br Safety br In total patients experienced severe
In total, 11 patients (3%) experienced severe day-of-treatment complications (Table 4). Severe adverse events within the first week after SIRT were rare; three patients experienced grade 3 fatigue in the 7 days after SIRT and one patient experienced grade 3 abdominal pain in the first week after SIRT. One hundred and forty-three patients experienced an adverse event. In total, 253 adverse events were recorded, of which 19 (8%) were grade 3 or above (Table 4). The most common events were mild fatigue and abdominal pain (grade 1e2). Relatedness of complications
Table 2 Treatment planning and procedure details
Parameter Data for 399 patients
Location of liver tumour(s)
Number of liver tumours
Bilirubin (mmol/l) prior to SIRT Median 9.0
Arteries embolised before SIRT
SIRT procedure target/type
in single session)
Number of administrations
SIRT microsphere brand
Percentage tumour to liver volume Median 15.0
Prescribed activity (GBq) for SIR- Median 1.64
Prescribed activity (GBq) for Median 3.91
Length of stay in hospital following
administered with SIRT
Table 2 (continued )
Parameter Data for 399 patients
Post-SIRT chemotherapy received
SIRT, selective internal Nigericin therapy; IQR, interquartile range.
and adverse events to the SIRT procedure were not routinely recorded. Events categorised as ‘other’ with a free-text description accounted for 53 (21%) of the total. Seven events of grade 3 or above were recorded in the ‘other’ category, which were as follows: acute kidney injury (grade 3; occurred 28 days after SIRT), bowel obstruction (grade 3; 21 days after SIRT); liver abscess (grade 3; 138 days after SIRT), skin rash (grade 3; 90 days after SIRT), delirium/de-mentia (grade 4; 79 days after SIRT), pulmonary emboli (grade 4; 47 days after SIRT); sepsis (grade 4; 18 days after SIRT). In total, 353 events were recorded as abnormal lab-oratory values (Table 4). The most common biochemical event categories were raised aspartate aminotransferase (22%), raised alanine aminotransferase (21%) and hypo-albuminaemia (19%). Eighteen of the 353 events (5%) were grade 3. No severe cases of radiation-induced liver disease (RILD), gastrointestinal ulceration, radiation pneumonitis, radiation cholecystitis or radiation pancreatitis were recorded.
Fig 1. KaplaneMeier curve of overall survival following selective in-ternal radiation therapy in patients with metastatic colorectal cancer liver metastases. Ninety-five per cent confidence intervals are shaded; the numbers at risk at 3-month intervals are displayed.
KaplaneMeier analysis and univariate Cox proportional hazards model of survival by baseline characteristics. Statistically significant P-values are shown in bold
Subgroup n (patients) n (events) Median overall Overall survival Hazard ratio P-value
survival (months) 95% confidence (95% confidence
Number of lines of previous chemotherapy (including biologics); Log-rank test P ¼ 0.098
Prior biological therapy (includes bevacizumab, cetuximab, aflibercept); Log-rank test P ¼ 0.783
Presence of extrahepatic metastases; Log-rank test P [ 0.021
Percentage tumour to liver volume (continuous)
Percentage tumour to liver volume; Log-rank test P < 0.001
ECOG, Eastern Cooperative Oncology Group.
* Reference category for univariate Cox regression analysis.
SIRT is reimbursed for the treatment of liver metastases from CRC in most developed countries at a national or regional level (Supplementary Table S2), but its effect on overall survival and cost-effectiveness in patients with CRC liver metastases has not been shown in prospective, rand-omised phase III studies. As prospective, randomised controlled clinical trials can take a decade or longer to address specific research questions , some health sys-tems have opted to study the treatment in a registry-based commissioning model to address specific deficiencies in the published literature and to accelerate advancement to full commissioning. This study was specifically commissioned to provide ‘real-world’ evidence on the survival of patients
treated with SIRT in a salvage setting to inform future commissioning policy decisions.
Patients with unresectable CRC liver metastases whose disease has progressed following chemotherapy have very few treatment options. New locoregional liver-directed therapies, such as SIRT or transarterial chemo-embolisation with drug-eluting beads, are emerging but have not yet become the standard of care. Patients in the control arm of clinical trials treated with best supportive care (BSC) in a salvage setting have a median overall survival ranging from 2.4 months  to 6.6 months . In this same population, NICE has recommended trifluridine-tipiracil on the basis of two randomised controlled trials (RCTs) that showed an improvement in overall survival by 2.0e2.4 months above BSC [11,12]. We carried out a systematic evidence review of