In patients with advanced colorectal cancer, liver metastases are the primary cause of morbidity and mortality. Unfortunately, only a minority of patients is a candidate for surgical resection with curative intent. Patients with irresectable disease will first receive palliative systemic therapy. Though, despite major advances in systemic treatment, overall survival remains disappointing in this subgroup of patients [13, 14].
Current standard of care in the systemic treatment of metastatic colorectal cancer is based on cytotoxic fluoropyrimidines, oxaliplatin, and irinotecan, as well as targeted therapy with the vascular endothelial growth factor (VEGF)-targeted monoclonal antibody bevacizumab [15–19]. There is no clear preference for sequential exposure to these drugs during consecutive lines of treatment or upfront combination therapy [20, 21]. When disease progression or intolerable toxicity occurs during first-line treatment, patients will subsequently receive another regimen as second-line treatment, with the choice of the regimen depending on the first chemotherapeutic agents. In the Netherlands, the first-line regimen of choice is CAPOX-B (capecitabine, oxaliplatin, and bevacizumab). After initial treatment with six cycles of CAPOX-B, maintenance therapy with capecitabine and bevacizumab is given until disease progression [22]. Subsequently an irinotecan-based regimen, e.g., FOLFIRI (leucovorin, fluorouracil (5-FU), and irinotecan) or irinotecan monotherapy, is indicated as second-line therapy. Only patients with a KRAS wild-type tumor may benefit from additional (third-line) treatment with an epidermal growth factor receptor (EGFR)-targeted monoclonal antibody (panitumumab or cetuximab).
Liver-directed therapy such as RE is increasingly applied as an alternative to achieve local disease control. Currently, two types of yttrium-90 microspheres are used in worldwide clinical practice: resin (SIR-Spheres, SIRTeX, Lane Cove, Australia) and glass (TheraSphere; BTG, Ottawa, ON, Canada) yttrium-90 (90Y) microspheres. In salvage patients with colorectal liver metastases, who have no regular treatment options left and an average life expectancy of less than 6 months, median overall survival after RE treatment with 90Y microspheres is around 12 months when given as monotherapy or in combination with chemotherapy [1]. Besides, treatment is generally well tolerated, with typical clinical toxicity being limited to mild symptoms of fatigue, abdominal pain, nausea, vomiting and/or fever during the first 2 weeks after treatment [23].
Despite these benefits of RE, there is still room for improvement. Unintentional deposition of radioactive microspheres in tissues other than the liver may cause serious treatment complications. Therefore, a safety procedure is performed in the week(s) before the actual treatment. During this procedure, coil embolization of extrahepatic branches may be performed and a strategic catheter position is chosen before administering a (harmless) scout dose of technetium-99 m-labelled macro-aggregated albumin (99mTc-MAA). Afterward, SPECT/CT and planar nuclear scintigraphy are obtained to exclude the presence of extrahepatic activity and significant liver-to-lung shunting. The treatment procedure is typically performed 1–2 weeks later, with the administration of 90Y microspheres from identical catheter positions, followed by posttreatment imaging with bremsstrahlung SPECT/CT or 90Y-PET/CT [24].
As a second topic of possible improvement, the intrahepatic distribution of therapeutic 90Y microspheres cannot be accurately predicted in advance. The scout dose of 99mTc-MAA particles differs markedly in embolic effect, size, weight, and number of particles infused [25], and therefore fails to predict the intrahepatic distribution of 90Y microspheres in most cases [26–28]. Besides, imaging of the 90Y microspheres biodistribution itself is already a challenge due to the lack of ?-radiation emission. Traditionally, bremsstrahlung SPECT/CT has been used for posttreatment imaging, but it suffers from a low spatial resolution. Internal-pair production-based 90Y-PET/CT has become available as an alternative for quantitative imaging, but the low count rate and inherent noise limit its applicability in daily clinical care [29–31].
Third, it is generally assumed that the preferential arterial vascularization of liver tumors will lead to a selective targeting of tumorous tissue following intra-arterial infusion of radioactive microspheres. It is known from pathological examinations of livers treated with 90Y RE that radioactive microspheres cluster preferentially within the peripheral tumor vasculature. The concentration of microspheres can be up to 200 times greater in the tumor periphery than in the tumor center and the healthy liver tissue [32]. Various studies have investigated dose-response relationships in RE. The majority of these studies found strong associations between T/N ratios, absorbed radiation doses, tumor response and overall survival [27, 33–39]. Yet, the degree of tumor targeting, as expressed by the T/N microsphere uptake ratio, shows marked interindividual variability in practice, with a reported range of 0.6–25.9 [40]. This heterogeneity in T/N uptake ratios is likely a result of various factors, including differences in tumor angiogenesis, microsphere characteristics, catheter position, and flow-bound distribution physics [25, 33]. A recent investigation demonstrated that up to 60 % of patients with liver metastases treated with 166Ho RE had at least one tumor that received less than or an equal amount of radioactivity as compared to the surrounding healthy liver tissue (T/N???1) [4]. Flamen et al., also reported similar findings, with 38 % of the metastatic liver lesions in their study having an unfavorable T/N uptake ratio (1) after RE with 90Y microspheres [39]. Since unfavorable T/N uptake ratios cannot be predicted and only become apparent after treatment, timely adjustments in treatment technique are not yet feasible.
The highly variable tumor targeting is an important clinical problem that may at least explain some of the inconsistencies in reported tumor response rates after RE [1, 41]. Considering the reported dose-response relationship, it can be expected that improved T/N ratios will positively affect tumor response after RE. It may also reduce hepatotoxicity, since healthy liver tissue absorbed dose has previously been correlated to biochemical toxicity [37]. Improvement of T/N ratios is especially important in CRLM, since metastases from this tumor type are relatively hypovascular compared with other tumor types (such as neuroendocrine tumors or uvea melanoma), and generally exhibit low T/N ratios.
The above outlined shortcomings of current RE practice are being addressed in the SIM trial, for which trial accrual has started as of November 2014. The distinctive imaging capacities and availability of an identical scout dose of 166Ho microspheres, combined with the promising effects on particle fluid dynamics facilitated by the ARC, may result in an optimized treatment technique of RE in patients with CRLM.
Trial status
Patient recruitment was ongoing at the time of submission.