LEADING ARTICLE

Does the choice of baseline liver imaging influence patient outcomeafter resection of colorectal liver metastases?John McCall

Department of Surgery, University of Otago, Dunedin, New Zealand

CorrespondenceJohn McCall, Department of Surgery, University of Otago, PO Box 913, Dunedin 9054, New Zealand. Tel: + 64 3 4740999. Fax:

+ 64 3 4747622. E-mail: john.mccall@otago.ac.nz

The surgical management of colorectal liver metastases (CLM)has evolved over recent decades to bring radical treatment andpotential cure within the reach of many more patients. Successfulstrategies include parenchyma-sparing segmental resection, com-bined resection and ablation, staged resection, portal vein embo-lization, repeat hepatectomy, resection of limited extrahepaticdisease, and systemic therapy to downstage initially inoperabledisease. Advances in cross-sectional imaging, such as multi-detector computed tomography (CT) and magnetic resonanceimaging (MRI), and nuclear medicine 18F-fluorodeoxyglucosepositron emission tomography (FDG-PET) have alsoimproved preoperative staging, with the result that the ‘open andclose’ laparotomy has, fortunately, become rare. Together, theseadvances have led to improvements in longterm survival, despiteextending the limits of radical treatment.1,2

In this issue of the journal, Knowles et al. examine the questionof baseline imaging in patients referred for resection of CLM. Theauthors conclude that liver-specific MRI, performed prior to neo-adjuvant chemotherapy, leads to better patient outcomes byreducing intrahepatic recurrence and the need for repeat hepate-ctomy. A crucial point of difference in this study is that it appliesimportant clinical endpoints rather than merely comparing sen-sitivity and specificity rates.

The study utilized a prospectively maintained database to iden-tify 242 patients who had undergone neoadjuvant chemotherapyfollowed by liver resection in a single institution and divided theminto those who had undergone baseline imaging with liver-specificMRI and those who had undergone CT, non-contrast MRI orgadolinium-enhanced MRI. Liver-specific MRI utilizes contrastagents that are taken up by normal hepatocytes and thereforeaccentuates the contrast differential between CLM and surround-ing hepatic parenchyma during dynamic sequences. All patientsunderwent liver-specific MRI between neoadjuvant chemo-therapy and surgery. FDG-PET scans were not routinely per-formed. The most important finding was that patients who did

not undergo liver-specific MRI at baseline showed more intrahe-patic recurrences (this increase was statistically significant whenresection margin recurrences were excluded). These patientsrequired more repeat hepatectomies to achieve longterm survivalequivalent to that in the group in which baseline imaging withliver-specific MRI had been performed. The implication is thatliver-specific MRI detected more lesions at baseline, particularlysmall lesions that are inclined to ‘disappear’ in response to che-motherapy. If these ‘disappearing’ lesions had been identifiedprior to chemotherapy, it is more likely they would have beenencompassed within the planned resection. Another importantfinding was that liver-specific MRI prior to chemotherapy led to achange in tumour staging in more than half of all patients, and innine patients one or more suspected liver lesions were benign. Anysurgeon who has carried out an unnecessary liver resection for asclerosed haemangioma or benign bile duct adenoma, believing itto be a CLM, will appreciate the value of fully characterizinglesions preoperatively.

If the findings of this study are robust, it will set a new standardin the workup of patients with resectable CLM. We should there-fore look critically at methodological aspects of the study. Theinvestigators had at their disposal a large, prospectively main-tained database with good longterm follow-up, but the study isessentially a post hoc analysis of a selected subset of patients (242of 715) drawn from the database. The patients were not prospec-tively selected and not randomized and unmeasured differencesbetween the groups may have existed and may have influenced thechoice of both baseline imaging and outcome measures. Of 419patients who underwent neoadjuvant chemotherapy, 155 wereexcluded because they received ‘suboptimal’ chemotherapy,defined as fewer than three cycles of oxaplatin or irinotecan-basedtreatment, or fluorouracil and leucovorin (FU/LV) alone.However, irinotecan has no additional benefit over FU/LV alone inthe adjuvant setting, for either primary or metastatic colorectalcancer,3,4 so there would appear to be no good rationale for

DOI:10.1111/j.1477-2574.2012.00458.x HPB

HPB 2012, 14, 281–282 © 2012 International Hepato-Pancreato-Biliary Association

selecting on this basis. It would have been more rigorous toinclude all patients going forward for neoadjuvant therapy, thusemulating an intent-to-treat analysis, or to include only thosetreated by a regimen reflecting the current standard of care, whichdoes not include irinotecan. Finally the primary endpoint ofintrahepatic recurrence was only significant when resectionmargin recurrences were excluded. However, this may be reason-able given that positive margin resections (15% and 13%, respec-tively) and margin recurrence rates (30% and 23%, respectively)were similar in both groups, which lends support to the hypoth-esis that lesions elsewhere in the liver that were missed on imagingexplain the higher recurrence rate in patients who did notundergo baseline imaging with liver-specific MRI.

Three months of chemotherapy before and after liver resectionhas become the standard of care for resectable CLM based on theEuropean Organization for Research and Treatment of Cancer(EORTC) study 409835. The study by Knowles et al. raises anothervery important issue, which is the management of lesions that‘disappear’ during neoadjuvant chemotherapy. We now know thatthe correlation between radiological disappearance and completepathological response is poor and that �80% of such lesions arelikely to recur within 12 months if they are left untreated.6 Indeed,this is part of the rationale for the ATTACHE trial, an AustralasianGastrointestinal Trials Group-sponsored study comparing pre-and postoperative adjuvant chemotherapy with postoperativechemotherapy for resectable CLM. As well as reducing lesion size,chemotherapy promotes hepatic steatosis (limiting the sensitivityof CT) and reduces tumour metabolic activity (limiting the sen-sitivity of FDG-PET), and thus MRI is likely to be superior in thissetting.7 Unlike the present study, another prospective study failedto demonstrate the superiority of liver-specific MRI overgadolinium-enhanced MRI for detection of CLM.8

So where does this leave us? It would be easy to say that thefindings of Knowles et al. need to be validated by a well-designedprospective study, but this is unrealistic because the minimumperiod needed to recruit participants and reach clinical endpointsis 5–10 years. During that time, imaging technology will inevitablymove forward. By contrast, there is little disadvantage to recom-mending liver-specific MRI for staging the liver in CLM.Gadolinium-ehhanced MRI has been shown to have advantagesover CT in this context7 and the report by Knowles et al. stronglysuggests, although does not prove, that liver-specific MRI is betterstill. There are no additional patient risks and the marginal costof liver-specific contrast is minute compared with the overall

treatment cost of combined chemotherapy and surgery. Theavoidance of even a few repeat hepatectomies and unnecessaryhepatectomies for benign disease will reduce both cost and patientmorbidity. A greater logistic challenge is to educate surgical andoncological colleagues to refer patients prior to the commence-ment of neoadjuvant chemotherapy so that thorough baselineassessment, including liver-specific MRI for the liver and PET-CTfor extrahepatic disease, is ensured. It may also be that selectedpatients are better served with surgery prior to chemotherapy,although we must await the results of the ATTACHE trial to betterdefine the respective risks and benefits of peri- and postoperativeadjuvant chemotherapy for resectable CLM.

References

1. Choti MA, Sitzmann JV, Tiburi MF, Sumetchotimetha W, Rangsin R,

Schulick RD et al. (2002) Trends in longterm survival following liver

resection for hepatic metastases. Ann Surg 235:759–766.

2. Kopetz SK, Chang GJ, Overman MJ, Eng C, Sargent DJ, Larson DW et al.

(2009) Improved survival in metastatic colorectal cancer is associated with

adoption of hepatic resection and improved chemotherapy. J Clin Oncol

27:3677–3683.

3. Van Cutsem E, Labianca R, Bodoky G, Barone C, Aranda E, Nordlinger B

et al. (2009) Randomized phase III trial comparing biweekly infusional

fluorouracil/leucovorin alone or with irinotecan in the adjuvant treatment of

stage III colon cancer: PETACC-3. J Clin Oncol 27:3117.

4. Ychou M, Hohenberger W, Thezanas S, Navarro M, Gascon P, Bokemeyer

C et al. (2008) Randomized phase III trial comparing infused 5-fluorouracil/

folinic acid (LV5FU) versus LV5FU+irinotecan (LV5FU+IRI) as adjuvant treat-

ment after complete resection of liver metastases from colorectal cancer

(LMCRC). (GPT-GMA-301). J Clin Oncol 26 (Suppl. 15):LBA4013.

5. Nordlinger B, Sorbye H, Glimelius B, Poston GJ, Schlag PM, Rougier P

et al. (2008) Perioperative chemotherapy with FOLFOX4 and surgery

versus surgery alone for resectable liver metastases from colorectal cancer

(EORTC Intergroup trial 40983): a randomized controlled trial. Lancet

371:1007–1016.

6. Benoist S, Brouquet A, Penna C, Julie C, El Hajjam M, Chagnon S et al.

(2006) Complete response of colorectal liver metastases after chemo-

therapy: does it mean cure? J Clin Oncol 24:3939–3945.

7. Floriani I, Torri V, Rulli E, Garavaglia D, Compagnoni A, Salvolina A et al.

(2010) Performance of imaging modalities in diagnosis of liver metastases

from colorectal cancer: a systematic review and meta-analysis. J Magn

Reson Imaging 31:19–31.

8. Mainenti P, Mancini M, Mainolfi C, Camera L, Maurea S, Manchia A et al.

(2010) Detection of colorectal liver metastases: a prospective comparison

of contrast-enhanced US, multidetector CT, PET/CT and 1.5 Tesla MR with

extracellular and reticulo-endothelial cell specific contrast agents. Abdom

Imaging 35:511–521.

282 HPB

HPB 2012, 14, 281–282 © 2012 International Hepato-Pancreato-Biliary Association