MULTIDISCIPLINARY EXPERT CONSULTATION
Due to the complexity of the case. It was presented at the multidisciplinary tumor board and a decision was made to proceed with microwave ablation of the liver lesions. Fluorodeoxyglucose (FDG) positron emission tomography (PET) CT obtained prior to microwave ablation demonstrated an increase in the size of the liver lesions (FDG avid) and new sub-centimeter FDG avid lesions in the peritoneum, suggestive of metastatic implants (Figure 3). Two of the three lesions, now measured more than 3 cm, prompting us to hold off on the microwave ablation. The patient was again presented to the multidisciplinary tumor board and was not considered a surgical candidate for hepatic metastasectomy or HIPEC due to general deconditioning, including severe weight loss. In light of the progressive hepatic and limited extrahepatic disease burden, the patient was offered selective internal radiation treatment or RE with Y90 and initiation of folinic acid, fluorouracil, and irinotecan (FOLFIRI) with bevacizumab, as a palliative measure. FOLFIRI was initiated prior to the Y-90 RE procedure; however, the bevaxizumab was not started until after the RE procedure.
Figure 3 Coronal and axial fluorodeoxyglucose positron emission tomography images show fluorodexoyglucose avid liver (white arrows) and peritoneal (black arrows) metastases.
A mapping angiogram was performed one week prior to the Y-90 RE treatment to evaluate the vascular anatomy and estimate the lung shunt fraction (LSF). As a part of the mapping angiogram, the gastroduodenal artery was embolized using 4 and 5 mm 0.018” Azur® CX coils (Terumo Interventional Systems). Based on the planar images obtained after injection of 99mtechnetium-labeled macroaggregated albumin (99mTcMAA) into the right hepatic artery, the shunt fraction toward the lung was calculated at approximately 2.5%.
Calculations of liver and tumor volumes were performed on a 3D segmentation application (TeraRecon Inc Foster City, CA, United States). The whole liver volume was 1325 mL. The right lobe measured 875 mL, the tumor volume was 150 mL, and the tumor to normal liver ratio (T:N) based off the 99mTcMAA injection during the mapping angiogram was 3:1. The prescribed activity was calculated using the DAVYR application (Dose and activity visualizer for Y-90 RE) (Figure 4) and the SIR sphere microsphere activity calculator (SMAC) chart provided by the manufacturer and a 1.0 Gigabecquerel (Gbq) activity of Y90 was prescribed for treatment of the right lobe metastatic lesions. Based on the body surface area (BSA) method recommended by the manufacturer, the dose to the tumor was estimated as 126 Gray (Gy) and dose to the healthy liver was estimated at 40.9 Gray (Gy) (Figure 4). The prescribed activity was successfully administered by positioning a Progreat 2.8 F 130 cm catheter (Terumo Interventional Systems) in the right hepatic artery distal to the origin of the cystic artery (Figure 5). Post procedure Bremsstrahlung planar and single photon emission computed tomography (SPECT) demonstrated appropriate distribution of Y90 spheres in the hepatic lobe without evidence of non-target embolization (Figure 6).
Figure 4 Dose and activity visualizer for Yttrium-90 radioembolization (DAVYR) application chart obtained after inputting the necessary fields, displaying the medical internal radiation dose, body surface area and partition models with prescribed activity Gigabecquerel, estimated dose to the tumor, liver and lungs Gray.
Figure 5 Hepatic angiogram.
Yttrium 90 administration to the right lobe. The tip of the catheter (purple arrow) is at the bifurcation of the sectoral branches of the right hepatic artery. Coils in the gastro-duodenal artery (yellow arrow).
Figure 6 Axial post procedure single photon emission tomography image shows appropriate distribution of Yttrium 90 particles in the right hepatic lobe with increased focal uptake in one of the right lobe lesions (white arrow).
The incidence of appendiceal cancer is 1.2 per 100000 people per year in the United States. Potentially due in part to this rarity, there are no accepted American Joint Commission on Cancer staging system or National Comprehensive Cancer Network evidence-based guidelines recommended for appendiceal cancer. Colon cancer workup, staging, and treatment is commonly applied to appendiceal carcinoma.
Appendiceal cancers can either present with or progress to peritoneal dissemination and the optimal therapy is complete surgical cytoreduction in combination with HIPEC. For patients not eligible for complete cytoreduction and HIPEC, as in this case, systemic chemotherapy with regimens utilized for metastatic CRC are generally considered the mainstay therapy. We embarked on a literature review to find potential treatment options for our patient.
A systematic review of the literature was performed using Pubmed for articles from inception to January 2019 using the key words “Y90”, “metastatic Appendix cancer” and “systemic chemotherapy”. We did not find any studies in which Y-90 RE was used to treat liver metastasis form appendix cancer. However we found several studies using Y 90 in the salvage setting following failure of chemotherapy with improved overall survival numbers[2-4]. Data was extracted by two extractors.
Interventional oncology treatments for liver dominant CRC metastasis include, RE, drug eluting beads irinotecan (DEBIRI) transarterial chemoemboization (TACE) and thermal ablation.
RE with Y-90 microspheres is based on the same principle as that of TACE, in that the hepatic artery provides the primary vascular supply to the tumor, while the portal vein provides the majority of blood flow to normal hepatic parenchyma. Furthermore, the microvascular density of hepatic tumors is much higher than the hepatic parenchyma, resulting in a proportionately higher exposure to microspheres. CRC cells are highly radiosensitive and even chemo-refractive lesions do not generally demonstrate any cross-resistance to radiation. Additionally, radiation works synergistically when used with radiation-sensitizing chemotherapeutic drugs.
The two commercial manufacturers of Y-90 microspheres are TheraSpheres (BTG International, ON, Canada) and SIR-spheres (Sirtex Medical Inc., Sydney, Australia). These microspheres, that have been specifically developed to deliver Y-90, measure 20-60 μm in size. Y-90 primarily emits beta radiation, with a half-life of 64.1 h and an average energy of 0.94 MeV. The beta particles have a range of 1.1 cm (average 2.5 mm in vivo) from the source, with 94% of the dose being deposited within the first 11 d of administration. Hence, RE works from “inside out” and inflicts minimal damage to the normal liver. This is in contrast to external beam radiation which exposes the normal liver parenchyma to excessive amounts of radiation, in the process of killing multifocal tumors, thus increasing collateral damage. One of the most critical components of the RE procedure is the dosimetry. There is one activity calculation method for Y-90 glass microspheres, however three alternatives exist for resin microspheres[11-13].
Both liver and tumor volumes, calculated based on CT or magnetic resonance imaging are used to determine resin microsphere activity. The empiric model recommends administration of 2.0 GBq for < 25% involvement, 2.5 GBq for 25%-50% involvement, and 3.0 GBq for > 50% involvement. This method has low safety margins[13,14] and is generally not recommended. The empiric method adjusted for patient BSA has been more widely utilized to calculate the activity needed to treat liver tumors. This model assumes that the patient’s BSA is proportional to liver volume, thus leading to a method of dose adjustment. Activity in GBq = (BSA - 0.2) + [VTumor / (VTumor + VLiver)]. Where VTumor is the tumor volume and VLiver is liver volume.
Lastly, the partition model is the only calculation method predicated on Medical Internal Radiation Dose principles (MIRD)[15-17]. The partition model identifies the lungs, tumor, and uninvolved liver parenchyma as three distinct vascular compartments that can be partitioned during RE and the dose to the individual components can be estimated for optimization and safety of the procedure. A slightly simplified MIRD based equation calculates the dose to the liver as follows: D liver (Gy) = Injected activity (GBq) × 50 × fractional uptake liver × [m liver (kg)] – 1.
With fractional uptake (FU) of the liver defined as: FU liver = (1-LSF) [m liver/TLR × m tumor+ m liver], where m liver is the liver mass, TLR is tumor to liver ratio and LSF is LSF.
This model usually works well with discrete tumor that can be separated from the surrounding parenchyma (generally hepatocellular cancer). The BSA activity planning is suited for small lesions with diffuse or infiltrative margins, which cannot be separated from the uninvolved liver parenchyma (generally metastasis).
A tumoricidal dose for SIR spheres is considered to be 120 Gy, so little benefit is gained when exceeding this dose, while parenchymal damage continues to increase with larger radiation doses[18,19]. The dose for uninvolved, normal parenchyma should preferably be at 50 Gy and should never exceed 70 Gy.
For TheraSphere® noncompartmental MIRD macrodosimetry proposed for by Salem et al is used. It uses the following equation for computation of the activity and assumes Y 90 distributes uniformly within the liver and decays completely in situ. Activity required (GBq) = [desired dose (Gy)] × [target liver mass (kg)]/50.
Of note, Therasphere dosimetry is independent of the tumor volume and depends solely on the infused tissue mass. For Therasphere, the recommended mean absorbed is between 80 and 150 Gy per lobe. The recommended tumoricidal doses is 120 Gy for HCC (due to the underlying cirrhosis) and 150 Gy for metastases. Maximum exposure to the lungs should be less than 30 Gy per session and 50 Gy cumulatively. Both manufacturers recommend prescribed activity reductions for patients with lung shunts higher than 10%[10,13].
Several investigators have studied the application of Y-90 RE in the salvage setting for liver dominant metastatic CRC[2,3]. Van Hazel et al reported results of combined Y-90 RE with systemic irinotecan therapy in patients who failed initial 5-FU chemotherapy. In a group of 25 patients, 11 (48%) had partial response (PR) and 9 (39%) had stable disease. Median survival was 12.2 mo.
DEBIRI TACE has been proposed by some as a possible alternative to RE, however evidence within the literature is currently limited. In our institute we consider it in cases where Y-90 RE is contraindicated or in the setting of failed RE therapy. TACE with DEBIRI involves two separate treatment sessions for each lobe, with each session occurring 4 wk apart. Based on this, patients with bilobar disease, may need four procedures. During each treatment session, as much as one vial of 100- to 300-μm LC Beads (Biocompatibles, Oxford, CT) loaded with 100-mg irinotecan is selectively delivered to treat the hepatic lesions. Furthermore, post-embolization syndrome is more severe after DEBERI as compared to Y-90, with moderate to severe abdominal pain observed in up to 30% of treatment sessions[20,21].
Thermal ablation has been successfully used in treatment of non rescectable CRC liver metastasis, with best outcomes for isolated liver lesions less than 3 cm without systemic spread. Although our patient initially appeared to be a good candidate for thermal ablation, his staging PET showed increased in size of the hepatic lesions and new lesions in the peritoneum, making him a less suitable candidate for liver ablation alone and prompted us to use systemic therapy and more aggressive liver directed therapy in the form of Y-90 RE to the right lobe.
Several studies in the past 2 years have reported that patients have worse survival outcomes with right sided colon cancer, and they may benefit less from standard therapies[23,24]. The subset of patients with right sided tumors in the SIRFLOX, FOXFIRE and SIRFLOX global trials had a better overall survival from addition of RE in combination with first line chemotherapy. Gene expression analyses have identified 4 consensus molecular subtypes (CMSs) which are different with right and left-sided CRCs, with greater proportions of the “microsatellite unstable/immune” CMS1 and the “metabolic” CMS3 subtypes found in right-sided colon cancers, which may explain their different behavior and natural history.