Splenic artery aneurysms: postembolization syndrome and surgical complications

Splenic artery aneurysms: postembolization syndrome and surgical complications

The American Journal of Surgery 193 (2007) 166 –170 Clinical surgery–International Splenic artery aneurysms: postembolization syndrome and surgical ...

667KB Sizes 0 Downloads 37 Views

The American Journal of Surgery 193 (2007) 166 –170

Clinical surgery–International

Splenic artery aneurysms: postembolization syndrome and surgical complications Gabriele Piffaretti, M.D.*, Matteo Tozzi, M.D., Chiara Lomazzi, M.D., Nicola Rivolta, M.D., Francesca Riva, M.D., Roberto Caronno, M.D., Patrizio Castelli, M.D., F.A.C.S. Vascular Surgery, Department of Surgery, University of Insubria-Varese, Ospedale di Circolo, 21100 viale Borri 57, Varese, Italy Manuscript received May 23, 2006; revised manuscript September 13, 2006

Abstract Background: This study assessed the endovascular embolization of splenic artery aneurysms and false aneurysms with special consideration given to postoperative complications. Methods: Fifteen patients (11 women; mean age, 56 y; range, 39 – 80 y) with splenic artery aneurysm (n ⫽ 13) or false aneurysm (n ⫽ 2) were treated with coil embolization. The lesion was asymptomatic in 9 patients, symptomatic in 5 patients, and ruptured in 1 patient. The mean aneurysm diameter was 33 ⫾ 23 mm (range, 15– 80 mm). Postoperative follow-up evaluation included a clinical visit and spiral computed tomography at 1, 4, and 12 months, and yearly thereafter. Results: Endovascular treatment was possible in 14 patients (93%) (1 failure: neck cannulation). Perioperative mortality was not observed. Morbidity included postembolization syndrome in 5 patients (30%). Neither pancreatitis nor spleen abscess occurred. The mean follow-up period was 36 months (range, 3– 60 mo). During follow-up evaluation we detected 1 sac reperfusion that was sealed successfully with additional coils. Surgical conversion or open repair were never required. Conclusions: At our institute, endovascular treatment represents the first-line treatment for splenic artery aneurysms. Postembolization syndrome and infarcts are common events but generally resolve without sequelae. © 2007 Excerpta Medica Inc. All rights reserved. Keywords: Splenic artery aneurysms; Endovascular

The splenic artery is the most frequent site of visceral arterial aneurysms [1,2]. Rupture is frequent, sometimes occurring as the first symptom, and sometimes is fatal [3]. Traditionally, open surgical repair has been performed [4]. Recently, feasible alternative approaches such as laparoscopic resection and endovascular exclusion have been suggested [5–7]. We report the results of our ongoing experience with endovascular treatment of 15 splenic artery aneurysms (SAAs), with special consideration given to the complications and consequences of postembolization syndrome. Patients and Methods Between January 1999 and May 2006, we observed 24 patients with SAAs. Five patients underwent surgical repair (4 open, 1 laparoscopic), and 4 patients are still under surveillance (asymptomatic aneurysm diameter ⬍ 20 mm); * Corresponding author. Tel.: ⫹39-332278226; fax: ⫹39-332278581. E-mail address: [email protected]

these aneurysms were excluded from this review. Hence, percutaneous endovascular treatment was attempted in 15 SAAs (13 aneurysms and 2 pseudoaneurysms). There were 11 women and 4 men (mean age, 56 ⫾ 13 y; range, 39 – 80 y). The lesion was asymptomatic and discovered fortuitously in 9 patients; 5 SAAs were symptomatic (upper abdominal pain), and 1 was ruptured (hemorrhagic shock). Comorbidities and associated risk factors included systemic hypertension (n ⫽ 4), chronic obstructive pulmonary disease (n ⫽ 3), atrial fibrillation (n ⫽ 3), chronic renal failure (n ⫽ 2), cancer (n ⫽ 3), ischemic heart disease (n ⫽ 1), portal hypertension (n ⫽ 1), diabetes (n ⫽ 2), rheumatoid arthritis (n ⫽ 1), and peptic ulcer (n ⫽ 1). Two patients had previous aortic surgery, 7 patients had previous major abdominal surgery, and 2 patients had chemotherapy. All patients underwent spiral computed tomography (CT) to confirm the diagnosis and assess the anatomic features and feasibility of the endovascular approach; in 6 patients, the diagnosis was made with ultrasonography before CT. The lesion was located as follows: proximal (n ⫽

0002-9610/07/$ – see front matter © 2007 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2006.09.007

G. Piffaretti et al / The American Journal of Surgery 193 (2007) 166 –170

1), in the intermediate segment (n ⫽ 8), near the hilus (n ⫽ 4), and intraparenchymal (n ⫽ 2). The mean aneurysm diameter was 33 ⫾ 23 mm (range, 15– 80 mm). Preoperative antibiotic prophylaxis was not used routinely. Endovascular exclusion was performed in the angiographic suite (n ⫽ 14), or in the operating theater (n ⫽ 1, ruptured aneurysm). Embolization was the only endovascular technique applied, and always was performed with coils, and sometimes associated with injection of N-butyl-2-cyanoacrylate (Histoacryl; Braun Medical, Boulogne, France). We defined technical success as exclusion of the aneurysm at the end of the procedure (Fig. 1). In the postoperative course, every patient underwent periodic blood tests for white blood cells (range, 4300 –11,000 cells/mm3), platelets (range, 150,000 – 450,000 cells/mm3), and lactate dehydrogenase (LDH) (range, 150 – 450 U/L) to control the clinical evolution of the spleen, and the so-called postembolization syndrome (PES). PES is defined as a combination of fever (38°C), abdominal pain, slowed transit, pleural effusion, and possibly also pancreatitis attacks after spleen infarction. Follow-up examination included a clinical visit and duplex ultrasonography and/or CT scan 1, 6, and 12 months after the procedure, and yearly thereafter.

Fig. 1. (A) Preliminary angiography shows a saccular aneurysm of the distal aspect of the splenic artery. (B) Final angiogram confirmed the complete sealing of the aneurysm and the preservation of the splenic vascularization.

167

Data for continuous variables were expressed as mean values. They were collected prospectively and analyzed in a retrospective manner. Early results were analyzed and compared by the chi-square or Fisher exact tests. Long-term results were analyzed by Kaplan-Meier curves, with dedicated software for Windows (SPSS 11; SPSS Inc., Chicago, IL). A P value of less than .05 was considered statistically significant. Results Technical success of endovascular treatment was achieved in 14 of the 15 patients (93%). One endovascular procedure failed: in this patient we were not able to cannulate the neck of a splenic pseudoaneurysm (the patient denied further treatment). Seven aneurysms were excluded with coils only; in the remaining 7 patients we used coils plus injection of N-butyl-2-cyanoacrylate. Perioperative mortality was not observed; intensive care unit and blood transfusion (4 units of packed red blood cells) were required in the patient with the ruptured SAA only. Access site complications did not occur; complications involved pleural effusion (n ⫽ 5) and early spleen infarction (n ⫽ 3; 2 partial, 1 massive). PES was noted in 5 patients (30%); in these patients serologic tests were totally silent. The mean duration of the PES was 3 days (range, 2– 8 d). Acute pancreatitis did not occur. No spleen abscess developed (Fig. 2). After the procedure, 12 patients remained asymptomatic; 2 patients complained of abdominal pain and were treated with nonsteroidal antiinflammatory drugs. In addition, in the 2 patients with massive spleen infarction, as a result of septic-like fever we punctured the splenic parenchyma under ultrasonographic guidance with no evidence of infection: in both patients all serologic test results and cultures were negative. In the postoperative course, the overall increase of white blood cells was 24% (preoperative mean, 7334 ⫾ 3967 cells/mm3; postoperative mean, 9137 ⫾ 3095 cells/mm3), and the overall increase of LDH was 26% (preoperative mean, 340 ⫾ 128 U/L; postoperative mean, 430 ⫾ 116 U/L); in contrast, platelets decreased by 16% (preoperative mean, 232 ⫾ 103 cells/mm3; postoperative mean, 195 ⫾ 66 cells/mm3). No statistically significant differences were noted among patients with infarcted spleens and those patients who did not develop spleen infarction (Table 1). Although the overall median hospitalization time was 3 days (range, 1–30 d), patients who developed spleen infarction required prolonged hospitalization (3.2 ⫾ 2 vs 21 ⫾ 14; P ⬍ .05). The mean follow-up period was 36 ⫾ 20 months (range, 3– 60 mo). At 4 months we observed 1 reperfusion of the sac of a giant aneurysm (diameter, 70 mm) originally treated with coils and N-butyl-2-cyanoacrylate without spleen infarction; this patient was scheduled for another endovascular approach and the aneurysm was sealed with additional coils. She developed a massive infarction that required prolonged hospitalization. At the follow-up evaluation, in those patients with a spleen infarction, the CT scan evaluation revealed that the spleen was decreased in volume but was partially enhanced by collaterals vessels, the splenic vein was permeable and the tail of the pancreas appeared

168

G. Piffaretti et al / The American Journal of Surgery 193 (2007) 166 –170

Fig. 2. (A) Preoperative CT angiography of a 7-cm aneurysm of the distal portion of the splenic artery. At 4 months the CT control revealed a (B) mild reperfusion (arrow) of the sac that was totally excluded with (C) adjunctive coil (arrow) embolization.

normal in all patients. All the patients are still alive; the event-free rate at 1 month as estimated by the Kaplan-Meier method was 93% and 85% at 12 and 36 months, respectively (Fig. 3). Comments Recently, the widespread use of diagnostic imaging modalities has increased the incidental detection of visceral artery aneurysms (VAAs). SAAs are the most frequent VAAs reported in the literature [1,2]; usually, they are single, twice as frequent in women as in men, and mostly

are located in the intermediate or distal segment of the splenic artery [8 –12]. Most of these aneurysms are small (diameter, ⬍3 cm), whereas giant aneurysms (diameter, ⬎10 cm) are rare [9,13]. All of these features have been confirmed in our experience. Various therapeutic options are available for patients with SAA. Traditional surgical management is the recommended repair in good-risk patients with asymptomatic aneurysms greater than 20 mm in diameter or with aneurysms with a rapid growth in size [1,2,4,10,13]. During the past decade, endovascular techniques have been advocated as an alternative to treat these aneurysms [4,5,14]. Among the 4 most recent series of endovascular exclusions of SAAs, Gabelmann et al [6] reported on 10 SAAs (3 ruptured) with a 92% primary technical success rate and a 10% mortality rate. Long-term success was 92%, with no evidence of sac reperfusion at the 46-month follow-up evaluation. Abbas et al [15] in 2002 obtained an 80% success rate with one celiac artery iatrogenic dissection that required open repair on 5 SAAs. In 2003, Guillon et al [9] reported a 75% primary exclusion on 12 aneurysms with no mortality, no complications, and no late reperfusion of the sacs at a mean follow-up period of 13 months. Saltzberg et al [16] described coil embolization for 11 lesions with no mortality and a 9% rate of late endoleaks. Our experience compares well with this series, reporting a 93% primary success rate with no mortality and 100% of secondary exclusions at a mean follow-up period of 36 months. Although most VAAs including SAAs are asymptomatic, as many as 22% of patients present with rupture and pose a therapeutic challenge in emergency situations [3,4,6]. Trastek et al [17] recorded 3 cases of acute rupture out of 100 SAAs. They also followed up 34 of these asymptomatic aneurysms (mean diameter, 1.8 cm) for an average period of 6.4 years: only 1 ruptured, giving a rupture risk of 3%. In symptomatic patients, the rupture rate ranges from 76% to 83%, but the time between the onset of symptoms and rupture was not reported in these studies. The mortality rate after rupture ranges between 10% and 25% [3,9,13,17]. Therefore, the application of endovascular therapy to SAAs has led to a more aggressive elective approach in treating asymptomatic aneurysms with smaller diameters (⬎1.5 cm) with the goal of preventing rupture, but also has been advocated for ruptured VAAs, especially when the aneurysm is in a poorly accessible location for surgical treatment, or when surgical risks are high because of poor general status [2]. When a VAA rupture is suspected, arteriography can be useful to confirm the diagnosis, determine the exact location of the aneurysm, screen for other lesions, assess the possibility of endovascular treatment, and allow temporary hemostasis with an inflatable balloon catheter before surgery if endovascular exclusion with embolization of the aneurysm is unfeasible or fails [2,6,9]. We report the embolization of a ruptured SAA in a young female patient who underwent emergency laparotomy for massive intraabdominal bleeding of undetected origin; in this patient, explorative laparotomy and intra-abdominal packing did not ensure hemodynamic stability. Thus, we decided to perform intraoperative angiography, which finally detected the origin of the active bleeding: a 25-mm ruptured saccular SAA. Coils were placed successfully on either side of the lesion to

G. Piffaretti et al / The American Journal of Surgery 193 (2007) 166 –170

169

Table 1 Postoperative trends of the biological markers used to detect PES Number of patients

Preoperative

Postoperative

Variation

P

6334 ⫾ 1670 6235 ⫾ 1812 6730 ⫾ 1315

9137 ⫾ 3095 8860 ⫾ 3454 9970 ⫾ 2371

Increase: 44% Increase: 42% Increase: 48%

NS NS NS

232 ⫾ 103 205 ⫾ 73 341 ⫾ 173

195 ⫾ 66 211 ⫾ 64 147 ⫾ 61

Decrease: 16% Increase: 3% Decrease: 57%

NS NS NS

340 ⫾ 128 360 ⫾ 83 268 ⫾ 276

430 ⫾ 116 401 ⫾ 90 517 ⫾ 185

Increase: 26% Increase: 11% Increase: 92%

NS NS NS

3

White blood cells, cells/mm Overall (n ⫽ 15) Noninfarcted spleen (n ⫽ 11) Infarcted spleen (n ⫽ 4) Platelets, cells/mm3 Overall (n ⫽ 15) Noninfarcted spleen (n ⫽ 11) Infarcted spleen (n ⫽ 4) LDH, U/L Overall (n ⫽ 15) Noninfarcted spleen (n ⫽ 11) Infarcted spleen (n ⫽ 4) NS ⫽ not significant.

prevent leakage or recurrent bleeding. Although the spleen developed a massive infarction, the patient did not develop abscess or superinfection, splenectomy was not required, and was discharged without further complications. She was last seen 48 months after the procedure; at that time, CT angiography confirmed the complete exclusion of the aneurysm. Clinical and radiologic follow-up evaluation may aid in the detection of early or delayed complications that include end-organ infarction, aneurysm recurrence, contrast nephropathy, embolism, and access site hematomas [12]. In fact, we must take into account the frequency of anastomoses, particularly in the splenopancreatic anatomic region [18]. In the literature, the recanalization rate after endovascular exclusion of VAAs has been reported in the 0% to 50% range [3]; in the most recent series, reperfusion after exclusion of SAAs averaged 4.5% [6,9,15,16]. These failures required repeat embolization or surgery in most patients. The only leakage (1 of 14; 7%) that we detected in the present experience involved a giant aneurysm of the distal aspect of the splenic artery; nevertheless, the reembolization led easily to the complete sealing of the an-

Fig. 3. Kaplan-Meier analysis: event-free evaluation of the endovascular approach.

eurysm. Presumably, the technical mistake was not inserting sufficient coils to make a more compact nest. Endovascular techniques confer many possible advantages. However, percutaneous coil embolization of SAAs is not without risks. One of the most common complications that potentially may require prolonged hospitalization, the so-called PES, which occurs in 30% of patients, combines fever (38°C), abdominal pain, slowed transit, pleural effusion, and possibly pancreatitis attacks after spleen infarction (Table 2) [2,9]. On the basis of their own work (4 patients) and literature reports (13 patients) McDermott et al found a successful outcome in 85% of patients, with postembolization pain with some fever in one third of patients (but no severe complications) [5]. Guillon et al [9] noted no important technical events, but in some patients (36%) they observed a transient increase of pancreatic enzyme levels, more frequently as the embolization was more proximal. They suggested that acute pancreatitis could occur if a complete arterial occlusion was performed, but it usually seems to resolve if the treatment objective is only arterial lesion exclusion (in contrast to a greater risk of pancreatitis in the case of massive parenchymal spleen embolization as a result of proximal embolization). Our experience was similar to this data. In fact, we observed PES in 4 patients (28%). Overall, statistical differences were not noted in the platelet levels and white blood cells counts, and in the LDH level between the preoperative and postoperative course. If we consider the 4 patients who developed a spleen infarction, between the preoperative and postoperative course we observed a significant increase of the LDH levels compared with noninfarcted patients (92% vs 11%), whereas white blood cell counts did not differ significantly (48% vs 42%). Interestingly, platelets decreased by 57% in infarcted spleens and increased by 3% in noninfarcted spleens; this condition could be assumed as a functional splenectomy. During coil embolization, the splenic vascularization must be preserved as much as possible through collateral gastric, omental, and pancreatic vessels because the risk of infection (1.4%) has been reported to be lower than for parenchymal embolizations, in which a severe splenic infarct is induced deliberately [9]. Gabelmann et al [6] reported that 20% of complications in their study required

170

G. Piffaretti et al / The American Journal of Surgery 193 (2007) 166 –170

Table 2 Literature review of the series specifically dealing with endovascular exclusion of SAAs Author, y

Patients/ruptured

PTS

Method

Mortality

Complication ⬎ intervention

PES

Follow-up period

Leak

Gabelmann et al,6 2002 Abbas et al,15 2002

10/3 5

92% 80%

Coils Coils

10% 0

20% 20%

46 ?

0 ?

Guillon et al,9 2003 Saltzberg et al,16 2005

12 11

75% 94%

Coils Coils

0 0

36% 36%

13 15

0 9%

Present study

15/1

93%

Coils

0

10% (abscess) ⬎ splenectomy 20% (celiac artery dissection) ⬎ ligation 0 36% (1 pancreatic, 3 spleen abscesses) ⬎ splenectomy (n ⫽ 1) 0

28%

36

7%

PTS ⫽ primary technical success.

open surgical repair. In 1 patient splenectomy was required after the development of an abscess, which led to postoperative multiple organ failure and death. Saltzberg et al [16] noted a 36% overall rate of abscesses (3 spleen, 1 pancreatic), but only 1 patient underwent open repair. In contrast, in one of the largest series in a trauma setting, Haan et al [19] reported the results of a multicenter review that performed splenic angioembolization in 140 injuries; distal embolization, generally reserved for patients with a small number of vascular injuries, resulted in a significant infarction rate of 27%, whereas nearly 20% of patients treated with main coil embolization had significant infarcts. In conclusion, they stated that splenic infarctions seem relatively common after embolization, but the vast majority of these patients are asymptomatic and can be managed nonsurgically. This also was confirmed in the series of Killeen et al [20]; they reported the results of 53 CT controls among 80 patients who underwent spleen embolization and found that splenic infarcts occurred in 63% of patients after proximal embolization and in 100% of patients after distal embolization. Most infarcts resolved without sequelae, 2 collections were drained and found to be sterile and 1 patient had a splenic abscess at laparotomy. We had similar observations. Although the 4 patients with spleen infarcts had a prolonged hospitalization, none required surgical repair, and CT scans did not reveal an abscess. This was confirmed in 2 patients by percutaneous echo-guided samples that were cultured and found to be sterile. Hence, if the result of embolization is immediately favorable and the clinical course is uneventful, we prescribe follow-up CT scans at 1 and 6 months, and yearly thereafter. Conclusions At our institute, endovascular treatment represents the first-line treatment for all SAAs because it offers several advantages over conventional surgical treatment, including in emergency cases, because of its accuracy, low morbidity, and excellent long-term results. PES and infarcts are common events but generally are resolved without sequelae.

References [1] Upchurch GR, Zelenock GB, Stanley JC. Splanchnic artery aneurysms. In Rutherford RB, ed. Vascular Surgery. 6th ed. Philadelphia: W.B. Saunders Co.; 2005:1565–9. [2] Berceli SA. Hepatic and splenic artery aneurysms. Semin Vasc Surg 1006;18:196 –201. [3] Carr SC, Mahvi DM, Hoch JR, et al. Visceral artery aneurysm rupture. J Vasc Surg 2001;33:806 –11. [4] Carr SC, Pearce WH, Vogelzang RL, et al. Current management of visceral artery aneurysms. Surgery 1996;120:627–33. [5] McDermott VG, Shlansky-Goldberg R, Cope C. Endovascular management of splenic artery aneurysms and pseudoaneurysms. Cardiovasc Intervent Radiol 1994;17:179 – 84. [6] Gabelmann A, Gorich J, Merkle EM. Endovascular treatment of visceral artery aneurysms. J Endovasc Ther 2002;9:38 – 47. [7] Reardon PR, Otah E, Craig ES, et al. Laparoscopic resection of splenic artery aneurysms. Surg Endosc 2005;19:488 –93. [8] Arca MJ, Gagner M, Heniford BT, et al. Splenic artery aneurysms: methods of laparoscopic repair. J Vasc Surg 1999;30:184 – 8. [9] Guillon R, Garcier JM, Abergel A, et al. Management of splenic artery aneurysms and false aneurysms with endovascular treatment in 12 patients. Cardiovasc Intervent Radiol 2003;26:256 – 60. [10] Mattar SG, Lumsden AB. The management of splenic artery aneurysms: experience with 23 cases. Am J Surg 1995;169:580 – 4. [11] Chiesa R, Astore D, Guzzo G, et al. Visceral artery aneurysms. Ann Vasc Surg 2005;19:42– 8. [12] Juszkat R, Rowinska Z, Oszkinis G, et al. Combined surgical-endovascular treatment of multiple splenic artery aneurysms. Eur Radiol 2004;14:2364 –5. [13] Huang IH, Zuckerman DA, Matthews JB. Occlusion of a giant splenic artery pseudoaneurysm with percutaneous thrombin-collagen injection. J Vasc Surg 2004;40:574 –7. [14] Sessa C, Tinelli G, Porcu P, et al. Treatment of visceral artery aneurysms: description of a retrospective series of 42 aneurysms in 34 patients. Ann Vasc Surg 2004;18:695–703. [15] Abbas MA, Stone WM, Fowl RJ, et al. Splenic artery aneurysms: two decades experience at Mayo clinic. Ann Vasc Surg 2002;16:442–9. [16] Saltzberg SS, Maldonado TS, Lamparello PJ, et al. Is endovascular therapy the preferred treatment for all visceral artery aneurysms? Ann Vasc Surg 2005;19:507–15. [17] Trastek VF, Pairlero PC, Joyce JW, et al. Splenic artery aneurysms. Surgery 1982;91:694 –9. [18] Moore E, Matthews MR, Minion DJ, et al. Surgical management of peripancreatic arterial aneurysms. J Vasc Surg 2004;40:247–53. [19] Haan JM, Biffl W, Knudson MM, et al. Western Trauma Association Multi-Institutional Trials Committee. Splenic embolization revisited: a multicenter review. J Trauma 2004;56:542–7. [20] Killeen KL, Shanmuganathan K, Boyd-Kranis R, et al. CT findings after embolization for blunt splenic trauma. J Vasc Intervent Radiol 2001;12:209 –14.