Survival and tolerability of liver radioembolization: a comparison of elderly and younger patients with metastatic colorectal cancer

Abstract

Aim

To evaluate the outcomes among elderly (≥70 years) and younger patients (<70 years) with liver-dominant metastatic colorectal cancer (mCRC) who received radioembolization (RE) as salvage therapy.

Methods

A retrospective review of 107 consecutive patients with unresectable mCRC treated with RE after failing first- and second-line chemotherapy.

Results

From 2002 to 2012, 44 elderly and 63 younger (<70 years) patients received RE. Patients had similar previous extensive chemotherapy and liver-directed interventions. Using modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria, either a stable or a partial radiographical response was seen in 65.8% of the younger compared with 76.5% of the elderly patients. RE was equally well tolerated in both groups and common procedure-related adverse events were predominantly grade 1–2 and of short duration. No significant difference was found with regard to overall median survival between younger [8.4 months; 95% confidence interval (CI) = 6.2–10.6] or elderly patients (8.2 months; 95% CI = 5.9–10.5, P = 0.667). The presence of extrahepatic disease at the time of RE was associated with a significantly worse median survival in both groups.

Conclusion

Radioembolization appears to be as well tolerated and effective for the elderly as it is for younger patients with mCRC. Age alone should not be a discriminating factor for the use of radioembolization in the management of mCRC patients.

Introduction

Colorectal cancer (CRC), a leading cause of cancer-related death worldwide, is predominately a disease of elderly people with a median age at onset exceeding 70 years.^1,2 In 2013, over 142 000 patients were diagnosed with CRC in the United States alone, and over 50 000 died from this disease.³ The liver is the most common site of metastases from CRC and 15–25% present with liver metastases at time of diagnosis.⁴ In a minority, a liver resection is feasible and represents the only chance for a cure.⁴ Even if a resection is attempted, two-thirds of the patients will develop recurrence and eventually succumb to their disease.⁵ Although patients with unresectable mCRC have a poor outcome, advances in chemotherapy and availability of biological agents have improved the median survival in those patients and even rendered some them resectable.⁴ Unfortunately, a significant number of patients have chemorefractory disease and their management becomes a clinical dilemma.

Radioembolization has emerged as a useful treatment option in the armamentarium against chemorefractory liver dominant metastatic (m)CRC.^6–8 In a recent systemic review, Saxena et al. evaluated 20 studies comprising 979 patients with mCRC who received radioembolization after failing conventional chemotherapy regimens and concluded that radioembolization is both safe and effective in the salvage setting and should be more widely utilized. Of interest, this review highlights the preferential treatment of younger patients with radioembolization.⁹ The median age of the patients enrolled in the studies was 10 years younger than the median age at first diagnosis of CRC in the wider patient population (for instance, 71 years in the United States³). Thus, the results from the various studies about tolerability and improved outcomes may not be generalizable to the elderly patients who are under-represented in the published studies.

As the life expectancy improves with the general population, discussions on the best ways to manage ageing CRC patients have become increasingly relevant. Several studies have shown similar cancer-specific and disease-free survival figures in all age groups receiving chemotherapy/biological agents or undergoing curative surgery for metastatic colorectal cancer.^10–12 The effects of advancing age on the tolerance and clinical outcomes after radioembolization in elderly patients are largely unknown. Therefore, a retrospective analysis was conducted to evaluate the clinical outcomes among elderly compared with younger patients with liver-dominant mCRC treated in the salvage setting with radioembolization.

Patients and methods

Patients and radioembolization technique

Patients with chemorefractory liver dominant mCRC who were treated consecutively with intra-arterial ⁹⁰Y microspheres between August 2002 and April 2012 at the University of Pittsburgh Medical Center were included. Prior to treatments, patients were evaluated by multidisciplinary teams. Patients eligible to receive RE were not candidates for a hepatic resection or ablation, had refractory disease after multiple chemotherapy regimens, an untreated life expectancy of >12 weeks, Eastern Conference Oncology Group (ECOG) performance status of 0–2, had adequate hepatic function (serum total bilirubin <2.0 mg/dl) and had adequate renal (creatinine <2.0 mg/dl) and haematological function (i.e. granulocyte count 1.5 × 10⁹/l, platelets 50 × 10⁹/l). Patients that were considered for RE had either exhausted or refused standard chemotherapy regimens. Patients with extrahepatic metastases were treated only if the tumour burden outside the liver was <10% of the total tumour burden, non-progressive extrahepatic disease in the past 3 months was evidenced by imaging and if chemotherapy options were not available. Data were recorded via an Institutional Review Board approved protocol.

The technical details and dosimetry of the process have previously been described¹³ but a modified partition model was employed for the calculation of the ⁹⁰Y microsphere activity to be administered to the patient.¹⁴ The prescribed activity was calculated to deliver 50 Gy to the targeted liver tissue. Prior to RE, all patients underwent selective visceral angiography, a technetium-99 m-labelled macroaggregated albumin (99mTc-MAA) study and a baseline computed tomography (CT) or positron emission tomography (PET)/CT. The selective visceral angiogram allows for definitive assessment of the arterial anatomy and possible embolization of vessels that may lead to extrahepatic ⁹⁰Y exposure. The 99mTc-MAA study allows pulmonary shunting to be evaluated. The use of RE is avoided if there is any uncorrectable extrahepatic shunting to the gastrointestinal tract, or >0.6 GBq (corresponding to a lung dose of 30 Gy) is shunted to the lungs.

RE was administered via unilobar treatments. When bilobar disease was present, the lobes were treated sequentially with approximately a 4-week interval. Some patients had multiple treatments to the same lobe. The determination to treat the same lobe repetitively was made by evaluating the performance status, liver function and the extent of extrahepatic disease. However, the determination to retreat patients was based solely on progression of disease as assessed via CT. Patients are observed in the hospital overnight and then discharged the next day with oral narcotics, proton pump inhibitors and antiemetics.

Follow-up and evaluation

Survival data were acquired from the electronic medical record and a search of the Social Security Index. All adverse events were graded using the National Cancer Institute Common Toxicity Criteria Adverse Events Version (CTCAE) 3.0. The nature and severity of all procedure-related adverse events and laboratory changes were evaluated from day 1 to month 3 after RE. Patients were followed with weekly laboratory data and routine office follow-up. A carcinoembryonic antigen (CEA) response was defined as a ≥50% decrease in post-treatment value from baseline values obtained at time of RE. Imaging was evaluated for a response to radioembolization using the modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria reviewing CT scans from 1–6 months after treatment. A radiographical response was graded according to mRECIST. A complete response was defined as disappearance of any intra-tumoural arterial enhancement in all target lesions. A partial response was defined as at least a 30% decrease in the sum of diameters of viable (enhancement in the arterial phase) target lesions, taking as reference the baseline sum of the diameters of target lesions. Progressive disease is an increase of at least 20% in the sum of the diameters of viable (enhancing) target lesions, taking as reference the smallest sum of the diameters of viable (enhancing) target lesions recorded since treatment started. Stable disease is any case that does not qualify for either partial response or progressive disease.¹⁵

Statistical analysis

All patients were followed up until death or September 2013. Statistical analysis was performed using SPSS 17.0 for Windows (SPSS Inc., Chicago, IL, USA). Descriptive statistics were performed on socio-demographic and disease-specific variables. Group comparisons on disease-specific variables were performed using chi-square analyses for categorical variables and Student's t-test for continuous variables. Overall survival (OS) was obtained by Kaplan–Meier survival analyses (log-rank). Survival was calculated from the time of first RE to the time of death. The association of multiple variables was examined with overall survival using Cox's proportional hazard regression analysis. Co-variables found on univariate analyses to have an association with survival were included in the model. The Cochran–Mantel–Haenszel was used to compare the CTCAE distribution between groups. For all comparisons, significance was at the 0.05 level.

Results

Patient and treatment characteristics

A total of 191 REs were performed in 107 patients during this time period and data are presented in Table 1. Both groups had similar baseline liver function tests performance status. Three patients beyond the inclusion criteria for radioembolization (bilirubin of 2–3 mg/dl) were treated at the discretion of the physician. Both groups (elderly vs. younger patients) had undergone extensive chemotherapy prior to radioembolization; 62.8% vs. 68% of patients received either bevacizumab or cetuximab, 17.1% vs. 28% both, and 22.8% vs. 36% had received prior capecitabine, respectively.

Table 1.

Baseline patient, disease and treatment characteristics and response among elderly (≥70 years) and younger patients (<70 years)

Characteristic	Parameter	Age ≥ 70 years	Age < 70 years	P-value
Gender	Male	33 (75%)	38 (60.3%)	0.143
Age, years	Mean ± SD	75.6 ± 5.4	55.7 ± 8.7	<0.001*
	Range	70.0–87.1	37.3–69.2
ECOG performance status	0	26/43 (60.5%)	33/60 (55%)	0.489
	1	14/43 (32.6%)	22/60 (36.7%)
	2	3/43 (6.9%)	5/60 (8.3%)
Prior procedures	Surgical	13/41 (31.7%)	19/61 (31.1%)	0.992
	RFA	11/41 (26.8%)	11/61 (18%)	0.331
Extra-hepatic disease	Yes	12/41 (29.3%)	29/62 (46.8%)	0.012*
Pulmonary nodules at Rx	Yes	14/41 (34.1%)	29/62 (46.8%)	0.052
Total bilirubin, mg/dl	Mean ± SD	0.75 ± 0.31	0.74 ± 0.46	0.419
Months from Dx of mCRC to 1st Rx	Median (Range)	20.11–175	24.52–92	0.205
Number of treatments	Total Number	79	112
	1 treatment, n (%)	16 (36.4%)	22 (34.9%)
	2 or more treatments, n (%)	28 (63.6%)	41 (65.1%)
Bilobar treatment, n (%)	Yes	23 (52.3%)	34 (53.9%)	1.000
Lung shunt (%)	Mean ± SD	3.1 ± 2.5	2.9 ± 2.0	0.599
Radiation dose to tissue per treatment	Mean ± SD (Gy)	43.6 ± 9.9	45.0 ± 7.0	0.429
Radiation activity given per treatment	Mean ± SD (mCi)	28.0 ± 13	30.8 ± 14.4	0.318
Treatments completed with ≥80% of prescribed dose administered, n (%)	≥80%	64 (81.0%)	94 (83.9%)	0.698
	<80%
CEA responsea,c n (%)	≥50% decrease	10/34 (28.6%)	13/48 (27.1%)	0.966
Radiographic Responsec,d, n (%)	Total available	34	41
	Partial response	4 (11.8%)	5 (12.2%)	0.998
	Stable disease	22 (64.7%)	22 (53.6%)	0.357
	Progressive disease	8 (23.5%)	14 (34.2%)	0.445

^*Statistically significant.

^aCEA response defined as a reduction in CEA ≥50% of the pretreatment level.

^bmRECIST criteria used to compare baseline measurement prior to 1st treatment with radiological response during 1–6 months follow-up imaging.

^cPercentages calculated based on available data.

A greater proportion of younger patients had extrahepatic disease (including portocaval or retroperitoneal lymphadenopathy, anastomotic recurrence or unresected primary colorectal tumour, peritoneal disease, bone metastases, and adrenal metastases) at the time of radioembolization, and a non-significant greater proportion of younger patients had pulmonary nodules at the time of treatment.

No patients were excluded from treatment owing to either an unacceptable level of pulmonary shunting or uncorrectable shunting to the extrahepatic gastrointestinal tract. There were no differences in the dose administered to the target lobe per treatment, the prescribed activity of ⁹⁰Y administered, and the activity that was actually delivered between the elderly and the younger patients. All those patients who received <80% of the prescribed dose had the treatment terminated owing to stagnation of flow secondary to the embolic process.

Treatment response

There was no significant difference in serial CEA levels available for review, and in CEA response. Although all patients had baseline and follow-up imaging obtained, only 34 (77%) elderly patients and 41 (65%) younger patients had imaging available for review. No patients had a complete radiological response. There was no significant difference in radiographical response in the elderly when compared with the younger patients (Table 1).

Safety and tolerability

Radioembolization was equally well tolerated in both the elderly and younger patients. Common procedure-related adverse events (fatigue, nausea and/or vomiting, abdominal pain, fever, and increased bilirubin) were predominately of mild-to-moderate intensity and of short duration (Table 2). Two patients in the elderly group and one patient in the younger group required readmission within 30 days. Reasons for readmission included an upper gastrointestinal tract bleed related to oesophageal varices 4 days after treatment in the younger patient, and unresolved abdominal pain and development of symptomatic brain metastases in two elderly patients. The post-treatment hepatic toxicity was assessed between the elderly and younger patients and found to be relatively mild in both groups. No patient had fulminant hepatic failure after treatment. One patient in each group developed a late (after 30 days) grade 4 toxicity that was related to a biliary stricture and resolved with endoscopic retrograde cholangiopancreatography and stenting. No patients developed post-treatment gastric or duodenal ulceration.

Table 2.

Main procedure-related adverse events in the elderly (≥70 years) and younger patients (<70 years) in the first 3 months post-treatment by severity (CTCAE v3)

CTCAE	CTCAE v3: number (%) of patients					P-value between sub-groups
Study sub-group	All Grades	Grade 1	Grade 2	Grade 3	Grade 4
Fatigue
Age ≥ 70 years	10 (22.7%)	8 (18.1%)	1 (2.3%)	1 (2.3%)	–
Age <70 years	15 (23.8%)	12 (19.0%)	2 (3.2%)	1 (1.6%)	–
Abdominal pain
Age ≥ 70 yearsr	7 (15.9%)	6 (13.6%)	1 (2.3%)	–	–
Age < 70 yearsr	12 (19.0%)	9 (14.3%)	2 (3.2%)	1 (1.5%)	–
Nausea and/or vomiting
Age ≥ 70 years	12 (27.3%)	9 (20.5%)	2 (4.5%)	1 (2.3%)	–
Age < 70 years	20 (31.7%)	16 (25.4%)	4 (6.3%)	–	–
Fever
Age ≥ 70 years	5 (11.4%)	4 (9.0%)	1 (2.4%)	–	–
Age <70 years	6 (9.5%)	5 (7.9%)	1 (1.6%)	–	–
Peak total bilirubina
Age ≥ 70 years	11 (29.7%)	7 (18.9%)	–	3 (8.1%)	1 (2.7%)
Age < 70 years	15 (26.8%)	9 (16.0%)	3 (5.4%)	2 (3.6%)	1 (1.8%)

CTCAE v3: Common Terminology Criteria for Adverse Events version 3.0.

^aPercentages calculated based on available data.

Overall survival

Forty-three (97.7%) and 60 (95.2%) patients died during the follow-up period in the elderly and younger group, respectively. Six elderly patients and eight younger patients died within 12 weeks of receiving their RE. Kaplan–Meier analysis revealed no significant difference in survival after radioembolization between elderly and younger patients [median 8.2 (95% confidence interval (CI) 5.9–10.5) months vs. 8.4 (95% CI 6.2–10.6) months, respectively; P = 0.351] (Fig. 1).

Figure 1.

Kaplan–Meier analysis of overall survival by age: No significant difference was found with regard to overall median survival between younger or elderly patients (8.4 vs. 8.2 months, P = 0.667)

Univariate and multivariate analyses for the entire cohort and the different groups are presented in Table 3 and Table 4. For the entire cohort, extrahepatic disease at the time of treatment was the only independent predictor of worse overall survival among the variables examined (Table 3). However, none of the variables were independent predictors of worse overall survival on multivariate analysis in each subgroup when analysed separately (data not shown).

Table 3.

Univariate and multivariate analysis of overall survival in all patients

Variable	Univariate analysis	Multivariate analysisa
	P-value	Hazard ratio	95% CI	P-value
Age	0.219
Gender	0.827
Extrahepatic spread	<0.001*	1.667	1.022–2.717	0.040*
Synchronous disease	0.143
Presence of pulmonary nodules	0.039*	1.215	0.775–1.908	0.395
Previous liver directed intervention	0.030*	1.591	0.990–2.556	0.055
CEA response	0.560
Number of treatments	0.004*	1.418	0.881–2.283	0.150

^*Statistically significant.

^aMultivariate hazard ratios calculated for the clinical variables that were found to be significant on univariate analysis.

CEA, carcinoembryonic antigen.

Table 4.

Comparison of survival by age in the elderly (≥70 years) and younger patients (<70 years)

Variable	Parameter	Age ≥ 70 years			Age <70 years
		Median survival (months)	95% CI	P-value	Median survival (months)	95% CI	P-value
Extrahepatic disease	Yes	4.2	1.5–6.9	<0.001*	6.0	3.7–8.4	0.022*
	No	10.8	7.8–13.8		9.0	7.9–11.1
CEA response	≥50% decrease	13.6	2.4–24.8	0.010*	10.3	7.1–13.7	0.725
	<50% decrease	3	2.8-3.2		9.7	7.1–11.9
Radiographical response	Partial	18.4	9.1–27.8	0.037*	26.1	9.3–43	0.273
	Stable	9.0	7.2–10.8		10.2	9.0–11.5
	Progressive	4.2	2.9-5.5		9.2	7.1–11.3
Number of treatments	1	5.2	3.1–10.2	0.117	5.6	2.5–9.4	0.203
	2+	10.8	7.4–13.6		10.2	7.6–12.2
Synchronous disease	Synchronous	9.0	5.3–11.9	0.707	9.5	6.0–12.2	0.093
	Metachronous	7.4	4.9–19.1		6.7	3.7–9.2
Presence of pulmonary nodules	Yes	5.1	2.9–14.5	0.388	7.4	5.6–9.4	0.246
	No	10.2	5.6–13.6		10.2	6.7–13.0
Previous liver-directed therapy	Yes	10.2	5.3–17.0	0.573	11.4	8.4–20.3	0.048*
	No	9.0	4.6–13.4		7.4	5.6–9.5
Gender	Male	9.0	5.6–11.9	0.505	8.4	5.3–9.5	0.090
	Female	5.1	3.0–17.0		10.2	6.0–18.3

^*Statistically significant.

CEA, carcinoembryonic antigen.

Further analysis by patient age revealed that the median survival in the very old (≥75 years, n = 19) and in patients <75 years (n = 88) was 10.7 (95% CI 2–19.6) and 8.0 (95% CI 6.7–9.2) months, respectively; P = 0.740. In patients ≥65 years (n = 55) and <65 years (n = 52), the median survival was 8.0 (95% CI 6.0–9.9) and 8.4 (95% CI 6.0–10.8) months, respectively; P = 0.452 (Supporting information Figs S1 and S2).

Discussion

In this study we attempted to evaluate whether elderly and younger patients with advanced chemorefractory mCRC undergoing radioembolization have similar outcomes. Golfieri et al. performed a similar comparison of the survival and tolerability of radioembolization in patients with unresectable hepatocellular carcinoma.¹⁶ They showed radioembolization to be equally well tolerated in both elderly and younger patients with no difference in overall survival. We similarly show that outcomes after radioembolization in terms of tolerability and efficacy are equivalent between the elderly and younger patients with mCRC. Our results suggest that elderly patients should be treated with the same strategy for chemorefractory mCRC as their younger counterparts, at least as long as there is no severe co-morbidity or contraindication to radioembolization.

Our study confirmed that the majority of patients were heavily pretreated irrespective of age. In studying some 15 000 patients, Egenvall et al. detected a trend towards a less frequent use of aggressive therapies in elderly patients, but other studies have shown that for patients who receive aggressive therapy, the treatments are equally tolerated and effective.^{10,11,17–20} Although the elderly group is presumed to poorly tolerate more intensive and invasive treatments, in reality, other factors such as functional performance rather than pure chronological age may come into play in the decision making.

The safety analysis addressed the question of whether elderly patients experienced greater levels of toxicity than younger patients. Common procedure-related adverse events were predominately low grade and of short duration but observed equally in elderly and younger patients. Safety and tolerability in our study population were similar to previous experiences with radioembolization for primary and metastatic malignancies.⁹ An increase in bilirubin to a grade ≥3 in the first 3 months after radioembolization was observed in a similar proportion between the elderly and younger patients. In our study, the observed grade ≥3 CTCAE bilirubin toxicity was similar to the rate reported (6–14%) in other large studies.^16,21 This early rise in bilirubin reflects some degree of radioembolization-induced liver disease (REILD) that has been described by Sangro et al. as a form of sinusoidal obstruction syndrome appearing 4–8 weeks after radioembolization that is not true liver decompensation.^22,23

Regarding efficacy, our study demonstrated a similar median survival, regardless of age after radioembolization in both groups. The median survival in our population was similar to other published reports comparing radioembolization to best supportive care.^6,24 Identifying prognostic factors is imperative to optimize patient selection and improve outcomes. In our study, we identified the presence of extrahepatic disease to be an independent predictor of worse overall survival in the entire cohort and associated with increased mortality in both the elderly and younger patients. To date, there is no generalized consensus on recommending radioembolization for patients with extrahepatic disease. Some centres do not accept any extrahepatic disease, whereas others accept patients with non-progressive extrahepatic deposits for treatment.^6,7 Limited extra-hepatic disease is currently not deemed a contraindication to radioembolization at our institution if this abides by the criteria listed above. In spite of this, further research is required to establish what burden of disease and which sites of extra-hepatic disease are most likely associated with a poor outcome. Of interest, we found that elderly patients had significantly less extra-hepatic disease at the time of radioembolization. This difference could be related to delayed tumour growth (less aggressive tumours) in elderly patients as we found that younger patients tended to have more advanced disease and a higher likelihood of presenting with synchronous disease at the time the cancer was diagnosed. However, this could also be explained by the fact that a more aggressive approach and more frequent utilization of radioembolization are adopted when treating younger patients with extrahepatic disease as compared with older patients.

Besides radioembolization, other liver-directed intra-arterial interventions have been evaluated as potential treatment options for unresectable, chemorefractory mCRC. Unlike radioembolization, bland transarterial embolization has shown no improvement in survival compared with best supportive care. Also, a Cochrane review concluded that there is minimal if any evidence to support TAE/TACE in treatment mCRC to the liver.²⁵

The main limitation of the present study is its retrospective single-centre design although many patients were followed prospectively. This study therefore gives information only about the patients who were treated. The results must be viewed in light of the fact that radioembolization was performed in elderly patients with good performance status and may not be extrapolated to a more frail elderly population with a poor performance status. Larger and prospective studies are needed to validate these findings.

In spite of these limitations, our data show that fit elderly patients should not be excluded from receiving radioembolization simply on the grounds of advanced age, as it is well tolerated and effective in this group. Age alone should not be a discriminating factor for the management of chemorefractory mCRC patients.

Conflicts of interest

None declared.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher*s web-site:

Figure S1 Kaplan–Meier analysis of overall survival in the very elderly patients ≥75 and patients <75 years

Figure S2 Kaplan–Meier analysis of overall survival between patients ≥75 and patients <65 years