Successful resection of pancreatic head cancer in a patient with circumportal pancreas: a case report with technical consideration

Case report

A 76-year-old woman with generalized pruritus visited the general practitioner. Her blood test showed elevated liver enzymes and bilirubin levels, and an abdominal ultrasound test revealed dilation of the intrahepatic bile ducts. She was referred to our hospital for further examination.

Dynamic computed tomography (CT) revealed an unenhanced mass at the head of the pancreas accompanied by dilation of the main pancreatic duct (MPD) and common bile duct (CBD), and most importantly, it revealed that the portal vein (PV) was surrounded by the pancreas. The dilated MPD was located behind the PV (Figure 1). Cytological and histological examinations were attempted by endoscopic retrograde cholangiopancreatography, and the cytology of the pancreatic juice was assessed as Class IV. Preoperative endoscopic retrograde biliary drainage was carried out.

Figure 1:

CT scan showed an unenhanced mass at the head of the pancreas (dotted circle) and dilated MPD and CBD, and it also showed that the PV was surrounded by two parts of the pancreas: the dorsal part and the ventral part.

MPD was located in the dorsal part of the pancreas. SMA, superior mesenteric artery; SpV, splenic vein.

A preoperative diagnosis was pancreatic head cancer with circumportal pancreas (CP). The patient underwent subtotal stomach-preserving pancreaticoduodenectomy (SSpPD). After transecting the pancreas above the superior mesenteric vein (SMV)/PV, we saw the dorsal part of the pancreas, which was the uncinated process communicating with the body of the pancreas behind the PV. After cutting the dorsal part of the pancreas, we saw two cutting planes of the pancreas located both above and below the PV (Figure 2A). We found a dilated MPD in the dorsal side of the cutting plane, but suturing the MPD and pancreas parenchyma with the gut in a deep operating field required great effort, so we mobilized the dorsal part of the pancreas from the PV and pulled through it above the PV to anastomose the MPD and pancreas parenchyma with the gut more easily (Figure 2B). The dorsal cutting plane was anastomosed with the jejunum in the way of modified Blumgart with a lost stent tube. No pancreatic duct was apparent at the ventral cutting plane, so we closed the plane with 3-0 Prolene continuous suturing and attached the ventral stump with the jejunum to prevent postoperative pancreatic fistula (POPF) (Figure 2C).

Figure 2:

The operative findings.

(A) After removing the head of the pancreas, we saw two cutting planes of the pancreas: the ventral part (VP) above the PV and the dorsal part (DP) behind the PV. The dilated MPD was located in the DP. (B) The DP was mobilized from the PV and then was located above the PV. A pancreatic tube was inserted in the MPD. (C) The DP was anastomosed with the jejunum in the way of modified Blumgart with a lost stent tube. The cutting plane of the VP was sutured and then attached to the jejunum to prevent POPF.

Histological examination revealed the moderately differentiated adenocarcinoma with lymph node metastasis (Figure 3). Although it took time to reach satisfactory oral feeding because of protracted nausea, the postoperative course was otherwise uneventful without any severe complications, such as POPF or postpancreatectomy hemorrhage, and she was discharged on postoperative day 36.

Discussion

CP is a rare pancreatic anomaly with a prevalence of 1.1% [1] to 2.5% [2]. CP itself is usually asymptomatic, so it is usually found only after performing abdominal CT. The pathogenesis of CP remains unknown, but the malformation of the pancreatic primordia or PV has been suggested [3].

Figure 3:

Resected specimen.

The forceps indicates the location of the PV and shows that the PV is encased between the VP and the DP of the pancreas.

CP was classified by Karasaki et al. [1] and otherwise by Joseph et al. [3]. Karasaki et al. classified the anomaly into (A) suprasplenic, (B) infrasplenic, and (C) mixed types based on the relationship between the splenic vein and the retroportal pancreas. Joseph et al. classified it into type 1 [retroportal MPD (RMPD)], type 2 [RMPD with pancreas divisum], and type 3 [anteportal MPD (AMPD)] based on the location of the MPD. Our case is a “suprasplenic” type in accordance with Karasaki’s classification and an “RMPD” type in accordance with Joseph’s classification. The prevalence of this type of CP is about 30% according to the report by Harnoss et al. [4].

CP casts problems when the surgical procedure for the pancreas is planned. When we cut the pancreas at the right side of the SMV, we encounter both the ventral pancreas and the dorsal pancreas. Therefore, anastomosing the pancreas with the gut may be challenging and the frequency of POPF may increase (it has been reported to be as high as about 40% [4]). To minimize the risk of POPF, several technical modifications have been suggested. One technical method is cutting the pancreas at the left side of the fusion part. In this way, we can get only one cutting plane of the pancreas that should be anastomosed with the gut [5]. However, this procedure has a disadvantage in that additional dissection and resection are necessary, and as a result, it might cause secondary damages to the pancreas or other tissues. It also has another disadvantage in that the resection of larger volumes of the pancreas might cause deterioration of the remnant pancreatic function. Considering these disadvantages, we did not choose this method.

Another technical approach is suturing or ligating the cutting plane without MPD (referred as a nondominant cutting plane) [6, 7]. Ohtsuka et al. [6] reported the method of suturing a nondominant cutting plane by a linear stapler. That case developed POPF unfortunately, but it was unclear whether the cause of POPF lay in the pancreaticojejunostomy or in the sutured site. Our case is similar to Ohtsuka’s report, except that suturing the nondominant cutting plane was performed by a hand-sewn technique and that the stump was attached to the jejunum, resulting in the successful surgical outcome without POPF or postpacreatectomy hemorrhage.

As far as we have investigated, we found 29 cases of pancreaticoduodenectomy for CP (Table 1). POPF occurred in seven cases; in five cases, the pancreas was transected above and below the SMV/PV, and in two cases, the pancreas was transected at the left side of the fusion part. Suturing or ligating the nondominant cutting plane was performed in three cases, and none of them developed POPF. Transecting the nondominant cutting plane by staplers was performed in five cases, and three of them developed POPF. Suturing or ligating the nondominant cutting plane may have an advantage in that we can check directly any pancreatic ducts that should be anastomosed or closed. However, the surgical method should be chosen depending on the type of anomaly.

In conclusion, CP is a rare, asymptomatic anomaly that matters in pancreaticoduodenectomy because of the higher risk of POPF. The first key to prevent POPF is that surgeons should keep this anomaly in mind and perform a careful preoperative image assessment. Although the best management remains unknown and more cases need to be investigated, the method of suturing a nondominant cutting plane and attaching it to the jejunum, as shown in the current case, may be reasonable.