Advertisement

CardioVascular and Interventional Radiology

, Volume 40, Issue 1, pp 69–80 | Cite as

Embolotherapy for Neuroendocrine Tumor Liver Metastases: Prognostic Factors for Hepatic Progression-Free Survival and Overall Survival

  • James X. Chen
  • Steven Rose
  • Sarah B. White
  • Ghassan El-Haddad
  • Nicholas Fidelman
  • Hooman Yarmohammadi
  • Winifred Hwang
  • Daniel Y. Sze
  • Nishita Kothary
  • Kristen Stashek
  • E. Paul Wileyto
  • Riad Salem
  • David C. Metz
  • Michael C. Soulen
Clinical Investigation

Abstract

Purpose

The purpose of the study was to evaluate prognostic factors for survival outcomes following embolotherapy for neuroendocrine tumor (NET) liver metastases.

Materials and Methods

This was a multicenter retrospective study of 155 patients (60 years mean age, 57 % male) with NET liver metastases from pancreas (n = 71), gut (n = 68), lung (n = 8), or other/unknown (n = 8) primary sites treated with conventional transarterial chemoembolization (TACE, n = 50), transarterial radioembolization (TARE, n = 64), or transarterial embolization (TAE, n = 41) between 2004 and 2015. Patient-, tumor-, and treatment-related factors were evaluated for prognostic effect on hepatic progression-free survival (HPFS) and overall survival (OS) using unadjusted and propensity score-weighted univariate and multivariate Cox proportional hazards models.

Results

Median HPFS and OS were 18.5 and 125.1 months for G1 (n = 75), 12.2 and 33.9 months for G2 (n = 60), and 4.9 and 9.3 months for G3 tumors (n = 20), respectively (p < 0.05). Tumor burden >50 % hepatic volume demonstrated 5.5- and 26.8-month shorter median HPFS and OS, respectively, versus burden ≤50 % (p < 0.05). There were no significant differences in HPFS or OS between gut or pancreas primaries. In multivariate HPFS analysis, there were no significant differences among embolotherapy modalities. In multivariate OS analysis, TARE had a higher hazard ratio than TACE (unadjusted Cox model: HR 2.1, p = 0.02; propensity score adjusted model: HR 1.8, p = 0.11), while TAE did not differ significantly from TACE.

Conclusion

Higher tumor grade and tumor burden prognosticated shorter HPFS and OS. TARE had a higher hazard ratio for OS than TACE. There were no significant differences in HPFS among embolotherapy modalities.

Keywords

Neuroendocrine tumor Liver metastases Embolization 

Introduction

Liver-directed embolotherapies for neuroendocrine tumor (NET) liver metastases have demonstrated efficacy for tumor growth reduction and symptom relief [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], and are supported by current consensus guidelines [12, 13, 14, 15, 16]. The optimal embolotherapy modality among conventional lipiodol and drug-eluting bead (DEB), transarterial chemoembolization (TACE), transarterial radioembolization (TARE), and transarterial embolization (TAE) remains unknown [3, 4, 6, 7, 17, 18]. Comparison of existing evidence has been confounded by heterogeneity of older pathologic grading systems [2, 4, 5, 7, 10, 19] and uncertain prognostic impact of patient demographics, different tumor primary sites, technical variations, and concurrent systemic therapy with somatostatin analogs [20, 21], cytotoxic chemotherapy, everolimus [22], sunitinib [23], and peptide receptor radionuclide therapy (PRRT) [24]. The lack of recommendations regarding embolotherapy technique in current guidelines has manifested in wide practice variation [25].

The purpose of this study was to evaluate the impact of disease and treatment-related factors on hepatic progression-free and overall survival of patients with NET liver metastases in current practice at multiple major cancer centers, to support the design of a prospective international multicenter Randomized Embolization Trial for NeuroEndocrine Tumor Metastases to the Liver (RETNET) (ClinicalTrials.gov Identifier NCT02724540).

Materials and Methods

Patient Cohort

This was an institutional review board-approved multicenter retrospective study of 202 patients with liver NET metastases treated with embolotherapy from 9/2004 to 2/2015 at eight institutions. Patients treated with prior hepatic embolotherapy at outside institutions or with unavailable pre- and post-treatment clinical and imaging data were excluded. Patients previously treated with local cytoreductive techniques including partial hepatectomy or thermal ablation were included in analysis.

Tumor Characteristics

Tumor grade was designated by the World Health Organization (WHO) 2010 classification. All cases submitted without WHO 2010 grade (n = 23) were excluded from analysis. Hepatic tumor burden was stratified by ≤50 and >50 % liver volume involvement using simple visual estimate. Cases without known tumor burden (n = 4) were excluded from analysis. No tumor primary sites were excluded from analysis, including cases of unknown primary site. The presence and location of extrahepatic metastases were determined on review of pretreatment imaging.

Patient Evaluation and Follow-Up Protocol

All patients were evaluated with contrast-enhanced computed tomography (CT) (triple phase or conventional) or magnetic resonance imaging (MRI). Follow-up clinical evaluation, lab tests, and imaging (CT or MRI) were performed at 4–6 weeks following embolotherapy, with 2- to 6-month intervals subsequently for the first 2 years, and at least annually thereafter with minor variations among institutions.

Embolotherapy Selection and Technique

Embolotherapies included conventional TACE, TARE, and TAE. No patients treated with DEB-TACE had complete data for analysis, and were therefore excluded (n = 20), so this technique could not be evaluated. The embolotherapy technique and materials used were at the discretion of the treating interventional radiologist. TACE was performed using lipiodol and doxorubicin alone or in combination with mitomycin C and cisplatin. TARE was performed using resin (SIR-Spheres, Sirtex Medical, Sydney, Australia) or glass (Theraspheres, BTG International, London, UK) 90Y microspheres, with administered activity according to manufacturers’ specifications. TAE was performed using microspheres (Embospheres, Merit Medical, Utah, USA; Embozene, CeloNova, TX, USA) or polyvinyl alcohol (PVA) in sizes ranging from <150 to 500 μm. Technical details have been described previously [7, 8, 26, 27].

Survival Endpoints

The primary endpoints of the study were overall survival (OS) and hepatic progression-free survival (HPFS) from the date of first embolotherapy. Hepatic progression was defined using response evaluation criteria in solid tumors (RECIST, version 1.1) [28]. The median imaging follow-up time for HPFS evaluation was 11 months (range 1–101 months). OS was defined as death from any cause. The date of death was determined by review of electronic medical records, and confirmed with Social Security Death Index (SSDI). The median follow-up time for OS evaluation was 24 months (range 2–125 months).

Embolotherapy Toxicity

Clinical and laboratory toxicities occurring post-embolotherapy were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 [29]. Grade 3 or higher events were considered severe.

Statistical Analysis

Baseline characteristics were compared between embolotherapy groups using χ 2 and Fisher’s exact tests for discrete variables, ANOVA methods for continuous variables, and Wilcoxon and Kruskal–Wallis methods for non-normal continuous variables. OS and HPFS were modeled using Kaplan–Meier method from the date of first embolotherapy, and comparisons between groups were tested using the log-rank test. To reduce confounding from variations in baseline characteristics, propensity score weighting, using inverse probability of treatment weight, was performed to balance heterogeneities between embolotherapy groups. The variables included in the propensity score model were ECOG performance score, tumor grade, tumor primary site, hepatic tumor burden, the presence of extrahepatic metastases, indication for embolotherapy, and systemic treatment during the follow-up period. Effects of patient-, disease-, and treatment-related covariates on OS and PFS were assessed with unadjusted and propensity score weighting-adjusted univariate and multivariate Cox proportional hazards models. A univariate p < 0.2 was used as cutoff for entry into multivariate Cox model analysis. Interactions between variables in multivariate analyses were tested using the Wald χ 2. The rate of clinical and laboratory toxicities between modalities was compared with χ 2 and Fisher’s exact tests. p < 0.5 was considered significant for all analyses. All statistical analyses were performed using Stata software (version 13; StataCorp, College Station, TX).

Results

Baseline Characteristics

Baseline characteristics of the 155 analyzed patients are shown in Table 1. There was a higher proportion of G3 tumors in the TACE group and G1 tumors in the TAE group (p = 0.001) as well as a higher proportion of extrahepatic metastases in the TAE group (p = 0.05). Patients undergoing TACE were treated more frequently for tumor burden and less frequently for tumor progression than the other modalities (p = 0.003).
Table 1

Baseline patient, tumor, and treatment characteristics by embolotherapy modality

Characteristic

Category

Total (n = 155)

TACE (n = 50)

TARE (n = 64)

TAE (n = 41)

p value

Sex

Male

89 (57 %)

30 (60 %)

39 (61 %)

20 (49 %)

0.43

Female

66 (43 %)

20 (40 %)

25 (39 %)

21 (51 %)

Age

Mean (range)

60.5

58.8

62.1

59.9

0.28

Years

(31–83)

(31–83)

(31–82)

(40–77)

ECOG score

0

90 (58 %)

24 (48 %)

38 (59 %)

28 (68 %)

0.14

≥1

65 (42 %)

26 (52 %)

26 (41 %)

13 (32 %)

Tumor grade

1

75 (48 %)

14 (28 %)

32 (50 %)

29 (71 %)

0.001

2

60 (39 %)

24 (48 %)

25 (39 %)

11 (27 %)

3

20 (13 %)

12 (24 %)

7 (11 %)

1 (2 %)

Tumor burden

≤50 %

116 (75 %)

38 (76 %)

52 (81 %)

26 (63 %)

0.12

>50 %

39 (25 %)

12 (24 %)

12 (19 %)

15 (37 %)

Site of primary tumor

Pancreatic

71 (47 %)

23 (46 %)

26 (40 %)

22 (54 %)

0.81

Gastrointestinal

68 (44 %)

22 (44 %)

30 (47 %)

16 (39 %)

Lung

8 (5 %)

3 (6 %)

3 (5 %)

2 (5 %)

Other/unknown

8 (5 %)

2 (4 %)

5 (8 %)

1 (2 %)

Extrahepatic metastases

Yes

81 (52 %)

24 (48 %)

29 (45 %)

28 (68 %)

0.05

 Lymph node

29 (35)

6 (25 %)

9 (30 %)

14 (50 %)

 Bone

16 (20 %)

4 (17 %)

9 (30 %)

3 (11 %)

 Peritoneum

9 (11 %)

4 (17 %)

2 (6.5 %)

3 (11 %)

 Other

5 (6 %)

1 (4 %)

2 (6.5 %)

2 (7 %)

 Multiple

23 (28 %)

9 (38 %)

8 (27 %)

6 (21 %)

Prior hepatic resection or ablation

Yes

35 (23 %)

14 (28 %)

13 (20 %)

8 (20 %)

0.62

Systemic therapy

Octreotide

121 (78 %)

37 (74 %)

50 (78 %)

34 (83 %)

0.58

Biologic

36 (23 %)

11 (22 %)

15 (23 %)

10 (24 %)

0.71

Cytotoxic

49 (32 %)

12 (24 %)

23 (36 %)

14 (34 %)

0.21

Indication for embolotherapy

Progression

103 (66 %)

25 (50 %)

50 (78 %)

28 (68 %)

0.003

Symptoms

29 (19 %)

11 (22 %)

12 (19 %)

6 (15 %)

Tumor burden

23 (15 %)

14 (28 %)

2 (3 %)

7 (17 %)

p values in bold indicate statistically significant

TACE transarterial chemoembolization, TARE transarterial radioembolization, TAE transarterial embolization, ECOG Eastern Cooperative Oncology Group

Hepatic Progression-Free Survival and Overall Survival

For the overall study cohort, the median HPFS was 13.2 months (95 % CI 10.2–16.2 months) and the median OS was 48.2 months (95 % CI 30.8–54.2 months). Unadjusted and propensity score-weighted univariate and multivariate Cox models for HPFS and OS by patient, disease, and treatment variables are shown in Tables 2 and 3.
Table 2

Univariate and multivariate Cox proportional hazards models for prognostic factors affecting hepatic progression-free survival

Variable

Category

N

PFS

Univariate

Multivariate

Univariate (P-weighted)

Multivariate (P-weighted)

Median, mo

1-year (%)

2-year (%)

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Sex

Male

89

15.1

57

37

1

  

1

  

Female

66

11.5

49

32

1.13 (0.78–1.65)

0.52

  

1.03 (0.65–1.63)

0.89

  

Age

<60 years

68

12.0

51

31

1

  

1

  

≥60 years

87

15.0

56

38

0.80 (0.55–1.17)

0.25

  

0.90 (0.56–1.43)

0.66

  

ECOG score

0

90

13.1

51

34

1

  

1

  

≥1

65

13.2

57

36

1.05 (0.72–1.53)

0.80

  

1.27 (0.78–2.05)

0.34

  

Grade

1

75

18.5

65

46

1

1

1

1

2

60

12.2

50

29

1.69 (1.11–2.56)

0.01

1.65 (1.05–2.61)

0.03

1.72 (1.04–2.84)

0.04

1.74 (1.07–2.87)

0.03

3

20

4.9

20

10

4.46 (2.52–7.87)

<0.001

2.77 (1.36–5.63)

0.005

5.40 (3.05–9.58)

<0.001

3.82 (1.90–7.70)

<0.001

Tumor burden

≤50 %

116

15.0

57

40

1

1

1

1

>50 %

39

9.5

44

21

1.66 (1.10–2.50)

0.02

1.89 (1.21–2.95)

0.005

2.01 (1.29–3.13)

0.002

2.25 (1.37–3.67)

0.001

Site of primary tumor

Pancreatic

71

12.4

51

31

1

1

1

1

Gastrointestinal

68

14.2

56

37

0.83 (0.56–1.24)

0.3 7

1.01 (0.66–1.55)

0.95

0.79 (0.47–1.34)

0.39

0.92 (0.56–1.52)

0.59

Lung

8

5.4

25

25

1.55 (0.70–3.42)

0.28

1.64 (0.72–3.73)

0.24

2.19 (1.08–4.45)

0.03

1.67 (0.86–3.25)

0.65

Other/unknown

8

28.0

86

67

0.42 (0.15–1.17)

0.10

0.49 (0.17–1.41)

0.18

0.42 (0.16–1.11)

0.08

0.40 (0.12–1.32)

0.07

Extrahepatic metastases

No

74

11.4

47

34

1

  

1

  

Yes

81

15.0

59

36

0.83 (0.57–1.20)

0.32

  

1.00

0.99

  

Prior hepatic resection or ablation

No

119

14.2

54

38

1

  

1

  

Yes

35

11.5

49

24

1.21 (0.78–1.87)

0.39

  

1.31 (0.83–2.06)

0.25

  

Indication for embolotherapy

Progression

103

13.1

50

34

1

  

1

  

Symptoms

29

13.2

59

32

0.85 (0.51–1.42)

0.54

  

0.86 (0.47–1.57)

0.61

  

Tumor burden

23

17.0

61

46

0.88 (0.52–1.49)

0.52

  

0.93 (0.56–1.55)

0.79

  

Embolotherapy

TACE

50

8.1

42

20

1

1

1

1

TARE

64

15.7

60

44

0.57 (0.37–0.88)

0.01

0.84 (0.52–1.37)

0.4

0.74 (0.41–1.35)

0.33

0.86 (0.51–1.45)

0.58

TAE

41

15.0

56

38

0.59 (0.36–0.96)

0.03

0.77 (0.44–1.34)

0.35

0.71 (0.35–1.44)

0.34

0.73 (0.38–1.41)

0.35

Systemic therapy during follow-up

No

88

17.0

66

43

1

1

1

1

Yes

67

8.1

39

26

1.77 (1.21–2.57)

0.003

1.42 (0.91–2.22)

0.12

1.78 (1.11–2.86)

0.02

1.24 (0.77–2.00)

0.37

p values in bold indicate statistically significant

ECOG Eastern Cooperative Oncology Group, TACE transarterial chemoembolization, TARE transarterial radioembolization, TAE transarterial embolization

Table 3

Univariate and multivariate Cox proportional hazards models for prognostic factors affecting overall survival

Variable

Category

N

PFS

Univariate

Multivariate

Univariate (P-weighted)

Multivariate (P-weighted)

Median, mo

1-year (%)

2-year (%)

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Hazard ratio (95 % CI)

p

Sex

Male

89

47.6

81

63

1

   

1

  

Female

66

52.3

86

80

0.74 (0.45–1.22)

0.24

  

0.67 (0.37 – 1.22)

0.19

  

Age

<60 years

68

48.1

82

70

1

0.92

  

1

  

≥60 years

87

52.1

84

71

0.97 (0.60–1.59)

   

1.20 (0.65 – 2.22)

0.55

  

ECOG score

0

90

52.3

91

79

1

 

1

1

1

≥1

65

26.7

71

58

2.05 (1.26–3.35)

0.004

2.10 (1.25–3.54)

0.005

2.02 (1.08–3.79)

0.03

2.04 (1.10–3.79)

0.02

Grade

1

75

125.1

92

80

1

 

1

1

1

2

60

33.9

84

71

2.28 (1.29–4.02)

0.004

2.28 (1.19–4.34)

0.01

2.68 (1.36–5.28)

0.004

2.32 (1.03–5.26)

0.04

3

20

9.3

48

30

8.84 (4.39–17.8)

<0.001

13.47 (5.37–33.8)

0.01

12.10 (5.32–27.6)

<0.001

13.3 (4.57–38.6)

<0.001

Tumor burden

≤50 %

116

52.3

86

76

1

 

1

1

1

>50 %

39

25.5

74

53

2.17 (1.30–3.63)

0.003

2.79 (1.62–4.80)

<0.001

2.00 (1.04–3.85)

0.04

2.64 (1.50–4.64)

0.001

Site of primary tumor

Pancreatic

71

48.2

81

69

1

 

1

1

1

Gastrointestinal

68

44.7

85

68

0.99 (0.59–1.67)

0.98

1.19 (0.66–2.14)

0.56

0.72 (0.37–1.41)

0.34

0.96 (0.43–2.18)

0.93

Lung

8

32.9

75

75

1.58 (0.62–4.06)

0.34

0.72 (0.25–2.06)

0.54

1.21 (0.33–4.44)

0.77

0.73 (0.18–3.03)

0.67

Other/unknown

8

125.1

86

86

0.17 (0.02–1.27)

0.08

0.15 (0.02–1.11)

0.06

0.08 (0.01–0.75)

0.03

0.07 (0.01–0.82)

0.03

Extrahepatic metastases

No

74

47.6

83

68

1

   

1

  

Yes

81

51.6

83

72

0.87 (0.53–1.42)

0.59

  

0.98 (0.54–1.79)

0.95

  

Prior hepatic resection or ablation

No

119

48.2

82

71

1

   

1

  

Yes

35

51.6

85

62

1.15 (0.66–2.02)

0.61

  

1.42 (0.77–2.63)

0.26

  

Indication for embolotherapy

Progression

103

51.6

83

73

1

   

1

  

Symptoms

29

28.1

77

61

1.26 (0.66–2.39)

0.49

  

1.32 (0.62–2.79)

0.47

  

Tumor burden

23

52.1

87

67

0.99 (0.49–1.98)

0.98

  

0.92 (0.46–1.84)

0.81

  

Embolotherapy

TACE

50

32.9

82

64

1

1

1

1

TARE

64

48.2

79

71

0.98 (0.56–1.69)

0.93

2.09 (1.12–3.90)

0.02

1.36 (0.65–2.82)

0.41

1.80 (0.87–3.74)

0.11

TAE

41

90

75

0.53 (0.26–1.08)

0.08

1.02 (0.45–2.33)

0.95

0.95 (0.37–2.46)

0.92

0.98 (0.40–2.41)

0.96

Systemic therapy during follow-up

No

88

52.3

86

71

1

   

1

  

Yes

67

48.2

79

69

1.29 (0.79–2.10)

0.30

  

1.49 (0.82–2.70)

0.19

  

p values in bold indicate statistically significant

ECOG Eastern Cooperative Oncology Group, TACE transarterial chemoembolization, TARE transarterial radioembolization, TAE transarterial embolization

Patient Factors

There were no significant differences in HPFS or OS by patient’s age or sex. ECOG score ≥1 prognosticated significantly worse OS versus ECOG score 0 (p = 0.004), without a significant difference in HPFS, concordant between unadjusted and propensity score-weighted analyses.

Tumor Factors

Higher tumor grade (Fig. 1A) and tumor burden >50 % liver volume (Fig. 1B) were associated with significantly shorter HPFS and OS. There were no significant differences in HPFS or OS by pancreatic or gastrointestinal primaries, which constituted 90 % of cases (Fig. 1C), concordant between unadjusted and propensity score-weighted analyses (Tables 2, 3). Better HPFS was seen in patients with lung primary tumor versus pancreas primary in propensity score weighted, but not unadjusted univariate Cox models (Table 2). The presence of extrahepatic metastases did not significantly impact HPFS or OS.
Fig. 1

Kaplan-Meier estimates of hepatic progression free survival (left) and overall survival (right) based on A Tumor grade, B Tumor burden, and C Primary tumor

Treatment Factors

TARE and TAE demonstrated significantly longer HPFS than TACE in unadjusted univariate analysis, but the significance was negated after propensity score weighting (Table 2). There was no significant OS difference between embolotherapy modalities in univariate Cox analysis. There were no significant differences in HPFS or OS by prior hepatic resection or thermal ablation, or by embolotherapy indication (Tables 2, 3). Pre-embolotherapy systemic therapies were not included in Cox model analysis due to significant interactions with tumor grade by Wald test (p < 0.001), reflecting more frequent octreotide therapy in G1 and G2 versus G3 tumors, and more frequent cytotoxic therapy in G3 versus G1 and G2 tumors. Patients treated with systemic therapy during the follow-up period demonstrated 8.9-month shorter median HPFS than those who did not undergo therapy (p = 0.003), without significant OS difference.

Multivariate Cox Models

The unadjusted and propensity score-weighted multivariate Cox models for HPFS are shown in Table 2. Significant prognostic factors for HPFS failure included higher tumor grade and tumor burden >50 %, concordant between unadjusted and adjusted models. Remaining variables, including embolotherapy modality, did not demonstrate statistical significance.

The unadjusted and propensity score-weighted multivariate Cox models for OS are shown in Table 3. Significant prognostic factors for increased risk of death in both unadjusted and propensity score-weighted models included ECOG score ≥1, higher tumor grade, and tumor burden >50 %. TARE demonstrated significantly higher HR than TACE in the unadjusted model (HR 2.1, p = 0.02), but not the propensity score-weighted model (HR 1.8, p = 0.11). Other/unknown primary tumor site demonstrated significantly lower HR than pancreatic primary in the propensity score-weighted model (HR 0.7, p = 0.03), but not in the unadjusted model (HR 0.15, p = 0.06).

Technical Factors

Survival outcomes by technical factors are shown in Table 4. No significant differences in HPFS or OS were observed among chemotherapy agents or 90Y microsphere agents. For TAE, there was significantly shorter HPFS, but not OS for the 300–500 μm group (3/40 cases). There were no significant differences in HPFS or OS between all other particle sizes (<150, 100–300 μm, or multiple). No significant differences in HPFS or OS were seen between total number of treatments for any modality.
Table 4

Effect of treatment-specific variables on hepatic progression-free survival and overall survival: univariate Cox proportional hazards models

Modality

Category

Hepatic PFS

Overall survival

Median (mo)

HR

p value

Median (mo)

HR

p value

TACE

Chemotherapy regimen

 Doxo (n = 6)

13.2

1

21.7

1

 Doxo + mito C (n = 16)

7.8

0.99

0.99

48.1

0.54

0.45

 Doxo + cisplatin (n = 4)

7.4

1.42

0.65

8.4

1.14

0.89

 Doxo + mito C + cisplatin (n = 13)

8.1

1.17

0.81

47.6

0.66

0.61

 

Total treatments

 1 (n = 11)

6.3

1

32.9

1

 2 (n = 26)

10.1

0.71

0.41

48.1

0.97

0.96

 ≥3 (n = 13)

7.1

0.74

0.52

25.5

1.09

0.89

TARE

Microsphere agent

 Resin (n = 43)

14.9

1

48.2

1

 Glass (n = 21)

23.4

0.89

0.72

51.6

0.91

0.82

 

Total treatments

 1 (n = 23)

26.7

1

53.4

1

 2 (n = 35)

14.9

1.61

0.18

48.2

1.65

0.24

 ≥3 (n = 6)

14.2

1.67

0.35

28.1

1.63

0.48

TAE

Embolic size (μm)

      

 <150 (n = 17)

13.1

1

1

 100–300 (n = 10)

15.1

1.56

0.34

40.2

2.44

0.22

 300–500 (n = 3)

3.2

40.07

<0.001

4.6

4.60

0.11

 Multiple (n = 10)

15.0

1.11

0.83

0.49

0.55

 

Total treatments

 1 (n = 7)

37.0

1

1

 2 (n = 18)

10.0

3.08

0.07

54.2

0.88

0.99

 ≥3 (n = 16)

16.2

1.28

0.67

1.09

0.76

p values in bold indicate statistically significant

PFS progression-free survival, Doxo doxorubicin, Mito mitomycin, TACE transarterial chemoembolization, TARE transarterial radioembolization, TAE transarterial embolization

Embolotherapy Toxicity

Adverse events are shown in Table 5. There was a significantly higher rate of pain (of any CTCAE grade) after TAE than TACE or TARE, without significant difference in rate of severe pain. There were three cases of radiation-induced liver disease (REILD) in the TARE group, classified under other clinical severe adverse event, including one patient who died from subsequent spontaneous bacterial peritonitis. All cases occurred in patients with hepatic burden <25 %, with 2–3 total TARE sessions. One of these patients had undergone prior intensity-modulated radiation therapy (IMRT) for a mesenteric tumor. None had undergone PRRT. Additional severe adverse events in the other clinical toxicity group included one case of aspiration-related respiratory failure requiring intubation (TACE, grade 3), gastric ulcer (TARE, grade 4), diarrhea (TAE, grade 3), and hypertension (TAE, grade 3). There was a significantly higher rate of biochemical toxicity (of any CTCAE grade and type) after TARE versus TACE or TAE, without significant difference in rate of severe biochemical toxicity.
Table 5

Total and severe (CTCAE grade 3 or higher) clinical and biochemical toxicities from embolotherapy

Adverse event

TACE (n = 45)

TARE (n = 67)

TAE (n = 43)

p value

Total clinicala

 Total

36 (80 %)

57 (85 %)

37 (86 %)

0.70

 Severe

3 (6.7 %)

7 (10.5 %)

4 (9.3 %)

0.79

 Pain

  Total

20 (44.4 %)

26 (38.8 %)

30 (69.8 %)

0.005

  Severe

0

2 (3.0 %)

1 (2.3 %)

0.52

 Fever

  Total

7 (15.6 %)

13 (19 %)

12 (27.9 %)

0.34

  Severe

0

1 (1.5 %)

1 (2.3 %)

0.62

 Nausea

  Total

15 (33.3 %)

22 (32.8 %)

16 (37.2 %)

0.89

  Severe

1 (2.2 %)

1 (1.5 %)

0

0.64

 Emesis

  Total

9 (20.0 %)

21 (31.3 %)

7 (16.3 %)

0.15

  Severe

1 (2.2 %)

0

0

0.29

 Fatigue

  Total

22 (48.9 %)

36 (53.7 %)

18 (41.9 %)

0.48

  Severe

1 (2.2 %)

1 (1.5 %)

1 (2.3 %)

0.94

 Weight loss

  Total

3 (6.7 %)

13 (19.4 %)

4 (9.3 %)

0.10

  Severe

0

1 (1.5 %)

0

0.52

 Other clinical

  Total

3 (6.7 %)

8 (11.9 %)

7 (16.3 %)

0.37

  Severe

1 (2.2 %)

4 (6.0 %)b

2 (4.7 %)

0.80

Total biochemicala

 Total

26 (57.8 %)

52 (77.6 %)

23 (53.5 %)

0.02

 Severe

5 (11.1 %)

5 (7.5 %)

4 (9.3 %)

0.80

 Bilirubin

  Total

11 (24.4 %)

10 (14.9 %)

5 (11.6 %)

0.24

  Severe

1 (2.2 %)

0

0

0.29

 Alkaline phosphatase

   Total

17 (37.8 %)

39 (58.2 %)

22 (51.2 %)

0.11

   Severe

1 (2.2 %)

0

0

0.29

  Aspartate aminotransferase

  Total

16 (35.6 %)

28 (41.8 %)

17 (39.5 %)

0.80

  Severe

2 (4.4 %)

1 (1.5 %)

4 (9.3 %)

0.16

 Alanine aminotransferase

 Total

11 (24.4 %)

27 (40.3 %)

13 (30.2 %)

0.20

  Severe

3 (6.7 %)

5 (7.5 %)

2 (4.7 %)

0.84

Total clinical and biochemicala

  Total

40 (88.9 %)

65 (97.0 %)

38 (88.4 %)

0.15

  Severe

8 (17.8 %)

11 (16.4 %)

7 (16.3 %)

0.98

p values in bold indicate statistically significant

aTotal number of patients with at least one complication

bIncludes one death resulting from radioembolization induced liver disease

TACE transarterial chemoembolization, TARE transarterial radioembolization, TAE transarterial embolization

Discussion

The optimal embolotherapy for NET liver metastases remains controversial, fueled by a paucity of high-level evidence [18] and inherent complexities of NET biology and multidisciplinary treatment algorithms. Patient factors like older age and male sex have shown conflicting prognostic impact on OS in prior studies [2, 3, 4, 5, 6, 8, 17, 18, 30, 31], and were not found to be significant factors in this study. ECOG performance score ≥1 was prognostic of shorter OS, concordant with a previous TARE series [5]. The prognostic impact of tumor grade was often confounded by interobserver heterogeneity in the WHO 2000/2004 grading system [2, 4, 5, 7, 8, 9, 10, 17, 18, 19, 32, 33]. The WHO 2010 system, while more reproducible, has not been as well evaluated in NET embolotherapy [6]. In this series, higher WHO 2010 grade prognosticated significantly shorter HPFS and OS. Larger volume hepatic tumor burden has been correlated with poorer survival, with proposed prognostic thresholds at >25 % [5], >50 % [8], and >75 % [4] liver volume, which is supported by the shorter HPFS and OS seen in tumor burden >50 % in this study. The presence of extrahepatic metastases did not significantly impact HPFS or OS, similar to several prior studies [6, 7], likely reflecting improved extrahepatic disease control from advances in systemic therapies. Previous studies have reported conflicting survival outcomes for gut versus pancreatic NETs [4, 6, 7, 8, 17, 19, 34]; the results of this study found no significant OS or HPFS difference between these tumors. No definitive conclusions could be made regarding lung or unknown/other primary sites, given the small sample sizes.

The survival outcomes by embolotherapy modality in this study were within range of previous single or multimodality series [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 22, 23, 24, 25, 34]. For OS, the multivariate Cox models suggested a trend toward worsened prognosis with TARE versus TACE with significantly higher HR in unadjusted model, but not in propensity score-weighted model. Prior comparative studies showed shorter and equivalent after TARE versus DEB-TACE [32], but TARE-only series have otherwise reported similar outcomes to TAE and TACE [1, 5, 33, 35]. The equivalent OS and HPFS between TACE and TAE in this study are supported by results of prior series [3, 6, 8, 17]. In this study, no significant differences in PFS or OS were found between TACE chemotherapy regimens, TARE microsphere agents, the most frequently used TAE embolic sizes (<150–300 μm), or total number of treatments.

In this study, the rate of severe adverse events was modest, ranging from 16 to 18 %, similar to rates from prior series [1, 3, 5, 6, 7, 8]. The significantly higher proportion of patients with pain following TAE (70 %) than TACE (44 %) or TARE (39 %) may be related to a higher degree of ischemia achieved by the TAE endpoint of complete stasis. Although the rate of all biochemical toxicity was higher in TARE (78 %) than TACE (58 %) or TAE (54 %), the rate of severe toxicities was not significantly different (8–11 %). TARE has been suggested to have less severe short-term toxicity than TACE or TAE [27], but these results suggest a similar safety profile. Notably, REILD occurred in three TARE patients (4.5 %), including one who ultimately died as a result of SBP, highlighting a pitfall to consider when selecting among modalities. No patients in the TACE or TAE groups developed liver failure.

Prior partial hepatectomy or thermal ablation did not demonstrate significant impact on OS or HPFS in this study, supporting the results of previous series [5, 8, 18], noting that the study was not tailored to evaluate details of these prior therapies including timing relative to embolotherapy. Treatment with systemic agents following embolotherapy was associated with shorter HPFS, reflecting the increased use of these therapies in patients with higher grade or progressive disease rather than a causal effect of systemic therapy on poorer outcomes.

This study was retrospectively designed and not powered for specific subgroup analyses, resulting in several limitations. Data were obtained from several centers to provide generalizability and adequate cohort size with each modality within the timeframe of modern practice; thus, unknown and uncontrolled biases in patient selection and technique were unavoidable. Propensity score weighting analysis was performed to reduce the effect of these differences and revealed largely concordant results with unadjusted analyses. Confounding from these variations will be substantially reduced in the upcoming RETNET trial (ClinicalTrials.gov Identifier NCT02724540) for which technical protocols will be standardized and randomized prospectively.

In conclusion, higher tumor grade and hepatic tumor burden >50 % were significant prognostic factors for shortened HPFS and OS. TARE trended toward worsened OS, while TAE demonstrated equivalent OS compared to TACE. There were no significant differences in HPFS between embolotherapy modalities. Safety profiles were similar between embolotherapy modalities, with low rates of severe adverse events. These results identify the most important stratification variables for design of prospective randomized controlled trials, such as the RETNET study, which remain necessary to elucidate if there is an optimal embolotherapy algorithm for NET liver metastases.

Notes

Compliance with Ethical Standards

Conflict of Interest

Steven C. Rose: consultant—SIRTeX; scientific advisory board—Surefire Medical. Sarah B. White: consultant—Guerbet, IO-rad, Grants—RSNA, SIR foundation, research support—Siemens. Nicholas Fidelman: Grants: BTG, GE Healthcare, Nordion. Daniel Y. Sze: consultant—Amgen, BTG, SirTeX Medical, W.L. Gore & Associates, Covidien, Guerbet, Cook, Codman; scientific advisory board—SureFire Medical, KoliMedical, Northwind Medical, TreusMedical, RadiAction Medical, EmboIX, Lunar Design, Jennerex Biotherapeutics. Riad Salem: consultant: BTG. David C. Metz MD: grants—Ipsen, Lexicon, AAA; consultant—Novartis, Takeda. Michael C. Soulen: grants—BTG, Guerbet; consultant—Guerbet, Merit. James X. Chen, Ghassan El-Haddad, Hooman Yarmohammadi, Winifred Hwang, Nishita Kothary, Kristen Stashek, E. Paul Wileyto, No disclosures.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. For this type of study formal consent is not required.

References

  1. 1.
    Barbier CE, Garske-Roman U, Sandstrom M, Nyman R, Granberg D. Selective internal radiation therapy in patients with progressive neuroendocrine liver metastases. Eur J Nucl Med Mol Imaging. 2016;43(8):1425–31.CrossRefPubMedGoogle Scholar
  2. 2.
    Dong XD, Carr BI. Hepatic artery chemoembolization for the treatment of liver metastases from neuroendocrine tumors: a long-term follow-up in 123 patients. Med Oncol. 2011;28(Suppl 1):S286–90.CrossRefPubMedGoogle Scholar
  3. 3.
    Fiore F, Del Prete M, Franco R, et al. Transarterial embolization (TAE) is equally effective and slightly safer than transarterial chemoembolization (TACE) to manage liver metastases in neuroendocrine tumors. Endocrine. 2014;47(1):177–82.CrossRefPubMedGoogle Scholar
  4. 4.
    Gupta S, Johnson MM, Murthy R, et al. Hepatic arterial embolization and chemoembolization for the treatment of patients with metastatic neuroendocrine tumors: variables affecting response rates and survival. Cancer. 2005;104(8):1590–602.CrossRefPubMedGoogle Scholar
  5. 5.
    Memon K, Lewandowski RJ, Riaz A, Salem R. Chemoembolization and radioembolization for metastatic disease to the liver: available data and future studies. Curr Treat Options Oncol. 2012;13(3):403–15.CrossRefPubMedGoogle Scholar
  6. 6.
    Pericleous M, Caplin ME, Tsochatzis E, Yu D, Morgan-Rowe L, Toumpanakis C. Hepatic artery embolization in advanced neuroendocrine tumors: efficacy and long-term outcomes. Asia Pac J Clin Oncol. 2016;12(1):61–9.CrossRefPubMedGoogle Scholar
  7. 7.
    Ruutiainen AT, Soulen MC, Tuite CM, et al. Chemoembolization and bland embolization of neuroendocrine tumor metastases to the liver. J Vasc Interv Radiol. 2007;18(7):847–55.CrossRefPubMedGoogle Scholar
  8. 8.
    Sofocleous CT, Petre EN, Gonen M, et al. Factors affecting periprocedural morbidity and mortality and long-term patient survival after arterial embolization of hepatic neuroendocrine metastases. J Vasc Interv Radiol. 2014;25(1):22–30 quiz 31.CrossRefPubMedGoogle Scholar
  9. 9.
    Sommer WH, Ceelen F, Garcia-Albeniz X, et al. Defining predictors for long progression-free survival after radioembolisation of hepatic metastases of neuroendocrine origin. Eur Radiol. 2013;23(11):3094–103.CrossRefPubMedGoogle Scholar
  10. 10.
    Strosberg JR, Choi J, Cantor AB, Kvols LK. Selective hepatic artery embolization for treatment of patients with metastatic carcinoid and pancreatic endocrine tumors. Cancer Control. 2006;13(1):72–8.PubMedGoogle Scholar
  11. 11.
    Bhagat N, Reyes DK, Lin M, et al. Phase II study of chemoembolization with drug-eluting beads in patients with hepatic neuroendocrine metastases: high incidence of biliary injury. Cardiovasc Intervent Radiol. 2013;36(2):449–59.CrossRefPubMedGoogle Scholar
  12. 12.
    Boudreaux JP, Klimstra DS, Hassan MM, et al. The NANETS consensus guideline for the diagnosis and management of neuroendocrine tumors: well-differentiated neuroendocrine tumors of the Jejunum, Ileum, Appendix, and Cecum. Pancreas. 2010;39(6):753–66.CrossRefPubMedGoogle Scholar
  13. 13.
    Kennedy A, Bester L, Salem R, et al. Role of hepatic intra-arterial therapies in metastatic neuroendocrine tumours (NET): guidelines from the NET-Liver-Metastases Consensus Conference. HPB (Oxford). 2015;17(1):29–37.CrossRefGoogle Scholar
  14. 14.
    Kulke MH, Anthony LB, Bushnell DL, et al. NANETS treatment guidelines: well-differentiated neuroendocrine tumors of the stomach and pancreas. Pancreas. 2010;39(6):735–52.CrossRefPubMedPubMedCentralGoogle Scholar
  15. 15.
    Kulke MH, Shah MH, Benson AB 3rd, et al. Neuroendocrine tumors, version 1.2015. J Natl Compr Canc Netw. 2015;13(1):78–108.PubMedGoogle Scholar
  16. 16.
    Pavel M, Baudin E, Couvelard A, et al. ENETS Consensus Guidelines for the management of patients with liver and other distant metastases from neuroendocrine neoplasms of foregut, midgut, hindgut, and unknown primary. Neuroendocrinology. 2012;95(2):157–76.CrossRefPubMedGoogle Scholar
  17. 17.
    Pitt SC, Knuth J, Keily JM, et al. Hepatic neuroendocrine metastases: chemo- or bland embolization? J Gastrointest Surg. 2008;12(11):1951–60.CrossRefPubMedPubMedCentralGoogle Scholar
  18. 18.
    Maire F, Lombard-Bohas C, O’Toole D, et al. Hepatic arterial embolization versus chemoembolization in the treatment of liver metastases from well-differentiated midgut endocrine tumors: a prospective randomized study. Neuroendocrinology. 2012;96(4):294–300.CrossRefPubMedGoogle Scholar
  19. 19.
    Ho AS, Picus J, Darcy MD, et al. Long-term outcome after chemoembolization and embolization of hepatic metastatic lesions from neuroendocrine tumors. AJR Am J Roentgenol. 2007;188(5):1201–7.CrossRefPubMedGoogle Scholar
  20. 20.
    Rinke A, Muller HH, Schade-Brittinger C, et al. Placebo-controlled, double-blind, prospective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors: a report from the PROMID Study Group. J Clin Oncol. 2009;27(28):4656–63.CrossRefPubMedGoogle Scholar
  21. 21.
    Caplin ME, Pavel M, Cwikla JB, et al. Lanreotide in metastatic enteropancreatic neuroendocrine tumors. N Engl J Med. 2014;371(3):224–33.CrossRefPubMedGoogle Scholar
  22. 22.
    Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med. 2011;364(6):514–23.CrossRefPubMedPubMedCentralGoogle Scholar
  23. 23.
    Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. N Engl J Med. 2011;364(6):501–13.CrossRefPubMedGoogle Scholar
  24. 24.
    Van Essen M, Krenning EP, De Jong M, Valkema R, Kwekkeboom DJ. Peptide receptor radionuclide therapy with radiolabelled somatostatin analogues in patients with somatostatin receptor positive tumours. Acta Oncol. 2007;46(6):723–34.CrossRefPubMedGoogle Scholar
  25. 25.
    Gaba RC. Chemoembolization practice patterns and technical methods among interventional radiologists: results of an online survey. AJR Am J Roentgenol. 2012;198(3):692–9.CrossRefPubMedGoogle Scholar
  26. 26.
    Memon K, Lewandowski RJ, Mulcahy MF, et al. Radioembolization for neuroendocrine liver metastases: safety, imaging, and long-term outcomes. Int J Radiat Oncol Biol Phys. 2012;83(3):887–94.CrossRefPubMedGoogle Scholar
  27. 27.
    de Baere T, Arai Y, Lencioni R, et al. Treatment of liver tumors with lipiodol TACE: technical recommendations from experts opinion. Cardiovasc Intervent Radiol. 2016;39(3):334–43.CrossRefPubMedGoogle Scholar
  28. 28.
    Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228–47.CrossRefPubMedGoogle Scholar
  29. 29.
    National Cancer Institute NIoH. Common terminology criteria for adverse events v4.03. 2010.Google Scholar
  30. 30.
    Lepage C, Rachet B, Coleman MP. Survival from malignant digestive endocrine tumors in England and Wales: a population-based study. Gastroenterology. 2007;132(3):899–904.CrossRefPubMedGoogle Scholar
  31. 31.
    Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”: epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol. 2008;26(18):3063–72.CrossRefPubMedGoogle Scholar
  32. 32.
    Whitney R, Valek V, Fages JF, et al. Transarterial chemoembolization and selective internal radiation for the treatment of patients with metastatic neuroendocrine tumors: a comparison of efficacy and cost. Oncologist. 2011;16(5):594–601.CrossRefPubMedPubMedCentralGoogle Scholar
  33. 33.
    Rhee TK, Lewandowski RJ, Liu DM, et al. 90Y Radioembolization for metastatic neuroendocrine liver tumors: preliminary results from a multi-institutional experience. Ann Surg. 2008;247(6):1029–35.CrossRefPubMedGoogle Scholar
  34. 34.
    Moertel CG, Johnson CM, McKusick MA, et al. The management of patients with advanced carcinoid tumors and islet cell carcinomas. Ann Intern Med. 1994;120(4):302–9.CrossRefPubMedGoogle Scholar
  35. 35.
    Kennedy AS, Dezarn WA, McNeillie P, et al. Radioembolization for unresectable neuroendocrine hepatic metastases using resin 90Y-microspheres: early results in 148 patients. Am J Clin Oncol. 2008;31(3):271–9.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media New York and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) 2016

Authors and Affiliations

  • James X. Chen
    • 1
  • Steven Rose
    • 2
  • Sarah B. White
    • 3
  • Ghassan El-Haddad
    • 4
  • Nicholas Fidelman
    • 5
  • Hooman Yarmohammadi
    • 6
  • Winifred Hwang
    • 7
  • Daniel Y. Sze
    • 7
  • Nishita Kothary
    • 7
  • Kristen Stashek
    • 8
  • E. Paul Wileyto
    • 9
  • Riad Salem
    • 10
  • David C. Metz
    • 11
  • Michael C. Soulen
    • 1
  1. 1.Division of Interventional Radiology, Department of RadiologyHospital of the University of PennsylvaniaPhiladelphiaUSA
  2. 2.Division of Interventional Radiology, Department of RadiologyUniversity of San Diego Medical CenterSan DiegoUSA
  3. 3.Division of Interventional Radiology, Department of RadiologyMedical College of WisconsinMilwaukeeUSA
  4. 4.Division of Interventional Radiology, Department of RadiologyMoffitt Cancer CenterTampaUSA
  5. 5.Division of Interventional Radiology, Department of RadiologyUniversity of San Francisco Medical CenterSan FranciscoUSA
  6. 6.Division of Interventional Radiology, Department of RadiologyMemorial Sloan Kettering Cancer CenterNew YorkUSA
  7. 7.Division of Interventional Radiology, Department of RadiologyStanford University Medical CenterStanfordUSA
  8. 8.Department of PathologyHospital of the University of PennsylvaniaPhiladelphiaUSA
  9. 9.Department of Biostatistics and EpidemiologyUniversity of PennsylvaniaPhiladelphiaUSA
  10. 10.Division of Interventional Radiology, Department of RadiologyNorthwestern Memorial HospitalChicagoUSA
  11. 11.Division of Gastroenterology, Department of MedicineHospital of the University of PennsylvaniaPhiladelphiaUSA

Personalised recommendations