T

to

Splenic Preserving Distal Pancreatectomy:Does Vessel Preservation Matter?

Joal D Beane, MD, Henry A Pitt, MD, FACS, Attila Nakeeb, MD, FACS,C Max Schmidt, MD, PhD, MBA, FACS, Michael G House, MD, Nicholas J Zyromski, MD, FACS,

homas J Howard, MD, FACS, Keith D Lillemoe, MD, FACS

BACKGROUND: Splenic preserving distal pancreatectomy (SPDP) can be accomplished with splenic artery andvein preservation or ligation. However, no data are available on the relative merits of thesetechniques. The aim of this analysis was to compare the outcomes of splenic preserving distalpancreatectomy with and without splenic vessel preservation.

STUDY DESIGN: From 2002 through 2009, 434 patients underwent distal pancreatectomy and 86 (20%) hadsplenic preservation. Vessel preservation (VP) was accomplished in 45 and ligation (VL) wasperformed in 41. These patients were similar with respect to age, American Society of Anesthe-siologists class, pathology, surgeons, and minimally invasive approach (79%). For comparison,a matched group of 86 patients undergoing distal pancreatectomy with splenectomy (DP�S)was analyzed.

RESULTS: The VP-SPDP procedure was associated with less blood loss than VL-SPDP or DP�S (224 vs508 vs 646 mL, respectively; p � 0.05). The VP-SPDP procedure also resulted in fewer gradeB or C pancreatic fistulas (2% vs 12% vs 14%; p � NS) and splenic infarctions (5% vs 39%;p � 0.01), less overall morbidity (18% vs 39% vs 38%, respectively; p � 0.05) and need fordrainage procedure (2% vs 15% vs 16%; p � 0.05), and shorter post-operative length of stay(4.5 vs 6.2 vs 6.6 days; p � 0.05).

CONCLUSIONS: This analysis suggests that outcomes are (1) best for VP-SPDP and (2) VL-SPDP provides noshort-term advantage over distal pancreatectomy with splenectomy. We conclude that splenicVP is preferred when SPDP is performed. (J Am Coll Surg 2011;212:651–657. © 2011 by the

American College of Surgeons)

dopw

prg

Distal pancreatectomy is the procedure of choice for be-nign, premalignant, and malignant disease of the body andtail of the pancreas. In recent years distal pancreatectomyhas become a safe procedure that is being performed morefrequently because imaging studies are identifying morebenign and premalignant lesions.1-4 For patients with ade-nocarcinoma of the pancreas, distal pancreatectomy withsplenectomy is recommended to assure that an adequateoncologic operation has been performed.5-7 However,he spleen has important immunologic functions, andverwhelming sepsis following splenectomy is a well-

Disclosure Information: Nothing to disclose.Presented at Southern Surgical Association 122nd Annual Meeting, PalmBeach, FL, December 2010.

Received December 6, 2010; Accepted December 14, 2010.From the Department of Surgery, Indiana University School of Medicine,Indianapolis, IN.Correspondence address: Henry A. Pitt, MD, Department of Surgery, Indi-

ana University School of Medicine, 535 Barnhill Dr RT 130D, Indianapolis,IN 46202. email: hapitt@iupui.edu

651© 2011 by the American College of SurgeonsPublished by Elsevier Inc.

escribed long-term complication.8,9 Thus, preservationf the spleen during distal pancreatectomy has been pro-osed to reduce the risk of post-splenectomy sepsis asell as late malignancies.10

Distal pancreatectomy with splenic preservation wasfirst described by Mallet-Guy and Vachon in 1943.11 Thisclassic technique preserves the splenic artery and vein byidentification and ligation of the multiple small, short vas-cular connections to the body and tail of the pancreas. In1988 Warshaw12 published an alternative technique inwhich the splenic artery and vein(s) are ligated proximallyand distally, with retention of the short gastric and leftgastroepiploic vessels to preserve blood flow to and from thespleen. Splenic preserving distal pancreatectomy (SPDP) hasbeen recommended by many surgeons because of fewer short-and long-term complications.13-18 In recent years splenic

reservation has received even more attention because lapa-oscopic distal pancreatectomy has been adopted by manyroups.2,4,19-25 To date, the relative merits of SPDP versus

distal pancreatectomy with splenectomy as well as open

ISSN 1072-7515/11/$36.00doi:10.1016/j.jamcollsurg.2010.12.014

wtsptmsfapam

tdttm

652 Beane et al Splenic Preserving Distal Pancreatectomy J Am Coll Surg

versus laparoscopic distal pancreatectomy have been re-ported. However, no data are available on the relativemerits of SPDP performed by the Mallet-Guy or War-shaw technique. The aim of this analysis was to comparethe outcomes of SPDP with and without splenic vesselpreservation (VP).

METHODSPatient populationPermission to review the outcomes of patients undergoingpancreatic surgery at the Indiana University Hospital wasobtained from the Indiana University School of MedicineIRB. A search of operative, pathology, and billing recordsfrom January 2002 through August 2009 identified 434patients who had undergone distal pancreatectomy.Splenic preservation (SPDP) was performed in 86 of thesepatients (20%). Of the patients undergoing SPDP, VP wasaccomplished in 45 (52%) and vessel ligation (VL) wasperformed in 41 (48%). The VP-SPDP and VL-SPDPgroups were similar with respect to age, American Societyof Anesthesiologists (ASA) class, pathology, surgeons, andminimally invasive approach (79%). For further compari-son, a group of 86 patients undergoing distal pancreatec-tomy with splenectomy (DP�S) matched for the same 5characteristics also was analyzed. None of the patients inthe VP-SPDP, VL-SPDP, or DP�S groups underwent dis-tal pancreatectomy for trauma or concomitantly with an-other major operation such as gastrectomy or colectomy.

Pancreatic operationsThe VP-SPDP11 or VL-SPDP12 procedure and DP�S

ere performed by standard techniques. Recent trends inhe relative proportion of pancreatic operations includingplenic preservation and laparoscopic and robotic distalancreatectomies have been reported elsewhere.4,26 In brief,he percentage of minimally invasive distal pancreatecto-ies has increased so that 45% were performed laparo-

copically or robotically in 2009. Seven surgeons per-ormed the 86 SPDP, whereas the 86 DP�S wereccomplished by 9 surgeons. However, the same 3 surgeonserformed 95% of the VP-SPDP, 90% of the VL-SPDP,nd 81% of the DP�S. The decisions regarding a mini-

Abbreviations and Acronyms

DP�S� distal pancreatectomy with splenectomySPDP � splenic preserving distal pancreatectomyVL � vessel ligationVP � vessel preservation

ally invasive versus an open approach, splenic preserva-

ion versus splenectomy, and VP versus VL were left to theiscretion of the surgeon. Similarly, the methods employedo transect the pancreas and close the pancreatic stump andhe decisions regarding drain placement and removal wereade by the individual surgeons.

Pancreatic pathologyAll pancreatic specimens were reviewed by members of theDepartment of Pathology at the Indiana University Schoolof Medicine. The length of the resected pancreas was re-corded for all specimens. The SPDP procedure was per-formed most commonly for premalignant lesions such asnoninvasive intraductal papillary mucinous neoplasms,mucinous cystic neoplasms, and neuroendocrine tumors,and benign pathology such as chronic pancreatitis and se-rous cystadenomas. Only 2 patients undergoing SPDP(2.3%) were determined at final pathology to have an ad-enocarcinoma of the pancreas. No significant differenceswere observed in the relative pathologic diagnoses betweenthe VP-SPDP and the VL-SPDP groups. Thus, the pathol-ogy of these 2 groups was combined during the process ofmatching SPDP with DP�S patients. In this process, therelative frequency of the lesions was maintained, and only 3adenocarcinomas were included in the DP-S group.

Surgical outcomesOperative data included blood loss and operative time.Procedure-specific post-operative morbidity included pan-creatic fistulas, surgical site infections, splenic infarctions,and hemorrhages. Pancreatic fistulas were graded A, B, orC according to the International Study Group for Pancre-atic Fistula.27 Only grade B and C pancreatic fistulas wereincluded in the calculation of overall morbidity. Superficialsurgical site infections included those above the fascial levelrequiring wound opening and drainage. Organ space infec-tions and uninfected peripancreatic fluid collections re-quiring percutaneous, endoscopic, or operative drainagewere recorded by the management procedure.The presenceof a splenic infarction was determined by review of com-puterized tomography scans performed post-operativelyfor concerns about undrained fluid collections, persistentfevers, and/or left upper abdominal or back pain. Post-operative hemorrhage was defined as the need for transfu-sion and/or reoperation. Other complications included inoverall morbidity were MI, respiratory failure requiringreintubation, acute renal failure requiring dialysis, deepvein thrombosis or pulmonary embolus requiring antico-agulation, Clostridium difficile colitis, and urinary tract in-fections. Post-operative length of hospital stay was recordedin days. Post-operative survival was determined by review

of the national Social Security database.

CDtint

a

653Vol. 212, No. 4, April 2011 Beane et al Splenic Preserving Distal Pancreatectomy

Statistical analysesContinuous variables are presented as means � SEM.

omparison of the 3 groups, VP-SPDP, VL-SPDP, andP�S, was performed by ANOVA followed by Student’s

, chi-square, or Fisher’s exact tests, as appropriate. Actuar-al survival was determined by the Kaplan-Meier tech-ique, and survival differences were analyzed by log-rankesting. Statistical significance was defined as p � 0.05.

RESULTSPatient populationDemographic data regarding the VP-SPDP, VL-SPDP, andDP�S groups are presented in Table 1. Because the groups

Figure 2. Pancreatic fistulas, grades A, B, and C. DP�S, distalpancreatectomy with splenectomy; VL-SPDP, vessel ligation–splenicpreserving distal pancreatectomy; VP-SPDP, vessel preservation–

Table 1. Patient PopulationVP-SPDP VL-SPDP DP�Sn % n % n %

n 45 41 86Age (y) 53.4 56.7 55.8Sex (% female) 78* 54 56Mean ASA class 2.66 2.78 2.84Pathology

IPMN 14 31 11 27 22 26Cystic tumors 9 20 13 32 22 26Pancreatitis 11 24 9 22 22 26NET 5 11 5 12 11 13Adenocarcinoma 2 4 0 0 3 3Other 4 9 3 7 6 7

*p � 0.03 versus VL-SPDP and DP�S.ASA, American Society of Anesthesiologists; DP�S, distal pancreatectomywith splenectomy; IPMN, intraductal papillary mucinous neoplasm; NET,neuroendocrine tumor; VL-SPDP, vessel ligation–splenic preserving distalpancreatectomy; VP-SPDP, vessel preservation–splenic preserving distalpancreatectomy.

(splenic preserving distal pancreatectomy.

were matched for age, ASA class, and pathology, no differ-ences were observed in these parameters. Significantlymore women (p � 0.03) were in the VP-SPDP group. Themost frequent pathologic diagnoses were intraductal pap-illary mucinous neoplasms, cystic tumors (mucinous cysticneoplasms and serous cystadenomas), and chronic pancre-atitis (Table 1). Only 5 adenocarcinomas of the pancreaswere included: 2 in the VP-SPDP and 3 in the DP�Sgroups. No differences were observed among the 3 groupswith respect to the mean length of pancreas resected (8.5 vs8.7 vs 8.6 cm; p � NS).

Operative outcomesThirty-two of 45 VP-SPDP (71%), 36 of 41 VL-SPDP(88%), and 57 of 86 DP�S (66%) were performed mini-mally invasively. Of these less invasive procedures, 11 VP-SPDP, 3 VL-SPDP, and 6 DP�S were performed roboti-cally, with the remainder being completed laparoscopically.Operative time and blood loss are presented in Figure 1.Operative times were approximately 30 minutes shorter forSPDP, but this difference did not reach statistical signifi-cance. Blood loss was 224 mL for VP-SPDP comparedwith 507 mL for VL-SPDP (p � 0.05) and 646 mL forDP�S (p � 0.05).

Post-operative outcomesGrade A pancreatic fistulas were common after each of the3 procedures (Fig. 2). However, only 1 VP-SPDP patient(2%) developed a grade B or C fistula (Fig. 3) comparedwith 5 VL-SPDP patients (12%) and 12 DP�S patients(14%) (Table 2). In addition, only 1 VP-SPDP patient(2%) required drainage of an infected peripancreatic fluidcollection compared with 6 VL-SPDP (15%; p � 0.05)nd 14 DP�S patients (16%; p � 0.05). Splenic infarcts

Figure 1. Operative time and blood loss. Blood loss was signifi-cantly less (p � 0.05) in the VP-SPDP patients. DP�S, distal pan-createctomy with splenectomy; VL-SPDP, vessel ligation–splenicpreserving distal pancreatectomy; VP-SPDP, vessel preservation–splenic preserving distal pancreatectomy.

Fig. 4) also were significantly less common after VP-SPDP

2pSpqrpwc(sc(1SDn

stpoi

ptmaH

pati

654 Beane et al Splenic Preserving Distal Pancreatectomy J Am Coll Surg

compared with VL-SPDP (2% vs 39%; p � 0.05) (Table). Reoperation was not required in any of the VP-SPDPatients, and all survived surgery. In comparison, 2 VL-PDP patients (5%) required reoperation, and 1 of theseatients, who bled from the splenic artery stump, subse-uently died. Whereas 5 DP�S patients (6%) requiredeoperations for infectious complications, all of the DP�Satients survived the second procedure. Overall morbidityas significantly less (p � 0.05) in the VP-SPDP (18%)

ompared with the VL-SPDP (39%) and DP�S patients38%). The mean length of post-operative stay also wasignificantly shorter (p � 0.05) for VP-SPDP (4.5 days)ompared with that of VL-SPDP (6.3 days) and DP�S6.6 days). Of the 167 patients without pancreatic cancer,57 (94%) were alive at 5 years after surgery. Only 1 VP-PDP (2%) compared with 3 VL-SPDP (7%) and 6P�S patients (7%) have died, but these differences were

ot statistically significant.

DISCUSSIONDistal pancreatectomy with splenic preservation can beperformed with splenic artery and vein preservation or li-

Table 2. Post-Operative Outcomes

Outcomes

VP-SPDP

VL-SPDP DP�S

n % n % n %

Pancreatic fistula* 1 2 5 12 12 14Splenic infarction 1† 2 16 39 NADrainage procedure 1† 2 6 15 14 16Reoperation 0 0 2 5 5 6Overall morbidity 8† 18 16 39 33 38Post-operative length of stay (d) 4.5† 6.3 6.6

*Grades B and C.†p � 0.05 versus VL-SPDP and DP�S.DP�S, distal pancreatectomy with splenectomy; VL-SPDP, vessel ligation–

Figure 3. Infected pancreatic fistulas (*) thatpreserving distal pancreatectomy and (B) vesspatient. Note the small splenic infarctions in both

splenic preserving distal pancreatectomy; VP-SPDP, vessel preservation–splenic preserving distal pancreatectomy.

gation. Although these 2 techniques were described de-cades ago, no good comparative data have been available.This analysis evaluated 45 patients undergoing VP-SPDP,41 patients who had VL-SPDP, and 86 matched patientswho underwent DP�S. Operative blood loss was signifi-cantly lower (p � 0.05) in the VP-SPDP patients. Post-operative outcomes including splenic infarcts, overall mor-bidity, the need for a drainage procedure, and length of staywere all significantly (p � 0.05) better in the VP-SPDPpatients. Outcomes for VL-SPDP were similar to those forDP�S.

With the proliferation of abdominal imaging in thepast decade, more small, asymptomatic benign or pre-malignant cystic and neuroendocrine tumors are beingdiscovered.3,4 During this same period, minimally inva-ive pancreatic surgery has emerged as a safe alternativeo open surgery.2,4,19-26 As a result, relatively more distalancreatectomies are being performed laparoscopicallyr robotically.4,26 With these trends, interest in preserv-ng the spleen has increased.16,17,28,29 However, relatively

little information comparing splenic preservation to dis-tal pancreatectomy with splenectomy has been pub-lished, and data regarding the relative merits of VP-SPDP versus VL-SPDP have not been available.

In 1991 Aldridge and Williamson13 compared 35 SPDPatients with 42 DP�S patients. The authors concludedhat SPDP can be accomplished safely, and they recom-ended more liberal application of splenic preservation. In1999 study of 235 distal pancreatectomies from Johnsopkins, Lillemoe and colleagues1 performed splenic pres-

ervation in only 37 patients (16%). In this report, mostshort-term outcomes were similar between SPDP andDP�S. In 2002 Shoup and associates16 from MemorialSloan-Kettering Cancer Center compared 46 SPDP with79 DP�S patients. They reported that SPDP patients hadfewer infections and severe complications as well as a

ed drainage in (A) vessel preservation–splenication–splenic preserving distal pancreatectomyents (arrows).

requirel lig

shorter length of stay. Similarly, in 2007 Carrère and asso-

sKuci

spf5

pr(

It

Tf8omisftti

qw

655Vol. 212, No. 4, April 2011 Beane et al Splenic Preserving Distal Pancreatectomy

ciates17 found that 38 VL-SPDP patients had fewer intra-abdominal abscesses and overall complications and shorterlength of stay than 38 matched DP�S patients.

In comparing the present series with these previous re-ports, differences include (1) a larger number of patients,(2) matching of SPDP and DP�S patients, and (3) differ-entiation of VP and VL patients. Similarities in observa-tions include reduced overall morbidity and shorter lengthof stay in patients undergoing VP-SPDP (p � 0.05). Ofnote, the length of stay was shorter in the present seriesthan in the prior reports. The potential benefit with respectto infectious complications was less clear in this report,although the need for drainage of an intra-abdominal col-lection was less (p � 0.05) following VP-SPDP. The newfindings in this report were that outcomes were better inVP-SPDP than in VL-SPDP and that VL-SPDP providedno obvious advantage over DP�S.

The observation that splenectomy increases post-operative morbidity has been reported following colec-tomy,30 and gastrectomy,31 and pancreatectomy. Experi-mental studies by Billiar and colleagues32 suggested thatplenectomy results in reduced activity of macrophages andupffer cells. In addition, studies in patients who havendergone splenectomy for trauma have suggested a de-rease in circulating T cells as well as an impaired primarymmune response.21 Thus, some data exist to support the

observations that splenectomy is associated with a short-term increase in infectious morbidity.

Another long-term risk of splenectomy is life-threatening sepsis. This risk is clearly greater in childrenthan in adults. Nevertheless, the risk in adults has beenestimated to be as high as 0.8% to 1.9%.8,9 However, whenplenectomy is planned pre-operatively, vaccinations forneumococcus, haemophilus, and meningococcus are per-ormed routinely. Thus, the observation in this report that

Figure 4. (A) Preserved splenic perfusion in a vectomy patient (arrow). (B) Large splenic infarctionatectomy patient (arrow).

-year survival did not differ between SPDP and DP�S w

atients is not surprising. However, a limitation of thiseport is that the exact cause of death in the 10 patients6%) who died is unknown.

The current report differs from prior studies in that thenternational Study Group for Pancreatic Fistula classifica-ion system27 was employed. With these guidelines, the

incidence of grade A fistulas, without clinical conse-quences, was quite high in both the SPDP and DP�Spatients. Considering that approximately 70% of patientshad cystic and/or neuroendocrine tumors in an otherwisenormal pancreas, this incidence of grade A fistulas is notsurprising. Both a soft pancreas and small pancreatic ductare known risk factors for pancreatic fistula. Interestingly,the incidence of grades B and C fistulas in the VL-SPDPand DP�S patients in this study (12% and 14%, respec-tively) are similar to those reported by Carrère and col-leagues17 (8% and 13%).

The observations with respect to splenic infarction re-ported in this series are reasonably unique. Splenic circula-tion after ligation of the splenic artery and vein has beenstudied33,34 and has generally been thought to be adequate.

he relatively high incidence of symptomatic splenic in-arction (27%) with VL-SPDP in this series may be due to8% of these procedures being performed laparoscopicallyr robotically. With open surgery, the pain from infarctionay be masked by the incisional pain. However, with min-

mally invasive surgery, after several days, the pain from theplenic infarction may become dominant and may lead tourther investigation by CT scan. Interestingly, none of thehese infarctions evolved to an abscess or became a long-erm pain issue. On the other hand, the influence of thesenfarctions on long-term splenic function is unknown.

As with all retrospective studies, this analysis raises newuestions and has limitations. One outstanding question ishether splenic function is normal following SPDP and

reservation–splenic preserving distal pancreate-vessel ligation–splenic preserving distal pancre-

ssel pin a

hether function is different with VP or VL. A related

656 Beane et al Splenic Preserving Distal Pancreatectomy J Am Coll Surg

question is whether SPDP is associated with a long-termincreased incidence of severe sepsis. A limitation is that dataregarding factors that may have predicted the risk of pan-creatic fistula, such as pancreatic texture, pancreatic ductsize, pancreatic fat,35 method of pancreatic closure, anddrain management, were not gathered. With respect toclosure techniques, 9 surgeons were involved and mini-mally invasive techniques and technology evolved signifi-cantly over the study period. As a result, multiple methodswere employed over time even by the same surgeon(s).

In summary, this analysis compared 86 patients under-going SPDP with 86 matched patients undergoing DP�S.The SPDP patients included 45 who had VP and 41 whohad VL. Operative and post-operative outcomes were bestfor VP-SPDP, and VL-SPDP provided no short-term ad-vantage over DP�S. We conclude that splenic vessel pres-ervation is preferred when SPDP is performed.

Author Contributions

Study conception and design: Beane, Pitt, Nakeeb, SchmidtAcquisition of data: BeaneAnalysis and interpretation of data: Beane, Pitt, Nakeeb,

Schmidt, House, Zyromski, Howard, LillemoeDrafting of manuscript: Beane, PittCritical revision: Nakeeb, Schmidt, House, Zyromski, How-

ard, Lillemoe

REFERENCES

1. Lillemoe KD, Kaushal P, Cameron JL, et al. Distal pancreatec-tomy: indications and outcomes in 235 patients. Ann Surg1999;229:693–698.

2. Kooby DA, Gillespie T, Bentrem D, et al. Left-sided pancreate-ctomy: a multicenter comparison of laparoscopic and open ap-proaches. Ann Surg 2008;248:438–446.

3. Spinelli KS, Fromwiller TE, Daniel RA, et al. Cystic pancreaticneoplasms: observe or operate? Ann Surg 2004;239:651–659.

4. Ziegler KM, Nakeeb A, Pitt HA, et al. Pancreatic surgery: evo-lution at a high-volume center. Surgery 2010;148:702–710.

5. Nordback IH, Hruban RH, Boitnott JK, et al. Carcinoma of thebody and tail of the pancreas. Am J Surg 1992;164:26–31.

6. Strasberg SM, Linehan DC, Hawkins WG. Radical antegrademodular pancreatosplenectomy procedure for adenocarcinomaof the body and tail of the pancreas: ability to obtain negativetangential margins. J Am Coll Surg 2007;204:244–249.

7. Balcom JH, Rattner DW, Warshaw AL, et al. Ten-year experi-ence with 733 pancreatic resections: changing indications, olderpatients, and decreasing length of hospitalization. Arch Surg2001;136:391–398.

8. Holdsworth RJ, Cuschieri A, Irving AD, et al. Postsplenectomysepsis and its mortality rate: actual versus perceived risks. Br JSurg 1991;78:1031–1038.

9. Cullingford GL, Watkins DN, Watts AD, et al. Severe late post-

splenectomy infection. Br J Surg 1991;78:716–721.

10. Mellesmkjoer L, Olsen JH, Linet MS, et al. Cancer risk aftersplenectomy. Cancer 1995;75:577–583.

11. Mallet-Guy P, Vachon A. Pancreatites Chroniques Gauches.Paris: Masson; 1943.

12. Warshaw Al. Conservation of the spleen with distal pancreatec-tomy. Arch Surg 1988;123:550–553.

13. Aldridge MC, Williamson RCN. Distal pancreatectomy withand without splenectomy. Br J Surg 1991;78:976–979.

14. Talamini MA, Pitt HA, Hruban RH, et al. The spectrum ofcystic tumors of the pancreas. Am J Surg 1992;163:117–124.

15. Benoist S, Duqué L, Sauvanet A, et al. Is there a role of preser-vation of the spleen in distal pancreatectomy? J Am Coll Surg1999;188:255–260.

16. Shoup M, Brennan MF, McWhite K, et al. The value of splenicpreservation with distal pancreatectomy. Arch Surg 2002;137:164–168.

17. Carrère N, Abid S, Julio CH, et al. Spleen-preserving distalpancreatectomy with excision of splenic artery and vein: a case-matched comparison with conventional distal pancreatectomywith splenectomy. World J Surg 2007;31:375–382.

18. Rodríguez JR, Madanat MG, Healy BC, et al. Distal pancreate-ctomy with splenic preservation revisited. Surgery 2007;141:619–625.

19. Fernández-Cruz L, Martínez I, Gilabert R, et al. Laparoscopicdistal pancreatectomy combined with preservation of the spleenfor cystic neoplasms of the pancreas. J Gastrointest Surg 2004;8:493–501.

20. Khanna A, Koniaris L, Nakeeb A, Schoeniger LO. Laparoscopicspleen-preserving distal pancreatectomy. J Gastrointest Surg2005;9:733–738.

21. Melotti GL, Butturini G, Piccoli M, et al. Laparoscopic distalpancreatectomy: results on a consecutive series of 58 patients.Ann Surg 2007;246:77–82.

22. Pryor A, Means J, Pappas TN, et al. Laparoscopic distal pancre-atectomy with splenic preservation. Surg Endosc 2007;21:2326–2330.

23. Bruzoni M, Sasson A. Open and laparoscopic spleen-preserving,splenic vessel-preserving distal pancreatectomy: indications andoutcomes. J Gastrointest Surg 2008;12:1202–1206.

24. Sa Cunha A, Rault A, Beau C, et al. A single-institution pro-spective study of laparoscopic pancreatic resection. Arch Surg2008;143:289–295.

25. Nau P, Melvin WS, Narula VK, et al. Laparoscopic distalpancreatectomy with splenic conservation: an operationwithout increased morbidity. Gastroenterol Res Pract 2009:846340.

26. Waters JA, Canal DF, Wiebke EA, et al. Robotic distal pancre-atectomy: cost effective? Surgery 2010;148:814–823.

27. Bassi C, Dervenis C, Butturini G, et al. International studygroup on pancreatic fistula definition. Postoperative pancreaticfistula: an international study group (ISGPF) definition. Sur-gery 2005;138:8–13.

28. Yoon YS, Lee HK, Han HS, et al. Patency of splenic vessels afterlaparoscopic spleen and splenic vessel-preserving distal pancre-atectomy. Br J Surg 2009;96:633–640.

29. Kimura W, Yano M, Sugawara S, et al. Spleen-preserving distalpancreatectomy with conservation of the splenic artery and vein:techniques and its significance. J Hepatobiliary Pancreat Surg2010;17:813–823.

30. Varty PP, Linehan IP, Boulos PB. Does concurrent splenectomyat colorectal cancer resection influence survival? Dis Colon Rec-

tum 1993;36:602–606.

3

3

3

3

657Vol. 212, No. 4, April 2011 Beane et al Discussion

31. Bonenkamp JJ, Hermans J, Sasako M, et al. Extended lymph-node dissection for gastric cancer. N Engl J Med 1999;340:908–914.

2. Billiar TR, West MA, Hyland BJ, et al. Splenectomy altersKupffer cell response to endotoxin. Arch Surg 1988;123:327–332.

3. Sato Y, Shimoda S, Takeda N, et al. Evaluation of spleniccirculation after spleen-preserving distal pancreatectomy bydividing the splenic artery and vein. Dig Surg 2000;17:519–522.

4. Miura F,TakadaT, AsanoT, et al. Hemodynamic changes of spleno-gastric circulation after spleen-preserving pancreatectomy with ex-cision of splenic artery and vein. Surgery 2005;138:518–522.

5. Mathur A, Pitt HA, Marine M, et al. Fatty pancreas: a factor inpostoperative pancreatic fistula. Ann Surg 2007;246:1058–1064.

Discussion

DR KEVIN BEHRNS (Gainesville, FL): This paper addresses thepractical issue of splenic preservation in the context of distal pancre-atectomy. The authors have reviewed their nicely balanced series of86 total patients, 45 with vessel preservation, and 41 without vesselpreservation. And they matched these patients to a group of patientswith distal pancreatectomy and splenectomy. The results unequivo-cally show that vessel preservation results in fewer complications interms of pancreatic fistulas, splenic infarction, infectious complica-tions, and overall morbidity. The conclusion that splenic vasculatureshould be maintained is straightforward.

However, I have several questions.What proportion of the 86 patientshad a preoperatively planned vessel ligation? That is, were all 41 patientswith vessel ligation intraoperative decisions based on anatomy and/ordisease, or is vessel ligation really a proxy for degree of difficulty of thecase? If a decision for vessel ligation is made intraoperatively, how manypatients had vessel ligation but required splenectomy during the indexoperation, so are not included in this series?

Obviously, if vessel ligation offers no advantage over distal pancre-atectomy and splenectomy, when should vessel ligation be performedversus just distal pancreatectomy and splenectomy? Finally, what isthe follow-up with patients with vessel ligation and splenic infarc-tion? Do they receive immunizations? And what are they told aboutthe risk of postsplenectomy sepsis?

DR DAVID ADAMS (Charleston, SC): My discussion, Mr VicePresident, Mr Secretary, members, and guests, will please you, I hope,for it begins in rhyme from Zeta, better known to most as Sir ZacharyCope:“The pancreas lies deeply at the backof the secluded lesser serous sac.The duodenum circles round its head,its narrow tail towards the spleen doth spread.The body lies upon the bony bed of vertebra L1, though in betweenthe cava and aorta intervene.If of this plain anatomy you make the fullest use

and note the mischief ferments make when once they may get loose

the symptoms of disease you can remember once for all.And at the needed moment, you can easily recall.”Two questions to these fine authors I will now ask,for they relate to the anatomy to which Cope put us to task.The serpentine splenic artery courses in and out of the superior

pancreatic bed.The splenic vein leaves the spleen and tunnels through the pancreas

from which it is fed.Preservation, in my view, is possible only in the most favorable

situationswhen the anatomy is safe and easy and the patient slender, fit, and low

risk for complications.Are the favorable results from splenic vein and artery preservation

that you reportrelated not to preservation anatomic, but to patient selection, a fre-

quent study tort?For are not patients in whom one saves the artery, vein, and spleen

trilogythose who have favorable anatomy and physiology?No low albumin, renal failure, duct dilation, pancreatic fibrosis, nor

cancerare found in patients whom vein and artery are spared is my answer.Are not the favorable results from saving vein and artery continuitypatients with good protoplasm and easy technical feasibility?Perhaps there is a multivariate analysis and data somewherethat releases me from my skepticism and needless care.And the second question, which I promised, is tell me, please,in which patients can the vessels be saved with ease,and which the spleen taken out without a second thought and one

replyrelinquish extra emotional and physical work and the 2/2 modifier

do not apply.Thank you for your fine work that pleases meand the opportunity to recite rhyme, and remember, this is not

poetry.

DR GARY VITALE (Louisville, KY): I have just 1 question. In mypractice, most of the distal pancreatectomies are for pancreatitis. And Inoticed, in your abstract, 20% of your patients had pancreatitis. I’minterested in the issue of long-term pain after subtotal pancreatectomyfor the pancreatitis patients; did you notice any difference in that subgroup?

I know it is anecdotal, but in those with pancreatitis who had asplenectomy and removal of the vasculature, as opposed to those whohad the spleen left in, I have had the impression that when I havedone vessel preservation, there has been more postoperative chronicpain. I have come to the conclusion that the peripancreatic inflam-mation and inflammatory response involving the peripancreaticnerves may be responsible for this. So I have been more inclined as oflate to do pancreatectomy and splenectomy in this subgroup of pa-tients. I wonder about your thoughts and what your data might showfor the chronic pancreatitis subgroup.

DR NORMAN MCSWAIN (New Orleans, LA): In the traumaworld, we have very little opportunity to preserve the splenic artery.But there’s always the question of whether the spleen should bepreserved by letting the short gastrics maintain it. The general phi-

losophy of trauma surgeons, I believe, is, it’s not worth their while to