Formation of lymphocele is a common surgical complication following kidney transplantation. The reported incidence varies from 0.6% to 51% (1–3). This large variability can be attributed to different definitions of lymphoceles, duration of follow-up, and the use of routine ultrasound (US). Lymphoceles usually occur 4 to 8 weeks after transplantation and are mostly asymptomatic and diagnosed incidentally at routine US examination (4, 5). The majority require no therapy. Lymphoceles may become symptomatic by causing compression of adjacent structures. This may lead to impaired kidney function by obstruction of urine outflow or transplant circulation, ipsilateral leg swelling including deep venous thrombosis, bladder outlet obstruction, or pain (6). Symptomatic lymphoceles require diagnostic imaging and surgical or radiological intervention leading to increased patient discomfort and prolonging hospital stay and thus increased cost. Means to minimize or prevent lymphocele formation are thus warranted.
In standard kidney transplantation, the graft is placed extraperitoneally in the iliac region (7). When kidney grafts are placed intraperitoneally, development of lymphocele is rare (8). Surgical treatment of lymphocele is usually performed by creating an internal drainage of the lymphocele to the peritoneal cavity by opening (fenestration) of the peritoneum (2). Prophylactic fenestration of the peritoneum may thus have a beneficial effect on lymphocele formation, which also has been reported in a retrospective study (6). This randomized controlled trial evaluated whether prophylactic fenestration of the peritoneum at the time of transplantation could reduce the risk of lymphoceles after kidney transplantation.
Between March 5, 2007, and May 31, 2009, a total of 130 patients were included in the study, 69 were randomized to peritoneal fenestration and 61 to the standard group (Fig. 1). One hundred twenty-eight patients were transplanted with kidneys from deceased donors, and received the same protocol-based quadruple immunosuppression consisting of induction with basiliximab 20 mg intravenously at day 0 and repeated at day 4, calcineurin inhibitor (cyclosporine 8 mg/kg/day=90 patients or tacrolimus 0.08 mg/kg/day=38 patients), methylprednisolone 540 mg at day 0 followed by 80 mg at day 1, tapered down to 20 mg by day 9, and mycophenolate mofetil 1500 to 2000 mg daily. Five patients, four in the standard group and one in the fenestration group, were later switched from calcineurin inhibitor to mammalian target of rapamycin inhibitors. None of these patients developed a symptomatic lymphocele.
One hundred twenty-four patients were treated according to the protocol and used in the analyses (66 prophylactic peritoneal fenestrations vs. 58 in the standard group, Fig. 1). Demographics and main laboratory findings are presented in Table 1. One patient in each group received a wound drain at the time of transplantation. In addition, four patients in the intervention group and six patients in the control group received a wound drain at reoperation in the study period.
In an intention to treat analysis for the primary endpoint, a total of 13 symptomatic lymphoceles were registered in the 130 patients, 10 of 61 (16.4%) in the standard group and 3 of 69 (4.3%) in the fenestration group (relative risk [RR]=0.27, 95%, confidence interval [CI]: 0.08–0.92, P=0.022).
In the patients treated per protocol, 11 of the 124 (8.9%) developed symptomatic lymphoceles, 9 (15.5%) in the standard group compared with 2 (3.0%) in the fenestration group (RR=0.20, 95% CI: 0.04–0.82, P=0.015). The symptomatic lymphoceles were diagnosed after a median of 38 days (range 17–306 days) posttransplantation. The indications for treatment of the lymphocele, type of procedure for first treatment, and results of the ultrasound examinations before treatment for lymphocele are presented in Table 2. Laparoscopic fenestration is the procedure of choice for treatment of lymphoceles in our institution, but despite this, only two patients received this as their first treatment. Five patients were treated with percutaneous drainage. The reason for choosing percutaneous drainage was recent myocardial infarction in one patient, preventing him to undergo anesthesia needed for surgery. In one patient, preoperative imaging indicated that laparoscopic surgical treatment would be difficult and drainage was preferred. In three cases, drainage was chosen because of busy schedule for the transplant surgeons. One patient was later treated with laparoscopic fenestration.
Regarding the four patients treated by open surgery, two operations were planned as open procedures (one because of previous surgery for ventral hernia and one because of simultaneous rupture of the wound). Two cases were converted from laparoscopic fenestration to open procedure.
Some patients had more than one treatment for their lymphocele. A total of seven major surgical procedures and seven percutaneous drainages were performed (standard group: seven operations/five drainages, fenestration group: no operations/two drainages).
Treatments of lymphoceles lead to 10 readmissions and a total of 43 days of hospitalization, all in the standard group. In addition, four treatments for lymphocele were carried out as an outpatient procedure, two in each group. In the diagnosis and follow-up of the lymphoceles, a total of 20 US examinations, 13 computed tomography (CT) examinations, and one magnetic resonance imaging (MRI) examination were performed (16 US/11 CT/1 MRI in the standard group compared with 4 US/2 CT/0 MRI in the fenestration group).
A total of 291 protocol-based US evaluations were performed (Table 3). Twenty-three patients (11 fenestration and 12 standard) had a reoperation involving the transplant area during the first 10 postoperative weeks, including operations/percutaneous drainage for lymphoceles, and were excluded from further US examinations as such procedures could influence the presence of fluid collections. For three of these patients in the fenestration group, fluid collections later proven to be urinomas were also excluded from analysis.
In the fifth postoperative week, the rate of patients with fluid collections was significantly higher in the standard group (29 of 44 patients, 66%) than in the fenestration group (19 of 51 patients, 37%; RR=0.57, 95% CI: 0.37–0.71, P=0.005). In addition, during the 1st and 10th postoperative week, there also was a clear trend toward higher rates of fluid collections in the standard group compared with the fenestration group (20 of 53 patients [38%] compared with 14 of 61 patients [23%]; RR=0.61, 95% CI: 0.34–1.08, P=0.085, in the 1st week, and 16 of 37 patients [43%] compared with 12 of 45 patients [27%]; RR=0.62, 95% CI: 0.34–1.13, P=0.12 in the 10th week, respectively). The majority of the fluid collections were small, but there was a trend toward higher number of large collections in the standard group (Table 3). In the fifth postoperative week, patients with fluid collections in the control group had a mean of 1.6 (1–4) collections versus 1.3 (1–3) in the intervention group (P=0.15).
Six of the 34 (17.6%) patients who had a perirenal fluid collection in the first postoperative week later developed a symptomatic lymphocele in relation to their kidney transplant compared with 4 of 80 (5.0%) of patients who did not have a fluid collection in the first postoperative week (RR=0.25, 95% CI: 0.06–0.94, P=0.063).
Graft function measured as estimated glomerular filtration rate (Cockroft and Gault formula) at 10 weeks and at 1 year postoperatively was not significantly different between the groups (Table 1).
In total, 31 patients underwent reoperations including operations for lymphocele, 17 of 58 (29%) in the standard group compared with 14 of 66 (21%) in the fenestration group (RR=0.72, 95% CI: 0.39–1.34, P=0.30; Table 4). There was a trend toward more intestinal complications in the fenestration group. One patient had no symptoms of obstruction, but CT taken 13 days postoperatively because of abdominal pain cased by a severe pancreatitis showed herniated intestine in a ruptured transplantation wound. In the one patient with perforation, a suture from closing of the wound went through a part of the intestine. In the third case, colon obstruction was cased by a large left polycystic left kidney.
Three patients in the standard group and one in the fenestration group died during the first postoperative year. There were two graft losses in the standard group because of thrombosis of the graft and one in the fenestration group because of rejection. During the first 10 weeks, two patients in the standard group and six in the fenestration group were treated for urinary tract infections (P=0.28). One patient in the fenestration group developed a severe pancreatitis.
To our knowledge this is the first randomized study to evaluate the effect of prophylactic peritoneal fenestration at time of kidney transplantation on postoperative lymphocele formation. The results indicate a reduced risk of both symptomatic lymphoceles and perirenal fluid collections. The mechanism behind this effect is probably the same as for surgical fenestration of symptomatic lymphocele, namely internal drainage of lymph into the peritoneal cavity, thus preventing accumulation of lymphatic fluid. Our findings are consistent with previous reports on the beneficial effect of prophylactic fenestration on lymphocele formation, which has been reported in nonrandomized studies in both adult and pediatric kidney transplant recipients (6, 9).
The clinical relevance of peritransplant fluid collections is largely determined by their location and size. Lymphoceles are the most common peritransplant fluid collections and usually occur 4 to 8 weeks after transplantation (4), which is in accordance with our findings. In a retrospective study, Layman et al. (6) found a reduced incidence of symptomatic lymphoceles in patients who underwent prophylactic fenestration but no decrease of the overall incidence of postoperative fluid collections. They defined fluid collections as collections larger than 2 cm that did not appear to be a hematoma at US examinations. In our study, any collection discovered during US examination was registered.
The incidence of symptomatic lymphocele in our study was relatively high compared with the reported incidence in the literature, which may limit the applicability of our results. However, there are publications with incidences of symptomatic lymphoceles as high as 26% and 12% (1, 10). In our study, one possible reason may be a relatively high dose of steroids. After autotransplantation, the occurrence of lymphoceles is rare, indicating that concomitant use of immunosuppression is an important pathogenetic factor (8). High steroid load is a risk factor for development of lymphoceles (11), and early withdrawal of steroids after kidney transplantation has resulted in a significantly lower incidence (12). On the other hand, we did not routinely use mammalian target of rapamycin inhibitors, which is associated with a higher frequency of lymphoceles compared with other immunosuppressive regimens (13–15).
Another factor in our material is that we did not routinely place a wound drain at the time of transplantation. Use of drains has been reported to reduce the incidence of fluid collections and lymphocele treatments (16).
The majority of lymphoceles are asymptomatic and incidentally discovered during routine ultrasound examinations (5). For evaluation of the effect of prophylactic fenestration, the findings of repeated ultrasound examinations in the postoperative period as in our study are important. There are no ultrasonographic criteria to differentiate a lymphocele from a seroma, urinoma, or even a hematoma (5). Therefore, it can be argued that we did not know what the fluid collections discovered by our ultrasound examinations represented. There are, however, no reasons to believe that the two groups differed regarding the true nature of the collections discovered. Previous studies indicate that between 30% and 50% of postoperative perirenal fluid collections discovered by ultrasound examinations represent lymphoceles (17–19).
There was a tendency toward more ureter complications in the fenestration group. However, we have not been able to identify a causal link between fenestration of the peritoneum and the occurrence of ureter complications. No increase in ureter complications has been reported by others who have described prophylactic fenestration (6, 9). A peritoneal fenestration could potentially torque or dislocate the kidney graft by intra-abdominal movements and increase the risk of intestinal obstruction by leaving a space for internal herniation. None of the fenestration patients had any sign of circulatory disturbance. In the patient with herniation of the intestine into a ruptured operation wound, the complication can probably be related to the wound rupture. Colon obstruction in one case in the fenestration group was caused by a large polycystic left kidney, and the situation was solved by mobilization of the colonic flexure. As this patient was transplanted with the kidney on the right side, the obstruction could not have been caused by the fenestration.
Finally, we found a trend toward increased incidence of symptomatic lymphoceles in patients who had fluid collections present in the first postoperative week. The presence of perirenal fluid in the first week might be a risk factor for the development of symptomatic lymphocele, and US control of early occurring fluid collections may be justified. This may also be an argument for the routine use of drains.
Lymphocele formation is a common complication and a significant cause of patient morbidity and discomfort after kidney transplantation. In addition, symptomatic lymphoceles also increase hospital cost. A significantly reduced incidence of symptomatic lymphoceles in the fenestration group could suggest a possible benefit of prophylactic fenestration. The results need to be confirmed in a population of transplant recipients on lower steroids and with the use of surgical drains.
MATERIALS AND METHODS
This was an open parallel-group randomized single-center study of recipients of kidney grafts from deceased donors. All kidneys used were from donors after brain death, and patients with combined transplantations were not included.
Eligible patients were adult (≥18 years of age) recipients of kidney grafts from deceased donors. Patients were prospectively and consecutively enrolled. The patients were randomized to undergo peritoneal fenestration during the transplantation or to be treated with standard surgical therapy. Patients who were included in other studies or previously had undergone extensive abdominal surgery were excluded.
The study took place at Oslo University Hospital from March 5, 2007, to June 1, 2010. The annual number of kidney transplants is 250 to 300. The rate of living donor transplantations is nearly 40%.
All transplantations were performed using standard extraperitoneal technique with the placement of the kidney graft in the left or right iliac region (7). End-to-side anastomosis of the kidney vessels to the external iliac artery/vein was preferred, and in most cases an extravesical ureterocystostomy was established. Stenting of the transplant ureter was performed only in selected cases. According to our standard procedure, wound drains were not routinely applied. In the fenestration group, a fenestration of the peritoneum was performed in a standardized way. An incision in the peritoneum parallel to the skin incision and with a length at least as long as the length of the kidney transplant was made after the transplant procedure. The peritoneal edges were not sutured to the edges to keep the fenestration open, and no interpositioning of omentum was done.
Protocol-based US examinations were performed in the 1st, 5th, and 10th postoperative week with a Siemens Sequoia (Siemens Acuson, Mountain View, CA) using a 4-MHz curved linear array transducer.
Any hypoechoic perirenal collection was registered and measured in three dimensions for volume calculation. At the time of US examination, the radiologist was blinded to the patient's randomization group.
Primary endpoint was incidence of symptomatic lymphoceles (defined as lymphoceles requiring surgical or radiological intervention) after 1 year. Secondary endpoint was prevalence of perirenal fluid collections in the 1st, 5th, and 10th postoperative week.
We also wanted to evaluate whether postoperative hematomas predispose to the formation of symptomatic lymphoceles. As hematomas cannot be reliably distinguished from other fluid collections by ultrasound (5), all identified collections in the first postoperative week were included in this analysis. Surgical complications during the first postoperative year and kidney function at 10th postoperative week and after 1 year were also registered.
Based on our experience with standard extraperitoneal technique at our institution, an incidence of 15% of symptomatic lymphocele in the standard group was expected (20). With intraperitoneal placement of the graft, lymphocele formation is rare (8), and opening of the peritoneum was thus expected to have a substantial effect on lymphocele formation. Given an anticipated effect of prophylactic fenestration with a reduction in the incidence of symptomatic lymphoceles from 15% to 1%, a sample size of 65 patients in each group would provide a power of 82% to detect this difference with P less than 0.05 (Fisher's exact test).
Block randomization was conducted in groups of 10, drawn from envelopes containing five notes from each group. The transplant surgeon on call enrolled the study participants and assigned them to fenestration or standard treatment after obtaining written informed consent.
For statistical data analyses, the SPSS software package was used (SPSS 16.0 for Windows; SPSS Inc., Chicago, IL). The incidence of symptomatic lymphoceles after 1 year and the prevalence of perirenal fluid collections in the two groups in the 1st, 5th, and 10th postoperative week were registered. The effect of fenestration on the risk of symptomatic lymphocele was calculated using RR and two-sided chi-square test. The same analysis was performed for the risk of fluid collections in the 1st, 5th, and 10th postoperative week.
In addition to the predefined endpoints, size and number of the fluid collections in the two groups were also compared using a two-tailed Mann-Whitney U test. To evaluate whether early occurrence of peritransplant fluid collections could predict development of symptomatic lymphoceles, the incidence of symptomatic lymphoceles in patients who had fluid collections in the first week were compared with the incidence of symptomatic lymphocele in the other patients using RR and two-sided Fisher's exact test. Finally, the rates of reoperated patients in the two groups were compared using RR and two-sided χ2 test. The statistical significance level was set to P less than 0.05.
The protocol and the informed consent were reviewed and approved by the by the local institutional review board (07/2885), the regional ethical committee (unique protocol ID: S-06342a) and the Norwegian data inspectorate. The transplant surgeon obtained written informed consent from each study subject before inclusion. The study is recorded at www.ClinicalTrials.gov, NCT01206868.
The authors thank Are Hugo Pripp, Unit for Biostatistics and Epidemiology, Oslo University Hospital, for valuable help with statistical considerations, and Michael Bretthauer for assistance in the preparation of the manuscript.
1. Atray NK, Moore F, Zaman F, et al. Post transplant lymphocele
: A single centre experience. Clin Transplant
2004; 18(suppl 12): 46.
2. Bailey SH, Mone MC, Holman JM, et al. Laparoscopic treatment of post renal transplant lymphoceles. Surg Endosc
2003; 17: 1896.
3. Goel M, Flechner SM, Zhou L, et al. The influence of various maintenance immunosuppressive drugs on lymphocele
formation and treatment after kidney transplantation
. J Urol
2004; 171: 1788.
4. Park SB, Kim JK, Cho KS. Complications of renal transplantation: Ultrasonographic evaluation. J Ultrasound Med
2007; 26: 615.
5. Tasar M, Gulec B, Saglam M, et al. Posttransplant symptomatic lymphocele
treatment with percutaneous drainage and ethanol sclerosis: Long-term follow-up. Clin Imaging
2005; 29: 109.
6. Layman RE, McNally M, Kilian C, et al. Does opening the peritoneum at the time of renal transplanation prevent lymphocele
formation? Transplant Proc
2006; 38: 3524.
7. Barry JM, Murray JE. The first human renal transplants. J Urol
2006; 176: 888.
8. Oyen O, Siwach V, Line PD, et al. Improvement of post-transplant lymphocele
treatment in the laparoscopic era. Transpl Int
2002; 15: 406.
9. Zaontz MR, Firlit CF. Pelvic lymphocele
after pediatric renal transplantation: A successful technique for prevention. J Urol
1988; 139: 557.
10. Zietek Z, Sulikowski T, Tejchman K, et al. Lymphocele
after kidney transplantation
. Transplant Proc
2007; 39: 2744.
11. Khauli RB, Stoff JS, Lovewell T, et al. Post-transplant lymphoceles: A critical look into the risk factors, pathophysiology and management. J Urol
1993; 150: 22.
12. Sandrini S, Setti G, Bossini N, et al. Steroid withdrawal five days after renal transplantation allows for the prevention of wound-healing complications associated with sirolimus therapy. Clin Transplant
2009; 23: 16.
13. Giessing M, Fischer TJ, Deger S, et al. Increased frequency of lymphoceles under treatment with sirolimus following renal transplantation: A single center experience. Transplant Proc
2002; 34: 1815.
14. Langer RM, Kahan BD. Incidence, therapy, and consequences of lymphocele
after sirolimus-cyclosporine-prednisone immunosuppression in renal transplant recipients. Transplantation
2002; 74: 804.
15. Valente JF, Hricik D, Weigel K, et al. Comparison of sirolimus vs. mycophenolate mofetil on surgical complications and wound healing in adult kidney transplantation
. Am J Transplant
2003; 3: 1128.
16. Derweesh IH, Ismail HR, Goldfarb DA, et al. Intraoperative placing of drains decreases the incidence of lymphocele
and deep vein thrombosis after renal transplantation. BJU Int
2008; 101: 1415.
17. Morley P, Barnett E, Bell PR, et al. Ultrasound in the diagnosis of fluid collections following renal transplantation. Clin Radiol
1975; 26: 199.
18. Pollak R, Veremis SA, Maddux MS, et al. The natural history of and therapy for perirenal fluid collections following renal transplantation. J Urol
1988; 140: 716.
19. Silver TM, Campbell D, Wicks JD, et al. Peritransplant fluid collections. Ultrasound evaluation and clinical significance. Radiology
1981; 138: 145.
20. Roine E, Bjork IT, Oyen O. Targeting risk factors for impaired wound healing and wound complications after kidney transplantation
. Transplant Proc
2010; 42: 2542.