Short-term outcomes of open liver lobectomy using an Endo GIA stapler: case series of five cats

Abstract

Case series summary

In this retrospective case series, open liver lobectomy using an Endo GIA stapler was evaluated in five client-owned cats. All procedures were performed via ventral midline celiotomy using the Endo GIA Ultra Universal Stapler with Curved Tip Articulating Vascular/Medium Tri-Staple cartridges. Three cats underwent total liver lobectomy of the left lateral lobe, and the remaining two underwent multiple lobectomies involving one or more divisions (left division, n = 1; left and central divisions, n = 1). In three cats, the lesions were adjacent to major hilar vascular and biliary structures. The median lobe resection and total surgical times were 21 and 51 mins, respectively. The median postoperative hospitalisation was 3 days. Intraoperative complications occurred in two cats (hypotension) and postoperative complications occurred in all cats, including mild anaemia (n = 5), ascites (n = 4), wound complication (n = 1) and transient hyperbilirubinemia (n = 1). Although one cat required an intraoperative transfusion for massive bleeding, none of the cats required revision surgery. All cats survived until discharge and were alive and asymptomatic at suture removal approximately 2 weeks postoperatively.

Relevance and novel information

To our knowledge, this is the first detailed report to evaluate the short-term outcomes of open liver lobectomy using the Endo GIA stapler in cats. Consistent with previous canine studies, no mortality or major complications occurred during the perioperative period, and the total surgical time was relatively short. These findings suggest that the Endo GIA stapler is a feasible option for open liver lobectomy in cats, particularly for lesions located in the left division, including selected lesions adjacent to the hilus. However, further studies are warranted to confirm its clinical relevance, define appropriate case-selection criteria based on lesion characteristics and compare it with conventional techniques.

Keywords

Complication Endo GIA stapler open liver lobectomy surgical procedure

Introduction

In recent years, several non-hand-sewn techniques have been reported for open liver lobectomy in dogs and cats.^1
–5 The feasibility of endoscopic gastrointestinal anastomosis (Endo GIA) staplers has also been evaluated for laparoscopic liver lobectomy in dogs, demonstrating their potential to facilitate minimally invasive surgery, particularly in small- to medium-sized patients.⁶

Endo GIA staplers allow for 360° axial rotation and tip articulation.⁷ These features might improve access to the hepatic hilus, even in cats with a small abdominal cavity. However, its application in cats has been reported only once, in a case report that did not focus on the surgical procedure or associated complications.⁸

To the best of our knowledge, no detailed study has evaluated the use of Endo GIA staplers for open liver lobectomies in cats. Therefore, the objective of this study was to describe the surgical procedure and short-term outcomes of open liver lobectomy using the Endo GIA stapler in cats.

Case series description

Materials and methods

Case selection

A retrospective observational case series was conducted at a single referral institution between 1 August 2023 and 31 October 2025. All the cases involved client-owned cats that underwent open liver lobectomy using an Endo GIA stapler.

Medical records review

Data including age, breed, body weight on the day of surgery, sex and clinical signs at the time of initial presentation were retrieved from medical records. Diagnostic and surgical procedures were reviewed from the records, as well as available intraoperative photographs and video documentation. Diagnostic evaluations included haematology, urine analysis, radiography, ultrasonography, CT and histopathology.

Outcome measures

The outcome variables included lobe resection time, total surgical time, duration of intensive care unit (ICU) stay, length of postoperative hospitalisation, change in packed cell volume (PCV) during hospitalisation, incidence and type of intraoperative and postoperative complications, timing of resumption of voluntary food intake and short-term survival. Lobe resection time was defined as the interval between skin incision and completion of the lobe resection, and the total surgical time was defined as the interval between skin incision and closure. Short-term survival was defined as survival until suture removal, approximately 14 days after surgery. Any surgical complications occurring until suture removal were classified according to the Veterinary Cooperative Oncology Group–Common Terminology Criteria for Adverse Events version 2.0.⁹

Anaesthesia and patient preparation

General anaesthesia was induced with propofol and maintained with isoflurane in oxygen. For analgesia, ropivacaine hydrochloride (2.0 mg/kg) and medetomidine hydrochloride (2.0 μg/kg) were administered for preoperative epidural anaesthesia. Fentanyl citrate (1–10 µg/kg/h) and ketamine hydrochloride (0.2–2 mg/kg/h) were administered as a continuous rate infusion (CRI) perioperatively. Perioperative antibiotics (cefmetazole sodium or ampicillin sodium) were administered at standard doses, with the choice determined according to the clinician’s preference and drug availability. The cats were positioned in dorsal recumbency, and the ventral abdomen was clipped and aseptically prepared using a chlorhexidine–alcohol solution. Portal pressure was measured via a 24 G over-the-needle catheter placed in a distal jejunal vein and connected through a fluid-filled extension set to a pressure transducer and monitor. The transducer was zeroed at the level of the right atrium, and pressure was recorded in mmHg.

Surgical procedure

All cats underwent a ventral midline celiotomy. The affected hepatic lobe was identified visually, and the adhesions were bluntly or sharply dissected when present. The triangular and coronary ligaments were transected using electrosurgical instruments if necessary. A vascular cartridge (Endo GIA Curved Tip Articulating Vascular/Medium Reload with Tri-Staple Technology; Medtronic) was loaded into a stapler (Endo GIA Ultra Universal Stapler; Medtronic). This cartridge provides six staggered rows of staples of varying heights (3.0, 2.5 and 2.0 mm). An elastic silicone guide, included in the stapler kit, was placed beneath the intended transection line on the hilus of the lobe (Figure 1). The jaw tip of the stapler was connected to the guide and introduced. After confirming that the position was appropriate, the jaws were closed and held for 10–30 s before firing. The cartridge was then fired to complete the transection. Based on intraoperative assessment by the attending surgeon, ligation clips (CLIP 9 VITALITEC; Peters Surgical) were applied for focal bleeding from identifiable vessels at the staple line, whereas oxidised regenerated cellulose (SURGICEL FIBRILLAR; Ethicon) was applied for diffuse parenchymal oozing to assist with haemostasis. After all procedures were completed and haemostasis was confirmed, the celiotomy incision was routinely closed in three layers.

Figure 1

Application of the elastic silicone guide. The guide was placed beneath the intended transection line at the hepatic hilus of the affected lobe

Postoperative care

All cats were admitted overnight to the ICU. The following day, the absence of re-bleeding was confirmed before transfer to the surgical ward based on a combination of abdominal ultrasonography (including assessment of the amount of abdominal effusion) and serial PCV measurements, together with overall clinical status. These assessments were continued during hospitalisation.

At the clinician’s discretion, additional postoperative analgesia was provided using buprenorphine hydrochloride (10–20 µg/kg IV q8h). To minimise excitement in cats, dexmedetomidine hydrochloride (0.5−1.0 μg/kg/h) was administered as a CRI for additional sedation and analgesia.

Results

Case characteristics

Five client-owned cats were included in the study (Table 1). The median age was 13.4 years (range 2.8–15.1) and the median body weight on the day of surgery was 3.94 kg (range 2.81–5.50). Scottish Fold was the most common breed (n = 3, 60.0%), followed by American Shorthair and Maine Coon (n = 1 each). One cat presented with intermittent vomiting and diarrhoea at the time of diagnosis. The other cats did not exhibit any clinical signs at the time of initial presentation.

Table 1

Signalment, haematological analyses and CT findings in five cats who underwent open liver lobectomy with the Endo GIA stapler

	Signalment				Haematological analyses										CT findings
Case	Breed	Sex	Age (years)	Weight (kg)	PCV(%)	TP(g/dl)	ALB(g/dl)	ALT(U/l)	AST(U/l)	ALP(U/l)	GGT(U/l)	T-BIL(mg/dl)	PT(s)	APTT(s)	Liver lobe	Appearance	Proximity to hilus*	Longest diameter (mm)	Others
1	American Shorthair	CM	15.1	2.96	38	8.3	2.9	973	119	160	5	0.6	10.5	16.8	LL	Solid	No	71	Solid nodules in LM and Q Small amount of ascites
2	Maine Coon	SF	2.8	5.50	47	7.2	3.2	50	24	26	2	0.2	7.2	56.4	LL	Cystic	No	109	–
2	Maine Coon	SF	2.8	5.50	47	7.2	3.2	50	24	26	2	0.2	7.2	56.4	RM and Q	Cystic	Yes	134	–
3	Scottish Fold	SF	14.4	3.94	34	7.5	2.5	55	26	40	6	0.1	6.9	19.7	LM	Solid	Yes	76	Hepatic lymph node enlargement
4	Scottish Fold	SF	13.4	2.81	31	7.1	2.9	32	15	30	4	0.1	6.3	47.5	LL	Solid	No	25	–
5	Scottish Fold	CM	12.0	4.60	32	8.0	3.4	40	24	40	4	0.3	6.0	24.4	LL	Solid	Yes	45	Multiple diffuse cystic nodules

Proximity to hilus indicates adjacency to major hilar vascular and biliary structures on CT

ALB = albumin (reference internal [RI] 2.3–3.5); ALP = alkaline phosphatase (RI <58); ALT = alanine aminotransferase (RI 22–84); APTT = activated partial thromboplastin time (RI 12.0–51.3); AST = aspartate aminotransferase (RI 18–51); CM = castrated male; GGT = gamma-glutamyl transpeptidase (RI 1–10); LL = left lateral lobe; LM = left medial lobe; PCV = packed cell volume (RI 30–52); PT = prothrombin time (RI 5.0–8.1); Q = quadrate lobe; RI = reference interval; RM = right medial lobe; SF = spayed female; T-BIL = total bilirubin (RI = 0.1–0.4); TP = total protein (RI 5.7–7.8)

Diagnostic procedures

All the cats underwent haematology, radiography, ultrasonography and CT; one cat underwent CT at another institution. The results of the haematological analyses are summarised in Table 1. Upon radiography and ultrasonography, three cats were diagnosed with urolithiasis and one with cholelithiasis. The details of CT findings are also shown in Table 1. The six main hepatic lesions were located in the left lateral lobe (n = 4) and left medial lobe (n = 1), and one lesion involved both the right medial and quadrate lobes (n = 1). Of these, three lesions were adjacent to major hilar vascular and biliary structures (Figure 2). The median longest lesion diameter was 73.5 mm (range 25–134). Urine analysis was performed in three cats and revealed a mild decrease in urine specific gravity in all three.

Figure 2

Postcontrast CT images in five cats. Yellow arrows indicate the portal vein at the hepatic hilus. Images are displayed in a soft tissue window (window width 350 Hounsfield units [HU]; window level 40 HU) with a slice thickness of 0.5 mm. (a) Dorsal plane in case 1; (b) transverse plane of the lesion of the right medial and quadrate lobes in case 2; (c) sagittal plane of the lesion of the left lateral lobe in case 2; (d) dorsal plane in case 3; (e) dorsal plane in case 4; (f) dorsal plane in case 5

Perioperative treatment and surgery

Until approximately 24–48 h postoperatively, fentanyl citrate (1 µg/kg/h) and ketamine hydrochloride (0.1–0.2 mg/kg/h) were administered as a CRI in two cats, dexmedetomidine hydrochloride (0.5–1.0 μg/kg/h) was administered as a CRI in one cat, and buprenorphine hydrochloride (10–20 µg/kg IV q8h) was administered in four cats.

The surgical details are summarised in Table 2. Three cats underwent total liver lobectomy of the left lateral lobe, and the remaining two underwent multiple lobectomies involving one or more divisions (left division, n = 1; left and central division, n = 1) (Figure 3).

Table 2

Specifics of surgical details and histopathology in five cats who underwent open liver lobectomy with the Endo GIA stapler

	Surgical detail							Histopathology
Case	Lobectomy site(s)	Size of Endo GIA Cartridge (mm)	Time of incision to lobe resection(mins)	Time of incision to closure(mins)	Additional procedure	Intraoperative complications	Ancillary haemostasis	Diagnosis	Margin of excision
1	LL	60	6	51	Liver nodules biopsy	Grade 1 hypotension	Ligation clips	Hepatocellular carcinoma with intrahepatic metastasis	NA
2	LL+LM+Q+RM	60	65	93	–	Grade 4 hypotension	–	Hepatic cyst	NA
3	LL+LM	60	22	58	Lymphadenectomy Oesophageal tube placement	–	Oxidised regenerated cellulose	Biliary adenoma	Complete
4	LL	60	21	47	–	–	Ligation clips	Biliary adenoma and carcinoma	Complete
5	LL	45	15	44	–	–	–	Biliary adenoma	Complete

LL = left lateral lobe; LM = left medial lobe; NA = not applicable or not available; Q = quadrate lobe; RM = right medial lobe

Figure 3

Application of the Endo GIA stapler in five cats. (a–c) Cases 1, 4 and 5 underwent a total liver lobectomy of the left lateral lobe; (d) case 3 underwent multiple lobectomies involving the left division; (e) case 2 underwent multiple lobectomies involving the left and central division. The stapler tip is introduced into the intra-abdominal space in two cats (b,c)

In case 2, one of the main lesions located at the base of the right medial and quadrate lobes was adjacent to the left hepatic vein, necessitating resection of the left and central divisions with concomitant cholecystectomy. The adhesions to the diaphragm and abdominal wall were dissected using electrosurgical instruments. Before resection, the vascular supply to both divisions was temporarily occluded en bloc by using a tourniquet. Although portal pressure increased from 12 to 18 mmHg, there were no changes in the colour of the pancreas or small intestine, and no hypotension was observed. After resection, the colour of the viscera and systemic arterial pressure did not change.

Except for this case, the lesions were macroscopically resected. Fluid accumulation was observed intraoperatively at celiotomy in one cat and was classified as a modified transudate. The staple cartridge sizes were 45 mm (n = 1) and 60 mm (n = 4). Ligation clips were applied along the staple line in two cats: to control bleeding from an identifiable hepatic arterial branch (n = 1) and at the surgeon’s discretion (n = 1). Oxidised regenerated cellulose was also applied to achieve haemostasis in the liver parenchyma of another cat (Figure 4). The median lobe resection time was 21 mins (range 6–65) and the median total surgical time was 51 mins (range 44–93).

Figure 4

Application of ligation clips or oxidised regenerated cellulose to the staple line in three cats. (a) Ligation clips are applied to stop bleeding from the hepatic artery in case 1; (b) ligation clips are applied at the surgeon’s discretion in case 4; (c) oxidised regenerated cellulose is applied to achieve haemostasis from liver parenchyma in case 3

Complications

Intraoperative complications occurred in two cats. One cat experienced transient grade 1 hypotension (mean arterial pressure <60 mmHg). Another cat developed grade 4 hypotension, which was suspected to have been caused by abdominal manipulation and intraoperative bleeding. This cat received a CRI of adrenaline (0.2–1.0 μg/kg/min) and a fresh whole blood transfusion before application of the Endo GIA device.

Postoperative complications occurred in all the cats. All cats experienced grade 1 or 2 anaemia that persisted until discharge (Figure 5). Four cats had grade 1 ascites. No fluid analysis was performed; however, the ascites resolved during hospitalisation. One cat had a grade 2 wound complication, for which the skin suture was partially revised under sedation 1 day postoperatively. The other cat developed grade 3 hyperbilirubinemia. There was no evidence of haemolytic anaemia and no obvious common bile duct obstruction at onset. The cat showed no clinical signs and improved with supportive care (Figure 6).

Figure 5

Packed cell volume trends (reference interval 30–52%) in five cats

Figure 6

Total bilirubin trends (reference interval 0.1–0.4 mg/dl) in case 2

Outcomes

None of the cats required a postoperative transfusion or revision surgery. The median duration of ICU stay was 20 h (range 20–30). Three cats resumed voluntary food intake on postoperative day 1, one on postoperative day 2 and the remaining one on postoperative day 3. All the patients survived until discharge. The median duration of postoperative hospitalisation was 3 days (range 2–8). All cats survived until suture removal at 2 weeks postoperatively without any clinical signs.

Histopathology

On histopathological examination, one cat had a malignant hepatic tumour, three cats had benign lesions and one cat had a mixed lesion (Table 2). Surgical margins were evaluated in three cats, all of which were histologically free of neoplastic cells. In the remaining two cats, margin re-evaluation was not possible because no residual tissue samples were available from the diagnostic laboratory.

Discussion

This case series demonstrates favourable short-term outcomes after open liver lobectomy using the Endo GIA stapler in cats, representing the first detailed report to evaluate this procedure. All cats survived for at least 2 weeks postoperatively without revision surgery. Most perioperative complications were minor. The median lobe resection and total surgical times were 21 and 51 mins, respectively. The median duration of postoperative hospitalisation was 3 days.

In previous feline studies of non-hand-sewn techniques, most deaths appeared to be related to postoperative anaemia and/or haemorrhage.^3,10 Postoperative transfusion was required in 33.3–61.1% of cats in these reports. In our study, all patients experienced grade 1 or 2 anaemia, but none required a postoperative transfusion. This difference might reflect the fact that most resections in the present series were performed on the left division, whereas previous reports included other lobes. As the transection line for resection of the left division lobes often appears relatively straight, the linear configuration of the Endo GIA stapler might be advantageous for a more straightforward transection and may have influenced perioperative blood loss.

In the feline report using the thoracoabdominal (TA) staplers, minor bleeding was observed in 43.8% of cats despite ancillary haemostasis being employed in 77.8% of cases. The authors suggested that staples smaller than the 3.5 mm cartridges most commonly used in that study might have been more appropriate.³ The Tri-Staple cartridges used in this study contain two sets of three staggered rows of staples with heights of 3.0, 2.5 and 2.0 mm towards the centre.¹¹ Accordingly, the lower staple heights might facilitate sealing of smaller vessels than TA staplers; however, compression of very small vessels may still be inadequate owing to the B-shaped morphology of the staples.^4,12,13 As arterial bleeding was also encountered in this study, particularly in cats, careful intraoperative inspection of the staple line and additional haemostasis, when necessary, are recommended.

In a previous canine study of open lobectomy,⁵ approximately half of the cases required additional transection methods, such as a vessel-sealing system or hand suturing. Furthermore, 38.5% of the patients required multiple cartridges. These variations were attributed to differences in the remaining peripheral liver parenchyma. In the present study, however, all procedures were completed using a single cartridge without additional transection methods. This likely reflects the species and size differences. As cats may require a shorter transection line than dogs, liver lobectomy can be completed using a single cartridge, which appears to be a potential advantage of the Endo GIA stapler in cats. Another advantage is its ability to rotate and articulate,⁷ which may facilitate access to the hepatic hilus within the limited feline upper abdominal space, unlike the TA stapler. In the present series, three cats had lesions adjacent to the hilus, and histologically clean margins were achieved in both cats (2/2) in which margin assessment was available. This suggests that Endo GIA use may be considered for selected hilar lesions.

One cat experienced grade 4 hypotension intraoperatively, and grade 3 hyperbilirubinemia and grade 1 ascites postoperatively. We confirmed that portal pressure did not exceed 18 mmHg or increase by more than 8 mmHg during temporary vascular occlusion, as these values are the recommended limits for congenital portosystemic shunts in cats.¹⁴ As prehepatic and posthepatic causes were considered unlikely at onset, mild transient portal hypertension might have occurred. An experimental study reported that normal dogs tolerated the acute removal of up to 70% of the total liver volume,¹⁵ whereas the resectable liver volume in cats remains unknown. To the best of our knowledge, there have been no cases of large-volume liver resection in cats like this one. Nevertheless, these perioperative complications appear to be more attributable to the complexity of multiple lobectomies involving two divisions rather than to the Endo GIA stapler itself.

This study has several limitations, including its retrospective design, small sample size, short follow-up period and possible inter-operator variability. Most lesions were located in the left division, and lobectomy was not performed in the right division. Although it seemed to be selection bias, there were no cases of cats that had liver lobectomy performed via other methods during the study period. Furthermore, the surgical margins were not reviewed in two cats, limiting the assessment of the transection line validity and our ability to evaluate the appropriateness of case selection for Endo GIA stapler application.

Conclusions

The Endo GIA stapler was successfully used in open liver lobectomies in cats. All cats survived for at least 2 weeks postoperatively. Although only one cat required intraoperative transfusion, most perioperative complications were minor. These findings suggest that the Endo GIA stapler is a feasible option for open liver lobectomy in cats, particularly for lesions located in the left division, including selected lesions adjacent to the hilus. Further studies are warranted to confirm its clinical relevance, define appropriate case-selection criteria based on lesion characteristics and compare it with conventional techniques.

Footnotes

Acknowledgements

The authors thank the owners of the cats and all the staff members of our institute. Artificial intelligence assistance (ChatGPT, OpenAI) was used for minor language refinement and formatting under the supervision of the corresponding author. The authors take complete responsibility for the scientific content of this manuscript.

Accepted: 13 March 2026

Author note

The data sets generated and analysed in the current study are available from the corresponding author upon reasonable request.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

The work described in this manuscript involved the use of non-experimental (owned or unowned) animals. Established internationally recognised high standards (‘best practice’) of veterinary clinical care for the individual patient were always followed and/or this work involved the use of cadavers. Ethical approval from a committee was therefore not specifically required for publication in JFMS. Although not required, where ethical approval was still obtained, it is stated in the manuscript.

Informed consent

Informed consent (verbal or written) was obtained from the owner or legal custodian of all animal(s) described in this work (experimental or non-experimental animals, including cadavers, tissues and samples) for all procedure(s) undertaken (prospective or retrospective studies). No animals or people are identifiable within this publication, and therefore additional informed consent for publication was not required.

ORCID iD

Shintaro Tomura

References

Cuddy

Risselada

Ellison

GW.

Clinical evaluation of a pre-tied ligating loop for liver biopsy and liver lobectomy.

J Small Anim Pract 2013; 54: 61–66.

Goodman

Casale

SA.

Short-term outcome following partial or complete liver lobectomy with a commercially prepared self-ligating loop in companion animals: 29 cases (2009–2012).

J Am Vet Med Assoc 2014; 244: 693–698.

Pavia

Kovak-McClaran

Lamb

Outcome following liver lobectomy using thoracoabdominal staplers in cats.

J Small Anim Pract 2014; 55: 22–27.

Risselada

Ellison

Bacon

, et al. Comparison of 5 surgical techniques for partial liver lobectomy in the dog for intraoperative blood loss and surgical time: comparison of techniques for partial liver lobectomy. Vet Surg 2010; 39: 856–862.

Szwec

Singh

Gatineau

Hepatic lobectomy in dogs using a stapling device with a vascular cartridge: a retrospective study of 13 cases.

N Z Vet J 2023; 71: 259–266.

Michael

Case

Massari

, et al. Feasibility of laparoscopic liver lobectomy in dogs. Vet Surg 2021; 50 Suppl 1: O89–O98.

Weaver

Barbur

LA.

Use of the Endo GIA™ stapler for lung lobectomy in dogs and cats undergoing open thoracic procedures (intercostal, transdiaphragmatic thoracotomy or median sternotomy): a retrospective study of 46 lung lobectomies.

Vet Surg 2026; 55: 327–334.

Guillen

Ressel

Finotello

, et al. Non-islet-cell tumour hypoglycaemia in a cat with hepatocellular carcinoma. JFMS Open Rep 2019; 5. DOI: 10.1177/2055116919856129.

LeBlanc

Atherton

Bentley

, et al. Veterinary Cooperative Oncology Group–Common Terminology Criteria for Adverse Events (VCOG-CTCAE v2) following investigational therapy in dogs and cats. Vet Comp Oncol 2021; 19: 311–352.

10.

Hanson

Pigott

Linklater

AKJ

. Incidence of blood transfusion requirement and factors associated with transfusion following liver lobectomy in dogs and cats: 72 cases (2007–2015). J Am Vet Med Assoc 2017; 251: 929–934.

11.

Covidien/Medtronic. Tri-staple 2.0: reloads and curved tip reloads. https://www.medtronic.com/content/dam/emanuals/mitg/PT00192710_Tri_Staple_2_reloads_ml_IFU_pg1.pdf (2024, accessed 16 October 2025).

12.

Lewis

Bellenger

Lewis

, et al. Hepatic lobectomy in the dog. A comparison of stapling and ligation techniques. Vet Surg 1990; 19: 221–225.

13.

Risselada

Polyak

MMR

Ellison

, et al. Postmortem evaluation of surgery site leakage by use of in situ isolated pulsatile perfusion after partial liver lobectomy in dogs. Am J Vet Res 2010; 71: 262–267.

14.

Tivers

Lipscomb

Congenital portosystemic shunts in cats: surgical management and prognosis: surgical management and prognosis.

J Feline Med Surg 2011; 13: 185–194.

15.

Szawlowski

Saint-Aubert

Gouttebel

, et al. Experimental model of extended repeated partial hepatectomy in the dog. Eur Surg Res 1987; 19: 375–380.