Objectives: Lymphocele is a well-known postoperative surgical complication after kidney transplant. In this study, our aim was to analyze incidence, risk factors, and outcomes of posttransplant lymphocele in a large cohort.
Materials and Methods: This observational study included 395 consecutive patients (219 males and 176 females) who underwent kidney transplant procedures from 183 living and 212 deceased donors in our center between January 2007 and 2014. A lymphocele was diagnosed with ultrasonography.
Results: The incidence of lymphoceles in our cohort was 31.9% (n = 126). There were no significant dif-ferences with regard to body mass indexes, age of donors, deceased donor ratios, acute rejection episodes, and history of abdominal surgery between those with and without lymphoceles. The pre-transplant serum albumin levels (3.29 ± 0.67 vs 3.48 ± 0.69 g/dL; P = .009) in the lymphocele group and diabetes mellitus ratios (15.9% vs 4.5%; P < .001) in the nonlymphocele group were lower than levels shown in the other group. The lymphocele ratio in patients who received cyclosporine was higher than that shown in patients who did not received it (37.5% vs. 27.4%; P = .032). There was no difference in lymphocele incidence between patients who were taking and those who were not taking mammalian target of rapamycin inhibitors, mycophenolate mofetil, or mycophenolate sodium. In regression analysis, presence of diabetes mellitus, transplant from deceased donors, older age of donors, and lower albumin levels were independent risk factors for posttransplant lymphocele occurrence.
Conclusions: Posttransplant lymphocele was a relatively common surgical complication in our cohort. We concluded that diabetes mellitus, use of kidneys from deceased donors, older donor age, and hypoalbuminemia were independent risk factors for lymphocele development.
Key words : Renal transplantation, Risk factors, Surgical complication
One of the most frequently occurring surgical com-plications after kidney transplant is development of lymphocele, which is a fluid collection between the allograft and the bladder.1 Lymphocele formation can occur secondary to leakage of lymph from the lymphatics surrounding the large pelvic vessels.
Its incidence varies and can range from 0.6% to 33.9%.2-5 The incidence of lymphocele formation remains high despite more accurate surgical techniques, reduction of other complications, and improvements in general outcomes.6,7 Several studies have investigated factors associated with the increased occurrence of clinically significant perinephric fluid collections and lymphocele after kidney transplant.2,8,9 In this study, we evaluated the incidence of lymphoceles, risk factors concerning lymphocele development, and outcomes according to lymphocele size in kidney transplant recipients.
Materials and Methods
For this study, we analyzed 395 consecutive reci-pients (219 males and 176 females with age ranging from 5 to 75 years, from 183 living and 212 deceased donors) who underwent kidney transplant in our center between January 2007 and December 2013. Patients demographics and clinical and laboratory data were obtained retrospectively from medical records.
Among 183 living donors, there was 1 living unrelated (altruistic) donor in the nonlymphocele group. This transplant was made after approval from the Regional Ethics Committee. There were 137 first-degree related donors (parents), which included 39 in the lymphocele group and 98 in the nonlymphocele group. There were 21 second-degree related donors (20 siblings and 1 grandmother), which included 4 in the lymphocele group and 17 in the nonlymphocele group. Of total donors, 21 were spouses who donated kidneys to husbands or wives, which included 5 in the lymphocele group and 16 in the nonlymphocele group. In addition, our donor cohort also included 1 uncle in the nonlymphocele group and 1 aunt and 1 cousin in the lymphocele group (P = .173).
All living donors underwent laparoscopic donor nephrectomy. Surgical operations were done in accordance with standard kidney transplant techniques. That is, the allograft was placed in the iliac fossa, its vessels were anastomosed to the iliac vessels, and the ureter was implanted into the bladder using the extravesical technique. All lymphatic vessels encountered during dissection of the iliac vessels were ligated. Stenting of the implanted ureter was used in all of the patients. A Jackson-Pratt drain (a closed negative pressure drain) was placed into the iliac fossa according to surgeon preference and amount of perioperative bleeding. The criterion for drain removal was a secretion of ≤ 20 mL/24 hours.
Initial immunosuppression was based on an induction therapy with basiliximab (perioperatively and on day 4) and a triple regimen with calcineurin inhibitors (cyclosporine or tacrolimus) or everolimus, mycophenolate (mofetil or sodium), and prednisolone.
Baseline ultrasonography was performed post-transplant and during rehospitalization or whenever complications occurred. If required, patients also underwent computed tomography. A symptomatic lymphocele was diagnosed with ultrasonography after patients showed presence or signs or symptoms of fluid collection causing compression of the graft ureter or the graft or iliac vessels, thrombosis of graft vessels or iliac vessels, urine retention, abdominal pain or discomfort, palpable mass, lower limb edema, or impaired graft function not being explained by other causes.
Initial aspiration was aseptically performed under ultrasonography guidance. The creatinine level of fluid aspirated from the collection was routinely measured to exclude the possibility of urinary leakage or urinoma. In addition, some patients underwent graft kidney scintigraphy, including late images, to evaluate suspicion of urinary leakage. Patients with asymptomatic lymphoceles were managed con-servatively. As the first treatment for large and symptomatic lymphoceles, percutaneous drainage placement was routinely performed under ultra-sonographic guidance. If needed, percutaneous nephrostomy was carried out for stabilization of renal functions. When percutaneous treatments failed or recurrence occurred, surgical intraperitoneal drainage with laparoscopic fenestration was performed.10 The lymphoceles were treated with conservative follow-up in 73 patients, aspiration in 26 patients, drainage in 17 patients, and fenestration in 10 patients. An open surgical drainage was not used in any of our patients.
Statistical analyses were performed with SPSS software (SPSS: An IBM Company, version 22.0, IBM Corporation, Armonk, NY, USA). The normality and homogeneity of data were evaluated by the Shapiro-Wilk test and Levine test, respectively. All numerical variables are presented as means and standard deviation. Continuous variables were compared using unpaired t test or Mann-Whitney U test in intergroup comparisons. Comparisons of numerical data among 3 groups according to lymphocele size were done with Kruskal-Wallis test, and pairwise comparison of significant variables was done with Mann-Whitney U test. Differences in proportion of categorical variables were compared with chi-square test or Fisher exact test. Lymphocele-related cate-gorical and continuous variables were determined and compared between patients with and without lymphoceles with univariate analysis. Binary logistic regression analysis was applied to all statistically significant variables to determine whether they were independent risk factors for lymphocele devel-opment. P values < .05 were considered to be significant.
The incidence of lymphocele in our cohort was 31.9% (n = 126). Lymphoceles developed in 55 patients (43.7%) within the first 2 weeks, whereas it developed at 2 to 4 weeks in 34 patients (27%) patients, at 5 to 12 weeks in 24 patients (19%), at 13 to 24 weeks in 7 patients (5.6%), and after 24 weeks in 6 patients (4.8%) posttransplant. In our cohort, 52 patients (41.3%) had asymptomatic lymphoceles. In addition, there was bleeding or hematoma in 10 patients (5.0%). The lymphocele and deceased transplant ratios increased from 16.6% and 40.4% in 2007 to 45.5% and 63.2% in 2013, respectively. However, there was no significant correlation in increased ratios of lymphocele and deceased transplant (r = 0.607, P = .148).
We observed no significant differences in age (38.5 ± 13.6 vs 37.4 ± 13.3 years), male-female dis-tribution (49 vs 127 females), body mass index (BMI) (23.3 ± 4 vs 23.5 ± 4.6 kg/m2), dialysis duration, donor age (50.8 ± 15.3 vs 48 ± 15 years), living-to-deceased donor ratio (50/76 vs 133/136), hemo-globin level (11.5 ± 1.9 vs 11.4 ± 2 g/dL), history of abdominal surgery (11.9% vs 16.3%), acute rejection ratio (11.1% vs 10.1%), rate of obesity (BMI ≥ 30 kg/m2; 5.6% vs 5.6%) and rate of being overweight (BMI of 25-29.9 kg/m2; 27.8% vs 25.7%), dialysis type, and primary disease between patients with versus without lymphoceles (Table 1). The lymph-ocele ratios were similar between patients with or without drainage system usage (P = .186). The pretransplant serum albumin levels (3.29 ± 0.67 vs 3.48 ± 0.69 g/dL; P = .009) in the lymphocele group and diabetes mellitus ratio (15.9% vs 4.5%; P < .001) in the nonlymphocele group were lower than levels shown in the other group. The pret-ransplant hemoglobin levels were comparable (11.5 ± 1.9 g/dL in the lymphocele group vs 11.4 ± 2 g/dL in the nonlymphocele group).
At the time of diagnosis, serum albumin levels in the lymphocele group (3.21 ± 0.61 g/dL) were comparable with their pretransplant levels (P > .05). Serum creatinine levels at time of development of lymphocele were 2.62 ± 2.49 mg/dL. The lymphocele ratio in the patients who received cyclosporine was higher than that shown in patients who did not receive it (37.5% vs 27.4%; P = .032). There was no difference in lymphocele incidence between patients who received or did not receive mammalian target of rapamycin (mTOR) inhibitors and mycophenolate mofetil or sodium. In regression analysis, presence of diabetes mellitus (odds ratio [OR] = 2.87; 95% confidence interval [95% CI], 1.28-6.43; P = .01), trans-plant from a deceased donor (OR = 1.99; 95% CI, 1.22-3.26; P = .006), older donor age (OR = 1.01; 95% CI, 1.00-1.03; P = .04), and lower albumin level (OR = 0.69; 95% CI, 0.49-0.98; P = .04) were indepen-dent risk factors for posttransplant lymphocele occurrence.
Patients were divided into 3 groups according to the maximum lymphocele size as measured by ultrasonography: group 1 had size < 5 cm (n = 50), group 2 had size of 5 to 10 cm (n = 66), and group 3 had size > 10 cm (n = 10). The median maximal lymphocele size of patients was 3 cm in group 1 (range, 0.8-4.8 cm), 6.5 cm in group 2 (range, 5-10 cm), and 13 cm in group 3 (range, 10.4-15). Among all 3 groups, age, sex, donor age, donor type, pretransplant albumin levels, time of lymphocele diagnosis, hemoglobin level, creatinine level, and albumin level at time of lymphocele development, history of diabetes mellitus, and treatment of lymphocele were similar (P > .05). Body mass index values of patients in group 3 (19.6 ± 3.0 kg/m2) were significantly lower than those of group 1 (22.9 ± 3.2 kg/m2; P = .013) and group 2 (24.1 ± 4.3 kg/m2; P = .005). There was a significant difference between calcineurin inhibitor types used in the groups, with 66% in group 1, 47% in group 2, and 20% in group 3 using cyclosporine (P = .013).
Lymphocele development following kidney trans-plant is a well-known surgical complication. Its frequency in our study was 31.9%, with 13.1% of patients being asymptomatic and 18.7% of patients being symptomatic. The incidence of symptomatic lymphocele after kidney transplant varies from 0.03% to 26% in the literature.1,10-12 A study from Turkey reported that the frequency of lymphocele in 362 recipients who underwent kidney transplant from living donors was 3.5% (n = 13) between 1983 and 2002.13 Another study found a lymphocele rate of 2.9% (n = 40) in 1359 kidney transplants performed between 1975 and 2001.14 These complication rates were lower than our rates. The experience of the surgeon and comorbidities play an important role in determining the risk of surgical complications, including lymphocele formation. In our study, we did not compare the surgical team or the surgeons because the same team usually performed all operations. However, if there are varied surgical teams or surgeons who perform the operation, thus performing the operation differently, this could be an important factor affecting the complication rates. The higher rates may also be associated with the surgical technique. A usually neglected factor is the ligation of donor lymphatics. Careful ligation of lymphatic vessels during graft preparation and its implantation together with postoperative drainage can signi-ficantly contribute to reducing incidence, with only 0.6% lymphocele rate in one study.3
In our study cohort, a drainage system was not used. We preferred a closed negative pressure drain (Jackson-Pratt drain) in all patients. Laparoscopic donor nephrectomy was performed in all living donors in our cohort. Laparoscopic donor nephrectomy can also decrease the number of lymphoceles, especially those who may need intervention.15 Donor neph-rectomy with an open or laparoscopic technique did not affect recipient lymphatic drainage in another study.16 Lymphatics should be sealed carefully with bipolar coagulation during laparoscopic donor nephrectomy, as well as lymphatics of the recipient site. A recent randomized controlled trial showed that bipolar cautery of lymphatic vessels can prevent asymptomatic or symptomatic lymphocele forma-tion in living- and deceased-donor kidney transplant recipients.17 Although there were no significant differences regarding these variables in both patients with and without lymphoceles in our study, another reason that patients may present with lymphocele formation can be previous operations like kidney transplants, history of peritoneal dialysis catheter, and presence of encapsulating peritoneal sclerosis. The risk of encapsulating peritoneal sclerosis in patients with chronic kidney disease who were treated with long-term peritoneal dialysis has been shown to be increased after kidney transplant.18
Different diagnostic techniques may also cause discrepancies in incidence. Most studies have reported symptomatic patients, and screening is not usually regularly done. The routine use of post-operative ultrasonography of grafts in kidney transplant recipients has increased the number of small and asymptomatic lymphoceles detected in our cohort. The lymphocele rate increased from 16.6% in 2007 to 45.5% in 2013 due to increased use of ultrasonography. However, some urinary leakages can be misdiagnosed as lymphoceles when only the lymphocele sample is evaluated. We excluded this possibility by using kidney scintigraphy in suspected cases together with analysis of aspiration fluid. Kidney scintigraphy is the best imaging modality for diagnosis of urinary leakage. However, because the urinoma is small and adjacent to kidney or bladder, planar images can be false-positive due to increased blood flow of abdominal wall. Use of single-photon emission computed tomography in differential diagnosis in selected cases can be useful for diagnosis of urinoma.19
In addition to surgical factors, medical reasons can also be associated with an increased risk of lymphocele development after transplant.12 In our study, the presence of diabetes mellitus, transplant from deceased donor, older donor age, and lower albumin levels were independent risk factors for posttransplant lymphocele formation. Diabetic microangiopathy is a risk factor for wound healing complications after transplant, and this pathologic mechanism can cause a high lymphocele incidence. Hypoalbuminemia has been reported to be an independent factor for lymphocele development.2 The decreased serum albumin concentrations at diagnosis of lymphocele in our population might be explained by loss of protein-rich lymphatic fluid and retrogression of oncotic pressure. Samhan and associates20 showed that the incidence of lymphocele was more when deceased allografts were used. This could be explained by increased inflammatory processes in these patients, where surgery is usually urgent and not routine as for living-donor transplant procedures. A retrospective study showed that postoperative lymphocele formation in grafts from donors older than 70 years was higher than in donors younger than 70 years (3.7% vs 0.5%; P = .011).21 Ulrich and associates2 showed a significant association of lymphocele formation with diabetes, tacrolimus therapy, and acute rejection in univariate analysis but only an association with diabetes after multivariate analysis.
Subclinical and clinical graft rejection and inflammation greatly enhance lymph production and leakage. This mechanism may partially mediate the effects of some immunosuppressive drugs on the incidence of lymphocele formation.7 In our series, acute rejection rate did not differ in patients with and without lymphoceles. The role of acute rejection as a risk factor for lymphocele development is still controversial. In a study from Goel and associates,4 multivariate analysis revealed acute rejection to be a risk factor for lymphocele development. Furthermore, warm ischemia time, duration of dialysis, obesity, recipient age, acute tubular necrosis, delayed graft function, and retransplant have also been associated with a greater risk of lymphocele.12 The incidence of postoperative lymphoceles in patients who were on dialysis > 15 years (20.8%) was significantly higher than that shown in patients who were on dialysis < 5 years (2.7%).22
Our present study detected no association between BMI and lymphocele development. Similarly, Behzadi and associates23 found no significant difference in the incidence of lymphocele between obese (BMI ≥ 30 kg/m2) and nonobese recipients. However, an analysis of 521 recipients showed that obese patients were more prone to development of lymphocele than recipients with high BMI (48.8% vs. 27.9%), with frequency of lymphocele occurrence increasing as BMI increased.24
Some immunosuppressive drugs such as anti-thymocyte globulin, high doses of mycophenolate mofetil (> 2 g/day), and steroids increase the risk of lymphatic complications.12 Consistent with the well-known inhibitory effects of steroids on several aspects of the wound healing process, Rogers and colleagues25 showed that steroid avoidance amel-iorated lymphocele formation and wound healing complications associated with sirolimus therapy. They also noted the likelihood of additive effects of steroids and sirolimus in inhibiting wound healing. As in many centers, we also discharge recipients on low-dose steroids, instead of the previous protocols of 30 to 40 mg/day, with oral prednisolone then tapered to 10 mg/day after 1 to 2 months and 5 mg/day after 3 to 6 months. However, because our study was retrospective, we did not evaluate the effects of this dose reduction on lymphocele frequency. Sandrini and associates26 also reported a lower incidence of wound healing problems and only 5% rate of lymphocele in patients with versus without steroid withdrawal 5 days after kidney transplant. This seems in accordance with incidence of lymphoceles reported in steroid-free patients on sirolimus regimen that ranged from 2.6% to 6.5%.25,27,28
The use of mTOR inhibitor (sirolimus and everolimus)-based treatment regimens in de novo kidney transplant can be associated with events such as lymphocele and surgical complications that affect wound healing.29,30 However, this matter is still debated.5,29,31 On the other hand, Tondolo and colleagues32 reported that the incidence of lymphocele was similar among multiple immunosuppressive regimens. In a study of 237 de novo kidney transplant recipients randomized to everolimus 1.5 or 3.0 mg or mycophenolic acid with cyclosporine and steroids,33 the 12-month incidence of lymphocele was 11.9%. Although the rate of lymphocele development was similar in the everolimus 1.5 mg and mycophenolic acid groups, patients randomized to everolimus 3.0 mg showed a significantly higher incidence of lymphocele development and need for intervention.33 In a study that compared different immunosup-pressive therapies, the incidence of lymphocele in the mTOR inhibitor group was greater than in those who received tacrolimus (19.2% vs 6.1%).34 In our study, the lymphocele ratio was higher in patients who received cyclosporine. Lymphocele development in patients who received sirolimus plus cyclosporine was shown to be significantly more often than development in patients who received sirolimus plus tacrolimus.35
Although the occurrence of perirenal fluid collections after kidney transplant is frequent, all occurrences are not symptomatic. In 26 of 118 patients (22%) with lymphocele diameter of 5 cm or greater, 8 patients had symptomatic lymphoceles requiring therapy. The incidence of symptomatic lymphoceles requiring treatment varies from 0.04 to 14.6%.7,10,36 Presently, the ultrasonography-guided percutaneous approach is the accepted initial diag-nostic modality.37 In our study, 73 patients received conservative care, 26 patients required aspirated ultrasonographic-guided percutaneous care, and 27 patients required drainage or laparoscopic fenes-tration. In our cohort, there was no graft loss. Although even a simple drainage offers a significant chance of success in lymphocele treatment, percu-taneous drainage procedures with the use of sclerosing agents have significantly improved treatment success. Tasar and associates treated 18 patients (6%) with symptomatic lymphoceles with percutaneous transcatheter ethanol sclerosis under ultrasonography guidance.38 This treatment option can be a simple, safe, and cost-effective method. On the other hand, surgical methods such as open or laparoscopic marsupialization still prevail. Recently, it has been possible to use fluoroscopic-guided lymphangiography to detect the origin of lymphatic leakage, allowing leakage to be rapidly treated with computed tomography-guided percutaneous embo-lization methods. In their study of 13 patients who received lymphangiography to investigate the location of lymphatic leakage and who did not respond to percutaneous ethanol sclerotherapy, Yildirim and associates39 treated 8 of 9 patients with confirmed lymphatic leakage with C-arm cone-beam computed tomography-guided percutaneous N-butyl-2-cyanoacrylate embolization. This procedure can be an effective treatment approach for patients with lymphoceles refractory to treatment before surgical intervention.
Treatment options for lymphoceles that develop after kidney transplant include simple aspiration under imaging control drainage with or without sclerotherapy and more invasive laparoscopic or open surgical options for fenestration of lymphocele into the peritoneal cavity. However, treatment decisions seem to depend on the center.40
Our study had several limitations. First, it was a retrospective study. Therefore, details on duration and amount of drainage could not be obtained from the electronic medical records. Of note, postoperative surgical site drainage may also be an important factor for lymphocele occurrence.
Lymphoceles are an uncommon surgical com-plication after kidney transplant that can form during the early posttransplant period. We observed that diabetes mellitus, use of deceased donors, older donors, and hypoalbuminemia were independent risk factors for lymphocele development.
DOI : 10.6002/ect.2018.0293
From the 1Department of Nephrology, the 2Department of Internal Medicine, and
the 3Department of Urology, Uludag University Faculty of Medicine, Bursa, Turkey
Acknowledgements: This paper was presented orally at 52nd Congress of the European-Renal-Association-European-Dialysis-and-Transplant-Assocation, which was held on May 28-31, 2015 in London, UK (abstract published in Nephrology Dialysis Transplantation 2015;30(Suppl 3): Meeting Abstract SP856; doi: 10.1093/ndt/gfv202.82). The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Emel Isiktas Sayilar, Uludag University Medical Faculty, Department of Nephrology, Bursa, Turkey
Phone: +90 507 964 80 90
Table 1. Characteristics of Study Recipients