A prospective comparison of transperitoneal and retroperitoneal laparoscopic nephrectomy in the extremely obese patient


Jihad Kaouk, Co-director of Robotic Surgery, Glickman Urological Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, A-100, Cleveland, OH 44195, USA. e-mail: kaoukj@ccf.org


Associate Editor

Ash Tewari

Editorial Board

Ralph Clayman, USA

Inderbir Gill, USA

Roger Kirby, UK

Mani Menon, USA


To prospectively compare the outcomes of transperitoneal laparoscopic nephrectomy (TLN) and retroperitoneal LN (RLN) in extremely obese patients, as LN in such patients (body mass index, BMI ≥ 40 kg/m2) is an accepted but technically challenging undertaking, and either approach to the kidney can be used.


Between July 1998 and August 2005, 51 consecutive patients with extreme obesity had 53 LNs at our institution (13 TLN and 40 RLN). Peri-operative data were collected prospectively in a database approved by the Institutional Review Board.


There were no statistically significant differences between the two approaches for several variables, including baseline BMI (44 vs 45; P = 0.23), muscle-splitting extraction incision length (7.0 vs 7.7 cm; P = 0.53), and intraoperative complications (none vs 5%; P = 0.99). RLN tended to cause less estimated blood loss (EBL, 150 vs 100 mL; P = 0.31), a shorter operation (190 vs 180 min; P = 0.11), larger specimen weight (682 vs 938 g; P = 0.078), lower intraoperative open conversion rate (15% vs 0%; P = 0.06), and a shorter hospital stay (53.6 vs 37.5 h; P = 0.33), although none of these variables was statistically significant at P < 0.05.


In the extremely obese patient, RLN tended to have advantages in EBL, operative duration, specimen weight, open conversion rate, and duration of hospital stay. RLN provides direct access to the renal hilum, and avoids the pannus and voluminous intra-abdominal fat encountered during TLN. These data and our experience support RLN as the technique of choice for LN in the extremely obese patient.


(transperitoneal) (retroperitoneal) laparoscopic nephrectomy


body mass index


estimated blood loss.


Extreme or morbid obesity (defined by the WHO as a body mass index, BMI, of ≥ 40 kg/m2) is a significant medical epidemic in the USA, with an increasing prevalence, to 5.1% of the adult USA population in 2002 [1]. In this population, laparoscopic nephrectomy (LN) is more challenging than in patients with a normal BMI [2], although morbidly obese patients gain a greater benefit than others from the laparoscopic over the open approach [3,4].

At our institution, both transperitoneal LN (TLN) and retroperitoneal LN (RLN) approaches to radical nephrectomy are used, and they provide an advantage over open surgery in estimated blood loss (EBL), hospital stay, analgesia requirements, and complication rates [5]. A prospective, randomized comparison of TLN vs RLN showed an advantage to RLN in the time to renal artery control, renal vein control, and overall operative duration [6]. That study showed no difference in EBL, hospital stay, complications, or analgesia requirements, but in the morbidly obese population there might be additional benefit to RLN. In patients with increased cardiopulmonary comorbidities, quicker surgery might lead to lower complication rates and a shorter hospital stay. Furthermore, RLN avoids the pannus and voluminous intra-abdominal fat that would be encountered in TLN, allowing for a more direct approach to the kidney. Thus in the present study we prospectively compared the results of TLN and RLN in morbidly obese patients.


Between July 1998 and August 2005, 51 consecutive patients with a BMI of ≥ 40 kg/m2 (range 40–77) had 53 LN procedures, 13 by standard TLN [7] and 40 by standard RLN [5]. The approach was selected at the time of surgery and was based on the surgeon’s preference. Positioning for these cases involved a 45° modified flank position for TLN, and the 90° flank position for RLN. In TLN, three 12-mm ports were used along the lateral margin of the ipsilateral rectus muscle. During the procedure, one 5-mm port was placed laterally to assist in retracting the kidney. In RLN, a 12-mm balloon port was placed off the tip of the 12th rib, while one 12-mm port was placed anteriorly and one was placed posteriorly along the costal margin. Extra-long ports were used as needed. Instruments used during surgery included a small-bowel grasper, electrocautery-equipped shears, and an electrocautery hook (Karl Storz, Kennesaw, GA, USA). Vascular control was obtained with a 10-mm endoscopic GIA stapler (Ethicon, Cincinnati, OH, USA) for the renal vein, and locking clips for the artery (Weck Closure Systems, Research Triangle Park, NC, USA).

Data before and during LN were collected prospectively by the operating surgeon in the operating room at the end of each procedure, and entered into a standard data sheet. All data after LN were collected by a research nurse before discharge from the hospital. The standard follow-up included a visit at 4–6 weeks after LN for an inspection of incision scars and to assess the serum creatinine level. CT of the abdomen, a chest radiograph, and serum creatinine level were checked at 6 months and then annually thereafter if there was no recurrence. These data were maintained in an Institutional Review Board-approved database.

Continuous data were summarized using the median and range, while categorical data were summarized using frequencies and proportions. A nonparametric Wilcoxon rank-sum test was used to compare continuous outcomes between the TLN and RLN groups; Fisher’s exact test was used to compare proportions. A two-sided significance level of 0.05 was used to test for statistically significant differences between the groups, with no adjustments for multiple outcomes. Outcomes for patients with more than one procedure were assumed to be statistically independent.


Demographic data showed no statistically significant difference in gender (five of 13 female in the TLN vs 65% in the RLN groups; P = 0.11) and mean BMI (44 vs 45, respectively, P = 0.23), but there was a significant difference in age (42 vs 61  years, respectively, P = 0.005). The tumour dimension on preoperative CT was 7.4 cm and 6.0 cm for the two groups, respectively.

The mean follow-up was 4.1 (0–51) months after TLN and 5.6 (0–19) months after RLN; none and one patient had a positive surgical margin, respectively. The one positive margin was in a grade 2, stage T3 specimen with microscopic fat invasion present at the inked fat margin. At 4 years of follow-up this patient had developed lung and liver metastatic disease.

Variables before, during and after LN are shown in Table 1; the median operative duration (from the time of incision to skin incision closure) was slightly less for RLN (180 min) than TLN (190 min) although this was not statistically significant (P = 0.11). The median EBL was 150 and 100 mL for TLN and RLN, respectively (P = 0.31). The median intraoperative i.v. fluid need was 5.0 and 3.5 L, (P = 0.22), and intraoperative urine output 425 and 300 mL, respectively (P = 0.36).

Table 1.  TLN vs RLN in 53 extremely obese patients
  • *

    number in sample if different from N.

n/N (%), or median (range) {n}*
N 13 40 
Gender, female  5 26 (65)0.11
Age, years 42 (30–73) 61 (33–76)0.005
BMI, kg/m2 44 (40–60) 45 (40–77)0.23
Operative duration, min190 (150–330) {9}180 (60–300) {30}0.11
EBL, mL150 (50–600) {11}100 (25–1800) {34}0.31
Blood transfusion  0  2 (5)0.99
Specimen weight, g682 (31–1204) {10}938 (177–1639) {27}0.078
Extract incision, cm  7.0 (2.0–9.0) {9}  7.7 (2.0–16.0) {30}0.53
With adrenalectomy  6/10 14/36 (39)0.29
Complications during LN  0  2 (5)0.99
Open conversion  2  00.057
Duration of stay, h 53.6 (15.5–96.0) {10} 37.5 (18.3–191) {32}0.33

There was no statistically significant difference in the need for transfusion, with no patients requiring transfusion after TLN and two (5%) after RLN (P = 0.99). Of these two patients, one was anaemic before RLN, requiring a transfusion of 1 unit of packed red blood cells during surgery. The other patient had bleeding from a lumbar vein that was controlled during RLN; this patient required a transfusion of 2 units of packed red blood cells during RLN.

The specimen weight had a borderline statistical significance, with a median of 682 and 938 g for TLN and RLN, respectively (P = 0.078), while the extraction incision was not statistically different (P = 0.53), at 7 and 7.7 cm, respectively.

Of those patients where the time to recovery of ambulation was known (eight TLN and 19 RLN), seven after TLN had 1 day to recovery of ambulation, vs 18 after RLN; one patient in each group took 2 days to recovery of ambulation. One patient resumed oral intake soon after RLN, five TLN and 17 RLN patients needed 1 day, two in each group needed 2 days and one after TLN needed 3 days. The mean convalescence after TLN and RLN was 3.5 and 4.8 weeks, respectively; oral and i.v. narcotic intake (mg morphine sulphate equivalents) was 79.3 and 43.5 mg, respectively.

There were no complications during TLN, although there were two during RLN (P = 0.99); these included one case of an avulsion of a right gonadal vein off of the inferior vena cava, requiring laparoscopic suturing, and one case of posterior port-site haemorrhage, managed during surgery with Carter-Thomson needle suturing. Two patients during TLN required conversion to open surgery for failure to progress, while none during RLN required open conversion. This difference was almost statistically significant (P = 0.06). Of the patients requiring open conversion, one failed to progress secondary to an inability to access the hilum due to the bulk of the centrally located tumour and lymphadenopathy, and the second was due to extensive scarring from previous episodes of pyelonephritis. The median hospital stay was similar, at 53.6 h for TLN and 37.5 h for RLN (P = 0.33).


LN provides a benefit over open surgery in obese patients [3,4]. At our institution, we are experienced in both approaches for radical nephrectomy, and we, as have other groups, have shown advantages of RLN in operative variables [6,8]. Although TLN has the advantages of a larger, clearer working space and easier anatomical landmarks, RLN provides direct access to the renal hilum and minimizes bowel handling. In the present prospective study assessing the outcomes in extremely obese patients, there were several important advantages in favour of RLN, i.e. a lower EBL, quicker surgery, larger specimen weight, lower open conversion rates and shorter stay. These trends were also apparent when the outcomes were compared excluding the two RLN patients who had two operations each, and the RLN patient who was converted to TLN. Although these trends were not statistically significant, this was not surprising as there were few patients in each group, giving a low statistical power to detect any clinically meaningful differences in outcomes. The small sample size and missing data for some patients also resulted in no adjustment for confounding factors (such as age).

Several factors give an advantage to RLN. First, quicker and easier access to the renal hilum allows for earlier vascular control and faster surgery. Second, the 90° flank positioning with ports placed away from the large pannus allow more direct port placement with improved manoeuvrability. Mobilizing the colon in TLN is also time-consuming, and might contribute to paralytic ileus. The present TLN patients required more time to resume oral intake. Last, by avoiding the peritoneum, the voluminous intraperitoneal fat can be avoided altogether.

Technical modifications are needed for either approach. In addition to the use of longer ports and instruments, RLN has several additional challenges. Finding the location of the 12th rib is difficult, and the use of an 18 G, 15-cm spinal needle assists in palpating the rib location and confirming optimal port-site placement. Also, after balloon dilating the retroperitoneal space, we often need to advance the outer portion of our 10-mm BluntipTM port (US Surgical, Menlo Park, CA, USA) below the level of the skin and into the s.c. tissue to achieve adequate port length. In addition, we occasionally need to maintain our carbon dioxide insufflation pressure at 18 mmHg, which is higher than our typical pressure of 15 mmHg, to maintain the retroperitoneal space [9].

A previous randomized, prospective study comparing TLN and RLN excluded patients with a BMI of ≥ 35 k/m2, as there was concern that morbid obesity would compromise an already limited working space in RLN [6]. This concern was warranted, although it proved to be unfounded, as the open conversion rate and operative time were less in RLN in the present series. Furthermore, other series compared LN in obese patients and found no difference in open conversion rates from those of patients with a normal BMI [10,11]. In both of these series, a threshold BMI of 30 kg/m2 was used. The present series focused on patients with morbid obesity (BMI ≥ 40 kg/m2), and therefore represents a more technically challenging group of patients.

A larger, randomized study with sufficient statistical power to detect clinically meaningful differences in outcomes between the groups would have eliminated any selection bias as to which approach was used, and in turn would have provided more definitive conclusions. However, the present series represents the largest published series of LN in the morbidly obese, and to our knowledge, is the only prospective study comparing TLN and RLN in these patients.

In the extremely obese patient, RLN tended to be advantageous in terms of EBL, operative duration, specimen weight, open conversion rate, and duration of stay. RLN provides direct access to the renal hilum, and avoids the pannus and voluminous intra-abdominal fat encountered during TLN. These data and our experience support RLN as the technique of choice in the extremely obese patient (BMI ≥ 40 kg/m2).


None declared.