HPB

DOI:10.1111/hpb.12437

ORIGINAL ARTICLE

Prospective diagnostic test accuracy comparison of computed tomography during arterial portography and Primovist magnetic resonance imaging in the pre-operative assessment of colorectal cancer liver metastases Jai S. Bagia1, Alan Chai2, Roger Chou2, Christopher Chu2, John Rouse2, Elizabeth Sinclair3, Leon Vonthethoff4 & Armando Teixeira-Pinto5 1 Department of Surgery, St George Hospital, Sydney, NSW, Australia, 2Department of Radiology, St George Hospital, Sydney, NSW, Australia, 3Department of Histopathology, Douglas Hanly Moir Pathology, Sydney, NSW, Australia, 4Department of Anatomical Pathology, South Eastern Area Laboratory Services (SEALS), Sydney, NSW, Australia, and 5School of Public Health, University of Sydney, Sydney, NSW, Australia

Abstract Objectives: To assess and compare the accuracy and inter-observer agreement for the detection of liver lesions using Primovist magnetic resonance imaging (pMRI) and computed tomography during arterial portography (CTAP). Methods: Patients evaluated at St George Hospital Liver Unit for colorectal liver metastases (CRCLM) underwent CTAP as part of standard staging. pMRI was added to the pre-operative assessment. Two radiologists reported CTAP and two reported pMRI. The sensitivity and specificity of CTAP and pMRI were calculated using histopathology as the gold standard. Results: Complete data were available for 62 patients corresponding to 219 lesions confirmed on histopathology. Agreement on the detection of lesions between the two radiologists that reported pMRI was higher than for CTAP (Kappa = 0.80 versus 0.74). Specificity of lesion detection for pMRI was 0.88 and 0.83 for CTAP (P = 0.112). Sensitivity for pMRI was 0.83 and 0.81 for CTAP. For patients who had chemotherapy before evaluation, pMRI had a significantly higher specificity than CTAP (0.79 versus 0.63, P = 0.011). Conclusions: pMRI is less invasive, has a good inter-observer agreement, has comparable sensitivity and specificity to CTAP in the pre-chemotherapy population and demonstrates better specificity in patients assessed post-chemotherapy. pMRI is a valid alternative to CTAP in the assessment of CRCLM. Received 10 February 2015; accepted 1 May 2015

Correspondence Jai S. Bagia, 410 Crown Street, Wollongong, 2500 NSW, Australia. Tel: +61 2 4227 3733. Fax: +61 2 4228 3563. E-mail: [email protected]

Introduction Background Colorectal cancer is responsible for 10% of solid tumour cancer deaths in the western world.1 Liver metastases occur in 60% of patients after resection of colorectal cancer and a liver resection remains the only curative option for these patients.

The clinical trial is registered with the Australia and New Zealand Clinical Trials Registry (ANZCTR.org.au), approved by the International Committee of Medical Journal Editors (ICMJE). The registration number is ACTRN1211000549921.

HPB 2015, 17, 927–935

Several large series report 5-year survival after a curative liver resection of 25% to 44%, with an operative mortality of 0% to 6.6%.2 Currently there is no consensus on the optimal approach to pre-operative staging of patients with colorectal liver metastases (CRCLM).2 Community equipoise exists regarding the best approach to assess CRCLM. The UK guidelines for resection of CRCLM allow that centres perform liver-specific imaging by local protocol,2 in the absence of clear evidence to support any one approach. Each imaging modality has advantages and disadvantages related to cost, the speed of acquisition, the use of

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

928

ionizing radiation, the risk of contrast reaction, and local availability and expertise. In the early 1990s, data from the USA and France showed significantly better detection rates for computed tomography during arterial portography (CTAP) at 93% compared to 75% for combined ultrasound and dynamic CT.3,4 CTAP involves selective angiographic catheterization of the superior mesenteric or splenic artery (Fig. 1). A bolus of contrast via the artery enhances the liver parenchyma when returning via the portal venous system. CRCLM appear hypodense as they are predominantly supplied by arterial blood while 70% of the liver’s blood supply is from the portal vein, and only 30% originates from the hepatic artery.5 Although sensitive, this technique is invasive and has a high false positive rate of 15% on a per patient basis.5 CTAP

Subsequent studies have acknowledged the improvements in helical CT with intravenous contrast and suggest that with the complementary benefit of intra-operative ultrasound (IOUS), these are equivalent to CTAP in the assessment of CRCLM.6,7 However, the range of management strategies currently available for CRCLM makes accurate pre-operative assessment of the full extent of disease important.2,8–10 In recent years, MRI has been used in liver imaging because of its high specificity and superior sensitivity regarding lesion detection and characterization in comparison to CT.11–13 The use of hepatobiliary-specific contrast has further improved the accuracy of MRI. Hepatobiliary-specific agents such as Primovist* (gadoxetic acid) are taken up by functioning hepatocytes and excreted in bile. Primovist provides early strong intravascular contrast, facilitating dynamic phase pMRI

(A)

(B)

(C) Figure 1 Perfusion defects seen on computed tomography during arterial portography (CTAP) (

arrows) can be clarified on magnetic

resonance imaging (MRI) where there is better intrinsic soft tissue contrast. Some of the lesions in the right lobe seen on hepatobiliary phase MRI/CTAP (

HPB 2015, 17, 927–935

arrows) are not well seen on the portal venous phase of CTAP owing to perfusional abnormality

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

929

imaging owing to its distribution in the extracellular fluid compartment. Approximately 50% is subsequently taken up by hepatocytes, and excreted into biliary canaliculi providing robust delayed hepatic biliary tree imaging at 20 min post injection, with persistent enhancement for more than 2 h,13–15 thus providing a dual function of dynamic imaging capability as well as delayed hepatobiliary imaging.15,16 Although pMRI for the assessment of CRCLM has been compared with other imaging modalities, there are no data comparing pMRI with CTAP.17–20 The enthusiastic adoption of pMRI by some units suggests a need for more accurate, non-invasive staging of CRCLM than has been available in most centres to date. pMRI has been shown to have sensitivity equivalent to CTAP and better specificity, although the two tests have never been directly compared. There is an appreciable false-positive rate associated with CTAP (Figs 1 and 2). This is important as it may result in a decision against curative surgery in favour of palliative chemotherapy, ablation or regional treatments, all of which while useful, are not as effective as resection in achieving long-term survival in appropriately selected patients.21 There is, therefore, a need to clarify the accuracy Primovist MRI can provide in CRCLM assessment compared with the current standard of CTAP. The aim of the study was to assess and compare the accuracy and inter-observer agreement for detection of liver lesions using Primovist MRI (pMRI) and CTAP. The secondary aim was to assess the accuracy of detection and characterization of liver lesions in patients who received neoadjuvant chemotherapy by comparing the sensitivity and specificity of Primovist MRI and CTAP in this subset.

ate risk of contrast-induced nephropathy, and these patients were prehydrated as per routine protocol for CTAP. Patients were also excluded if they had poor English and were unable to use the help of an interpreter. All patients had the routine workup for CRCLM, including CT chest with a CTAP of the abdomen as well as a PET scan. pMRI was added to the pre-operative assessment. More details on the CTAP and pMRI protocols are given in the supporting information for online publication (SIOP) 1 and 2. At operation, intra-operative evaluation with inspection, palpation and IOUS was performed. Raters Four radiologists experienced in hepatobiliary reporting participated in the study. The two more familiar with liver MRI were asked to report the pMRI, whereas the other two radiologists reported CTAP, and they were blind to the other radiology exam.

Patients and methods

Gold standard and data collection protocol Demographic data and clinical history were collected for all patients. Each radiologist reported the lesions identifiable by the allocated radiology exam, as well as characterization and segmental location in the liver, according to Couinaud classification.22 The gold standard for lesion identification was histopathology. The details on pathology protocol are given in the SIOP 3. Pathologists were required to examine specimens with a fine-cut protocol to identify all small lesions to provide information for assessment of the false-negative rate of the two radiology tests. In patients where lesions were not resected after identification, but rather treated with radiofrequency ablation at surgery, histopathology is not available but for the purpose of diagnostic test accuracy evaluation these lesions were assumed to be CRCLM.

Patients Patients with CRCLM considered suitable for a liver resection aged between 18 and 85 years were eligible for inclusion. Patients with renal impairment (eGFR < 30 ml/min) were excluded, as this is a contraindication to CTAP and pMRI. For CTAP, patients with eGFR 30–60 ml/min carries an intermedi-

Neoadjuvant treatment Where staged or delayed surgery occurred, histopathology still available, allowing correlation of imaging to assess accuracy of imaging characterization. Where surgery delayed 3 months for neoadjuvant chemotherapy, CTAP

CTAP

was the was and

pMRI

Figure 2 Cyst versus metastasis. Some cysts especially the small ones may appear as indeterminate low attenuation lesions and yet a

have typical fluid signal (arrow on CTAP) and hyperintensity in T2-weighted MRI sequences (arrow on MRI)

HPB 2015, 17, 927–935

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

930

pMRI were repeated prior to surgery to evaluate any change in disease and ongoing resectability. This was not a significant deviation from usual practice (Fig. 3). The group that did not proceed to surgery received systemic chemotherapy for palliation and were not included in the analysis. Statistical analysis To demonstrate a difference of 10% in the false-positive rate from the 15% reported for CTAP to 5% for pMRI, with 80% power and a significance level of 0.05, 160 lesions were needed. Assuming an average of two lesions per patients, it was estimated 80 patients would need to be recruited. Data are summarized as the median and interquartile range for continuous variables and percentages for categorical variables. As patients can have multiple lesions, sensitivity and specificity for CTAP and pMRI were computed using lesion as the unit of analysis. Because lesions are clustered at the patient’s level, and each patient is observed by two radiologists, some correlation between observations is expected. Thus, generalized estimation equations (GEE) with logistic regression formulation were used to calculate sensitivities and specificities. An independent working correlation matrix was used for the GEE. The diagnostic performance of both radiological tests was also evaluated for the subgroups of patients who had chemo-

therapy within 6 months prior to scans and patients who had no chemotherapy in the same time period as well as for patients with a larger number of lesions (> 6) and patients with low number of lesions (< 6). The inter-observer agreement for CTAP and pMRI was measured using kappa statistics. The significance level was set at 0.05 for all hypothesis tests. The software IBM SPSS (V22.0; IBM Corp., Armonk, NY, USA) and R 3.1.023 were used for the statistical analysis. Ethics Ethics approval from the Central Network Human Research Ethics Committee (HREC) was granted for the study: HREC/ 10/STG/124. Patient’s consent was obtained in accord with ethical standards of our HREC.

Results Eighty-eight patients were initially recruited from July 2011 to May 2013. Of these, only 62 patients underwent an operation and thus, were eligible for the analysis (Fig. 4). Two patients had two operations; therefore 64 operations were performed, and a total of 219 lesions were confirmed by histopathology. Three patients had radiofrequency ablation of the liver lesion, so histopathology was not available.

CRCLM

CTAP/primovist MRI

Resectable

Surgery histopathology

Neoadjuvant chemotherapy

Not resectable

Repeat CTAP/ primovist MRI

Resectable

Surgery histopathology

Palliative chemotherapy

Not resectable

Repeat imaging

Compare change in size of lesions on radiology

Figure 3 Flowchart of study participants and tests. Colorectal liver metastases (CRCLM), computed tomography during arterial

portography (CTAP), primovist magnetic resonance imaging (pMRI)

HPB 2015, 17, 927–935

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

931

Table 1 Demographic and clinical characteristics of the patients included in the study (n = 62)

Patients enrolled: 88

Demographic data

Evaluations for CRCLM: 91

No operation: 26

Age, median (IQR)

64 (56–70)

Male, n (%)

39 (63%)

Number of lesions, median (IQR)

2 (1–5)

Chemotherapy within 6 months of scans, n (%)

22 (36%)

Type of chemotherapy

Palliative chemo: 17

5-Flurouracil Xeloda (with Avastin)

Neoadjuvant chemo: 5 Refused treatment: 2 No disease seen: 1

6 (2)

5-Flurouracil Oxaliplatin-based (with Avastin)

13 (6)

5-Flurouracil Irinotecan-based (with Avastin)

2 (1)

Regorafenib (with Avastin)

Not fit for surgery: 1

1 (0)

Planned for 2 stage procedure, n (%)

3 (5%)

Received staged procedure, n (%)

2 (3%)

IQR, interquartile range.

Patients analysed: 62 Table 2 Kappa statistics for inter-observer agreement. Computed tomography during arterial portography (CTAP), primovist magnetic resonance imaging (pMRI)

(64 evaluations)

Figure 4 Description and reasons of patients excluded from current analysis. Colorectal liver metastases (CRCLM)

Demographics and details of chemotherapy regimens are shown in Table 1. The median age of the patients was 64 years, and 63% were male. Twenty-two patients received chemotherapy within 6 months prior to scans. The details of the chemotherapy are given in Table 1. Inter-observer agreement on lesions reported by the four radiologists is summarized in Table 2.

Radiologist

pMRI 2

CTAP 1

CTAP 2

pMRI 1

0.80(0.74–0.85)

0.75(0.70–0.81)

0.64(0.57–0.70)

pMRI 2



0.77(0.72–0.82)

0.66(0.60–0.72)

CTAP 1





0.74(0.68–0.79)

The sensitivity for lesion detection was similar for pMRI and CTAP (0.83 versus 0.81, P = 0.642). The specificity of lesion detection was higher on pMRI, 0.88 than CTAP, 0.83 but did not reach statistical significance (P = 0.112).

Table 3 Sensitivity and specificity with the respective 95% confidence intervals based on 219 lesions in 62 patients, confirmed by histopathology

No of patients

No of lesions

CTAP

pMRI

CTAP

pMRI

62

219

0.81 (0.76; 0.86)

0.83 (0.76; 0.87)

0.83 (0.79; 0.87)

0.88 (0.84; 0.91)

Rater 1

60

214

0.81 (0.73; 0.87)



0.84 (0.75; 0.90)



Rater 2

59

213

0.82 (0.73; 0.89)



0.81 (0.72; 0.88)



Rater 3

60

212



0.82 (0.75; 0.88)



0.88 (0.81; 0.92)

Rater 4

62

219



0.84 (0.78; 0.88)



0.88 (0.82; 0.92)

Chemotherapy

22

131

0.82 (0.73; 0.88)

0.83 (0.77; 0.87)

0.63 (0.51; 0.73)

0.79 (0.70; 0.86)

No chemotherapy

40

88

0.80 (0.71; 0.86)

0.82 (0.74; 0.88)

0.93 (0.89; 0.95)

0.92 (0.88; 0.95)

6 or less lesions

52

115

0.80 (0.72; 0.86)

0.84 (0.77; 0.89)

0.86 (0.79; 0.90)

0.89 (0.85; 0.93)

More than 6 lesions

10

104

0.83 (0.74; 0.90)

0.82 (0.76; 0.87)

0.54 (0.40; 0.68)

0.74 (0.58; 0.85)

All patients

Sensitivity (95% CI)

Specificity (95% CI)

Radiologista

Neoadjuvant treatment

Number of lesions

a

There were some missing observations.

HPB 2015, 17, 927–935

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

932

When stratifying by neoadjuvant chemotherapy treatment, the sensitivities of CTAP and pMRI were identical in the 22 patients (131 lesions) who had chemotherapy within 6 months prior to scans (0.82 and 0.83, respectively) and the 40 patients (88 lesions) who did not have chemotherapy in the 6 months period prior to scanning (0.80 and 0.82, respectively). However, the specificity of pMRI was significantly higher than CTAP for patients who had chemotherapy (0.79 versus 0.63, P = 0.011). For patients with no chemotherapy, both CTAP and pMRI showed high specificity (0.93 and 0.92, P = 0.859). When stratified by the number of lesions identified in the histopathology, sensitivity and specificity of CTAP and pMRI were identical for the 52 patients (115 lesions) with 6 or fewer lesions. However, the specificity of both tests decreased in the

CTAP

10 patients (104 lesions) with more than six lesions. In this subgroup, pMRI had a significantly higher specificity than CTAP (0.74 versus 0.54, P = 0.042) (Table 3).

Discussion Several findings should be highlighted from this present study: pMRI showed a good agreement between the two radiologists; operational characteristics of pMRI for the detection of a liver lesion seem to be, as good as CTAP; for the subgroup of patients that had chemotherapy within 6 months of scans, pMRI showed a better specificity compared with CTAP; and for patients with a large number of lesions (> 6), an improvement was seen in specificity for pMRI compared with CTAP.

pMRI

(A)

(B)

(C) Figure 5 Post-chemotherapy computed tomography during arterial portography/magnetic resonance imaging (CTAP/MRI). Lesions after

treatment are much more conspicuous on MRI than on CT. Arrows to CRCLM on CTAP and primovist (p)MRI

HPB 2015, 17, 927–935

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

The St George unit was uniquely placed to perform such a comparison as CTAP remained the units standard in the pre-operative evaluation of patients with CRCLM. The cumulative institutional experience has meant the procedure was done with a low complication rate, and radiologists were experienced in the interpretation of results minimizing false-positives owing to perfusion defects and inexperienced reporting. Although this is the first study comparing pMRI and CTAP using histopathology as a gold standard in the context of CRCLM, several other studies have shown evidence of good operational characteristics of pMRI in different contexts.17–19,21,24,25 Interobserver agreement for pMRI was also reported before (kappa = 0.66)21 but lower than what was found in this study (kappa = 0.88). This may be explained by the fact that in this study readers were not blinded to patient’s history of CRCLM, and this may have increased the accuracy of the radiologists. This study found a higher specificity of pMRI compared to CTAP in the group of patients who received chemotherapy pre-operatively (and, therefore, pre-scan). There was already some expectation that pMRI could be more accurate in this group of patients as hepatobiliary contrast behaviour will be less affected by steatosis and vascular changes related to chemotherapy, providing better parenchyma/tumour contrast owing to hepatocyte uptake of Primovist.26–29 This is clinically relevant given that CT and CTAP are more difficult to interpret in patients who have had neoadjuvant chemotherapy for their CRCLM with irinotecan or oxaliplatin. Therefore, the resulting steatosis, post sinusoidal changes and peliosis affect liver parenchyma making a contrast-related differentiation between the liver and tumour more difficult, limiting the accuracy of interpretation.30 (Fig. 5) The incidence of disappearing CRCLM ranges from 5 to 38%.31–33 The inability to observe disappearing CRCLM after chemotherapy on MRI is strongly associated with a true, complete response at histology.31 Conversely, the ability to detect the site of ‘disappearing CRCLM’ at surgery is more common in patients without pre-operative evaluation with pMRI suggesting these lesions are not true, complete responders.31–34 Variability of detection of disappearing CRCLM at surgery is multifactorial, but the quality of the pre-operative imaging is a significant factor. These data suggests pMRI may be most suitable for patients with disappearing CRCLM after neoadjuvant chemotherapy to assess residual disease and delineate patients with a true radiological complete response.35 There are several limitations to the present study. First, it is difficult to assess true and false-negative rates in the unresected liver, as only the resected liver is subjected to histopathology assessment.36,37 Ideal assessment of the true and false-negative rate is possible when the liver is explanted for histopathology, clearly only possible in a liver transplant population.38 In the context of CRCLM, association of

HPB 2015, 17, 927–935

933

findings with radiology, intra-operative assessment and histopathology on a per lesion basis, provide the best available information for the assessment of sensitivity and specificity of the two tests compared. However, intraoperative inspection, palpation and IOUS will not detect deep and small lesions. IOUS is also an imperfect standard of reference with sensitivity ranging from 80% to 98%.21,39 False-negative results occur mainly with superficial subcapsular regions of the entire liver.25 Follow-up imaging with CT or MRI is also imperfect, as it requires broad assumptions regarding tumour doubling times to distinguish between new and pre-existing micrometastases.21 One important aspect that was not addressed in this study is the economic evaluation of pMRI. The cost of CTAP is rebated by Medicare at $A1000perCTAP.40 Primovist MRI is currently not rebated by Medicare. Patients are asked to pay for this investigation. The current fee for this procedure varies from A 450 to $A 750, but other factors need to be considered. There is, therefore, a need for formal costeffectiveness analysis of this procedure to address the potential cost saving related to the shift from CTAP as a standard diagnostic test to pMRI as the first-line assessment for CRCLM. In conclusion, pMRI is a useful diagnostic study in the assessment of CRCLM and may be a valid alternative to CTAP given that pMRI is less invasive, of comparable sensitivity and specificity in the pre-chemotherapy population, demonstrates better specificity in lesion detection in patients assessed postchemotherapy and shows good reproducibility. Acknowledgments The authors would like to thank the radiology staff at St George Hospital who made every attempt to expedite the MRI required in study patients, and oncologists and surgeons of the Liver Unit at St George Hospital who recruited patients to the study. Thank you also to Jon Koea for his constructive comments in the final stages of this manuscript and to Michael Solomon for his guidance and support throughout the work on this study. The authors also thank the HPB editor and reviewers for their constructive comments which contributed to this final version of the manuscript. Thank you also to Bayer for provided funding for Primovist used in patients recruited to the study. Armando Teixeira-Pinto was supported by NHMRC program grant 633003 to the Screening & Test Evaluation Program (STEP).

Conflicts of interest Dr Bagia received a grant from Bayer to fund the Primovist contrast used in this study. For the remaining authors: none declared.

References 1. Geoghegan JG, Scheele J. (1999) Treatment of colorectal liver metastases. Br J Surg 86: 158–169.

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

934

2. Garden OJ, Rees M, Poston GJ, Mirza D, Saunders M, Ledermann J et al. (2006) Guidelines for resection of colorectal cancer liver metastases. Gut 55(Suppl 3): iii1–iii8. 3. Billingsley K. (2006) Comparison of helical CT and CT arterial portography for colorectal liver metastastes. World J Surg 30: 1900–1901. 4. Soyer P, Levesque M, Elias D, Zeitoun G, Roche A. (1992) Preoperative

tases from colorectal cancer: a systematic review and meta-analysis. J Magn Reson Imaging 31: 19–31. 19. Niekel MC, Bipat S, Stoker J. (2010) Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment. Radiology 257: 674–684.

assessment of resectability of hepatic metastases from colonic carci-

20. Seo HJ, Kim MJ, Lee JD, Chung WS, Kim YE. (2011) Gadoxetate dis-

noma: CT portography vs sonography and dynamic CT. AJR Am J

odium-enhanced magnetic resonance imaging versus contrast-en-

Roentgenol 159: 741–744.

hanced

5. Soyer P, Bluemke DA, Fishman EK. (1994) CT during arterial portography for the preoperative evaluation of hepatic tumors: how, when, and why? AJR Am J Roentgenol 163: 1325–1331. 6. Fusai G, Davidson BR. (2003) Management of colorectal liver metastases. Colorectal Dis 5: 2–23. 7. Valls C, Andia E, Sanchez A, Guma A, Figueras J, Torras J et al. (2001) Hepatic metastases from colorectal cancer: preoperative detection and assessment of resectability with helical CT. Radiology 218: 55–60.

18F-fluorodeoxyglucose

positron

emission

tomography/

computed tomography for the detection of colorectal liver metastases. Invest Radiol 46: 548–555. 21. Bluemke DA, Sahani D, Amendola M, Balzer T, Breuer J, Brown JJ et al. (2005) Efficacy and safety of MR imaging with liver-specific contrast agent: U.S. multicenter phase III study. Radiology 237: 89–98. 22. Sutherland F, Harris J. (2002) Claude Couinaud: a passion for the liver. Arch Surg 137: 1305–1310. 23. Ihaka RGR. (1996) A language for data analysis and graphics. J Comput Graph Stat 5: 299–314.

8. Folprecht G, Grothey A, Alberts S, Raab HR, Kohne CH. (2005) Neoadju-

24. Donati OF, Hany TF, Reiner CS, von Schulthess GK, Marincek B,

vant treatment of unresectable colorectal liver metastases: correlation

Seifert B et al. (2010) Value of retrospective fusion of PET and MR

between tumour response and resection rates. Ann Oncol 16: 1311–

images in detection of hepatic metastases: comparison with 18F-

1319.

FDG PET/CT and Gd-EOB-DTPA-enhanced MRI. J Nucl Med 51:

9. Hillingso JG, Wille-Jorgensen P. (2009) Staged or simultaneous resection of synchronous liver metastases from colorectal cancer – a systematic review. Colorectal Dis 11: 3–10. 10. Garcea G, Ong SL, Maddern GJ. (2008) Inoperable colorectal liver metastases: a declining entity? Eur J Cancer 44: 2555–2572.

692–699. 25. Huppertz A, Balzer T, Blakeborough A, Breuer J, Giovagnoni A, HeinzPeer G et al. (2004) Improved detection of focal liver lesions at MR imaging: multicenter comparison of gadoxetic acid-enhanced MR images with intraoperative findings. Radiology 230: 266–275.

11. Frydrychowicz A, Lubner MG, Brown JJ, Merkle EM, Nagle SK, Rofsky

26. Aloia T, Sebagh M, Plasse M, Karam V, Levi F, Giacchetti S et al.

NM et al. (2012) Hepatobiliary MR imaging with gadolinium-based con-

(2006) Liver histology and surgical outcomes after preoperative

trast agents. J Magn Reson Imaging 35: 492–511.

chemotherapy with fluorouracil plus oxaliplatin in colorectal cancer liver

12. Bhattacharjya S, Bhattacharjya T, Baber S, Tibballs JM, Watkinson AF,

metastases. J Clin Oncol 24: 4983–4990.

Davidson BR. (2004) Prospective study of contrast-enhanced computed

27. Ryan P, Nanji S, Pollett A, Moore M, Moulton CA, Gallinger S et al.

tomography, computed tomography during arterioportography, and

(2010) Chemotherapy-induced liver injury in metastatic colorectal can-

magnetic resonance imaging for staging colorectal liver metastases for

cer: semiquantitative histologic analysis of 334 resected liver specimens

liver resection. Br J Surg 91: 1361–1369.

shows that vascular injury but not steatohepatitis is associated with

13. Lencioni R, Donati F, Cioni D, Paolicchi A, Cicorelli A, Bartolozzi C.

preoperative chemotherapy. Am J Surg Pathol 34: 784–791.

(1998) Detection of colorectal liver metastases: prospective comparison

28. Slim K, Nini E, Forestier D, Kwiatkowski F, Panis Y, Chipponi J.

of unenhanced and ferumoxides-enhanced magnetic resonance imag-

(2003) Methodological index for non-randomized studies (minors):

ing at 1.5 T, dual-phase spiral CT, and spiral CT during arterial portog-

development and validation of a new instrument. ANZ J Surg 73:

raphy. MAGMA 7: 76–87.

712–716.

14. Seale MK, Catalano OA, Saini S, Hahn PF, Sahani DV. (2009) Hepatobil-

29. Vauthey JN, Pawlik TM, Ribero D, Wu TT, Zorzi D, Hoff PM et al. (2006)

iary-specific MR contrast agents: role in imaging the liver and biliary

Chemotherapy regimen predicts steatohepatitis and an increase in 90-

tree. Radiographics 29: 1725–1748.

day mortality after surgery for hepatic colorectal metastases. J Clin

15. Kitao A, Zen Y, Matsui O, Gabata T, Kobayashi S, Koda W et al. (2010) Hepatocellular carcinoma: signal intensity at gadoxetic acid-enhanced

Oncol 24: 2065–2072. 30. Bipat S, van Leeuwen MS, Comans EF, Pijl ME, Bossuyt PM, Zwinder-

MR Imaging–correlation with molecular transporters and histopathologic

man AH et al. (2005) Colorectal liver metastases: CT, MR imaging, and

features. Radiology 256: 817–826.

PET for diagnosis – meta-analysis. Radiology 237: 123–131.

16. Weinmann HJ, Schuhmann-Giampieri G, Schmitt-Willich H, Vogler H,

31. Auer RC, White RR, Kemeny NE, Schwartz LH, Shia J, Blumgart LH

Frenzel T, & Gries H. (1991). A new lipophilic gadolinium chelate as a

et al. (2010) Predictors of a true complete response among disappear-

tissue-specific contrast medium for MRI. Magn Reson Med, 22, 233–

ing liver metastases from colorectal cancer after chemotherapy. Cancer

237; discussion 42.

116: 1502–1509.

17. Zech CJ, Korpraphong P, Huppertz A, Denecke T, Kim MJ, Tanomkiat

32. Elias D, Youssef O, Sideris L, Dromain C, Baton O, Boige V et al. (2004)

W et al. (2014) Randomized multicentre trial of gadoxetic acid-en-

Evolution of missing colorectal liver metastases following inductive

hanced MRI versus conventional MRI or CT in the staging of colorectal cancer liver metastases. Br J Surg 101: 613–621. 18. Floriani I, Torri V, Rulli E, Garavaglia D, Compagnoni A, Salvolini L et al. (2010) Performance of imaging modalities in diagnosis of liver metas-

HPB 2015, 17, 927–935

chemotherapy and hepatectomy. J Surg Oncol 86: 4–9. 33. van Vledder MG, de Jong MC, Pawlik TM, Schulick RD, Diaz LA, Choti MA. (2010) Disappearing colorectal liver metastases after chemotherapy: should we be concerned? J Gastrointest Surg 14: 1691–1700.

ª 2015 International Hepato-Pancreato-Biliary Association

HPB

935

34. Elias D, Goere D, Boige V, Kohneh-Sharhi N, Malka D, Tomasic G et al.

39. Schmidt J, Strotzer M, Fraunhofer S, Boedeker H, & Zirngibl H. (2000).

(2007) Outcome of posthepatectomy-missing colorectal liver metas-

Intraoperative ultrasonography versus helical computed tomography

tases after complete response to chemotherapy: impact of adjuvant

and computed tomography with arterioportography in diagnosing col-

intra-arterial hepatic oxaliplatin. Ann Surg Oncol 14: 3188–3194.

orectal liver metastases: lesion-by-lesion analysis. World J Surg, 24,

35. Adams RB, Aloia TA, Loyer E, Pawlik TM, Taouli B, Vauthey JN. (2013) Selection for hepatic resection of colorectal liver metastases: expert consensus statement. HPB 15: 91–103.

43–47; discussion 8. 40. Medicare Australia. MBS online. Available at http://www9.health.gov.au/ mbs/search.cfm?q=56507&sopt=I (last accessed 1 June 2011).

36. Abe K, Abe R, Abe T, Adachi I, Ahn BS, Aihara H et al. (2002) Observation of double cc production in e+e

annihilation at squareroot[s]

approximately 10.6 GeV. Phys Rev Lett 89: 142001. 37. Young JM, Solomon MJ. (2009) How to critically appraise an article. Nat Clin Pract Gastroenterol Hepatol 6: 82–91.

Supporting information Additional Supporting Information may be found in the online version of this article:

38. Hardie AD, Nance JW, Boulter DJ, Kizziah MK. (2011) Assessment of

Appendix S1. CTAP protocol.

the diagnostic accuracy of T2*-weighted MR imaging for identifying

Appendix S2. pMRI protocol.

hepatocellular carcinoma with liver explant correlation. Eur J Radiol 80:

Appendix S3. Pathology protocol.

e249–e252.

HPB 2015, 17, 927–935

ª 2015 International Hepato-Pancreato-Biliary Association

Prospective diagnostic test accuracy comparison of computed tomography during arterial portography and Primovist magnetic resonance imaging in the pre-operative assessment of colorectal cancer liver metastases.

To assess and compare the accuracy and inter-observer agreement for the detection of liver lesions using Primovist magnetic resonance imaging (pMRI) a...
NAN Sizes 0 Downloads 8 Views