, historical median overall survival (OS) and intrahepatic progression free survival (PFS) rates after chemotherapy alone remain poor [41]. Therefore, after induction chemotherapy, consolidative local treatment should be considered for unresectable intrahepatic cholangiocarcinoma to improve local control and intrahepatic PFS and mitigate tumour-related liver failure. Liver failure results from obstruction of the portal or hepatic veins and/or bile ducts due to tumour progression [42]. Therefore, local ablative or locoregional image-guided procedures with different therapeutic goals have gained importance in the treatment of intrahepatic cholangiocarcinoma. Although local ablative procedures, such as microwave ablation or RFA, interstitial HDR brachytherapy or SBRT aim at complete local control in the sense of full remission, these locoregional procedures usually achieve only partial remission. They are therefore mostly used in palliation or, less frequently, neoadjuvantly. Examples include both TACE and SIRT. In patients in whom surgery is not an option due to tumour size, location, multifocal disease or restricted performance status, local ablative procedures are an alternative local treatment option and interstitial HDR brachytherapy could increase OS. Retrospective studies support the efficacy of brachytherapy as consolidation therapy with excellent local control rates. Brachytherapy is able to achieve adequate local control rates independent of the tumour diameter, even in larger tumours (> 4 cm), as long as a good coverage with a therapeutic dose could be achieved [43]. Another option is the use of HDR brachytherapy as a salvage option in case of recurrences after surgical therapy [44]. An important component of multidisciplinary discussion is the choice, timing and sequencing of planned systemic therapy in combination with brachytherapy. Oligometastastic disease The concept of OMD is today supported by a growing number of high-quality trials. Three randomized trials reported an

improvement in PFS [45] or OS [46,47] by the addition of local metastases-directed therapy to standard-of-care systemic therapy. Based on these positive studies, the concept of radical local treatment with curative intent in OMD has been rapidly implemented by the oncology community. While the exact definition of oligometastatic disease remains controversial, there is general agreement that more aggressive local treatments are desirable in patients with low metastatic burden. Moreover, the development of highly potent systemic therapies has contributed to a paradigm shift from the exclusive palliative status of metastasized disease to a potentially curable condition in an increasing number of patients with solid tumours. This has renewed interest in locally ablative treatment options for patients with limitedmetastatic burden in combination with systemic treatments. For unresectable liver metastases, a variety of other locally ablative treatments, the so-called "toolbox of locally ablative treatments," can be used [48]. A phase II study reported longer survival for patients who underwent thermal ablation of unresectable liver metastases compared with systemic treatment alone [47]. Two established radiotherapy methods are currently available for the treatment of liver metastases: either external beam SBRT or minimally invasive HDR interstitial brachytherapy [2]. While SBRT is a widely used technique that is frequently employed in the treatment of various anatomical regions, including brainmetastases, lung metastases, liver metastases, or bone metastases - interstitial HDR brachytherapy of the liver is generally underutilized, although it is a well-established radiation technique. Brachytherapy is an excellent tool to deliver very high doses directly to liver lesions with low morbidity [49].

7. TUMOUR AND TARGET VOLUMES

Target delineation includes gross tumour volume (GTV) with an additional margin of 3 – 5 mm for the clinical target volume (CTV), depending on visualization quality of the GTV. The GTV is contoured using additional pre-treatment diagnostic imaging (CTwith i.v. contrast or liver specificMRI with hepatocyte specific contrast). In individual cases, image fusion with diagnostic PET or MRI images may be necessary for proper target definition.

Figure 3: Exemplar target definition and OAR contouring. Red: PTV, dark red: liver, grey: portal vein, blue: truncus coeliacus

Primary and secondary liver malignancies

8

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

Figure 4: An example of the set-up (left image) and material (right image) used for CT-guided catheter implantation using the Seldinger technique. Exemplar materials used for implant: 1) sterile swaps and compresses, 2) sterile draping material, 3) scalpel and hollow puncture needle, 4) numbered angiography sheaths, 5) brachytherapy catheters.

Figure 5: Schematic workflow of the catheter placement using the Seldinger technique (left) and image of the final implant (right). The blue brachytherapy catheter can be seen inside the green/white angiography sheath. The transfer tube of the afterloader (not included in the image) will be connected directly to the blue brachytherapy catheter.

8. TECHNIQUE

Although no relative movements between the target and the applicator are expected, the expected overall uncertainty should be considered when creating the PTV. The largest uncertainty in dose delivery is the potential change of patient anatomy and catheter location between simulation imaging and dose application [50]. Additional spatial uncertainties are caused by offsets of the afterloader dwell positions, uncertainties in image registration, image resolution, target volume andOAR delineation and catheter reconstruction. Such influences, in addition to dose calculation uncertainties (described below under “Treatment planning”) must be quantified or estimated to determine the PTVmargin (typically in the range of 1-3 mm). Since large PTV margins may increase the total PTV volume significantly, the choice for the PTVmargin should be carefully considered and frequently, no PTV margin is applied at all in clinical practice (CTV = PTV). In any case, the target volume should be fully containedwithin the liver. In addition, OAR such as liver, stomach, duodenum, colon, small intestine, kidneys and heart are delineated (see Figure 3).

Materials Amongst several different techniques to perform liver implants, the direct puncture method and the Seldinger technique are the ones most commonly used. The catheter-in-catheter Seldinger technique is frequently used by interventional radiologists [51]. Catheter placement via the Seldinger technique (usually under fluoroscopy-CT-guidance), requires sterile swabs and compresses, a scalpel, a hollow puncture needle, an angiography sheath, a stiff guide-wire, the brachytherapy catheters, suture material for suturing the sheath to the skin and sterile solution for preparing the procedural site. An example is given in figure 4. Hydrophilic coated sheaths allow an easy insertion and reduce pain for the patient during skin and liver capsule puncture. After the dose delivery, catheters are removed. Small pieces of gelatine foam are used to prevent bleeding. The gelatine foam is cut out, rolled and

Primary and secondary liver malignancies

9

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

inserted through each angiography sheath during removal, sealing the puncture tract [2]. Patient preparation Interstitial brachytherapy is typically performed under local anaesthesia and conscious sedation (e.g. midazolam [0.5 mg up to 2.5 mg stepwise, patient-specific] and fentanyl [50-75 µg up to 200 µg stepwise, patient-specific]). Premedication with antiemetics (e.g. ondansetron 8 mg i.v. and dexamethasone 8 mg i.v., directly before catheter insertion) is recommended. To enhance patient comfort during catheter placement, treatment planning, treatment application, and catheter removal, a urethral catheter may be placed. Vital monitoring (non-invasive blood pressure, heart rate, oxygenation and ECG) is essential during the entire treatment. The patient should be positioned carefully in order to have the best possible access to the tumour. Typically, a supine setup with both arms elevated above the head is chosen. When identifying possible puncture tracts, the required length of the puncture tracts should be considered as well as the proximity to possible structures at risk (e.g. arteries, ribs, bowel, lung). Patient comfort is of utmost importance, as the patient must remain in this position for possibly several hours to ensure unchanged patient anatomy and to prevent catheter dislocation. The patient has to remain in this position for the entire duration of the treatment until catheter removal. During catheter placement, the interventional radiologist usually stands on the right side of the patient (see figure 4). Access for the application of sedative-analgesic medications is extended to the back of the CT gantry, which allows adjustments even during the intervention. Due to the respiration depressant effect of analgesics and sedatives, the patient is usually supplied with oxygen peri- and post-intervention (nasal tube). Preplanning Prior to actual catheter placement, it may be useful to generate a pre-plan to assess the optimal catheter position for full dose coverage and OAR sparing. Based on available diagnostic 3D-imaging, the PTV and OAR are delineated and a proposal for the number and placement of catheters can be developed in close collaboration with the medical physicist and the radiation oncologist or interventional radiologist, who will perform the implant. Ideally, the catheters should be arranged in parallel, as this configuration provides good and reproducible dose coverage of the PTV. However, this is not practical in most cases, since practically accessible catheter trajectories are limited and avoidance of OARs should be prioritized. With some experience, the pre-planning steps can be omitted in standard cases, but will remain helpful in complex target lesions. Catheter placement The catheters should be implanted by a trained radiation oncologist or interventional radiologist.The intervention is always conducted under sterile conditions. The catheters are usually implanted under CT-fluoroscopic guidance. However, some centres also have the possibility to perform the procedure in an open bore MRI, which presupposes the availability of MR-compatible materials. Ultrasound-guided implantation is rarely used, however it is a very good option where staff are trained in this approach. For 3D treatment planning, a subsequent simulationCT orMRI is required. The brachytherapy catheter is sealed at the tip and usually has a millimetre scale, which is helpful for correct positioning within the angiography sheath. The brachytherapy catheters are then numbered to allow the catheters to be reconstructed correctly later. Adhesive skin badges are helpful for this purpose and can

be attached to the brachytherapy catheter in a way that the final position within the angiography sheath is also marked and that a shift of the brachytherapy catheter relative to the angiography sheath is prevented. The angiography sheath should be sutured to the skin of the patient for fixation (see Figure 5). It is important, that the brachytherapy catheters are always inserted to a depth that is at least equal or exceeds the length of the angiography sheath within the liver, as the angiography sheath is radiopaque and the exact position of the brachytherapy catheter tip would not be visible on CT imaging for catheter reconstruction if it was located within the angiography sheath. The number of catheters depends on the size and shape of the tumour. Irregularly shaped and larger tumours require more catheters in order to adequately cover the target with the prescribed dose and to ensure good sparing of healthy liver tissue and adjacent organs at risk. Another method, favoured by some radiation oncologists is the direct puncture technique. Here, plastic catheters and steel obturators are used. However, this technique does not allow sealing of the puncture tract after catheter removal [51]. Furthermore, if a catheter position correction is required in cases where it was not adequately implanted, the direct puncture technique causes more trauma compared to a needle rearrangement using the Seldinger technique. After catheter placement, simulation imaging (planning CT/MR) is performed. In case of a planning CT, the use of contrast agent should be considered. The image set is then transferred to the treatment planning system (TPS). A maximum slice thickness of 2 mm is preferable, in order to achieve precise target volume delineation and catheter position definition. After target delineation, the next step is the reconstruction of the catheters in the TPS. This can be done already before the target structure and OAR delineation, or even simultaneously if the TPS provides such functionality.The planning goal is to achieve a robust, high quality treatment plan in a short time, in order to minimize the overall treatment time (from catheter implantation to catheter removal) for improved patient comfort. Each catheter must be correctly identified and reconstructed and the first dwell position in each catheter must be marked. If imaging reference markers (e.g. CT markers of known length) are used inside the catheters, in order to define a reference position, it should be noted that the positional uncertainty of the dwell positions tends to increase with increasing distance from this reference position. Additionally, a smaller catheter curvature radius increases the systematic offset of the dwell positions due to slackness of the source inside the catheter. This effect is usually small in HDR liver brachytherapy due to large curvature radii. After catheter reconstruction, all reconstructed catheters should be checked for plausibility, e.g. via visual inspection of a 3D rendering of the catheters.The 3Dmodel should show a smooth shape of the catheters without any kinks. It may be helpful to have a schematic drawing of how the catheters exit through the skin of the patient. In addition, if catheter insertion is performed using the Seldinger technique, the excess end of the brachytherapy catheter from the angiography sheath should be measured outside the patient as a second validation (see Figure 6). Using this information, the internal excess of the brachytherapy 9. TREATMENT PLANNING

Primary and secondary liver malignancies

10

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

catheter is known and can be checked against the manual catheter reconstruction, namely the distance of the end of the angiography sheath to the (radiopaque) tip of the brachytherapy catheter on the CT.The angiography sheath itself can be seen with great reliability, as it is fully radio-opaque (see Figure 6). The catheter reference point is commonly defined by radio- opaque markers within the tip of the brachytherapy catheter or alternatively CT/MR markers are used, which are inserted into the brachytherapy catheters prior to the simulation imaging. After catheter reconstruction, dwell position activation is performed. All dwell positions inside any target volume should be automatically activated if the TPS offers this functionality. Additionally, all dwell positions should be activated, which are for example outside of all PTVs, but which might be helpful to reach distant parts of the PTV not efficiently coverable by other dwell positions or which might help improve OAR sparing. For this purpose the 3D view of catheters and regions of interest can be very helpful. A 2 mmdwell position spacing is commonly used for liver HDR brachytherapy. Dose optimization is either performedmanually or via semiautomatic inverse dose optimization or as a combination of both. It is feasible to generate a first version of the plan via automated dose planning and then proceed with manual fine adjustment of dwell times, especially in the case of large PTVs with several catheters. When optimizing the dose, as a first step it should be aimed to achieve full coverage of the PTVwith the prescribed dose. In the following steps, the reduction of overall dwell time and sparing of OARs will be achieved. Compared to other entities, where a template-based needle placement is possible (e.g. interstitial multi-catheter HDR brachytherapy of the breast [52]), the dose distribution in liver brachytherapy plans is usually more inhomogeneous with very high central doses. This is the reason why homogeneity indices are not a useful metric in HDR brachytherapy of the liver and should not be used to judge the quality of a treatment plan. It is more appropriate to aim for complete dose coverage of the PTV (important DVH parameters are the minimum dose D100%, or near minimum dose D98% [as recommended in ICRU reports 83 [53]] and 89 [54], in addition to D95% and D90%), while meeting all OAR constraints (see table 2). In other words, dose heterogeneity is accepted to maintain optimal conformity. In further plan optimisation steps, attempts can be made to reduce the overall dwell time. Neighbouring dwell positions should not exhibit extreme variations of dwell time in order to increase the plan robustness against spatial uncertainties but the dwell time modulation restrictions should also not be too strict [55]. To increase homogeneity, the insertion of numerous catheters would be necessary, depending on the volume of the treated lesion. However, due to the potential risk of procedural complications during catheter placement, usually only a limited number of catheters are inserted as compared to other anatomic districts (e.g. breast, prostate). The trade-off between the loss of full dose coverage and efficacy of the treatment or the acceptance of possible higher risks for side effects must be evaluated in close collaboration with the physician in each specific case. However, usually, the sparing of OAR has priority over full dose coverage. The puncture tract is frequently irradiated with ~5 Gy to the catheter surface up to the skin of the patient to avoid seeding along the puncture tract (see Figure 7) [56]. During dose optimisation the planning physicist should be aware of the underlying dose calculation algorithm and its limitations and uncertainties. In case of the TG43 formalism, larger dose calculation uncertainties are present in the lower dose region and tissue heterogeneity effects can cause large errors at tissue-lung or tissue-air interfaces (patient surface) due to different

scatter conditions. This will result in an overestimation of dose inside the tissue close to a tissue-air interface for example. Larger TG43 dose calculation errors also occur in high-Zmaterial. In the mid- and high-dose region inside the liver, the TG43 formalism yields accurate results. After plan approval, the technical plan data is transferred to the brachytherapy control computer. The integrity of this data transfer must be validated for each plan and plausibility checks should be performed. All relevant parameters of the treatment planmust be cross checked with the ones of the brachytherapy device, e.g. the current source activity or overall irradiation time. The duration of the irradiation is determined by the size of the target volume, the number of catheters, the planned dose and the activity of the source (diminishing with time due to radioactive decay); typically, it is between 5 and 40 minutes for a source with an activity of 370 GBq, but can even exceed 90 minutes in cases of very large target volumes. The dose is usually applied in a single fraction. The prescribed dose depends on the tumour histology. An overview is given in table 2. The dose prescription aims to achieve a full target coverage with the aim that the prescribed dose should encompass 100% of the PTV volume. Therefore, the D100%, D98%, D95% and D90% dose values of the PTV are usually of primary interest when evaluating the target coverage and should be reported in addition to the median PTV dose. Generally, the dose reporting principle is based on the recommendations for IMRT treatments of ICRU report 83 [53] and on the recommendations for brachytherapy of the cervix of ICRU report 89 [54]. Most metrics can be directly used in HDR brachytherapy of the liver, but for example the near maximum dose D2% of the PTV loses its validity, since extremely high dose values can occur close to the source inside the target volume. Since small regions of very high dose can considerably influence the PTVmean dose, the PTVmedian dose is favourable, although the PTV mean dose might also be of prognostic value and can be reported additionally [57]. For the same reason, the relative hyperdose volumes inside the PTV (e.g. V150%, V200% or V300%) can be reported optionally. Regarding constraints for organs at risk, most constraints are D0.1cm 3 or D1.0cm 3 dose constraints (e.g. esophagus, stomach, duodenum, bowel, heart, vessels, bile ducts, spinal canal), while for parallel organs (e.g. liver, kidney) usually volumetric constraints are applied. Level 2 dose reporting (DVH parameters [see table 2], treatment technical data and additional information about the TPS) is the current standard of care in modern HDR brachytherapy. The typical dose rate of a newly delivered iridium-192 source with a source strength of 40 kU (ca. 370GBq) is > 7Gy/minute at 1 cm from the source and thus, radiobiological effects are comparable to that of strongly hypofractionated or even single fraction external beam radiotherapy treatments, such as stereotactic body radiotherapy or stereotactic radiosurgery. It should be noted that the average gamma energy emitted by an iridium-192 brachytherapy source is approximately by a factor 15 lower than the average photon energy of a 6 MV beam from a linear accelerator. 10. DOSE, DOSE RATE, FRACTIONATION

Primary and secondary liver malignancies

11

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

Figure 6: The excess end of the brachytherapy catheter to the angiography sheath outside the patient can be measured with a sterile ruler as a second validation and can be used to validate the internal excess of the brachytherapy catheter.

Figure 7: Examplar cases with dose optimized to 1 × 20 Gy (green isoline). The puncture tract is treated with ~5 Gy to the catheter surface (see right image). Left image: Oligometastasis (breast cancer), prescription dose 1×20 Gy. Blue isoline: 15 Gy, green isoline: 20 Gy, yellow isoline: 30 Gy, white isoline: 60 Gy, GTV volume: 10.5 cm 3 , D100%: 20.1 Gy, D98%: 23.5 Gy, D95%: 25.8Gy, D90%: 29.7 Gy; Dmean: 59.7 Gy, OAR: Healthy liver volume: 1088 cm 3 , V5Gy: 22%, V10Gy: 9%. Right image: Oligometastasis (duodenal carcinoma), prescription dose 1×20 Gy. Dark blue isoline: 10 Gy, blue isoline: 15 Gy, green isoline: 20 Gy, yellow isoline: 30 Gy, red isoline: 40 Gy, white isoline: 60 Gy, GTV volume: 540 cm 3 , D100%: 18 Gy, D98%: 20Gy, D95%: 21.4 Gy, D90%: 23 Gy; Dmean: 42.8 Gy, OAR: Heart D0.1 cm 3 : 17.6 Gy, D1.0cm 3 : 16.5 Gy, Stomach D0.1 cm 3 : 11.2 Gy, D1.0 cm 3 : 9.9 Gy; Healthy liver volume: 1981 cm 3 , V5Gy: 65%, V10Gy: 32%.

11. MONITORING

image set and the deviations quantified. Tracking devices (e.g. electromagnetic tracking) allow for the validation of the correct source positions during dose delivery [60]. For each different brachytherapy technique, possible emergency scenarios must be considered. An advantage of the Seldinger technique is that with a catheter-in-catheter technique, the emergency procedure is easy to handle, even in the very rare case of a detached source (broken weld seam between source and Bowden drive cable) inside any catheter. In this case, the (source containing) brachytherapy catheters can be quickly removed from the angiography sheaths. Immediate measures must be taken to prevent bleeding after catheter removal. Catheter removal Since both, hepatic metastases and liver parenchyma are typically well perfused, complications may arise from bleeding along the

Treatment delivery After approval of the treatment plan and transfer to the afterloader software, the brachytherapy catheters are connected to the afterloader via transfer tubes and the dose is delivered to the target (see Figure 8). Verification of the correct catheter localization before initiating the dose delivery is essential. To do so, an additional low dose CT or equivalent imaging modality of the treatment region including all catheters immediately prior to the plan application should be considered, in order to ensure that no significant changes in catheter position and patient anatomy occurred between planning CT / MR and starting the dose delivery. For this purpose, imaging must be possible without moving the patient, since this could again lead to unexpected changes of anatomy or even catheter displacements.The freshly acquired CT / MR can then be registered to the planning

Primary and secondary liver malignancies

12

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/07/2022

catheter path after removal. When the Seldinger technique is used, it is possible to perform catheter tract embolisation by applying gelatine sponge sealing (see chapter "technique") during stepwise withdrawal of the angiography sheath. In addition, standardised follow-up is important for early detection of complications (e.g. ultrasound after 1–2 hours) and routine monitoring (noninvasive blood pressure, heart rate, oxygenation and ECG) should be ensured over the first 4 post-interventional hours. Ideally, patient monitoring after removal of the catheters can be performed in a dedicated and adequately staffed monitoring area and patients are not brought to ward prior to documentation of normal vital observations over 2 hours.

Figure 8: Connection of the transfer tubes to the brachytherapy catheters prior to dose delivery.

TABLE 2 PLANNING AIM AND ORGANS AT RISK (OAR) CONSTRAINTS [1,49,58,59] Planning aim D 100% OAR constraints D 0.1 cm³

D 1.0 cm³

Metastases of colo-rectal cancer 25 Gy

Oesophagus

15 Gy 15 Gy 15 Gy 15 Gy 15 Gy 22 Gy

12 Gy 12 Gy 12 Gy 12 Gy 12 Gy

Cholangiocellular Carcinoma Hepatocellular Carcinoma

Stomach

20 Gy 15 Gy 20 Gy

Duodenum Small bowel Large bowel

Other metastases

Heart

Great vessels (Aorta, Vena cava inferior) 37 Gy Gall bladder 20 Gy Spinal canal 12 Gy

10 Gy

Skin surface Main bile duct

10 Gy 20 Gy

18 Gy

V 10Gy < 33% of healthy liver volume V 5Gy < 66% of healthy liver volume

Liver

Liver

TABLE 3 PUBLISHED LITERATURE OF PRIMARY LIVER LESIONS

Treated lesions (number)

Median overall survival (range)

D 100% (range)

Study (Year)

Lesion size (range) 4.4 cm (1-15 cm) 7.1 cm (5-12 cm) 5 cm (1.8-12 cm) 4.5 cm (3-6.5cm) 5.2 cm (1-18 cm)

Entity

Study design Patients (% male)

Local control

Mohnike (2010) [63] Collettini (2012) [79] Collettini (2015) [40] Mohnike (2019) [80] Schnapauff (2012) [61] Jonczyk (2018) [43]

Prospective, single-center 83 (79.5%) 140

15 Gy (12-15 Gy) 15.8 Gy (15-20 Gy)

Hepatocellular

95% at 12 m

19.4 m

Retrospective, single-center 35 (83%)

93.3% at 12.8 m mean FU 91.5% at 20 m median FU Time to untreated progression: 67.5% at 12 m

35

15.4 m

Retrospective, single-center 98 (83.7%) 192

16.5 Gy

29.2 m

Randomized phase II, single-center

37 (84%)

n/a

15 Gy

78.4% at 12 m

Cholangiocellular carcinoma

Retrospective, single-center 15 (33%) Retrospective, single-center 61 (62%)

20 Gy (15-20 Gy)

22

10 m (1-25 m)

14 m

2 cm (1-3.8 cm) versus 6.9 cm (4-14.8 cm)

15.5 m (<4 cm) versus 10 m (>4 cm)

142

20 (15-20 Gy) 87% (<4 cm) versus 78% (>4 cm) at 12 m