Portal thrombosis

Surg Endosc (1998) 12: 1173–1176

© Springer-Verlag New York Inc. 1998

Portal thrombosis A rare complication of laparoscopic splenectomy A. Valeri,1 F. Venneri,1 L. Presenti,1 F. Nardi,2 A. Grossi,3 D. Borrelli1 1

U.O. Chirurgia Generale e Vascolare, Azienda Ospedlaiera Careggi, Firenze, Italy Servizio di Radiologia, Azienda Ospedaliera Careggi, Careggi-Firenze, Italy 3 Servizio Ematologia, Azienda Ospedaliera Careggi-Firenze, Viale Morgaent, 80 50135, Firenze, Italy 2

Received: 18 March 1997/Accepted: 18 September 1997

Abstract. Portal thrombosis is a rare complication of splenectomy. We performed 12 laparoscopic splenectomies and observed this complication only in one patient with idiopathic thrombocytopenia (ITP). The right branch of the portal vein presented a partial thrombosis, while the left branch was completely obstructed by thrombi. Abdominal ultrasonography and an ultrasound doppler exam allowed us to diagnose this event and a retrograde angiography performed afterward confirmed our diagnosis. A 48-h intravenous heparin treatment was promptly begun, followed by anticoagulant drugs (dicumarol). The patient was dismissed 5 days afterward, presenting a steady-state ultrasound doppler pattern and a complete normalization of liver parameters. An ultrasound doppler exam performed 1 month after anticoagulant therapy showed a complete resolution of portal thrombosis. We believe that early diagnosis of this rare complication, prompt beginning of anticoagulant therapy, and care in surgical procedures may reduce patient lifethreatening risks and assure complete remission. Key words: Splenectomy — Laparoscopy — Portal thrombosis — Treatment

The first laparoscopic splenectomy was performed in France by Bernard DeLaitre in 1991. The evolution of surgical procedures since then has improved this approach, assuring safety and an acceptable operating time. The main indications for laparoscopic splenectomy are idiopathic thrombocytopenic purpura, autoimmune hemolytic anemia (an affliction in which the spleen is usually not enlarged), hereditary spherocytosis and splenic masses of unknown

Correspondence to: F. Venneri, Via F. Gioli, 23, 50018 Scandicci, FI, Italy

origin, and we believe, staging and restaging of lymphomas in which splenectomy is necessary [11, 13, 21]. This procedure may be performed with the patient on a right lateral position [4, 7, 8] or supine [11]. We prefer the right lateral position with a slight anti-Trendelenburg inclination. During a 20-month period, we performed 12 laparoscopic splenectomies in six patients affected by thrombocytopenic purpura, four patients with hemolytic anemia, one patient with spherocytosis, and 1 patient with a nonHodgkin’s lymphoma presenting involvement of the spleen following chemotherapy. We also performed a combined laparoscopic cholecystectomy for gallstones in a female patient affected by spherocytosis. The average operating time was 135 min and the average hospital stay 4 days. All patients received an antipneumococcic vaccination and perioperative antibiotic prophylaxis. Immunoglobulins, cortisone, and platelets were subministered in patients with thromocytopenia in order to reach an acceptable 40,000–60,000 platelet count before surgery. In two of these last patients postoperative bleeding occurred and was treated by a conservative approach. In another female patient we observed a partial thrombosis of portal vein branches treated successfully with anticoagulants, and this case is the objective of this paper.

Case report G.O., a 32-year-old female patient affected by idiopathic thrombocytopenia (ITP) not responsive to cortisone therapy was submitted to laparoscopic splenectomy. Preoperative intravenous IgG (0.4 mg/kg for 5 days) were subministered and the patient presented a good coagulative pattern. Antipneumococcus vaccine was also administered before surgery. Perioperative calcic heparin (5,000 I.U. twice daily) was routinely subministered along with antibiotic prophylaxis using cephazolin 1 g/twice a day i.v. until discharge. Laparoscopic splenectomy was performed according to the procedure proposed by Carroll and Delaitre [4, 7]: the patient is placed in a right lateral position on a ‘‘bean bag’’; four to five trocars are placed; splenocolic and splenophrenic ligaments are tied and cut; a laparoscopic suture device is used to clip spleen vessels; short gastric vessels are cut

1174

Fig. 1. Retrograde angiography. Patency of main portal and mesenteric veins. between clips; the spleen is then removed within an endobag, first being fragmented to facilitate extraction through the umbilical incision according to the ‘‘open’’ technique using a Hasson trocar; a drainage tube is then placed and removed on the 2nd postoperative day. Surgery lasted nearly 1 h 45 min, requiring a 3-d hospital stay [3, 6, 9, 17]. The patient was discharged with these peculiar blood exam values: WBC 17,600, PLT 538,000. She was advised to continue heparin treatment and amoxycillin 1 g twice a day until the 7th postoperative day upon discharge, but she did not follow these instructions. The patient returned to our observation on the 5th postoperative day because of the onset of a sudden abdominal pain referred to the epigastric region and upper right quadrant. She immediately underwent ultrasonography, which excluded any intraabdominal liquid gain, while signs of a complete obstruction of portal vein left branch and a partial obstruction of its right branch were seen. Patency of the main portal and mesenteric veins with a clear flow toward the liver is confirmed. These data were afterward confirmed by an ultrasound doppler exam and by retrograde angiography. No thrombosis was demonstrated to be within the splenic vein (Fig. 1). Laboratory findings revealed leucocytosis, increase in platelets (600,000), and alterations of hepatic parameters as follows: SGOT 50,SPGT 92, total bilirubin 1.2. Intravenous heparin was begun at 1,000 I.U./h in order to yield doubling of PTT value. Cephotriaxone 2 g i.v./day was administered and then stopped 2 days after WBC normalization. Antispastic drugs were also prescribed to complete the therapeutic pattern. Pain disappeared within a few hours, while liver parameters returned within normal ranges after 48 h (SGOT 36, SGPT 37, total bilirubin 0.92). WBC count returned normal after 4 days (4,700 mm3). An ultrasound doppler exam was performed 48 h after admission and did not show any significant change relative to the one performed upon admission. Anticoagulants were then administered in combination with intravenous heparin; this regimen was later stopped once PT values ranged between 30 and 40%. Patient was discharged 8 days afterward. Anticoagulants were continued in order to maintain PT values within the ranges mentioned above. An ultrasound doppler exam was performed before discharged and there was no substantial variation relative to the preceding exam. A control ultrasound doppler exam was taken 24 days after and showed patency of right portal branch and nearly complete obstruction of the left portal branch (Figs. 2 and 3). Anticoagulants were confirmed. Another ultrasound doppler scan was performed 45 days after, showing a partial thrombosis still present within the left portal branch, while the right branch seems patent with normal blood flow towards the liver.

Discussion Portal thrombosis following splenectomy represents a very rare complication in itself (0.3% over 1,738 laparoscopic

Fig. 2. Ultrasound doppler exam. Patency of right portal branch. Fig. 3. Ultrasound doppler exam. Obstruction of left portal branch.

splenectomies reviewed by Pinna et al. in 1982 [18]) [5, 10, 12, 17, 19]. Probably portal thrombosis is much more frequent than reported. In our case, for example, the symptoms complained of by the patient such as epigastric upset and pain and fever may lead to a mistaken diagnosis of postoperative pancreatitis or pulmonary embolism or biliary sepsis [20]; nevertheless, portal vein thrombosis should be considered in patients with fever and abdominal pain after splenectomy. This complication was never observed following laparoscopic splenectomies. According to a review of literature, laparoscopic splenectomy seems to be a safe procedure as the laparotomic approach, with the only disadvantages represented by a prolonged surgical time lapse and high costs [3, 11, 15, 21, 22, 29]. In particular, the incidence of the most frequent complications reported in the literature are on average 0.14% [3, 6, 11, 15, 21, 29] and are represented by intraoperative and postoperative bleeding, subphrenic abscess, hyperthermia, respiratory infections, and pancreatitis [6, 16]. No article referred in the literature reports portal thrombosis and overwhelming sepsis as complications of laparoscopic splenectomy whereas these are reported to have a 2.5–13.5% incidence in laparotomic procedures [23]. Some comparative studies, though they refer to a limited number of cases, seem to prove a much lesser incidence of intra- and postoperative complications after laparoscopic

1175

splenectomy compared to the laparotomic approach (lesser incidence of respiratory infections, sepsis, and intraoperative bleeding) [28]. The higher costs of the laparoscopic procedure are surely balanced by the reduced hospital stay [12, 15, 28]. All the literature data reviewed along with our personal experience lead us to affirm that laparoscopy accounts all the advantages of this technique proving safeness and reliability mainly in absence of an overwhelming splenomegaly [3, 11, 15, 16, 21, 29]. We believe that portal thrombosis following splenectomy may be caused by these three major pathogenetic factors: 1. Variations in blood components (platelet increase) 2. Blood stasis 3. Surgical maneuvers with probable endothelial lesions [1]. Thrombocytosis is very frequent after splenectomy for nonmalignant blood disorders, with a 75% incidence rate [2], though in patients affected by persistent thrombocytosis no increase in thromboembolic disorders has been reported [1, 26]. In many myeloproliferative disorders in which, today, splenectomy is rarely performed, the incidence of thrombocytosis is the same but is accompanied by a major onset of thromboembolic complications also in the splanchnic territory [14]. Thrombocytosis may also be most frequently associated to mesenteric and portal thromboembolism if antithrombin III deficiency is revealed [17]. Our patient was affected by thrombocytopenic purpura, and no variation in antithrombin III or thrombophylic status was reported. Regarding the second factor, postsplenectomy thrombosis (blood stasis), portal or mesenteric thrombosis was proved to be very frequent in patients affected by preexisting portal hypertension [1, 24]. A probable cause of blood stasis within the portal-mesenteric route following laparoscopic splenectomy may be the positive intraabdominal pressure due to pneumoperitoneum. We performed laparoscopic splenectomy without going over 12 mmHg of pneumoperitoneum. Normal portal pressure is nearly 10–12 mmHg and a review of the literature revealed that significant hemodynamic variations do not occur for a pneumoperitoneum maintained at 12 mmHg in elderly patients [9]. In experimental swine models, significant hemodynamic changes were demonstrated for pneumoperitoneum at values over 14 mmHg. Normally in the pig, portal pressure is usually 4–5 mmHg [27]; these data are therfore not comparable to the results obtained in man. As for the third factor that causes portal thrombosis, surgical manuveurs, in our case we feel that thrombocytosis played a fundamental role, though we may not exclude that the surgeon’s abilities also had a significant part. In particular, ligature of the splenic vein close to the parenchyma itself with a particularly long splenic vein stump residue may favor blood stasis. We have indeed performed coagulation of splenic vessels very close to spleen parenchyma using clips and an endoGIA without trauma or insults to the main branch. Some authors in fact have proposed ligature of the splenic vein as close as possible to the to the point where it encounters the superior mesenteric vein [1] as a preventive manuveur for portalmesenteric thrombosis, besides postoperative platelet antiaggregant agent treatment. The nucleus of the thrombus developed very close to the clips, and a sudden onset of a

peripheral embolization may have caused a partial obstruction of the left portal branch and a total obstruction of the right branch. Normal serum amylase values do not account for pancreatitis, which may itself be a pathogenetic factor in portal thrombosis. The use of clips and an endoGIA, rather than normal sutures, does not seem to favor this complication. We believe that a particularly long splenic vein stump residue may be a predisposing factor for portal thrombosis, associated to a significant postoperative thrombocytosis not treated with antiaggregants. We believe that an antiaggregant therapy using ASA (aspirin) upon discharge and continued for a few weeks at home may be an efficient prophylaxis of thrombosis compared to calcium heparin alone [1]. Our patient did not take ASA but neither did she continue calcium heparin; thus, this may have favored the onset of thrombosis, along with her significant platelet status (600,000). Our therapeutic pattern upon patient readmission, consisting of immediate intravenous heparin subministration and accompanied by anticoagulants (dicumarol), surely avoided evolution of portal thrombosis and favored a complete recovery of the right branch. An ultrasound doppler scan performed within 3 months proved this patency, with persistence of complete left portal branch obstruction which does not yield any significant hemodynamic alterations. Once portal-mesenteric thrombosis is diagnosed, an anticoagulant and/or thrombolytic drug treatment is required in order to preserve liver and bowel integrity. In the case of late diagnosis where an appropriate medical treatment is therefore delayed, the patient may die. Rattner, in his series of seven patients affected by portal vein thrombosis following splenectomy, reports that diagnosis in two patients was made over a 3-day period, and they died upon onset of symptoms [20].

References 1. Boissel P, Bresler L, Trackeon G, Poussot D, Gosdidier J (1991) Mesenteric and portal vein thrombosis following splenectomy in hemolytic diseases. A proposal of 3 cases. Chirurgie 117: 591–595 2. Boxer MA, Braun J, Ellman L (1978) Thromboembolic risk of postsplenectomy thrombocytosis. Arch Surg 113: 808–809 3. Brunt LM, Langer JC, Quasebarth MA, Whitman ED (1996) Comparative analysis of laparoscopic versus open splenectomy. Am J Surg 172: 596–599 4. Carroll BJ, Phillips EH, Semel CJ, Fallas M, Morgenstern L (1992) Laparoscopic splenectomy. Surg Endosc 6: 183–185 5. Chistopher Ellison E, Fabri Peter J (1985) Complicazioni della Splenectomia. Etiologia, prevenzione e trattamento. Clin Chir N Am 16: 1490–1509 6. Daly B, Sukumar SA, Krebs TL, Wong JJ, Flowers JL (1996) Non biliary laparoscopic gastrointestinal surgery: role of CT in diagnosis and management of complications. AJR Am J Roentgenol 167: 455– 459 7. Delaitre B (1995) Laparoscopic splenectomy. The ‘‘hanged spleen’’ technique. Surg Endosc 9: 1490–1509 8. Delaitre B, Maignen B (1991) Sple´nectomie par voie coelioscopique: 1 observation (Letter). Presse Med 20: 22–63 9. Dhoste K, Lacosete L, Karayan J, Lehuede MS, Thomas D, Fuscardi J (1996) Haemodynamic and ventilatory changes during laparoscopic cholecystectomy in elderly ASA III patients. Can J Anaesth 43–8: 783–878 10. Emmerman A, Zurnig C, Peiper M, Weh HJ, Broelsch CE (1995) Laparoscopic splenectomy. Technique and results in a series of 27 cases. Surg Endosc 9: 924–927 11. Flowers JL, Lefor AT, Steers J, Heyman M, Graham SM, Imbembo

1176

12.

13. 14.

15.

16. 17. 18. 19. 20.

AL (1996) Laparoscopic splenectomy in patients with hematologic diseases. Ann Surg 224: 19–28 Friedman RL, Fallas MJ, Carroll BJ, Phillips EH (1996) Laparoscopic splenectomy for ITP: the Gold Standard. Abstract 1996 Scientific Session of the 5th World Congress of Endoscopic Surgery. Surg Endosc 10: 177–217 Gigot JF, Healey ML, Ferrant A, Michaux JL, Njinou B, Kesten PJ (1994) Laparoscopic splenectomy for idiopathic thrombocytopenic purpura. Br J Surg 81: 1171–1172 Gordon DH, Schaffner D, Bennett JM, Schwartz SI (1978) Postsplenectomy thrombocytosis: its association with mesenteric, portal and/or renal vein thrombosis in patients with myeloproliferative disorders. Arch Surg 113: 713–715 Hashizume M, Ohta M, Kishihara F, Kawanaka H, Tomikana M, Neno K, Tanone K, Higashi H, Kitano S, Suginachi K (1996) Laparoscopic splenectomy for idiopathic thrombocytopenic purpura: comparison of laparoscopic surgery and conventional open surgery. Surg Laparosc Endosc 6: 129–135 Janu PG, Rogers DA, Lobe TE (1996) A comparison of laparoscopic and traditional open splenectomy in childhood. J Pediatr Surg 31: 109–113 Peters TG, Lewis JD, Flip DJ, Morris L (1997) Antithrombin III deficiency causing postsplenectomy mesenteric venous thrombosis coincident with thrombocytopenia. Ann Surg 185: 229–231 Pinna CD, Gambarella R, Toia N (1982) La Splenectomia nell’Eta’ Pediatrica. Arch. ed Atti 84° Congresso Societa` Italiana Chirurgia Generale—Roma Poulin EC, Thibault C, Mamazza J (1995) Laparoscopic splenectomy. Surg Endosc 9: 172–177 Rattner DW, Ellman L, Warshaw AL (1993) Portal vein thrombosis after elective splenectomy. An underappreciated, potentially lethal syndrome. Arch Surg 128: 565–569

21. Rhodes M, Rudd M, O’Rourke N, Nathanson L, Fielding G (1995) Laparoscopic splenectomy and lymph node biopsy for hematologic disorders. Ann Surg 222: 43–46 22. Schlinkert RT, Mann D (1995) Laparoscopic splenectomy offers advantages in selected patients with immune thrombocytopenic purpura. Am J Surg 170: 624–626 23. Shatney CH (1987) Complications of splenectomy. Acta Anaesthesiol Belg 38: 333–339 24. Suzuki S, Nakamura S, Baba S, Sakaguchi S, Ohnuki Y, Yokoi Y, Nishiyama R (1992) Portal vein thrombosis after splenectomy successfully treated by an enormous dosage of fibrinolytic agent in a short period: report of two cases. Surg Today 22: 464–469 25. Targarona EM, Ardid J, Sugranes G, Mateos G, Balague` C, LopezGuillermo A, Trias M (1996) Long term assessment of laparoscopic splenectomy or autoimmune and hemolytic diseases. Abstracts 1996 Scientific Session of the 5th World Congress of Endoscopic Surgery. Surg Endosc 10: 177–217 26. Visudiphan S, Ketsa-Ard K, Piankijagum A, Tumliang S (1985) Blood coagulation and platelet profiles in persistent post-splenectomy thrombocytosis. The relationship to thromboembolism. Biomed Pharmacother 39: 264–271 27. Voltz J, Koster S, Weiss M, Schnidt R, Urbaschek R, Melchert F, Albrecht M (1996) Pathophysiologic features of a pneumoperitoneum at laparoscopy: a swine model. Am J Obstet Gynecol 174(1 patient 1): 132–401 28. Watson DI, Conventry BJ, Chin T, Gell PG, Malycha P (1997) Laparoscopic versus open splenectomy for immune thrombocytopenic purpura. Surgery 121: 18–22 29. Zamir O, Szold A, Matzner Y, Ben-Yehuda D, Seror D, Deutsch I, Freund HR (1996) Laparoscopic splenectomy for immune thrombocytopenic purpura. J Laparoendosc Surg 6: 301–304