Open Access

Utility of clinical and MR imaging parameters for prediction and monitoring of response to capecitabine and temozolomide (CAPTEM) therapy in patients with liver metastases of neuroendocrine tumors

Maria Ingenerf
Maria Ingenerf
, 
Christoph Auernhammer
Christoph Auernhammer
, 
Roberto Lorbeer
Roberto Lorbeer
, 
Michael Winkelmann
Michael Winkelmann
, 
Shiwa Mansournia
Shiwa Mansournia
, 
Nabeel Mansour
Nabeel Mansour
, 
Nina Hesse
Nina Hesse
, 
Kathrin Heinrich
Kathrin Heinrich
, 
Jens Ricke
Jens Ricke
, 
Frank Berger
Frank Berger
 and 
Christine Schmid-Tannwald
Christine Schmid-Tannwald
   | Apr 14, 2024
Radiology and Oncology's Cover Image
About this article
Previous Article
Next Article
Cite
Article Category: Research Article
Published Online: Apr 14, 2024
Page range: 196 - 205
Received: Dec 08, 2023
Accepted: Feb 20, 2024
DOI: https://doi.org/10.2478/raon-2024-0024
Keywords
© 2024 Maria Ingenerf et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.

FIGURE 1.

Flow-chart of including process of patients.CAP/TEM = capecitabine and temozolomide; DWI = diffusion-weighted imaging
Flow-chart of including process of patients.CAP/TEM = capecitabine and temozolomide; DWI = diffusion-weighted imaging

FIGURE 2.

A 72-year-old man with liver metastasis of pancreatic NET classified as responder with a PFS of 38 months. The baseline axial contrast-enhanced T1- weighted image (hepatobiliary phase) (A) shows hypointense lesions (arrows) in segment 8 and exophytic in segment 1. The metastases show (B) restricted diffusion (arrows) with high signal on axial DW-MR image b = 800 s/mm2 and dark signal (arrows) on ADC map (C). After initiation of CAPTEM, the metastases (arrows) exhibited a decrease in size (D) On the axial DW-MR image b = 800 s/mm2, the metastasis (arrow) (E) demonstrated less hyperintense signal to liver and predominantly hyperintense signal (circle) on the ADC map (F) indicating less restricted diffusion compared to the preinterventional image.ADC = apparent diffusion coefficient; CAPTEM = capecitabine and temozolomide; DW-MR = diffusion-weighted magnetic resonance; NET = neuroendocrine tumor; PFS = progression-free survival; PR =partial remission; TARE = transarterial radioembolization
A 72-year-old man with liver metastasis of pancreatic NET classified as responder with a PFS of 38 months. The baseline axial contrast-enhanced T1- weighted image (hepatobiliary phase) (A) shows hypointense lesions (arrows) in segment 8 and exophytic in segment 1. The metastases show (B) restricted diffusion (arrows) with high signal on axial DW-MR image b = 800 s/mm2 and dark signal (arrows) on ADC map (C). After initiation of CAPTEM, the metastases (arrows) exhibited a decrease in size (D) On the axial DW-MR image b = 800 s/mm2, the metastasis (arrow) (E) demonstrated less hyperintense signal to liver and predominantly hyperintense signal (circle) on the ADC map (F) indicating less restricted diffusion compared to the preinterventional image.ADC = apparent diffusion coefficient; CAPTEM = capecitabine and temozolomide; DW-MR = diffusion-weighted magnetic resonance; NET = neuroendocrine tumor; PFS = progression-free survival; PR =partial remission; TARE = transarterial radioembolization

FIGURE 3.

A 56-year-old man with liver metastasis of pancreatic NET classified as nonresponder with a PFS of 3 months. The baseline axial contrast-enhanced T1- weighted image (hepatobiliary phase) (A) shows a hypointense lesion (arrow) in segment 4A. The metastasis shows a strong artrerial enhancement (B) and restricted diffusion (arrow) with high signal on axial DW-MR image b = 800 s/mm2 (C) and dark signal (arrow) on ADC map (D). After 3 months under CAPTEM, the metastasis (arrow) (E) exhibited an increase in size; however, it shows less arterial enhancement (F). On the axial DW-MR image b = 800 s/mm2, the metastasis (arrow) demonstrated hyperintense signal to liver and increasing hypointense signal on the ADC map indicating increasing restricted diffusion compared to the baseline image ADC, apparent diffusion coefficient.ADC = apparent diffusion coefficient; CAPTEM = capecitabine and temozolomide; DW-MR = diffusion-weighted magnetic resonance; NET = neuroendocrine tumor; PFS = progression-free survival
A 56-year-old man with liver metastasis of pancreatic NET classified as nonresponder with a PFS of 3 months. The baseline axial contrast-enhanced T1- weighted image (hepatobiliary phase) (A) shows a hypointense lesion (arrow) in segment 4A. The metastasis shows a strong artrerial enhancement (B) and restricted diffusion (arrow) with high signal on axial DW-MR image b = 800 s/mm2 (C) and dark signal (arrow) on ADC map (D). After 3 months under CAPTEM, the metastasis (arrow) (E) exhibited an increase in size; however, it shows less arterial enhancement (F). On the axial DW-MR image b = 800 s/mm2, the metastasis (arrow) demonstrated hyperintense signal to liver and increasing hypointense signal on the ADC map indicating increasing restricted diffusion compared to the baseline image ADC, apparent diffusion coefficient.ADC = apparent diffusion coefficient; CAPTEM = capecitabine and temozolomide; DW-MR = diffusion-weighted magnetic resonance; NET = neuroendocrine tumor; PFS = progression-free survival

FIGURE 4.

(A) Survival analysis for ∆ size of NELM with a cut-off of ≤ 0% for responder. This cut-off revealed a slightly longer median PFS time of 12.2 vs. 3.6 month (p = 0.062). (B) The median cut-off for ∆ ADCmean/ADCmean liver showed a significantly longer median PFS time of 15.3 compared to 4.1 month (p = 0.024). Both the combination of ∆ size of NELM > 0% and ∆ ADCmin < − 2.9% and the combination of ∆ size of NELM > 0% and ∆ CgA > 12.6% could differentiate patients with a longer median PFS time. Median PFS of the group with ∆ size of NELM > 0% and ∆ ADCmin < −2.9% was 3.6 m compared to 12 months (p = 0.021) in the group not fulfilling these criteria or a maximum of one criterion. Median PFS of the group with ∆ size of NELM > 0% and ∆ CgA > 12.6% was 3.6 m compared to 11.3 months (p = 0.072) in the group not fulfilling these criteria or a maximum of one criterium.ADC = apparent diffusion coefficient; CgA = chromogranin A; NELM = neuroendocrine liver metastasis; PFS = progression-free survival
(A) Survival analysis for ∆ size of NELM with a cut-off of ≤ 0% for responder. This cut-off revealed a slightly longer median PFS time of 12.2 vs. 3.6 month (p = 0.062). (B) The median cut-off for ∆ ADCmean/ADCmean liver showed a significantly longer median PFS time of 15.3 compared to 4.1 month (p = 0.024). Both the combination of ∆ size of NELM > 0% and ∆ ADCmin < − 2.9% and the combination of ∆ size of NELM > 0% and ∆ CgA > 12.6% could differentiate patients with a longer median PFS time. Median PFS of the group with ∆ size of NELM > 0% and ∆ ADCmin < −2.9% was 3.6 m compared to 12 months (p = 0.021) in the group not fulfilling these criteria or a maximum of one criterion. Median PFS of the group with ∆ size of NELM > 0% and ∆ CgA > 12.6% was 3.6 m compared to 11.3 months (p = 0.072) in the group not fulfilling these criteria or a maximum of one criterium.ADC = apparent diffusion coefficient; CgA = chromogranin A; NELM = neuroendocrine liver metastasis; PFS = progression-free survival

ROC analysis of the previously selected imaging and clinical parameters

AUC Cut-off (Median) Sensitivity (%) Specificity (%) Youden-Index
Ki-67% 0.72 > 15 69 59 0.28
Hepatic tumor burden 0.73 < 10 84 72 0.56
∆ CgA 0.73 > 12.6 67 64 0.31
∆Size NELM 0.71 > 0 69 65 0.34
∆ Size PNET - > −2.7 100 50 0.50
∆ ADCmin 0.71 < −2.9 65 63 0.28
∆ ADCmean/ADCmean liver 0.76 < 6.9 76 75 0.51
∆ Size NELM+ ∆ CgA 0.79 > 0/> 12.6 78 60 0.38
∆ Size NELM+ ∆ ADCmin 0.70 > 0/< −2.9 88 60 0.48
∆ Size NELM+ ∆ ADCmean/ ADCmean liver 0.77 > 0/< 6.9 78 58 0.36

Patients characteristics

Baseline N = 44 Follow-Up N = 33
Age (years) 60.4 (50.5; 70.2)
Males 33 (75.0%)
Time initial diagnosis – therapy start 685 (199; 1230)
Clinical parameter
  Hepatic tumor burden (%) 10 (5 ;40)
  CgA (ng/ml) 610 (119; 2093) 647 (261; 2357)
  Bilirubin (mg/dl) 0.6 (0.4; 0.8) 0.7 (0.6; 0.9)
  Grading
    1 1 (2.4%)
    2 32 (76.2%)
    3 6 (14.3%)
    NEC = 4 3 (7.1%)
Ki-67 primary tumor (%) 15 (8;20)
Localization primary tumor
    Pancreas 37 (84.1%)
    Lung 7 (15.9%)
MRI parameter
NELM
    Size (mm) 28 (19;36) 24.5 (18;38.5)
    T1 non-contrast/T1 liver 0.62 (0.53;0.68) 0.68 (0.56;0.75)
    T2/T2 liver 1.63 (1.16;2.07) 1.66 (1.21;2.17)
    ADCmin 448.5 (242.5;628.5) 549 (341;848)
    ADCmean 903 (708.5;1069.5) 969 (764;1250)
    ADCmin/ADCmin liver 0.80 (0.60;0.93) 0.85 (0.51;1.32)
    ADCmean/ADCmean liver 0.82 (0.74;0.96) 0.99 (0.65;1.32)
    % arterial vascularization 42.5 (15;80) 22.5 (5;74.5)**
PNET
    Size (mm) 43 (32;70) 43 (29.5;52)
    T1 non-contrast /T1 pancreas 0.63 (0.59;0.76) 0.68 (0.61;0.84)
    T2/T2 pancreas 1.38 (0.85;1.67) 1.08 (0.83;1.34)
    ADCmin 604.5 (237;648) 628 (499.5;758.5)
    ADCmean 985 (810;1150) 1042.5 (939;1167)
    ADCmin/ADCmin pancreas 0.69 (0.41;1.11) 0.73 (0.58;0.85)
    ADCmean/ADCmean pancreas 1.01 (0.78;1.19) 0.89 (0.72;0.97)
    % arterial vascularization 15 (10;80) 7 (5;45)

Differences in change of clinical and imaging tumor parameters between responder and non-responder

Change between baseline and follow-up (%) Non-responder (< 6 months PFS) N = 17 Responder (≥ 6 months PFS) N = 16 p-value
Clinical parameter
  CgA 61.2 (−8.3;251.9) −1.5 (−69.3;19) 0.036
  Bilirubin 0 (−20;40) 8.3 (−15.3;133.3) 0.312
MRI parameter
NELM
  Size (mm) 20 (−4.7;50) −8.0 (−20.1;2.2) 0.038
  T1 non-contrast/T1 liver 5.4 (−3.8;32.6) −6.8 (−13.6;11.2) 0.078
  T2/T2 liver 1.6 (−9.2;24.1) −5.7 (−26.2;32.8) 0.589
  ADCmin −22.8 (−41.1;40.2) 49.7 (−6.7;146.4) 0.037
  ADCmean −3.5 (−18.4;14.1) 11.7 (−3.4;75.4) 0.056
  ADCmin/ADCmin liver −32.3 (−46.2;70.8) 47.5 (12.7;251.7) 0.113
  ADCmean/ADCmean liver −16.3 (−30.6;6.9) 30.0 (6.9;90.4) 0.011
  % arterial vascularization −16.7 (−75; −5.9) −16.7 (−50.0;11.8) 0.298
PNET
  Size (mm) 2.3 (−5.4;20) −55 (−60; −17.8) 0.013
  T1 non-contrast /T1 pancreas 7.4 (−3.8;36.7) −5 (−19.7;1.9) 0.116
  T2/T2 pancreas −16.6 (−22;1.2) −36.1 (−40.3; −10.1) 0.229
  ADCmin 14.4 (−13.7;260.8) 18.7 (−33.2;48.9) 0.782
  ADCmean 8.3 (−4.5;29.3) 4.0 (−26.3;4.6) 0.405
  ADCmin/ADCmin pancreas −3.6 (−29;76.6) 53 (−18.4;80.9) 0.518
  ADCmean/ADCmean pancreas −23.2 (−35.5;4.5) −5.7 (−14.3;0.2) 0.518
  % arterial vascularization −50 (−80;0) −50 (−80;0) 0.851

Differences in baseline clinical and imaging tumor parameters between responder and non-responder

Non-responder (< 6 months PFS) N = 23 Responder (≥ 6 months PFS) N = 21 p-value
Age 57.8 (44.1;71.1) 61.7 (55.8;68.8) 0.953
Males 16 (69.6%) 17 (81.0%) 0.494
Time ID – Therapy start (d) 851 (426;1552) 396 (153;1004) 0.115
Clinical parameter
  Hepatic tumor burden (%) 5 (5;20) 20 (10;40) 0.007
  CgA 592 (116;2031) 616 (156.5;2745) 0.706
  Bilirubin 0.6 (0.4;0.8) 0.6 (0.3;0.9) 0.859
  Grading 0.234
    1 0 (0%) 1 (5%)
    2 15 (68.2%) 17 (85%)
    3 4 (18.2%) 2 (10%)
    NEC = 4 3 (13.6%) 0 (0%)
  Ki-67 primary tumor (%) 16.5 (10;30) 10.0 (5;15) 0.013
  Localization primary tumor 0.232
    Pancreas 21 (91.3%) 16 (76.2%)
    Lung 2 (8.7%) 5 (23.8%)
MRI parameter
NELM
    Size (mm) 25.5 (17;33.5) 29.8 (21.8;37.5) 0.348
    T1 non-contrast/T1 liver 0.60 (0.53;0.68) 0.64 (0.54;0.74) 0.263
    T2/T2 liver 1.62 (1.2;2.07) 1.69 (1.12;2.06) 0.903
    ADCmin 506 (228;639) 424 (243;606) 0.827
    ADCmean 852.5 (674;1059) 911 (790.5;1082.5) 0.495
    ADCmin/ADCmin liver 0.80 (0.63;0.93) 0.74 (0.51;1.03) 0.846
    ADCmean/ADCmean liver 0.82 (0.68;0.93) 0.86 (0.78;1.02) 0.342
    % arterial vascularization 45 (15;85) 36.3 (15;72.5) 0.494
PNET
    Size (mm) 38 (30;44) 75.5 (65;85.5) 0.024
    T1 non-contrast /T1 pancreas 0.60 (0.58;0.71) 0.71 (0.63;0.8) 0.258
    T2/T2 pancreas 1.38 (0.84;1.67) 1.38 (1.11;1.5) 0.777
    ADCmin 604.5 (237;648) 527 (316.5;698) 1.000
    ADCmean 893 (789;1055) 1084 (996.5;1256) 0.157
    ADCmin/ADCmin pancreas 0.79 (0.41;1.18) 0.63 (0.44;0.8) 0.480
    ADCmean/ADCmean pancreas 1.09 (0.66;1.31) 0.97 (0.96;1.1) 0.888
    % arterial vascularization 10 (5;70) 65 (30;85) 0.130
eISSN:
1581-3207
Language:
English
Publication timeframe:
4 times per year
Journal Subjects:
Medicine, Clinical Medicine, Internal Medicine, Haematology, Oncology, Radiology
Journal RSS Feed