Pancreas - Carcinoma

CT Assessment of Resectability

Otto van Delden and Robin Smithuis

From the Radiology Department of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands


Pancreatic adenocarcinoma has a poor prognosis.
Complete resection of the tumor is the only curative treatment.
About 10-15% of all patients with a pancreatic carcinoma will finally undergo resection and only in half of these cases the resection will prove to be radical.
In this article we will focus on the CT-findings that are used to select patients with probable resectable tumors.
As the clinical presentation, staging and treatment of other types of pancreatic neoplasms is distinctly different from adenocarcinomas, these are not discussed in this article.

by Otto van Delden and Robin Smithuis


  • Pancreatic carcinoma
Non-resectable pancreatic head tumor obstructing the common bile duct and pancreatic duct. Tumor surrounds the superior mesenteric vein at the junction with the splenic vein. Paraaortic and celiac lymphnodes and a small liver metastasis.

Pancreatic carcinoma is a relatively common tumor with an incidence of 7,6 per 100.000 per year in Western-Europe.
It comprises about 2,5 % of all newly diagnosed tumors and 5% of all cancer.
The majority of pancreatic cancers (85%) are adenocarcinoma of ductal origin.
It is more common in men (men:woman 1,5:1) between the age of 60 and 70 years [1-4].

In spite of the limited tumor size the majority of pancreatic head cancers (80%) are not eligible for resection at the time of diagnosis.
This is due to advanced local tumor extension (40%) or the presence of distant metastatic disease (40%) mostly due to liver metastases of para-aortic lymphadenopathy.

  • Treatment


The only curative treatment option is surgical resection.
Out of every hundred patients with pancreatic carcinoma only 20 patients will be sceduled for explorative laparotomy.
Out of these 20 only about 13-14 patients will undergo resection of the tumor, but only half of these resections will finally prove to be radical at pathologic examination of the resected specimen..
The resection consists of a partial pancreaticoduodenectomy according to Whipple or the modern variant, the so-called 'pylorus-preserving' pancreaticoduodenectomy.

When the tumor proves to be unresectable during exploratory laparotomy, a so-called 'double bypass' (gastro-enterostomy en hepaticojejunostomy) is usually performed for palliative reasons.
When curative resection is not considered an option, based on preoperative imaging and cytology or histology, palliation consists of endoscopic or percutaneous biliary stenting and celiac plexus block for relief of pain.
Patients with a relatively short life expectancy (e.g. patients with extensive hepatic metastases), are probably best served by palliation by means of endoscopic bile duct stenting.
In patients with a longer life expectancy (e.g. patients with a small, but locally unresectable tumor without distant metastatic disease), a double bypass is generally also considered acceptable palliation.

  • Pancreatic cancer distribution

Most pancreatic cancers occur in the head of the pancreas (75%).
A minority is found in the body (15%) and tail (10%).
At the time of diagnosis a pancreatic head carcinoma is usually a little larger than 3 cm.
When tumors of the pancreatic body and tail are diagnosed, they are usually much larger, because they present late with aspecific symptoms.
These tumors are usually irresectable.
Tumors originating in the distal common bile duct or ampulla may also grow into the pancreatic head and together with pancreatic head carcinoma these tumors are often grouped together under the name periampullary tumors.
This has some practical value as diagnostic imaging, staging and treatment of all these periampullary tumors is the same.

Imaging Work-up

  • Ultrasound
LEFT: small liver lesion with typical appearance of metastasisRIGHT: Hypoechoic pancreatic head tumor obstructing pancreatic duct.


The most striking clinical symptom leading to diagnostic imaging is painless obstructive jaundice, which is caused by compression or ingrowth of the distal common bile duct.
US is the first line imaging test for the evaluation of these patients.
US can determine the level of obstruction in most cases (sensitivity >90%).
In patients with a pancreatic head tumor, typically dilatation of the common bile duct and pancreatic duct (double duct sign) is seen, which is very suggestive for a mass in the pancreatic head, even in the absence of a visible mass.
The tumor itself usually presents as a hypoechoic mass (figure).

In the detection of pancreatic cancers US has an overall sensitivity of 75% and a specificity of 75%.
However in many cases US will suffice as the only imaging test for diagnosis and staging.
This is particularely true in patients with tumors > 3 cm and liver metastases > 2 cm.
The overall sensitivity and specificity for determining resectability of all pancreatic carcinomas however is only 63% and 83% respectively.


If the cause of a distal bile duct obstruction is not revealed by US and there is a high suspicion for a pancreatic or periampullary tumor, the next diagnostic test is CT.
ERCP (or MRCP) is only the next step when there is a high suspicion of bile duct stones.
Whenever a pancreatic tumor is detected with US and no definite signs of unresectability are found, the next step is CT.
CT should be done before ERCP and insertion of an endoprosthesis, because artifacts and post-ERCP pancreatitis may hamper the diagnostic accuracy of CT.

As pancreatic carcinoma is a hypovascular tumor, it presents as a hypodens mass on a CECT.
The mass is usually ill-defined. In 10 - 15% the tumor is isodens and therefore may be difficult to detect.
Tumors smaller than 2 cm. may also be difficult to detect on CECT.
In these cases

indirect signs may be helpful such as the presence of the double duct sign, atrophy of the pancreatic tail, or fullness of the pancreatic head (loss of the lobular appearance of the pancreatic parenchyma).

  • mrcp
Double duct sign indicating pancreatic head carcinoma


CT and MRI both have a higher sensitivity than ultrasound for the detection of small ( MRI-sequences should involve at least T2W-images en dynamic T1W-images after intravenous administration of gadolinium.
MRCP is also very sensitive for detecting a periampullary mass, but offers no significant additional staging information [9].


Many patients in whom a pancreatic head tumor is detected by ultrasound still undergo ERCP.
Although ERCP has a high sensitivity for detecting pancreatic head tumors, it is nowadays no longer indicated because the diagnosis can usually be made with non-invasive tests.
ERCP offers no usefull tumor staging information.
It is doubtfull whether pre-operative bile duct drainage by ERCP is beneficial for the patient [12].
Pre-operative biliary drainage may potentially even increase the risk for post-operative infectious complications.

  • Endoscopic ultrasound
Endoscopic US of small pancreatic head tumor obstructing the common bile duct

Endoscopic ultrasound

Endoscopic ultrasound is generally accepted as the most sensitive imaging test for the detection of small pancreatic head tumors, particularly when smaller than 2 cm [10].
These pancreatic head tumors can be missed even on a technically excellent CT and therefore a 'negative' CT-scan in a patient with a strong suspicion for pancreatic head cancer requires additional imaging with endoscopic ultrasound.
Unfortunately, there are only a few centers in The Netherlands with sufficient experience in this operator-dependent-technique.
Endoscopic ultrasound has also been used for local tumor staging, but is currently not frequently used as such in the Netherlands.

  • Diagnostic laparoscopy
LEFT: laparoscopic biopsy of superficial liver metastasis.RIGHT: peritoneal metastases.

Diagnostic laparoscopy

Diagnostic laparoscopy, sometimes complemented by laparoscopic ultrasound has been advocated by some as a staging tool.
Laparoscopy is much more sensitive than any other technique for the detection of peritoneal implants and superficial liver metastases.
Local staging is also feasible by laparoscopic ultrasound.
However, a large series has shown, that the yield of laparoscopy after CT is not high enough to justify using this technique routinely [19,20].
It may be usefull in selected cases where there is doubt about resectability or when suspected metastatic disease cannot be proven otherwise.

CT protocol

  • CT protocol

Water should be used as oral contrast material.
A precontrast scan of the pancreas can be performed to look for calcifications within the pancreas, which may indicate the presence of a focal pancreatitis.
On a contrast enhanced CT (CECT) these calcifications may not be appreciated as they will be confused with enhancement.

Depending on the type of multidetector CT, 120 - 150 ml contrast is given at an injection rate of 3-5 ml/s.
Slice thickness depends on the type of scanner that is used, but should be preferentially 2-3 mm or less.
An early arterial phase-scan (delay 20 sec) does not add significant information on the staging of the pancreastumor, since there is not enough contrast in the pancreas [8].
Only if the surgeons want to get optimal pre-operative 3D-information on the anatomy of the mesenteric arteries this phase is included.

Early portal phase

The early-portal phase is also called the pancreatic phase.
It has a scan-delay of 40-50 sec.
This is the most important phase for detecting and staging a pancreatic tumor.
At that moment the normal pancreatic parenchyma will enhance optimally, because it gets all of its bloodsupply through the arterial and capillary system.
In this phase there is optimal attenuation difference between the hypodense tumor and the normal enhancing pancreatic parenchyma.
This phase helps in the differentiation of liverlesions and usually the mesenteric arteries and veins are well opacified [7].

Late portal phase

The 'late portal' or hepatic phase has a scan-delay of 70-80 sec.
At that moment the normal liverparenchyma will enhance optimally, because normal livercells get 80% of their bloodsupply through the portal venous system.
Liver metastases do not get their bloodsupply from the portal venous system and will be seen in this phase as hypovascular or hypodense lesions.
This phase is performed for the overall assessment of the abdomen to look for liver metastases, lymphnodes and peritoneal implants.
This phase is also helpfull for local staging of the tumor and detection of venous ingrowth.

Resectable or Irresectable that's the question

It is of the utmost importance to stage a pancreatic tumor correctly as the clinical consequences of this are enormous.
Overstaging will lead to undertreatment if a laparotomy is not performed in a patient with a potentially resectable tumor.
Understaging will lead to an unnecessary laparotomy with all the associated risks.
To withhold the chance for curative resection from as few patients as possible, it is important to determine unresectability with a very high specificity, even if this means a lower sensitivity.
Some patients will therefore get the benefit of the doubt and undergo a negative exploratory laparotomy.
In some series the unresectability rate at laparotomy may be as high as 30%.

  • Local Tumorspread
The pancreatic head is surrounded by the portal vein, SMA, mesenteric root, duodenum, IVC and aorta

Local Tumorspread

Since the pancreas has no capsule, pancreatic tumor will easily spread into adjacent structures (figure).
Because the confluens of the portal and superior mesenteric vein is in direct continuity with the pancreatic head, ingrowth into this vessel will often be the first sign of tumor extension outside the pancreas.
Ingrowth into the celiac axis or superior mesenteric artery is always considered a criterium for unresectability.
Although partial resection of the portal vein or superior mesenteric vein are technically possible and are being performed, ingrowth into these vessels is considered a criterium for unresectability by most oncologic surgeons in the Netherlands.
Some centers in the US and Japan will resect part of the portal vein in case of tumor ingrowth.

  • resectable


Although associated with a worse prognosis, the presence of peripancreatic lymphnode metastases does not constitute a definite contraindication for resection.
Limited ingrowth into the peripancreatic fat, ingrowth into the duodenum or the gastroduodenal artery does not render a tumor unresectable as this vessel and the duodenum can be resected en-bloc with the tumor.
When there is contiguity between the tumor and the portal or superior mesenteric vein, but the vessel is surrounded by tumor for less than half the circumference (

  • vessel <180?

On the left two cases of pancreatic tumors with tumor-vessel contiguity These patients generally will be given the benefit of the doubt and will be sceduled for operation.

  • Peritoneal metastases
Peritoneal metastases (arrowheads) in a patient with a pancreatic tumor

Not resectable

Tumor ingrowth into stomach, colon, mesocolon, inferior vena cava or aorta constitute definite criteria for unresectability.
Also the presence of hepatic metastases, peritoneal metastases or para-aortic lymfnode metastases is an absolute sign of unresectability.
Mesenteric lymph node metastases, not immediately adjacent to the pancreas usually also indicate unresectability.
Liver metastases and distant lymph node metastases should allways be proven by means of cytologic or histologic biopsy before refraining from exploratory laparotomy.
Ingrowth into the celiac axis, hepatic artery or superior mesenteric artery also preclude resection.

  • Vessel ingrowth 2

Vessel ingrowth

When a fatplane or normal pancreatic parenchyma is visible between the tumor and the vessel, the tumor is usually locally resectable.
When there is tumor-vessel contiguity, but the vessel is surrounded by tumor for less than half the circumference ( This group of patients will usually get the benefit of the doubt and undergo exploratory laparotomy.

  • Tumor in direct contiguity with the confluens
Tumor in direct contiguity with the confluens >180?

On the left a pancreatic tumor in direct contiguity with the confluens of the portal and superior mesenteric vein.
The tumor surrounds the confluens for more than half the cirumference (>180?).
This tumor was regarded as unresectable.

  • More than 180?
LEFT: Irresectable tumor surrounding the AMS >180?.RIGHT: Irresectable tumor totaly surrounding the AMS.

When the tumor surrounds the vessel for more than half the cirumference (>180?), the tumor will nearly allways be unresectable. Most surgeons will consider this a solid criterium for unresectability [13-16]. Flattening of the vessel or irregular vascular contours are also indicative of ingrowth. When the tumor surrounds the portal vein or superior mesenteric vein completely (360?) or occludes the vessel, the tumor is allways unresectable [13-16].

  • teardrop

Other criteria for vascular ingrowth have been described, such as dilatation of the gastrocolic trunk (a sidebranch of the superior mesenteric vein) and the 'mesenteric teardrop sign'.
These signs are not more sensitive or specific than the abovementioned criteria and therefore probably do not have much additional value as other criteria for unresectability are usually also present in these cases[17,18].

  • tumor thrombus
Tumor thrombus is present in the lumen of the superior mesenteric vein

On the left a tumor thrombus is present in the lumen of the superior mesenteric vein.
This is also a sign of unresectability.

  • encasement of the hepatic artery

On the left a pancreatic carcinoma with encasement of the hepatic artery.
The pancreatic duct is obstructed with subsequent atrophy of pancreatic tail.
This tumor is not resectable.

Differential diagnosis

The differential diagnosis of a pancreatic head tumor includes carcinoma, focal pancreatitis, lymphoma and metastasic disease.
Sometimes it is difficult to differentiate between a pancreatic head tumor and focal pancreatitis in the pancreatic head.
There are no solid imaging criteria to decide this with certainty in all cases.
Although a double duct lesion may be seen in cases of pancreatitis, this finding should always lead to a strong suspicion for pancreatic carcinoma.
Imaging guided biopsy has a limited value in these cases because false-negative results frequently occur and can therefore not be used to rule out cancer.
PET has currently not proven it?s use in the differentiation as tumor and inflammation may both show increased uptake of the radiofarmacon. In cases where differentiation is impossible, laparotomy or imaging follow-up may be performed, depending on the specific clinical circumstances.
Most large surgical series show, that in about 5% of the patients, who undergo resection for suspicion of pancreatic head cancer only pancreatitis is eventually found in the resected specimen. [22].

  1. Ahlgren JD. Epidemiology and risk factors in pancreatic cancer. Semin Oncol 1996;23:241-50.

  2. Rosewicz S, Wiedenmann B. Pancreatic carcinoma. Lancet 1997;349:485-9

  3. Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for 100,313 patients diagnosed from 1985-1995, using the National Cancer Database. J Am Coll Surg 1999;189:1-7.

  4. Di Magno EP, Reber HA, Tempero MA. AGA technical review on the epidemiology, diagnosis, and treatment of pancreatic ductal adenocarcinoma. American Gastroenterological Association. Gastroenterology 1999;117:1464-84

  5. Karlson BM, Ekbom A, Lindgren PG, Kalskogg V, Rastad J. Abdominal US for diagnosis of pancreatic tumor: prospective cohort analysis. Radiology 1999;213:107-11.

  6. Sheridan MB, Ward J, Guthrie JA, Spencer JA, Craven CM, Wilson D, et al. Dynamic contrast-enhanced MR imaging and dual-phase helical CT in the preoperative assessment of suspected pancreatic cancer: a comparative study with receiver operating characteristic analysis. AJR Am J Roentgenol 1999;173:583-90

  7. Boland GW, O?Malley ME, Saez M, Fernandez-del-Castillo C, Warshaw AL, Mueller PR. Pancreatic-phase versus portal vein-phase helical CT of the pancreas: optimal temporal window for evaluation of pancreatic adenocarcinoma. AJR Am J Roentgenol 1999;172:605-8.

  8. Graf O, Boland GW, Warshaw AL, Fernandez-del-Castillo C, Hahn PF, Mueller PR. Arterial versus portal venous helical CT for revealing pancreatic adenocarcinoma: conspicuity of tumor and critical vascular anatomy. AJR Am J Roengtenol 1997;169:119-23

  9. Adamek HE, Albert J, Breer H, Weitz M, Schilling D, Riemann JF. Pancreatic cancer detection with magnetic resonance cholangiopancreatography and endoscopic retrograde cholangiopancreatography: a prospective controlled study. Lancet 2000;356:190-3.

  10. Wiersema MJ. Accuracy of endoscopic ultrasound in diagnosing and staging pancreatic carcinoma. Pancreatology 2001;1:625-35.

  11. Rose DM, Delbeke D, Beauchamp RD, Chapman WC, Sandler MP, Sharp KW, et al. 18F-fluorodeoxyglucose-positron emission tomography in the management of patients with suspected pancreatic cancer. Ann Surg 1999;229:729-38

  12. Sewnath ME, Karsten TM, Prins MH, Rauws EJ, Obertop H, Gouma DJ. A meta-analysis on the efficacy of pre-operative biliary drainage for tumors causing obstructive jaundice. Ann Surg 2002;236:17-27.

  13. Lu DS, Reber HA, Krasny RM, Kadell BM, Sayre J. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic-phase, thin-section helical CT. AJR Am J Roentgenol 1997;168:1439-43.

  14. O?Malley ME, Boland GW, Wood BJ, Fernandez-del-Castillo C, Warshaw AL, Mueller PR. Adenocarcinoma of the head of the pancreas: determination of surgical unresectability with thin-section pancreatic-phase helical CT. AJR Am J Roentgenol 1999;173:1513-8.

  15. Phoa SS, Reeders JW, Rauws EA, Wit LT de, Gouma DJ, Lam?ris JS. Spiral computed tomography for preoperative staging of potentially resectable carcinoma of the pancreatic head. Br J Surg 1999;86:789-94.

  16. Phoa SS, Reeders JW, Stoker J, Rauws EA, Gouma DJ, Lam?ris JS. CT criteria for venous invasion in patients with pancreatic head carcinoma. Br J Radiol 2000;73:1159-64.

  17. Hough TJ, Raptopoulos V, Siewert B, Mattews JB. Teardrop superior mesenteric vein: CT sign for unresectable carcinoma of the pancreas. AJR Am J Roentgenol 1999;173:1509-12.

  18. Hommeyer SC, Freeny PC, Crabso LG. Carcinoma of the head of the pancreas: evaluation of the pancreaticoduodenal veins with dynamic CT-potential for improved accuracy in staging. Radiology 1995;196:233-8.

  19. Nieveen van Dijkum EJ, Romijn MG, Terwee CB, Wit LT de, Meulen JH van der, Rauws EA, et al. A multicenter trial of laparoscopic staging and subsequent palliation in patients with periampullary carcinoma. Ann Surg 2002

  20. Barreiro CJ, Lillemoe KD, Koniaris LG, Sohn TA, Yeo CJ, Coleman J, et al. Diagnostic laparoscopy for periampullary and pancreatic cancer: what is the true benefit? J Gastrointest Surg 2002;6:75-81.

  21. Tilleman EH, Phoa SS, Delden OM van, Rauws EA, Gulik TM van, Lam?ris JS, Gouma DJ. Reinterpretation of radiological imaging in patients referred to a teriary care referral center with a suspected hepatobiliary or pancreatic malignancy; impact on treatment strategy. Eur Radiol 2002

  22. Gulik TM van, Moojen TM, Geenen R van, Rauws EA, Obertop H, Gouma DJ. Differential diagnosis of focal pancreatitis and pancreatic cancer. Ann Oncol 1999;10(Suppl 4):85-8.

  23. Diederichs CG, Staib L, Vogel J, Glasbrenner B, Glatting G, Bramts HJ, et al. Values and limitations of 18F-fluorodeoxyglucose-positron-emission tomography with preoperative evaluation of patients with pancreatic masses. Pancreas 2000;20:109-16.

  24. Fr?hlich A, Diederichs CG, Staib L, Vogel J, Beger HG, Reske SN. Detection of liver metastases from pancreatic cancer using FDG PET. J Nucl Med 1999;40:250-5.

  25. Gouma DJ, Tilleman EH, Benraadt J, Bossuyt PM. Diagnostiek en behandeling van het distale galweg- of pancreascarcinoom; richtlijn van de Integrale Kankercentra Amsterdam en Stedendriehoek Twente. Ned Tijdschr Geneeskd