Evaluation of Liver Biomarkers as Prognostic Factors for Outcomes to Yttrium-90 Radioembolization of Primary and Secondary Liver Malignancies

Abstract

The objective of this study was to examine indicators of liver function and inflammation for prognostic value in predicting outcomes to yttrium-90 radioembolization (RE). In a retrospective analysis, markers of liver function and inflammation, biomarkers required to stage liver function and inflammation, and data regarding survival, tumor response, and progression after RE were recorded. Univariate regression models were used to investigate the prognostic value of liver biomarkers in predicting outcome to RE as measured by survival, tumor progression, and radiographic and biochemical tumor response. Markers from all malignancy types were analyzed together. A subgroup analysis was performed on markers from patients with metastatic colorectal cancer. A total of 31 patients received RE from 2004 to 2014. Median survival after RE for all malignancies combined was 13.6 months (95% CI: 6.7–17.6 months). Results from an exploratory analysis of patient data suggest that liver biomarkers, including albumin concentrations, international normalized ratio, bilirubin concentrations, and the model for end-stage liver disease score, possess prognostic value in predicting outcomes to RE.

Introduction

Many cancers affect the liver in the form of a primary or secondary liver malignancy. Hepatocellular carcinoma (HCC) often results from chronic hepatocellular injury and regeneration due to various factors, including viral hepatitis infection, alcohol abuse, and nonalcoholic fatty liver disease, and is associated with primary liver malignancies. Recently, there has been an increase in hepatitis C virus-related HCC incidence with models predicting the current incidence doubling by 2024.¹ Although less common, intrahepatic cholangiocarcinoma (ICC) may also produce primary liver malignancies.² Secondary liver malignancies are often the result of metastatic colorectal cancer (mCRC), but can also arise due to metastatic neuroendocrine tumors (mNET) and metastatic breast cancer.^3,4 Colorectal cancer, the third most prevalent cancer in the world, develops into metastatic disease in an estimated 40% of patients.^5,6 Primary and secondary liver metastases can often be surgically unresectable and refractory to chemotherapy requiring different treatment modalities.

Radioembolization (RE) is a type of brachytherapy that has shown promise in the treatment of unresectable primary and secondary liver malignancies.⁷ During RE, glass or resin microspheres impregnated with the β-emitting radionuclide yttrium-90 (⁹⁰Y) are infused into the hepatic artery through catheterization. The microspheres become lodged in the arterioles of the tumor resulting in internal irradiation of the tumor by the ⁹⁰Y β-particle emission (Fig. 1).⁸ Normal liver parenchyma receives blood from two sources, including the portal vein and the hepatic artery; however, malignancies within the liver are primarily vascularized by the hepatic artery alone. Administration of the radioactive microspheres into only the hepatic artery upstream of the malignancy allows for selective delivery of radioactivity to tumor cells while keeping the radiation dose to the noncancerous portions of the liver within tolerable levels.⁹

FIG. 1.

Yttrium-90 microspheres in the vasculature of a tumor. Compliments of Sirtex Medical, Inc.

Despite a growing interest in investigating prognostic factors associated with outcome to RE, there remains no consensus on which factors are of the greatest value and use. Although biomarkers and scoring systems, which assess liver function and inflammation, have long established prognostic value and clinical use in the treatment of liver disease, limited data are available on how they might correlate to outcomes after RE.^10

–14 Further identification of possible prognostic factors, specifically those relating to markers of liver function and inflammation, is needed to guide larger trials. The objective of this study was to perform an exploratory analysis to examine liver biomarkers and scoring systems for prognostic value in predicting outcomes to RE.

Materials and Methods

This study was a single-institution, retrospective exploratory data analysis of patients who had received ⁹⁰Y microsphere (SIR-Spheres^®) RE from January 2004 to July 2014 at the University of New Mexico Hospital. Patient electronic charts were reviewed for demographics as well as data relevant to disease history and the RE procedure. Demographic data recorded included sex, age, and race. Data regarding tumor disease recorded included tumor type, extrahepatic tumor presence, and hepatic lobe involvement. Procedural data recorded included amount of radioactivity administered, technical success of RE procedure, number of RE procedures performed, and toxicities after RE. Other data recorded include liver function tests drawn before RE, serum markers of tumor progression, number and lines of treatments previously received, as well as laboratory values and medical history needed to calculate Child-Pugh, aspartate aminotransferase to platelet ratio index (APRI), and model for end-stage liver disease (MELD) scores. Outcome data recorded included date of death, date of disease progression, tumor maximum diameter as determined by computed tomography before and after RE, and biochemical markers of tumor progression recorded before and after RE. MELD, APRI, and Child-Pugh scores were calculated according to the methods outlined in Figure 2 for each patient.

FIG. 2.

Calculation of the Child-Pugh, AST to APRI, and MELD scores.^10
–12 AST, aspartate aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; INR, international normalized ratio; MELD, model for end-stage liver disease; ULN, upper limit of normal.

Resin ⁹⁰Y microspheres (SIR-Spheres; Sirtex Medical Limited, North Sydney, NSW, Australia) were delivered by an interventional radiologist according to established guidelines and institutional protocol for RE. Preprocedural scans and examinations, including administration of Technetium-99m Macroaggregated albumin, were performed for all patients to determine the suitability of the patient's vascular anatomy for RE administration. The dosage of radioactivity was calculated according to Sirtex guidelines using a body surface area model. Sequential RE administration was performed as opposed to simultaneous administration for those patients receiving treatments to multiple lobes. This study was approved by the University of New Mexico Human Research Review Committee. Informed consent was waived due to the retrospective nature of this study.

The primary outcome of this study was overall survival (OS). OS was defined as time from first RE to death from any cause. Secondary outcomes included progression-free survival (PFS), biochemical response, and radiographic response. PFS was defined as time from first RE to either disease progression or death, whichever occurred first. Biochemical response was defined as a negative percent change in carcinoembryonic antigen (CEA) from pre- to post-RE levels (mCRC only). All CEA levels were measured within 3 months before or after RE. Radiographic response was determined by computed tomography and defined by response evaluation criteria in Solid Tumors (RECIST) 1.1 criteria. Follow-up occurred by retrospective chart review until a patient passed away or until the end of July 2014, whichever occurred first. If patients were alive at the end of the study or were lost to follow-up, their data were included and censored at the time of the last known visit.

Descriptive measures of statistics were performed by calculating median, range, and frequency as appropriate. Median survival was calculated using the Kaplan–Meier method. Liver function tests, liver scoring systems, and patient characteristics, including age, sex, extrahepatic tumor involvement, presence of bilobar disease, and number of prior therapies, were correlated to survival (OS and PFS) in an exploratory analysis using the log-rank test and univariate Cox proportional hazards model. Model assumptions of proportional hazards were checked using Schoenfeld residuals. These factors were analyzed for correlation to biochemical and radiographic response using Fisher's exact tests and univariate logistic regression models. Data were analyzed for all malignancy types combined. A subgroup analysis was performed on data from patients with mCRC. All statistical analyses were performed using R, version 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

A total of 31 patients received RE from 2004 to 2014 for liver tumors caused by mCRC (12), HCC (8), mNET (6), and ICC (5). Successful delivery of the planned dose of microspheres occurred during all but two procedures. There were 13 patients who underwent a second RE procedure. Treatment was generally well tolerated. Toxicities were usually mild and typically consisted of nausea and abdominal pain, which was well managed with supportive care. Results from descriptive analysis of patient characteristics are presented in Table 1. Results from descriptive analysis of liver biomarkers and scoring systems are presented in Table 2.

Table 1.

Baseline Patient Demographics

	All malignancy types, n = 31	mCRC, n = 12	HCC, n = 8	mNET, n = 6	ICC, n = 5
Age	62 (38–77)	60 (38–73)	61.5 (57–77)	62.5 (45–71)	66 (42–75)
Sex
Male	24 (77.4%)	7 (58.3%)	7 (87.5%)	5 (83.3%)	5 (100%)
Female	7 (22.6%)	5 (41.7%)	1 (12.5%)	1 (16.7%)	0 (0%)
Race
White	19 (61.2%)	7 (58.3%)	7 (87.5%)	2 (33.3%)	3 (60%)
Hispanic	5 (16.2%)	1 (8.3%)	1 (12.5%)	3 (50%)	0 (0%)
American Indian or Alaskan Native	2 (6.4%)	0 (0%)	0 (0%)	1 (16.7%)	1 (20%)
Other	5 (16.2%)	4 (33.3%)	0 (0%)	0 (0%)	1 (20%)
Weight (kg)	81.4 (50.2–115.6)	79.4 (50.2–115.6)	85.9 (71.6–104.1)	69.8 (58.6–108.2)	76.8 (60–115)
Dose administered (GBq)	1.22 (0.74–1.73)	1.14 (0.91–1.48)	1.36 (0.74–1.67)	1.54 (0.74–1.73)	1.22 (1.05–1.56)
Lobe involvement
Bilobal	11 (35.5%)	6 (50%)	1 (12.5%)	3 (50%)	1 (20%)
Unilobal	20 (64.5%)	6 (50%)	7 (87.5%)	3 (50%)	4 (80%)
No. of prior therapies	3 (1–7)	4 (3–7)	1.5 (1–4)	2 (1–3)	2 (1–5)

Values are presented as number (percentage) or median (range).

mCRC, metastatic colorectal cancer; HCC, hepatocellular carcinoma; mNET, metastatic neuroendocrine tumor; ICC, intrahepatic cholangiocarcinoma.

Table 2.

Indicators of Liver Function

	All malignancy types, n = 31	mCRC, n = 12	HCC, n = 8	mNET, n = 6	ICC, n = 5
Albumin (g/dL)	3.6 (2.8–4.2)	3.75 (3.2–4.2)	3.4 (2.9–3.7)	3.6 (3.2–3.9)	3.5 (2.8–4.1)
Alkaline phosphatase (ng/mL)	140 (59–1074)	116 (71–265)	170.5 (79–403)	148.5 (59–270)	160 (110–1074)
Alanine aminotransferase (U/L)	38 (12–361)	34 (15–49)	92.5 (12–361)	35.5 (22–45)	36 (20–131)
Aspartate aminotransferase (U/L)	37 (18–446)	29.5 (18–49)	105.5 (31–446)	36.5 (18–109)	53 (18–349)
INR	1.07 (0.91–1.33)	1.06 (0.93–1.16)	1.07 (0.96–1.22)	1.02 (0.91–1.16)	1.18 (1.03–1.33)
Total bilirubin (mg/dL)	0.8 (0.3–1.7)	0.5 (0.3–0.8)	0.95 (0.8–1.7)	0.75 (0.3–1.4)	0.9 (0.6–1.3)
Total protein (mg/dL)	7.2 (6–8)	7.25 (6.5–8)	7.3 (6–8)	7.25 (7–7.7)	6.9 (6.4–7.6)
Creatinine (mg/dL)	0.9 (0.46–1.7)	0.94 (0.46–1.2)	0.78 (0.7–1.3)	0.85 (0.7–1.1)	1.35 (0.55–1.7)
Platelet count (×1000)	205 (46–576)	230 (143–576)	137 (74–210)	216 (46–452)	192 (62–382)
MELD score	7 (6–13)	7 (6–8)	8.5 (6–13)	7 (6–9)	11 (7–13)
APRI score
<1.5	22 (71%)	12 (100%)	2 (25%)	5 (83.3%)	3 (60%)
≥1.5	9 (29%)	0 (0%)	6 (75%)	1 (16.7%)	2 (40%)
Child-Pugh class
A	27 (87.1%)	12 (100%)	5 (62.5%)	5 (83.3%)	5 (100%)
B	4 (12.9%)	0 (0%)	3 (37.5%)	1 (16.7%)	0 (0%)

Values are presented as number (percentage) or median (range).

MELD, model for end-stage liver disease; APRI, aspartate aminotransferase to platelet ratio index; INR, international normalized ratio.

Median OS for all malignancies combined was 13.6 months (95% CI: 6.7–17.6 months), while median OS for only those patients with mCRC was 10.4 months (95% CI: 3.6–17.8 months). Median PFS for all malignancies combined was 4.3 months (95% CI: 2.9–7.5 months), while median PFS for only those patients with mCRC was 4.1 months (95% CI: 1.6–8.1 months). Survival curves for all malignancies combined and mCRC are presented in Figures 3 and 4, respectively. After RE, there were five patients (16%) who had progressive disease, and the remainder had stable disease per RECIST 1.1 criteria. There were four patients (33%) with mCRC that had a CEA response.

FIG. 3.

OS and PFS for all malignancy types combined. OS, overall survival; PFS, progression-free survival.

FIG. 4.

OS and PFS for metastatic colorectal cancer.

Univariate regression of predictor variables and OS resulted in two significant associations. Increased albumin was found to be significantly associated with increased OS for all malignancies combined (p = 0.049, HR = 0.21, 95% CI: 0.05–0.99) as well as for mCRC alone (p = 0.02, HR = 0.02, 95% CI: 0.001–0.52). No other liver function tests, liver scoring systems, or patient characteristics were found to be significantly associated with OS.

Univariate regression of predictor variables and PFS resulted in three significant associations. Increased international normalized ratio (INR) was found to be significantly associated with decreased PFS for all malignancies combined (p = 0.01, HR = 874, 95% CI: 4.62–165337). A higher MELD score was significantly associated with decreased PFS for all malignancies combined (p = 0.03, HR = 1.27, 95% CI: 1.03–1.56). Increasing total bilirubin concentrations in the range of 0.3–0.8 mg/dL were found to be significantly associated with increased PFS in mCRC patients (p = 0.03, HR = 0.001, 95% CI: 0.001–0.53). No other liver function tests, liver scoring systems, or patient characteristics were found to be significantly associated with PFS.

Univariate regression of predictor variables to CEA response and to RECIST criteria did not indicate any significant associations.

Discussion

As albumin concentrations before RE increased, OS after RE increased for patients with all malignancy types combined and for patients with mCRC. A growing body of evidence suggests that albumin concentrations are an important prognostic factor for RE. In one recent study, albumin concentrations >3 g/dL were found to be significantly associated with increased OS in patients with mCRC who underwent RE.¹⁵ In another recent study, albumin concentrations >2.8 g/dL were significantly associated with increased survival in patients treated with RE for HCC.¹⁶

As INR increased before RE, PFS after RE decreased for patients with all malignancies combined. This finding did not remain significant in mCRC subgroup analysis, indicating that this may only be pertinent to the tumors analyzed other than mCRC. This finding is similar to that found in a recent study, which determined that an INR >1.2 was significantly associated with decreased survival in patients with HCC who were treated with RE.¹⁷

A higher MELD score was found to be associated with decreased PFS for all malignancy types combined. This finding did not remain significant during mCRC subgroup analysis. This suggests that the prognostic value of the MELD score may only be applicable for tumors arising from HCC, ICC, or mNET. In patients receiving RE treatment for HCC tumors, a recent study determined that the MELD score was a significant prognostic factor for OS.¹⁴ It is important to note that INR is a component in the calculation of the MELD score. As such, some of the significance seen in the MELD score trend was likely due to the influence of INR.

This study indicated that mildly elevated total bilirubin in patients with mCRC was associated with increased PFS. Bilirubin is sometimes evaluated in RE candidates as it has been recognized that high bilirubin concentrations are associated with poorer OS.^9,18 Although this finding would appear contradictory to this established trend, it is important to examine the range of bilirubin concentrations in this sample. The correlation was established in patients with mCRC who had a range of bilirubin concentrations from 0.3 to 0.8 mg/dL, and conclusions drawn from extrapolation of the trend outside of this range may be invalid. Most studies report that values above 1.3 mg/dL were associated with poorer outcomes. This finding would suggest that bilirubin concentrations closer to 1 mg/dL, but not necessarily higher than 1 mg/dL, may be associated with better outcomes.

The Child-Pugh and APRI scoring systems did not demonstrate a significant predictive value for the outcomes of this study. However, individual components of these scoring systems did on occasion demonstrate significant predictive value (for example, albumin). While these scoring systems have demonstrated clinical value for evaluating hepatic function and inflammation outside of the context of liver malignancies, their role in evaluating response to RE may be limited.

The strengths of this study include analysis of a large number of potential prognostic factors as well as the inclusion of data from tumors other than mCRC. Other strengths include retrieval of data from a single institution with a standardized protocol, reporting of RE use primarily as salvage therapy, and data from a population with diversified demographics. There were several limitations to this study. The small sample size used in this study did not permit for a multivariate analysis due to the risk of overfitting the regression models. Small sample sizes also prohibited subgroup analysis of each malignancy type. This resulted in a combination of all malignancies into a single analysis that limits the applicability of some of these findings to a single malignancy type. Another limitation includes the study's retrospective observational design, which permits for possible selection bias.

Conclusion

There may be prognostic value in biomarkers and scoring systems that assess liver function and inflammation for predicting outcomes to RE. Albumin concentrations, INR, bilirubin concentrations, and MELD scores were found to each have a significant association to OS and PFS in univariate analysis. As such, they warrant further investigation by larger trials and meta-analysis to validate their prognostic value and clinical usefulness.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Rein

, Wittenborn

, Weinbaum

, et al. Forecasting the morbidity and mortality associated with prevalent cases of pre-cirrhotic chronic hepatitis C in the United States. Dig Liver Dis, 2011; 43:66.

Mouli

, Memon

, Baker

, et al. Yttrium-90 radioembolization for intrahepatic cholangiocarcinoma: Safety, response, and survival analysis. J Vasc Interv Radiol, 2013; 24:1227.

Vyleta

, Coldwell

. Radioembolization in the treatment of neuroendocrine tumor metastases to the liver. Int J Hepatol, 2011; 2011:785315.

Pentheroudakis

, Fountzilas

, Bafaloukos

, et al. Metastatic breast cancer with liver metastases: A registry analysis of clinicopathologic, management and outcome characteristics of 500 women. Breast Cancer Res Treat, 2006; 97:237.

Siegel

, Ma

, Zou

, et al. Cancer statistics, 2014. CA Cancer J Clin, 2014; 64:9.

Haddad

, Bani Hani

, Pawlik

, et al. Colorectal liver metastases. Int J Surg Oncol, 2011; 2011:285840.

Zurkiya

, Ganguli

. Beyond hepatocellular carcinoma and colorectal metastasis: The expanding applications of radioembolization. Front Oncol, 2014; 4:150.

Kennedy

, Nutting

, Coldwell

, et al. Pathologic response and microdosimetry of (90)Y microspheres in man: Review of four explanted whole livers. Int J Radiat Oncol Biol Phys, 2004; 60:1552.

Kennedy

. Radioembolization of hepatic tumors. J Gastrointest Oncol, 2014; 5:178.

10.

Wai

, Greenson

, Fontana

, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology, 2003; 38:518.

11.

Kamath

, Weisner

, Malinchoc

, et al. A model to predict survival in patients with end-stage liver disease. Hepatology, 2001; 33:464.

12.

Pugh

, Murray-Lyon

, Dawson

, et al. Transection of the oesophagus for bleeding oesphageal varices. Br J Surg, 1973; 60:646.

13.

Fahmueller

, Nagel

, Hoffmann

, et al. Predictive and prognostic value of circulating nucleosomes and serum biomarkers in patients with metastasized colorectal cancer undergoing Selective Internal Radiation Therapy. BMC Cancer, 2012; 12:5.

14.

Weng

, Ertle

, Zheng

. A new model to estimate prognosis in patients with hepatocellular carcinoma after Yttrium-90 radioembolization. PLoS One, 2013; 8:e82225.

15.

Lewandowski

, Memon

, Mulcahy

, et al. Twelve-year experience of radioembolization for colorectal hepatic metastases in 214 patients: Survival by era and chemotherapy. Eur J Nucl Med Mol Imaging, 2014; 41:1861.

16.

Memon

, Kulik

, Lewandowski

, et al. Comparative study of staging systems for hepatocellular carcinoma in 428 patients treated with radioembolization. J Vasc Interv Radiol, 2014; 25:1056.

17.

Sangro

, Carpanese

, Cianni

, et al. Survival after yttrium-90 resin microsphere radioembolization of hepatocellular carcinoma across Barcelona clinic liver cancer stages: A European evaluation. Hepatology, 2011; 54:868.

18.

Dunfee

, Riaz

, Lewandowski

, et al. Yttrium-90 radioembolization for liver malignancies: Prognostic factors associated with survival. J Vasc Interv Radiol, 2010; 21:90.