|
Incidence of benign pathology in patients undergoing hepatic resection for suspected malignancy
R.A.E. Clayton
D.L. Clarke
E.J. Currie
K.K. Madhavan
R.W. Parks
O.J. Garden
Department of Clinical and Surgical Sciences
(Surgery) Royal Infirmary of Edinburgh
Correspondence to: Mr RW Parks, Department of Clinical and Surgical Sciences, University of Edinburgh, Royal Infirmary of Edinburgh, Lauriston Place, Edinburgh EH3 9YW
Keywords: Hepatectomy, metastasis, biliary stricture, cholangiocarcinoma, Klatskin’s
tumour,
hepatoma, biopsy, staging
Surg J R Coll Surg Edinb Irel., 1 February 2003, 32-38
Background: Confirming the presence of hepatic or proximal bile duct malignancy pre-operatively remains difficult and some patients may undergo surgical resection for suspected malignant lesions which subsequently turn out to be benign. The aim of this study was to establish whether improvements in pre-operative staging might better identify this patient population. Methods: Analysis of a prospectively collected database, which has been maintained in our unit since 1988. Results: Of 250 consecutive patients undergoing hepatic resection for presumed malignancy, 18 (7.2%) were shown to have benign pathology. These “false positive” rates were 4 out of 160 (2.5%) resections for colorectal metastases, 4 out of 49 (8.2%) resections for other solid hepatobiliary tumours and 10 out of 41 (24.4%) resections for hilar cholangiocarcinoma. Four of the 18 patients (22%) developed post-operative complications but there was no postoperative mortality. Conclusion: Although hepatic resection remains a potentially curative procedure for patients with tumours involving the liver parenchyma or proximal bile ducts, pre-operative confirmation of malignancy remains difficult. Despite appropriate investigation a subset of patients with benign disease will still be subjected to major hepatic resection which should be undertaken in a specialist unit
INTRODUCTION
Hepatic resection for hilar cholangio carcinoma and parenchymal intrahepatic malignancies,
when undertaken in specialist units, is now well established as an effective treatment with
acceptable morbidity and mortality rates. Five-year survival after hepatic resection for hilar
cholangiocarcinoma has been reported in up to 35% of patients, compared with a 5-year
survival rate of just 4% if no surgical procedure is undertaken.1,2
In our unit, 3-year survival for patients undergoing resection of colorectal
metastases is 65%; peri-operative mortality is 5% and the major complication rate is 20%
following hepatic resection for all indications.3 These figures are comparable to those of other
specialist units.
However, confirming a pre-operative diagnosis of malignancy in the setting of a suspicious intra-hepatic or hilar lesion is difficult. Brush cytology from biliary lesions is specific but relatively insensitive.4 Fine needle aspiration or biopsy of solid intra-hepatic lesions may potentially seed tumour cells and carries a small but definite risk of complication such as bile leakage or bleeding.5 Consequently, many surgeons do not pursue a rigorous preoperative histological or cytological diagnosis as part of the pre-operative work-up. As a result a subset of patients may be subjected to surgical resection for what eventually turns out to be benign disease. The aim of this study was to ascertain how many patients in our series underwent major liver resection for benign pathology and to consider the diagnostic workup and outcome of surgical intervention in these patients.
PATIENTS AND METHODS
Data were extracted from a prospective database which has been maintained in our
unit since October 1988. Patients whose liver lesions were diagnosed by conventional
radiology to be specifically benign, and subsequently underwent resection were excluded; these included patients with cysts,
adenomas and focal nodular hyperplasia. Important clinical features such as jaundice,
abdominal pain, weight loss and pruritus, a past history of gallstones, sclerosing
cholangitis, inflammatory bowel disease or a previous malignancy in those with suspected
metastatic disease were noted.
In all patients liver function tests, including a coagulation screen, were measured and in appropriate patients tumour marker levels such as alpha fetoprotein (AFP) and carcinoembryonic antigen (CEA) levels were also measured. Pre-operative assessment included a computerised tomography (CT) scan in all patients. Other imaging techniques were used depending on the clinical presentation and CT findings. Patients with biliary strictures underwent endoscopic retrograde cholangiopancreatography (ERCP) and biliary stenting if delay to surgery was anticipated. If ERCP and stenting were unsuccessful, a percutaneous transhepatic cholangiogram (PTC) and stent insertion was undertaken. If lesions encroached on major blood vessels, angiography was performed in order to ensure resectability. Magnetic resonance imaging (MRI) scanning was not available in our unit during the study period and was not used. All radiological images were reviewed by a consultant surgeon and radiologist. Brush cytology was not performed routinely, but any external biliary drain fluid was routinely sent for cytological analysis. In order to minimise the risk of tumour seeding, it is not our policy to perform pre-operative biopsy.
Figure 1: Histological section showing an abscess cavity, neutrophil infiltration and fibrosis and adherent gallbladder wall in a hepatic resection specimen from a patient presenting with an apparently malignant biliary stricture
RESULTS
Two-hundred and fifty patients underwent hepatectomy for presumed neoplasms
between 1988 and 2001. The breakdown of this group, with their pre-operative working
diagnoses and eventual false-positive rates, are summarised in Table 1. Final histological
analysis revealed no evidence of malignancy in 18 (7.2%) patients (13 females, 5 males),
ranging from 2.5% in patients with suspected colorectal metastases to 24.4% in those with
suspected malignant biliary strictures. The 18 patients had a mean age of 50 years (range
21-77 years). The clinical, biochemical and
radiological findings which led to the
preoperative diagnosis of malignancy in the 18 patients are summarised in Table 2, along with
the final histology.
| TABLE 1. BREAKDOWN OF HEPATIC RESECTIONS UNDERTAKEN SHOWING INCIDENCE OF BENIGN PATHOLOGY | ||||
|
Presumed pre-operative diagnosis |
Final pathology |
Total resections |
“False positive” (%) |
|
| Benign | Malignant | |||
| Malignant biliary strictures | 10 | 31 | 41 | 24.4% |
| Secondary liver tumours | 4 | 156 | 160 | 2.5% |
| Gallbladder carcinoma | 2 | 13 | 15 | 13.3% |
| Hepatocellular carcinoma | 2 | 32 | 34 | 5.9% |
| Total | 18 | 232 | 250 | 7.2% |
Ten patients had presumed malignant biliary strictures, of whom eight presented with obstructive jaundice. Four also had weight loss. Of the other two patients, one was known to have sclerosing cholangitis and presented with sepsis, a rise in serum alkaline phosphatase and gamma glutamyltransferase and a CT scan which revealed a 1.5cm lesion resembling cholangiocarcinoma at the hepatic duct confluence. The remaining patient presented with intra-hepatic abscesses for which resection was indicated. In all but one patient at least two investigations were suggestive of malignancy (Table 2). Additionally, two patients had brush cytology which revealed cells consistent with cholangiocarcinoma. Figures 1-3 demonstrate pre-operative CT and ERCP films suggesting malignancy from one of these patients, along with the final histology which revealed chronic inflammation.
Of the four patients with suspected metastatic carcinoma, three had previously undergone resection for primary colorectal carcinoma and their hepatic lesions were detected on routine follow-up CT scanning. The fourth patient presented with abdominal pain, obstructive jaundice and weight loss. She had no previous history of malignant disease but CT revealed a lesion suspicious of a metastatic deposit in segment four.
The two patients suspected of having hepatocellular carcinoma (HCC) presented with right upper quadrant pain. In both patients, AFP levels were within normal limits but CT scanning revealed features consistent with neoplasia. One had a 20cm abdominal mass and in the other patient laparoscopic ultrasonography also revealed features consistent with HCC. Both patients with suspected gallbladder carcinoma presented with abdominal pain and weight loss.
The hepatic resections undertaken in the 18 patients are summarised in Table 2. All patients with suspected hilar cholangiocarcinoma also underwent radical bile duct excision and lymphadenectomy. The mean blood loss was 1642ml (240-5560ml) but only four patients (22%) required blood transfusion (median transfusion 5 units). One patient required admission to the intensive therapy unit due to blood loss during resection of a 20cm diameter lesion but made a good recovery without further intervention. There were no deaths in this series but four patients (22%) had significant post-operative complications: three patients developed subphrenic or subhepatic collections requiring percutaneous (2) or surgical (1) drainage and one patient developed pneumonia and pulmonary embolism. Median post-operative hospital stay was 10 days (range 7 - 26 days).
| TABLE 2. SUMMARY OF PRE-OPERATIVE INVESTIGATIONS, OPERATIVE INTERVENTIONS AND FINAL PATHOLOGY ON THE 18 PATIENTS SHOWN TO HAVE BENIGN DISEASE WHO UNDERWENT HEPATIC RESECTION FRO PRESUMED MALIGNANCY | |||||||||||||
| Pre-operative working diagnosis | Past history | Suspicious symptoms | Suspicious ultrasound | Suspicious CT | Suspicious cholangiogram | Suspicious cytology or histology | Suspicious laparoscopic ultrasound | Suspicious laparotomy findings | Operation | Final pathology | |||
| Hilar biliary strictures | |||||||||||||
| 1 | Cholangiocarcinoma | 54 | F |
Yes |
Yes | Yes | Yes | Yes | Yes | Inconclusive | Left hepatectomy & excision extrahepatic biliary tree | Abscess, sclerosing cholangitis | |
| 2 | Cholangiocarcinoma | 46 | F | Yes | No | No | Yes | No | Inconclusive | Yes | Right hepatectomy (at second laparotomy) | Benign inflammatory pseudotumour | |
| 3 | Cholangiocarcinoma | 59 | M | Cholecystectomy | Yes | Yes | No | Yes | No | No | Yes | Right hepatectomy & excision of caudate & lymphadectomy | Cholangitis, fibrotic hilar nodule |
| 4 | Cholangiocarcinoma | 51 | F | Yes | Not done | Yes | Yes | No | Not done | Yes | Left hepatectomy & excision extrahepatic biliary tree, lymphadenectomy | Impacted stone, periductal fibrosis | |
| 5 | Cholangiocarcinoma | 77 | F | Yes | Not done | Inconclusive | Yes | No | Yes | Yes | Extended right hepatectomy | Inconclusive, but no malignancy | |
| 6 | Cholangiocarcinoma | 67 | M | Sclerosing cholangitis | No | No | Yes | Yes | Not done | No | Yes | Extended right hepatectomy | Fibrosis, sclerosing cholangitis |
| 7 | Cholangiocarcinoma | 51 | F | Sclerosing cholangitis | Yes | Not done | Inconclusive | Yes | Yes | Not done | Unknown | Left hepatectomy | Fibrosis, sclerosing cholangitis |
| 8 | Cholangiocarcinoma | 50 | F | Yes | Not done | Yes | No | Not done | Not available | No | Left hepatectomy & excision of extrahepatic biliary tree | Chronic inflammation and fibrosis | |
| 9 | Cholangiocarcinoma | 47 | F |
DVT |
Yes | No | Yes | Yes | No | Yes | Yes | Extended right hepatectomy & radical bile duct excision | Reactive inflammatory changes, fibrosis |
| 10 | Cholangiocarcinoma | 34 | M |
Known intra-hepatic abscess |
Not known | Not available | Yes | Not available | Not available | Not available | Yes | Right hepatectomy | Benign inflammatory pseudotumour |
| Solid liver lesions | |||||||||||||
| 11 | Hepatoma | 21 | F |
Huge abdominal mass arising from liver |
Yes | Not done | Yes | Not done | Not done | Not done | Yes | Extended right hepatectomy & excision of caudible lobe | Leiomyoma of the liver |
| 12 | Hepatoma | 24 | F |
Hereditary haemorrhagic telangiectasia |
No | Inconclusive | Yes | Not done | No | Yes | Yes | Extended right hepatectomy | Post-necrotic regenerative cirrhotic nodule |
| 13 | Colorectal metastases | 41 | F |
Dukes B colorectal carcinoma, left hemicolectomy |
No | Not done | Inconclusive | Not done | Not done | No | Yes | Posterior segmentectomy (segs IV, VII) | Haemangioma |
| 14 | Metastases or leiomyoma | 61 | F | Cholesystectomy | Yes | Yes | Yes | No | No | Not done | Yes | Left hepatectomy & lymph adenectomy, excision extra-hepatic biliary tree | Benign hepatobiliary cystadenoma |
| 15 | Colorectal metastases | 65 | F | Dukes C colorectal carcinoma, anterior resection, Breast carcinoma |
No |
Not done | Yes | Not done | Not done | Not done | No | IVB, V segmental resection | Haemangioma, focal nodular hyperplasia |
| 16 | Colorectal metastases | 44 | M | Anterior resection |
Not known |
Not available | Not available | Not available | Not available | Yes | Yes | Right hepatectomy | Focal nodular hyperplasia |
| 17 | Gallbladder carcinoma | 54 | M | Gastritis, erosive duodenitis |
Yes |
No | Inconclusive | Inconclusive | Inconclusive | Yes | Yes | Atypial wedge resection (segs IV, V) | Chronic cholecystitis, ductal proliferation |
| 18 | Gallbladder carcinoma or hepatic primary | 61 | F | Hiatus hernia |
No |
Inconclusive | Inconclusive but CTAP suspicious | Not done | Not done | Inconclusive | Yes | Extend left hepatectomy | Cystadenoma, haemangioma |
DISCUSSION
Differentiation of benign and malignant focal solid liver lesions or biliary strictures
at the hepatic hilum is difficult and a major concern in the management of such lesions.
The differential diagnosis of a stenotic lesion of the proximal bile duct includes
cholangiocarcinoma, metastatic lymph node deposits from a wide range of primary visceral
malignancies, Hodgkin’s and Non-Hodgkin’s lymphoma and a number of benign conditions
such as Mirizzi syndrome, granulomas, benign fibrosis and tuberculosis. However,
the ideal of a pre-operative tissue diagnosis is controversial in solid lesions and difficult to
achieve in hilar lesions. Lesions of the biliary system are often not suitable for biopsy, and
for this reason cytological techniques are most widely used. Cytology from biliary aspirates
is often unhelpful as the exfoliated tumour cells degenerate quickly in bile and are often
poorly preserved. Brush biopsy performed at ERCP is frequently employed, yields well
preserved cells and has a low complication rate. However, whilst it is highly specific for the
diagnosis of malignancy, poor sensitivity limits its applicability. Stewart et al (2001) recently
reviewed 448 satisfactory specimens obtained from 406 patients and found the specificity of
the technique to be 98% but the sensitivity was only 59.8%.4 There were three false positives
in this group, two of whom were subjected to surgery from which one patient ultimately
succumbed. Inadequate specimens were obtained in 5.3% of patients. This experience
included all proximal and distal biliary strictures and it could be suggested that the
sensitivity for lesions at the biliary confluence may well be even lower. Bile cytology was
positively misleading in two cases in our own series as the reports indicated cellular atypia
or overt malignancy. In the remaining five cases, the clinical and radiological appearances
of malignancy were so overwhelming that we regarded the negative cytology report as
unhelpful.
There are a number of reasons for the poor sensitivity of brush cytology. Submucosal tumour spread or stricturing secondary to external compression by malignant lymph nodes may spare the epithelium. Interpretation of well differentiated carcinomas or certain malignant subtypes such as papillary or mucinous tumours may be difficult. New techniques, including identification of p53 immunoreactivity or digital image analysis, might improve diagnostic accuracy.6,7
Figure 2: Pre-operative CT scan from patient in Figure 1 showing a 2cm hilar lesion and dilation of the right intrahepatic ductal system
However, these techniques are still evolving and are not part of routine practice. The inference remains that whilst a positive cytological diagnosis is helpful, negative cytology cannot reliably exclude the presence of malignancy. A negative cytology result should not be regarded as proof of the absence of a malignant lesion, and as complete surgical resection of cholangiocarcinoma remains the only therapy which is potentially curative, confirmatory cytological proof should not be mandatory prior to considering these patients for hepatic resection. However, in the absence of preoperative cytology, a subset of patients may undergo radical liver resectional surgery for what is ultimately shown to be benign disease.
This approach is supported by several other authors. Despite the extensive pre-operative investigation of patients suspected of having a malignant biliary stricture, in all reported series there is a subgroup of patients who eventually turn out to have benign disease following resection. None of these authors actively pursued a pre-operative cytological diagnosis. Hadjis et al (1985) reported 8 patients ultimately shown to have benign disease in a series of 104 who had been subjected to bile duct excision for what was thought to be a malignant hilar lesion.8 Similarly Wetter et al (1991) reported a series of 98 patients with a presumed pre-operative diagnosis of cholangiocarcinoma of the common hepatic duct.9 The final diagnosis was not sclerosing cholangiocarcinoma in 30 of this group, and included 22 other malignant lesions and 8 benign lesions (Mirizzi syndrome [2], benign granulomas [3], idiopathic benign focal stenosis [3]. Gerhards et al (2001) reported a false positive rate of 15% for resections of presumed hilar cholangiocarcinomas.10 In their series, 132 patients were subjected to a radical excision of a presumed proximal cholangiocarcinoma, and in 20 of these final histology demonstrated benign disease despite pre-operative assessment including ultrasound (US), doppler flow assessment, CT, ERCP, PTC, brush cytology, diagnostic laparoscopy and laparoscopic ultrasound. They also found that intra-operative assessment was accurate in assessing the lesion in only three patients, revealing bile duct stones (2), and Mirizzi syndrome (1). In the remaining 17 cases, operative assessment was unhelpful and possibly misleading. In 11 patients the lesion was thought to be obviously malignant at operation and in the remaining 6 the operating surgeon was unsure whether the lesion being excised was benign or malignant. In all cases, radical bile duct excision was performed and in three patients this was combined with an extended hepatectomy.
Figure 3: Pre-operative ERCP in the same patient showing a hilar biliary stricture
In the assessment of focal solid liver lesions, radiological investigations may not be sufficient to ascertain an accurate diagnosis. A focal liver lesion, especially in the setting of cirrhosis, may represent a HCC, metastatic deposit, haemangioma, focal fatty change, adenoma, focal nodular hyperplasia or a lipoma. Typically, HCC appears hypo-echoic, however, central necrosis and interstitial fibrosis may cause the lesion to appear hyperechoic making it difficult to distinguish HCC from a haemangioma.11 Elevated tumour markers (aFP or CEA) are helpful in distinguishing malignant from benign lesions, but when levels are marginally elevated or low the situation may still be unclear. Ca19-9 can be similarly misleading and is not routinely measured in this unit. It is not the policy in our unit to obtain a pre-operative percutaneous biopsy of focal lesions of the liver, as cases of needle tract seeding following percutaneous liver biopsy have been reported.5 The use of core needle biopsy (18-19 gauge) has been shown to be associated with a 2-5% risk of seeding of malignant cells and a bleeding risk of 1%. The use of fine needle aspiration (20-21 gauge) of such lesions has been advocated in an attempt to reduce these complications. 12,13 It is recognised that it can be difficult to interpret histological specimens obtained from a core biopsy or fine needle aspiration. This is a particular problem when dealing with well-differentiated HCC, which may appear benign. Furthermore, incorrect placement of the needle may lead to inappropriate sites being biopsied. Patients with solid hepatic lesions in this series were investigated using ultrasonography (US), helical CT, laparoscopic US and depending on the clinical setting either a lipiodol scan or a computed tomography angio-portogram (CTAP). If a firm diagnosis of a benign lesion cannot be established with these investigations, resectional surgery is advocated.
Torzilli et al (1999) reported a series of 160 patients with 255 focal solid lesions of the liver who were investigated without the use of pre-operative fine needle biopsy.14 Ultrasonography, spiral CT, visceral angiography, MRI and lipiodol CT scanning were employed. The pre-operative diagnosis in 221 of the 225 (98.2%) lesions was correct and only 4 patients (2.5%) out of 160 had unnecessary surgery. The final= diagnosis in these four patients was inflammatory pseudotumour (2), focal nodular hyperplasia (1) and a dysplastic nodule (1). The authors concluded that fine needle biopsy is unnecessary in the assessment of focal solid liver lesions and that the potential of needle tract seeding with the procedure limited its role to one of obtaining a tissue diagnosis when palliative non-surgical therapy was being planned. Similarly, Levy et al (2001) did not pursue preoperative biopsy in their series of 65 liver resections undertaken in 60 patients suspected of having HCC.12 They confirmed HCC in 63 cases (96.9%). However, in the subset of lesions less than 3cm in diameter with low aFP levels, the false positive rate was much higher (20%). The authors stated that modern radiological techniques remain inaccurate in the assessment of small hepatic nodules and determining whether such a nodule is benign or malignant remains difficult. The benign liver lesion most likely to cause diagnostic confusion is the small focal solid lesion in the setting of cirrhosis with low or marginally elevated tumour marker levels. Any stricture or mass at the hepatic hilum must be considered to be malignant and treated appropriately. Hepatic and bile duct resection is a safe procedure when performed in a specialist centre.15-18 While it is undesirable to subject patients to unnecessary major surgery, the alternative is to deny a subset of patients potentially curative surgery for life-threatening malignancy. This situation is analogous to that seen in distal biliary and pancreatic strictures. Moosa and Gamagami (1995) stated that for pancreatic and distal biliary strictures, “radical resection is the only way to avoid the tragedy of not resecting a potentially curable lesion. Every surgeon should be prepared to accept an occasional benign pathology report.”19
Despite extensive investigation a subset of patients will still be subjected to major hepatic and bile duct resection for what will ultimately be benign disease. The only way to keep this at a low and acceptable level is to have the various investigative modalities and expertise available to the specialist surgeon.
REFERENCES
1. Kosuge T, Yamamoto J, Shimada K, Yamasaki S, Makuuchi M. Improved surgical results for
hilar cholangiocarcinoma with procedures including major hepatic
resection. Ann Surg 1999; 230: 663-71
2. Farley DR, Weaver AL, Nagorney DM. “Natural history” of unresected
cholangiocarcinoma: patient outcome after noncurative intervention. Mayo Clin Proc 1995;
70: 425-29
3. Finch MD, Crosbie JL, Currie E, Garden OJ. An 8-year experience of hepatic resection: indications
and outcome. Br J Surg 1998; 85: 315-19
4. Stewart CJ, Mills PR, Carter R, O’Donohue J, Fullarton G, Imrie CW, Murray WR. Brush cytology
in the assessment of
pancreaticobiliary strictures: a review of 406 cases. J Clin Pathol 2001; 54: 449-55
5. John TG, Garden OJ. Needle track seeding of primary and secondary liver carcinoma after percutaneous
liver biopsy. HPB Surg 1993; 6: 199-203
6. Stewart CJ, Burke GM. Value of p53 immunostaining in
pancreaticobiliary brush cytology specimens. Diagn Cytopathol 2000; 23: 308-13
7. Rumalla A, Baron TH, Leontovich O, Burgart LJ, Yacavone RF, Therneau TM, de Groen PC, Sebo
TJ. Improved diagnostic yield of endoscopic biliary brush cytology by digital image analysis. Mayo Clin Proc
2001; 76: 29-33
8. Hadjis NS, Collier NA, Blumgart LH. Malignant masquerade at the hilum of the liver. Br J Surg 1985;
72: 659-61
9. Wetter LA, Ring EJ, Pellegrini CA, Way LW. Differential diagnosis of sclerosing cholangiocarcinomas of
the common hepatic duct (Klatskin tumors). Am J Surg 1991; 161: 57-62; discussion 62-63
10. Gerhards MF, Vos P, van Gulik TM, Rauws EA, Gouma DJ. Incidence of benign lesions in patients resected
for suspicious hilar obstruction. Br J Surg 2001; 88: 48-51
11. Caturelli E, Pompili M, Bartolucci F, Siena DA, Sperandeo M, Andriulli A, Bisceglia M.
Hemangioma-like lesions in chronic liver disease: diagnostic evaluation in patients.
Radiology 2001; 220: 337-42
12. Levy I, Greig PD, Gallinger S, Langer B, Sherman M. Resection of hepatocellular carcinoma
without preoperative tumor biopsy. Ann Surg 2001; 234: 206-09
13. Huang GT, Sheu JC, Yang PM, Lee HS, Wang TH, Chen DS. Ultrasound-guided cutting biopsy
for the diagnosis of hepatocellular carcinoma--a study based on 420 patients. J Hepatol 1996; 25: 334-38
14. Torzilli G, Minagawa M, Takayama T, Inoue K, Hui AM, Kubota K, Ohtomo K, Makuuchi M. Accurate
preoperative evaluation of liver mass lesions without fine-needle biopsy. Hepatology 1999; 30: 889-93
15. Hanazaki K, Kajikawa S, Shimozawa N, Mihara M, Shimada K, Hiraguri M, Koide N,
Adachi W, Amano J. Survival and recurrence after hepatic resection of 386 consecutive patients with
hepatocellular carcinoma. J Am Coll Surg 2000; 191: 381-88
16. Buell JF, Rosen S, Yoshida A, Labow D, Limsrichamrern S, Cronin DC, Bruce DS, Wen M,
Michelassi F, Millis JM, Posner MC. Hepatic resection: effective treatment for primary and secondary
tumors. Surgery 2000; 128: 686-93
17. Strong RW, Lynch SV, Wall DR, Ong TH. The safety of elective liver resection in a special unit. Aust N Z J Surg
1994; 64: 530-34
18. Pol B, Campan P, Hardwigsen J, Botti G, Pons J, Le Treut YP. Morbidity of major hepatic
resections: a 100-case prospective study. Eur J Surg 1999; 165: 446-53
19. Moossa AR, Gamagami RA. Diagnosis and staging of pancreatic neoplasms. Surg Clin North Am
1995; 75: 871-90
Copyright: 27 November 2002