Trends in laparoscopic splenectomies (LSs).
Operative outcome by spleen size. LS indicates laparoscopic splenectomy; OS, open splenectomy; and HALS, hand-assisted LS.
Percentages of conversion from laparoscopic splenectomy to open splenectomy by spleen size and time. *P<.05.
Perioperative outcomes of estimated blood loss (A), intraoperative blood transfusion (B), reoperation for bleeding (C), and average operative time (D) for laparoscopic splenectomy vs open splenectomy by spleen size. Error bars indicate SD. *P<.05 for laparoscopic splenectomy vs open splenectomy.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Grahn SW, Alvarez J, Kirkwood K. Trends in Laparoscopic Splenectomy for Massive Splenomegaly. Arch Surg. 2006;141(8):755–762. doi:10.1001/archsurg.141.8.755
Copyright 2006 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2006
During the past 10 years, expertise with minimally invasive techniques has grown, leading to an increase in successful laparoscopic splenectomy (LS) even in the setting of massive and supramassive spleens.
Retrospective series of patients who underwent splenectomy from November 1, 1995, to August 31, 2005.
Academic tertiary care center.
Adult patients who underwent elective splenectomy as their primary procedure (n = 111).
Main Outcome Measures
Demographics, spleen size and weight, conversion from LS to open splenectomy, postoperative length of stay, and perioperative complications and mortality. Massive splenomegaly was defined as the spleen having a craniocaudal length greater than 17 cm or weight more than 600 g, and supramassive splenomegaly was defined as the spleen having a craniocaudal length greater than 22 cm or weight more than 1600 g.
Eighty-five (77%) of the 111 patients underwent LS. Of these 85 patients, 25 (29%) had massive or supramassive spleens. These accounted for 40% of LSs performed in 2004 and 50% in 2005. Despite this increase in giant spleens, the conversion rate for massive or supramassive spleens has declined from 33% prior to 1999 to 0% in 2004 and 2005. Since January 2004 at our institution, all of the massive or supramassive spleens have been removed with a laparoscopic approach. Patients with massive or supramassive spleens who underwent LS had no reoperations for bleeding or deaths and had a significantly shorter postoperative length of stay (mean postoperative length of stay, 3.8 days for patients who underwent LS vs 9.0 days for patients who underwent open splenectomy; P<.001).
Despite conflicting reports regarding the safety of LS for massive splenomegaly, our data indicate that with increasing institutional experience, the laparoscopic approach is safe, shortens the length of stay, and improves mortality.
Laparoscopic splenectomy (LS) was first described in 1992 by Delaitre et al.1 Since that time, several studies2-4 have demonstrated the advantages of the laparoscopic approach over open splenectomy (OS), including shorter hospital stay, decreased blood loss, faster recovery, and better quality-of-life outcomes. Success with normal-sized and mildly enlarged spleens has led to increasing use of LS in patients with splenomegaly. Historically, splenomegaly was considered a contraindication for LS, but recent studies5-7 have shown that LS is feasible with enlarged spleens. Targarona et al5 showed a reduced transfusion requirement and shorter length of stay following laparoscopic vs open resection of enlarged spleens (>1000 g) with a 20% conversion rate, albeit with the expected increase in operative time.
Laparoscopic resection of massive spleens (>600 g) has several inherent challenges, including limited working space, difficulty with retrieval, adherence of adjacent organs, and potential trauma to enlarged veins or the splenic capsule resulting in bleeding. Any of these problems may necessitate conversion to an open approach. For patients with massive spleens (20-30 cm), Poulin and Mamazza8 reported a 17% conversion rate and found that 83% of patients required transfusion. In a recent study,9 7 patients with massive spleens underwent successful laparoscopic resection; however, the safety of this approach has been disputed.10 Patel et al10 reported a 23% conversion rate, a 10-fold increase in morbidity, and prolongation of the hospital stay with the laparoscopic approach in 27 patients with massive spleens.
Our study describes a 10-year experience with LS and compares the laparoscopic and open approaches, with special consideration to spleen size, diagnosis, and changing patterns over time.
Between November 1, 1995, and August 31, 2005, 189 patients underwent splenectomy for a variety of indications. There were 111 patients included in this study. Patients who were younger than 18 years or had splenectomy in the setting of trauma, bleeding varices, or another major procedure were excluded. After receiving approval from our institutional review board, a retrospective review of the medical records was conducted. Data collection included patient demographics, diagnosis, American Society of Anesthesiologists classification scores, operative details, and postoperative morbidity and mortality.
Splenic size was determined by preoperative imaging with computed tomography or ultrasonography or by intraoperative measurement. Splenic weight was the aggregate weight of the morcellated and aspirated splenic tissue as measured in the pathology laboratory. Patients were classified according to the greater of spleen size or weight. Since our laparoscopic extraction technique included vacuum suction, weight of the specimen may have been underestimated. Among patients who received transfusions immediately preoperatively, the length estimated by elective computed tomography was often inaccurate when compared with intraoperative measurement using an umbilical tape and ruler. The normal spleen weighs 150 to 200 g and has a craniocaudal length shorter than 12 cm; thus, spleens weighing less than 600 g or with a craniocaudal length shorter than 17 cm were considered nonmassive. Spleens weighing 600 to 1600 g or with a craniocaudal length of 17 to 22 cm were classified as massive, and spleens weighing more than 1600 g or with a craniocaudal length greater than 22 cm were considered supramassive.7
Postoperative length of stay was calculated from the number of days until discharge or transfer from the surgery service. Cases that began laparoscopically but were converted to an open approach were included in the laparoscopic cohort in the final analysis on an intention-to-treat basis.
Statistical analysis was performed using analysis of variance with Holms t test or χ2 analysis, with P<.05 indicating statistical significance.
Patients typically received Haemophilus influenza type B, meningococcal, and pneumococcal vaccines at least 2 weeks prior to surgery in elective cases or prior to discharge if splenectomy was done on an urgent basis. For patients with enlarged spleens on physical examination, preoperative evaluation typically included a computed tomographic scan to determine the spleen size and position relative to the midline to plan port placement and patient positioning. All of the patients had a preoperative complete blood cell count, and in patients with infiltrative diseases such as myelofibrosis or lymphoma, coagulation study results were obtained to exclude impaired liver synthetic function.
In cases of massive splenomegaly, the patient was placed in the supine position with Velcro straps and wide cloth tape across the chest and pelvis to allow for steep changes in table position. The stomach was decompressed with an orogastric tube, and a Foley catheter was placed. Local anesthesia was instilled into the umbilical fold, and a 10-mm port was placed with a semiopen technique. If the spleen crossed the midline, a Hasson cannula was inserted instead in the right midabdomen and the scope was inserted through it. A 5-mm port was placed in the left midaxillary line at the inferior pole of the spleen. A rolled 4 × 4-in radiopaque gauze sponge tied at both ends with umbilical tape (Kittner roll gauze) was introduced via the 10-mm port and used with a grasper to elevate the inferior pole of the spleen and expose the hilum. A 10-mm port was placed in the left upper quadrant in a direct line with the splenic hilum, typically in the left midclavicular line. Care was taken to keep the angle shallow enough to permit introduction of a stapler or clips across the hilar vessels. An additional 5-mm port was placed in the midepigastrium. Four ports were typically required, but a fifth port occasionally may have been necessary to move the scope port up to the costal margin for better visualization of the superior short gastric vessels or diaphragmatic attachments.
The short gastric vessels were divided using the LigaSure (Valleylab, Boulder, Colo) or harmonic scalpel beginning at the level of the inferior pole of the spleen. Using the Kittner roll gauze to gently and atraumatically elevate the inferior pole, the lienocolic ligament was then taken down, working closely to the spleen to avoid injury to the colon. Division of the superior short gastric vessel may have been deferred until after the hilar division to facilitate exposure. Occasionally, the left lateral segment of the liver was scarred to the superior pole of the spleen and required division using a stapler if a plane could not be developed.
The splenic artery was identified, preferably just proximal to its bifurcation, and carefully skeletonized for about 2 cm. A large locking hemoclip was placed on the artery both proximally and distally followed by 11-mm titanium clips inside the hemoclips to prevent slippage, and the artery was divided. This allowed for better visualization of the splenic vein and the pancreatic tail. A plane was developed underneath the distal artery, and the vein was carefully skeletonized and divided in a similar fashion. This approach minimized the risk of injury to the pancreas. With massive spleens, the splenic vein was sometimes too large to be occluded by clips, in which case we used the vascular load (2.0 mm) of a stapler. It is important to recognize that the vein wall can be quite thin and staples larger than 2.0 mm may allow staple line leakage to occur. In patients with lymphoma or other infiltrative disorders, enlargement of the hilar lymph nodes and scarring to the vessels may have precluded safe dissection of the individual vessels. In such patients, we circumferentially dissected the hilar structures distal to the pancreatic tail and used a stapler across the pedicle of hilar tissue (2.5-3.0 mm depending on the extent of adenopathy). Suspicion that the tail of the pancreas may have been included in the staple line prompted placement of a closed-suction drain at the conclusion of the procedure. The retroperitoneal and diaphragmatic attachments, and sometimes the superior short gastric artery, were then taken down using electrocautery and the harmonic scalpel. We used argon beam coagulation for any oozing from the perisplenic fat, which may occur in patients with severe thrombocytopenia, or for oozing from the splenic capsule or diaphragm.
A hand port may facilitate dissection in situations where adjacent organs are scarred onto the spleen and planes must be developed, for manipulation of a supramassive spleen, or if troublesome bleeding develops. We preferred the upper midline location to avoid interference with camera angles.
Once the spleen was been mobilized, an umbilical tape was introduced into the abdomen and the spleen was measured to determine the size of the specimen bag required. The umbilical tape was then guided around the “waist” of the spleen and a snug knot was tied on the hilar side for use as an atraumatic handle during spleen manipulation. This was especially helpful for massive spleens in which the attached tissue was too weak to allow for controlled manipulation of the specimen. The spleen was grasped and moved toward the pelvis, all of the Kittner roll gauze was removed, and a thick-walled specimen bag was introduced and partially unfurled in the left upper quadrant with the end of the bag flipped up along the diaphragm. The patient was placed in the Trendelenburg position with the left side down, a grasper was locked onto the lower center lip of the bag to keep the posterior edge down, and the spleen was maneuvered into the bag using a second grasper locked onto the knot of the umbilical tape. Any identified accessory spleens were placed into the specimen bag. The mouth of the bag was then “flowered” out through the 10- to 12-mm port (or the hand port). The spleen was extracted using a combination of mechanical crushing with a ring clamp and vacuum extraction using the Synevac Vacuum Curettage suction aspirator (Berkeley Medevices, Inc, Richmond, Calif) with a 10- to 14-mm wand under continuous visualization from inside the abdomen to prevent spillage and resultant splenosis. Once the spleen was removed, the abdomen was irrigated and reinspected for homeostasis, with careful attention paid to the short gastric vessels and the splenic vessels.
For patients with tumors in which surgical margins were of interest or for echinococcal cysts, we placed the specimen in a thick-walled bag and retrieved it through a Pfannenstiel incision.
A baseline complete blood cell count was checked in the recovery room. Most patients resumed a normal diet the morning following surgery.
One hundred eleven adult patients underwent splenectomy for a range of indications. Eighty-five (77%) underwent an LS, and of those 85 patients, 25 (29%) had massive or supramassive splenomegaly. Patient demographics for those who underwent LS vs OS according to spleen size are shown in Table 1. Patients with nonmassive spleens who underwent OS had significantly higher American Society of Anesthesiologists classification scores and lower preoperative hematocrits than patients who underwent LS. This reflects the higher acuity among patients with thrombotic thrombocytopenic purpura who are sick at the time of the procedure. The majority of patients (12 of 25 patients) with massive or supramassive spleens had malignancy as compared with only 4 of 60 patients with nonmassive spleens (Table 2).
Despite thrombocytopenia and anemia, preoperative transfusion was avoided unless the platelet count was less than 5 × 109/L or the hematocrit was less than 24%. Early in our experience, we observed that preoperative transfusion of platelets and red blood cells seemed to increase the friability and size of the spleen, thereby increasing the operative difficulty, without increasing the circulating cell mass. Transfusion was typically delayed until the splenic artery was occluded, and in most cases, it proved unnecessary. Similarly, we observed in a few early cases that preoperative splenic artery embolization seemed to induce an inflammatory response in the hilum, reduce the integrity of the vascular wall, and render the splenic capsule more friable; thus, routine use of embolization was abandoned.
The fraction of patients with massive or supramassive spleens undergoing LS in our practice increased from 0% to 43% during the study period (Figure 1). The percentage of splenectomies attempted laparoscopically for massive and supramassive spleens increased from 28% in 2000 to 100% in 2005. Of those, 69% from 1995 to 1999 and 88% from 2000 to 2005 were able to be completed laparoscopically.
The conversion rate from LS to OS was significantly higher among patients with supramassive spleens compared with those with massive and nonmassive spleens (P<.001 and P<.009, respectively) (Figure 2). All of the patients with massive and supramassive spleens who underwent splenectomies that were converted from LS to OS had a malignant diagnosis (Table 2). The conversion rate for massive and supramassive spleens decreased over time (from 11% to 0% and 75% to 25%, respectively) (Figure 3). In fact, no case of any spleen was converted during the past 2 years.
Seventy-five percent of the LSs for all of the spleens and 70% of the LSs for the massive and supramassive spleens were performed by 2 surgeons who averaged 5 to 10 LS cases per year. Although neither of the surgeons have completed minimally invasive technique fellowships, both individuals have 10 to 15 years' experience performing advanced laparoscopic procedures. The rest of the LSs were performed by 9 other surgeons who each averaged fewer than 2 LS cases per year. The 2 surgeons who performed the majority of LSs had a 15% conversion rate for massive and supramassive spleens, whereas the conversion rate was 66% for those surgeons who performed fewer than 2 LSs per year. For patients with nonmassive spleens, the conversions typically occurred during the first years that LS was performed or when LS was done by surgeons who performed fewer than 2 LSs per year, suggesting that the learning curve may have impacted these conversion rates. Use of the hand port for removal of massive spleens peaked in 2004. The hand port was significantly more likely to be used for cases involving massive spleens than for cases involving nonmassive spleens (P<.009). The fraction of cases in which the hand port was used was actually higher for massive spleens than for supramassive spleens (33% vs 14%, respectively) (Figure 2).
In general, patients who underwent OS had a 2.5-fold increase in blood loss and were 3 times more likely to require transfusion than were those patients who underwent LS (Figure 4A and B). Among patients with supramassive spleens, there was no difference in estimated blood loss among patients treated with LS vs OS. The data for patients who underwent LS were skewed by 1 patient with myelodysplasia and sarcoma with coagulopathy in whom the blood loss was estimated at 4.7 L. Excluding that patient, the blood loss in the 2 cohorts was similar (mean blood loss, 707 mL for the LS cohort vs 770 mL for the OS cohort). This is consistent with prior reports.5 We also found that patients who underwent LS were much less likely to require reoperation for bleeding than those who underwent OS (1% vs 12%, respectively; P<.02) (Figure 4C). None of the patients with massive or supramassive spleens who underwent LS in this series required reoperation. As expected, operative times were longer for cases completed laparoscopically (mean operative time, 233 minutes for LS in supramassive spleens vs 128 minutes for OS in supramassive spleens) (Figure 4D).
Laparoscopy reduced hospital stay by an average of 5 days for the entire series of patients (mean ± SD hospital stay, 2.9 ± 2.1 days for LS vs 8.3 ± 9.1 days for OS; P<.05). Patients with massive splenomegaly treated with LS stayed in the hospital for an average of 8 days fewer than those who underwent OS (Table 3). The length of stay was shorter with LS as compared with OS in comparable spleen size groups, even in those patients who had procedures converted from LS to OS. This may reflect surgeon preference and patient selection.
No patients died after LS. Among patients who underwent LS, 2 (3%) underwent reoperation, one for bleeding and the other for a colonic perforation. Both patients had nonmassive spleens and were treated early in the series. Of the patients who underwent LS, 2 (3%) had postoperative pneumonia and 1 (1%) had an intra-abdominal fluid collection requiring drainage. Among patients who underwent OS, 3 (12%) died during their hospital stay. Thrombotic thrombocytopenic purpura was the cause of death in 2 patients who had cerebral edema and herniation, and the third patient had profound acute hemolysis and multiple organ failure in the setting of myelofibrosis. Of those patients who underwent OS, 3 (12%) required reoperation for bleeding, 2 (8%) had deep venous thrombosis, and 1 (4%) developed postoperative pneumonia.
During the past 10 years, the practice of removing the spleen at our institution has evolved so that 40% of spleens removed during the past year were massive or supramassive. This trend reflects both expansion of spleen size indications related to increased surgeon experience and concomitant referral bias. Our data show that laparoscopy is a safe and effective technique for the removal of massive and supramassive spleens. Although operative times were longer among patients who underwent LS, the length of stay, transfusion requirements, and mortality were reduced. Our current practice is to attempt all elective splenectomies laparoscopically regardless of spleen size.
Reduced blood loss remains an advantage of laparoscopy, particularly for splenic surgery in which there is a significant risk of massive bleeding. Bleeding risk during splenectomy arises from technical factors related to resection of a highly vascularized organ with thin-walled veins and to the underlying thrombocytopenia in the majority of patients.11 As a whole, we found that patients treated with LS had reduced volume of blood loss, reduced transfusion requirements, and fewer reoperations for bleeding as compared with patients who underwent OS. These advantages likely result from the entire dissection being performed under direct vision using electrocautery without blunt dissection, which may lead to capsular tears. Laparoscopy also offers the advantage of a magnified and illuminated view of the hilum, a “clean, well-lit room” in which to work.
As expected, the hand port was used more commonly for massive spleens than for nonmassive spleens, but surprisingly, it was used less commonly for supramassive spleens. The hand port was frequently used during the period of transition toward larger spleens that were approached laparoscopically. The hand port provides an extra measure of control of the splenic hilar vessels and facilitates manipulation of very large spleens.12,13 Use of the hand port did not increase the length of stay in this series.
The largest spleens were among patients with myelofibrosis and malignancies. Comorbid conditions such as liver insufficiency may adversely affect outcome after LS in these patients.14 Careful preoperative evaluation with particular attention to hepatic synthetic function is essential in reducing perioperative complications.
In summary, the laparoscopic approach should be considered for patients requiring elective splenectomy regardless of spleen size. For patients with nonmassive and massive splenomegaly, LS has the advantages of reduced blood loss, transfusion requirements, reoperations, length of stay, and mortality. The trends in our practice over time suggest that with sufficient experience, even supramassive spleens can be removed safely using minimally invasive techniques.
Correspondence: Kimberly Kirkwood, MD, Department of Surgery, University of California, San Francisco, Box 0790, 521 Parnassus Ave C341, San Francisco, CA 94143-0790 (email@example.com).
Accepted for Publication: April 13, 2006.
Previous Presentation: This paper was presented at the 77th Annual Meeting of the Pacific Coast Surgical Association; February 19, 2006; San Francisco, Calif; and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.
Paul D. Hansen, MD, Portland, Ore: The authors have produced a thoughtful review of the data from the UCSF [University of California, San Francisco] experience with splenectomy, and the manuscript provides a straightforward technique for performing a safe, laparoscopic splenectomy as well as a hand-assisted splenectomy. They provide further details regarding the size of the spleen and its relationship to the indication for the procedure, alterations in surgical technique, perioperative complications, and the overall outcomes.
The authors have described what I believe is a classic transition from the practice of open splenectomy to a practice of preferred minimally invasive splenectomy with selected utilization of hand assistance or open procedures. A review of the literature today will show strong support for and general agreement with the findings of this paper.
When the laparoscopic approach to splenectomy was first introduced in the early 1990s, it was performed with great caution and careful patient selection. There was a long list of contraindications that were preferred, which included obesity, prior surgery, portal hypertension, thrombocytopenia, and splenomegaly, to name a few.
Over the years, with increasing experience and improved technologies, these contraindications have not fallen off of the list, but they have been placed into a broader context. Splenic size will have some obvious as well as less obvious impacts on the performance of a safe and efficient procedure. Manipulation of the spleen, access to the ligamentous attachments, increasing size of the vasculature, and extraction all become more difficult to manage. In addition, the consequences of bleeding and capsular tear can become much more difficult to manage.
A good deal of judgment is required in considering whether the goals of the operation could be safely met with a minimally invasive technique, and we would stress that an insightful knowledge of our own capabilities and limitations must be used when we are deciding whether to perform a purely laparoscopic approach, a hand-assisted approach, or an open approach and which will best serve the good of the patient. We are not wed to the technique but to the optimal treatment of the patient. We see all 3 techniques as part of our armamentarium, and we carefully need to select which one is going to serve the patient at any given time.
I believe in the discussion and conclusion of this paper; Dr Kirkwood and her colleagues have come to that same conclusion. With all of this in mind, I have tried to evaluate this paper from the perspective of a hands-on surgeon asking, “How can I actually apply the information in this paper to my own practice?” While the paper brings up a number of possible topics worth discussing, I would like to make 4 specific comments about splenectomy and ask the author questions with each of those comments.
The first, the decision to convert from a laparoscopic procedure to a hand-assisted or an open procedure would depend upon the operating surgeon's skill set and experience. With that in mind, can you comment on a specific finding or events that would systematically instigate a conversion to a hand-assisted or an open procedure?
Second, along those same lines, in a large group practice such as UCSF or many of the practices up and down the West Coast, there are different levels of interest in this operation. There are different skill sets coming into the procedure. Did you make any attempt to correlate the management decisions with the experience of the lead surgeon, meaning, were there some surgeons who were more likely to open and what specific factors might have led them to do so?
Third, with a supermassive spleen, the 2-kg spleen and up, intraoperatively, splenic manipulation is difficult and standard techniques for morcellation become increasingly impractical. We use an infraumbilical hand port at the outset of selected cases and have found it to facilitate both a safe dissection and the extraction process while not significantly impairing the recovery process, and that is consistent with the data that you have provided here in your paper.
Is there a size or a condition in which it simply makes sense to start with a hand in the abdomen? There has got to be some size at which you just say, you know what, this is just too big. And, I agree that we will typically start every procedure with a scope first.
Finally, in our hands, the presence of portal and splenic hypertension with vascularized ligaments or splenic or gastric varices can be a challenging obstacle to safely overcome. You have not commented directly on venous hypertension. Did you encounter this, and if so, what steps were used to manage it?
John A. Ryan, MD, Seattle, Wash: I know the brunt of the paper was to compare laparoscopic vs open, but I was just interested because you had such a large experience in these big spleens, did you have problems with splenic vein thrombosis going into the portal system afterwards?
Sunil Bhoyrul, MD, La Jolla, Calif: Is this significantly different regarding your technique for a normal-sized spleen? Do you still do the normal-sized spleens supine, or would you put a normal-sized patient lateral, and then finally, I guess you sort of answered this question, but is there a point when it is too big with that spleen in the right iliac crest? Are you really going to try that laparoscopically?
Dr Kirkwood: The first question pertained to routine indications to either start open or to convert. I would say the 1 routine indication to start open is where time is of the essence, so for us, that would be a trauma spleen where you are losing blood. We didn't include those patients in this series because we felt it was a different patient population with different issues. Having said that, I have been called to the operating room during, for example, a laparoscopic fundoplication in which there was a splenic injury. I have typically completed those procedures laparoscopically. I think it is a matter of sort of how quickly are you losing blood. Is the patient unstable? You clearly lose something in terms of time to complete the procedure laparoscopically, so that would be the 1 indication to start open.
That sort of goes to Sunil's question as well and yours [Dr Hansen’s] about whether or not there is a too-big size. Anymore, I would say quite nearly, no. If you can get a scope in, even if it has to be an open cannulation in the right midabdomen, then I usually will put a scope in. What I find is that even if you can just get the short gastrics down laparoscopically, you have saved yourself a world of hurt when you open because that is usually such an incredibly narrow little space in a deep hole. So we start with the notion that virtually all the time, if we can get a port in safely anywhere in the abdomen, then we will start laparoscopically. We will go as far as we can go, and then typically we will put in a hand port if necessary. Routine indications to convert to a hand port then would typically be either we have caused some sort of vascular injury which has not been easy to work with laparoscopically and we need to get a hand in to control the hilum, essentially Pringle the hilum, or in cases where the patient has malignancy, the spleen can be so heavy that we don't have minimally invasive techniques available, even using a cigarette or a paddle, to really effectively manipulate the spleen.
If I need to put a hand port in, I usually use a gel port and I don't typically put it in first because I find that it is in the way. Although, the infraumbilical use, as you suggested, can make that less of a burden. I usually prefer the upper midline location since it affords excellent hand control of both the hilum and the apical attachments to liver, stomach, and diaphragm, which can be quite scarred. But, once you put the hand port in, it's kind of a barrier and can be in the way, and so I put it in only when we actually need it.
You asked whether certain surgeons, due to varying experience, converted at different times. Well, clearly, that answer is yes, and you know this to be true from your own experience. In our practice, most of the massive and supermassive spleens are targeted to 1 of a very few practitioners with a lot of experience doing those laparoscopically. For patients who have nonmassive spleens that need to be removed, there is a wider variety of practitioners who will do those, and that does contribute to a higher conversion rate. I think that the big difference you see is in the likelihood of converting for bleeding because over time, you develop a lot of tricks for being able to deal with bleeding laparoscopically, including use of a variety of techniques to improve exposure, rapid irrigation devices, and argon, etc. So, where I saw the biggest effect of procedural volume was in the likelihood of converting for bleeding. Then, I think there was an occasional bowel injury or something like that, and if the practitioner is not comfortable suturing intracorporeally, then that should of course be an indication to convert the procedure.
The other question you asked was regarding venous hypertension. I didn't attempt these initially because I was a little concerned about the possibility of sudden and terrifying hemorrhage. Over time, however, we have done more of these, and I would say the biggest difference I see in those patients is that the pressure in the short gastrics is so much higher that you can't count on the harmonic scalpel or even the LigaSure to taking them effectively, so I typically use a mechanical means of taking the vessels. So, we use plastic Hemolock clips or, if necessary, a stapler. Because that pressure in that system tends to be high and the vessels are dilated and very thin walled, we are more inclined to control them mechanically rather than using the cautery.
Dr Ryan's question regarding splenic vein thrombosis: this has not been a clinical problem, presumably due to adequacy of venous runoff through the pancreas.
I will tell you a sort of variation on that theme that I have seen that has become an issue is with preoperative arterial embolization. We find that typically the vessels in the hilum then disintegrate and there is an inflammatory response in the hilum that becomes a big problem. We actually initially started with arterial embolization thinking that it would reduce blood loss in removing massive spleens and found that it created much more havoc than it did save us anything. So, that would be the corollary to that I would raise and just point out that I would avoid the arterial embolization preoperatively in these patients because it seems to not only cause an inflammatory response in the hilum but also reduces the integrity of the spleen and the splenic capsule, precipitating bleeding.
Sunil, 1 other question you asked that I wanted to address was the differences in doing the normal-sized spleens and the massive spleens, and I think when you were with us we were doing the normal spleens right lateral decubitus left side up, and we still do, and so the majority of spleens that we take out for ITP [idiopathic thrombocytopenic purpura], for example, are done in decubitus position. It really does offer you the great advantage of having the spleen be the obvious target in the field. You don't have to go search through the fat, and you can elevate the spleen off of the pancreatic tail and thereby reduce the likelihood of injury to the tail of the pancreas, which I think is an advantage.