The impact of body mass index on surgical outcomes of robotic partial nephrectomy

Authors


Jihad H. Kaouk, Zegarac-Pollock Professor of Surgery, Institute Vice Chair for Surgical Innovations, Director, Center for Laparoscopic and Robotic Surgery, Glickman Urological and Kidney Institute, Cleveland Clinic, 9500 Euclid Avenue/Q-10, Cleveland, OH 44195, USA. e-mail: kaoukj@ccf.org

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

“What's known on the subject?” and “What does the study add?”

Obesity is associated with higher incidence of renal cell carcinoma. Laparoscopic and robotic partial nephrectomy (RPN) was shown to be technically feasible in the obese population. In the present study we evaluated the impact of obesity on outcome of RPN, in a large cohort of patients.

In the present study, obese patients had a higher American Society of Anesthesiologists score and larger tumour size. We evaluated obesity as a categorical and a continuous variable, and we adjusted for confounding factors. We categorized obesity based according to the WHO classification of obesity. We described our technical modifications to overcome difficulties that can be encountered during the surgery. Obese patients had a higher estimated blood loss, but no difference in blood transfusion rate, operation duration or warm ischaemia time.

OBJECTIVE

  • • To assess the impact of body mass index (BMI) on the surgical outcomes of robotic partial nephrectomy (RPN).

PATIENTS AND METHODS

  • • Medical charts of 250 consecutive patients who underwent RPN at our institution between 2006 and 2010 were reviewed.
  • • Patients were categorized based on their BMI into four groups per international classification of obesity into: normal (BMI < 25 kg/m2), overweight (25–29.9), obese (30–39.9) and morbidly obese (≥40).
  • • Preoperative characteristics as well as perioperative and postoperative outcomes were analysed and compared between the groups.

RESULTS

  • • Of the 250 patients, 43 (17.2% of the entire cohort) were non-obese, 104 (41.6%) were overweight, 75 (30%) were obese, and 28 (11.2%) were morbidly obese.
  • • Groups were similar in terms of age, gender, history of previous surgery and nephrometry score (P= 0.5).
  • • Patients with higher BMI had a higher American Society of Anesthesiologists (ASA) score (median 3 for obese and morbidly obese groups vs 2 for non-obese groups; P= 0.002) and tumour size (median 3.6, 2.9, 2.5 and 2.3 cm in those who were morbidly obese, obese, overweight and with normal BMI, respectively; P= 0.005).
  • • Patients within the morbidly obese group had a higher estimated blood loss (median 250 mL) than other groups (median: 200, 200, 150 mL, respectively) (P= 0.03).
  • • No significant difference was detected between the groups in terms of operation duration, warm ischaemia time, transfusion rate and postoperative complications.

CONCLUSION

  • • Robotic partial nephrectomy represents an effective treatment modality for renal tumours providing equivalent surgical outcomes even for patients with BMI up to 60 kg/m2.
Abbreviations
ASA

American Society of Anesthesiologists

CCI

Charlson comorbidity index

EBL

estimated blood loss

IQR

interquartile range

RPN

robotic partial nephrectomy

WIT

warm ischaemia time

INTRODUCTION

Obesity represents a major health problem in industrialized countries, where its prevalence has dramatically increased over the last two decades [1]. A higher risk of developing RCC [2] and of mortality from RCC [3] has been found in obese patients than in non-obese patients. On the other hand, improved survival after partial nephrectomy has been reported in obese patients with organ-confined disease [4,5].

Renal surgery is steadily evolving towards nephron-sparing minimally invasive approaches [6]. In referral centres, laparoscopic partial nephrectomy largely duplicated the outcomes of the gold standard open technique with the added advantage of faster recovery [7]. However, the implementation of laparoscopic partial nephrectomy has been significantly limited secondary to the requirement of advanced laparoscopic skills and a considerable learning curve [8].

Robotic technology is being increasingly applied in urological surgery [9], because of unique features that are helpful in overcoming the learning curve for demanding procedures such as partial nephrectomy. Indications for robotic partial nephrectomy (RPN) have significantly expanded to include even complex renal tumours [10–12]. Early comparative studies have shown that RPN offers equivalent [10] or potentially better [11] perioperative outcomes than laparoscopic partial nephrectomy.

Although it has been suggested that the degree of obesity could be associated with increased difficulty during laparoscopic kidney surgery [13], laparoscopic partial nephrectomy had similar or even better perioperative outcomes than open partial nephrectomy in obese patients [14]. RPN has been recently shown to be safe and feasible in obese patients with a trend towards higher blood loss and greater operation duration and warm ischaemia time (WIT) [15].

It is widely recognized that BMI provides a reliable indicator of obesity and is used to screen for weight categories that might lead to health problems. The aim of the present study was to assess the impact of BMI on the surgical outcomes of patients undergoing RPN in a tertiary care institution.

PATIENTS AND METHODS

Our ongoing institutional review board-approved prospectively maintained database was used to identify the study population. The medical records of 250 consecutive patients treated with RPN from 2006 to 2010 at our institution were retrospectively reviewed. Patients' selection for RPN was non-randomized, and patients were counselled about treatment options when deemed appropriate for surgery. No randomization was performed.

Patients were categorized into four groups according to their BMI, as follows: group 1, normal weight (BMI < 25 kg/m2); group 2, overweight (BMI 25–29.9 kg/m2); group 3, obese (BMI 30–39.9 kg/m2); and group 4, morbidly obese (BMI ≥ 40 kg/m2) [16,27] .

Our standard technique for RPN has been described elsewhere [17] . Briefly, the patient is placed in the modified flank position (Fig. 2). All pressure points are padded, and use of a double arm board or taping at the ipsilateral arm to the patient's side is left to the discretion of the surgeon. The abdomen is insufflated to 15 mmHg with a Veress needle at the lateral border of the rectus at the level of the 12th rib and serves as the site of the 12 mm camera port. An 8 mm port is placed at the lateral border of the ipsilateral rectus muscle, about 3 cm below the costal margin. A second 8 mm port is placed ≈5–7 cm cephalad and medially to the anterior superior iliac spine.

Figure 2.

Picture showing the port placement in a patient with a BMI of 40 kg/m2, on the left side, note the lateral placement of the ports to avoid the shift of the pannus. C, camera port; R1, first robotic arm; R2, second robotic arm; A, assistant port; C, position of placement of the camera port in a normal BMI patient, R'1, usual placement site for first robotic arm; R'2, usual placement site for second robotic arm.

An assistant 12 mm port is placed along the lateral border of the rectus muscle in the ipsilateral lower abdominal quadrant. For right-sided cases, a 5 mm port is placed in the sub-xiphoid area for liver retraction. The robot (da Vinci surgical system, Intuitive Surgical Inc., Sunnyvale, CA, USA) is docked and placed over the patient's shoulder. A 30° down scope is utilized along with the pro-grasp robotic grasper in the left hand for retraction and dissection with either the monopolar scissors or hook in the right hand. The colon is mobilized and reflected medially and the ureter is identified, dissected and elevated anteriorly.

Dissection is carried cranially to the renal hilum which is cleared of fat and adventitial tissue; the renal vessels are identified and dissected free to allow for occlusion separately with bulldogs or Satinsky clamps. Gerota's fascia is opened and dissection is carried out along the renal capsule until the mass is exposed. Fat overlying the mass is left intact to allow for histological staging, and the area around the mass is cleared to allow for reconstruction of the kidney when the mass is removed. The laparoscopic ultrasonography probe is used to identify the depth and extent of the renal mass, and outline of dissection is identified and marked (Fig. 1). Before hilar occlusion, 12.5 g mannitol i.v. is given. The tumour is then resected along the previously scored margin using cold scissors.

Figure 1.

Renal mass of right kidney; patient's BMI = 60 kg/m2.

The base of the tumour vessels and collecting system is sutured using a running 2–0 vicryl suture with a hem-o-lock clip at the free end. The suture is brought through the renal capsule with the final throw and secured with sliding hem-o-lock clips and a knot. The renal capsule is re-approximated using a continuous, horizontal mattress 0-vicryl suture with a sliding hem-o-lock clip placed after each suture pass through the capsule. The renal hilum is unclamped and the renorrhaphy is inspected for haemostasis. The specimen is placed in a laparoscopic entrapment sac and removed from an extended lower quadrant port site. A Jackson–Pratt drain is placed through a lower lateral port.

In obese groups, extra padding is used for pressure points. The patient is placed in flank position at about 80° (vs 45–60° which is used in patients with a normal BMI), to allow the abdominal pannus and bowel loops to be displaced medially. The trocars are placed more laterally (Fig. 2), the double arm board is placed more superiorly to avoid clashing with robotic arms, and use of longer trocars is adopted as needed.

Baseline patient information (age, gender, American Society of Anesthesiologists [ASA] score, side), tumour characteristics (Fig. 1) (size, histology characteristics, nephrometry score [nephrometry score is a scoring system calculated by including five different variables, including size, endophytic or exophytic, nearness to the deepest part of the tumour to the collecting system, anterior or posterior, relation to the polar lines; each factor is given a score of 1–3 while the anterior and posterior orientation of the tumour given the letter a, p]), operative variables (operation duration, estimated blood loss [EBL], WIT, intraoperative complications) and early postoperative outcomes (length of stay, transfusion rate, change in estimated GFR (eGFR), postoperative complications, margin status) were analysed and compared between groups. Complications were recorded using the Clavien classification system [18]. Change in eGFR was recorded from baseline to latest postoperative follow-up using the Modification of Diet in Renal Disease equation [19].

The mean and SD for normal distribution and the median and interquartile range (IQR) for non-normal distributions were used. For the descriptive ANOVA between BMI and variables, the Kruskal–Wallis test (continuous variables) and Fisher's exact test (categorical variables) were used.

For continuous outcomes, the logarithms of the values were mostly used, as their distributions were not normal, the exception being change in eGFR, and a multivariable linear model was built (adjusted for age, ASA, Charlson comorbidity index (CCI), size of the tumour and nephrometry score).

For the categorical variable (i.e. postoperative complications), a multivariable logistic regression model was used and adjusted for the same variables.

RESULTS

Of 250 patients, 43 (17.2% of the entire cohort) were non-obese, 104 (41.6%) were overweight, 75 (30%) were obese, and 28 (11.2%) were morbidly obese. There was a higher ASA score in morbidly obese patients than in others (P= 0.002).

Tumour size was larger in the morbidly obese patients, with a median (IQR) of 3.6 cm (2.5–4.5), than in the normal BMI (P= 0.003), obese (P= 0.01) and overweight groups (P= 0.01).

Twelve patients had a nephrometry score of 1–4, 69 patients had nephrometry score of 5–6 and 97 patients had a nephrometry score ≥7. There was no difference between renal score categories with regard to BMI (P= 0.8). The median (IQR) nephrometry score per group was 7(5–8) for normal BMI, 6 (5–8) for overweight patients, 7 (6–9) for obese and 7 (6–9) for markedly obese patients (P= 0.5) (Table 1).

Table 1. Baseline characteristics
VariableNormal (BMI < 25 kg/m2) (N= 43)Overweight (BMI ≥ 25 to <30 kg/m2) (N= 104)Obese (BMI ≥ 30 to <40 kg/m2) (N= 75)Morbidly obese (BMI ≥ 40 kg/m2) (N= 28) P
  1. P evaluates differences among four BMI groups and from the Kruskal–Wallis test (continuous variables) and Fisher's exact test (categorical variables).

Mean (SD) age, years56.8 (14.3)60.6 (11.1)57.5 (10.8)58.1 (12.4)0.2
Male (n[%])24 (56)67 (64)42 (56)11 (39)0.1
Median (IQR) CCI0 (0–2)0 (0–1)0 (0–1)1 (0–2)0.3
Median (IQR) ASA score2 (2–3)2 (2–3)3 (2–3)3 (2–3)0.002
Median (IQR) tumour, cm2.3 (1.9–2.8)2.5 (1.8–3.4)2.9 (2.2–3.9)3.6 (2.5–4.5)0.005
Median (IQR) nephrometry score7 (5–8)6 (5–8)7 (4–9)7 (6–9)0.5

With regard to perioperative outcomes (Tables 2–4), there was no significant difference in log (operation duration) between BMI categories, after adjustment for age, ASA, CCI and tumour size and nephrometry score. Estimated blood loss was higher in morbidly obese patients than in all other groups (mean 250 mL vs 150, 200 and 200 mL for normal weight, overweight and obese patients, respectively, P= 0.04); another factor that contributed to increased blood loss was tumour size (P= 0.002).

Table 2. Perioperative outcomes
OutcomeNormal (BMI < 25 kg/m2) (N= 43)Overweight (BMI ≥ 25 to <30 kg/m2) (N= 104)Obese (BMI ≥ 30 to <40 kg/m2) (N= 75)Morbidly obese (BMI ≥ 40 kg/m2) (N= 28) P
  1. P evaluates differences among four BMI groups and from the Kruskal–Wallis test (continuous variables) and Fisher's exact test (categorical variables).

Median (IQR) EBL, mL150 (100–200)200 (100–300)200 (100–350)250 (150–375)0.1
Median (IQR) operation duration, min180 (140–200)180 (150–210)180 (150–210)200 (180–240)0.1
Median (IQR) WIT, min17 (14–24)17.5 (13.3–23.8)19 (12–26)20 (16–26)0.5
Any postoperative complications (n[%])9 (21)27 (26)27 (36)11 (39)0.2
Postoperative transfusion (n[%])5 (12)9 (9)8 (11)5 (18)0.6
Mean (SD) change in GFR (ml/min/1.73 m2) (latest – preoperative)−9.8 (19.8)−7.7 (15.6)−6.7 (16.6)−11.5 (15.8)0.6
Median (IQR) hospital stay, days3 (2.8–4.0)3 (3–4)3 (3–4)4 (3–4.3)0.4
Table 3. Association between BMI categories and log (operation duration) after adjustment for age, ASA, CCI and tumour size
VariableCoefficient (SE) P value
  1. Formula: log (OR time) =4.80+ one of the following (0 if BMI < 25; 0.13 if BMI ≥ 25 to <30; 0.07 if BMI ≥ 30 to <40; 0.17 if BMI ≥ 40) +0.004*Age − 0.007*ASA − 0.005*CCI + 0.006*Tumour size.

Intercept4.80 (0.12)<0.001
BMI, kg/m2  
 <2500.03
 ≥25 to <300.13 (0.06)
 ≥30 to <400.07 (0.07)
 ≥400.17 (0.08)
Age0.004 (0.002)0.02
ASA−0.007 (0.04)0.9
CCI−0.005 (0.02)0.8
Tumour size0.006 (0.01)0.6
Table 4. Association between BMI as a continuous variable and log EBL, WIT, hospital stay, postoperative complications, change in eGFR after adjustment for age, ASA, CCI and tumour size
VariableEBLWITHospital stayPostoperative complicationsChange in eGFR
CoefficientP valueCoefficient P valueCoefficient P valueOdds ratio P valueCoefficient (SE) P value
Intercept5.27 (0.46)<0.00012.90 (0.23)<0.0010.25 (0.39)0.5  −1.83(11.9)0.9
BMI0.10 (0.10)0.04−0.005 (0.006)0.40.006 (0.007)0.41.09 (0.9–1.1)0.4−0.09(0.22)0.7
Age−0.002 (0.007)0.4−0.003 (0.004)0.90.002 (0.004)0.71 (0.99–1.01)0.90.11 (0.15)0.5
ASA−0.15 (0.14)0.4−0.08(0.08)0.30.06 (0.08)0.52.3 (0.8–6.9)0.1−0.72 (2.65)0.8
CCI−0.01 (0.07)0.1−0.06(0.05)0.2−0.01 (0.05)0.81 (0.5–1.8)0.93.31(1.55)0.04
Tumour size0.13 (0.04)0.020.11 (0.03)<0.0010.03 (0.02)0.21.4 (1.1–2.0)0.02−1.50(0.79)0.1
RENAL score−0.019 (0.04)0.8−0.01(0.02)0.60.008 (0.02)0.70.8 (0.6–1.1)0.3−0.95(0.75)0.2

Warm ischaemia time was not significantly higher with increased BMI (P= 0.8). Mean (IQR) hospital stay was longer in the morbidly obese group, at 4 (3–4.3) days, than in all other groups (mean = 3 days), although this association did not reach statistical significance (P= 0.4).

Increase in BMI was not significantly associated with the occurrence of postoperative complications (P= 0.3), while increased tumour size (P= 0.004) and longer operation duration (P= 0.001) were correlated with the occurrence of complications on multivariate analysis. Intraoperative and postoperative complications are listed in Table 5.

Table 5. Postoperative complications according to Clavien–Dindo. Data are expressed as n (%).
ComplicationGroup 1 (N= 43)Group 2 (N= 104)Group 3 (N= 75)Group 4 (N= 28)
  • *

    Rise in serum creatinine by 50% or more, or rise in serum creatinine of ≥0.3 mg/dL.

Clavien I    
 Wound infection1 (0.96)000
 Prolonged drainage1 (2.3)3 ( 2.8)1 (1.3)1 (3.5)
 Ileus1 (2.3)2 (1.9)3 (4)1 (3.5)
 Urine retention001 (1.3)1 (3.5)
 Respiratory problems01 (0.96)5 (6.6)2 (7.1)
 Rhabdomyolysis0001 (3.5)
 Acute kidney injury+*002 (2.6)1 (3.5)
Clavien II    
 UTI01 (0.96)1 (1.3)1 (3.5)
 Thromboembolic1 (2.3)3 ( 2.8)1 (1.3)1 (3.5)
Clavien IIIb    
 Postoperative bleeding002 (2.6)0
 Thromboembolic1 (2.3)3 ( 2.8)1 (1.3)1 (3.5)
 Hernia01 (0.96)01 (3.5)

In relation to functional outcome (Table 2), there was no significant association between changes in eGFR (ml/min/1.73 m2) between groups (P= 0.3)

DISCUSSION

Obesity has been associated with increased incidence of several cancers, including of the oesophagus, pancreas, colon, breast and kidney [20]. There is no direct explanation as to the role of obesity in the development of cancer but it has been related to chronic tissue hypoxia, insulin resistance, compensatory hyperinsulinaemia, obesity-induced inflammatory response and lipid peroxidation [21], an increased concentration of adipokines that support tumour growth and a lower concentration of the tumour suppressor adiponectin [22]. In the present series there was a significant difference in tumour size between morbidly obese patients and those in other groups (P= 0.005). One explanation is that patients sent to our teritiary referral centre with higher BMI tend to be observed for longer by referring physician due to higher surgical risk. It is of note that, although the tumour size was larger in morbidly obese patients, the RENAL nephrometry score was not statistically different (P= 0.5).

Robotic surgery in obese patients is likely to be more technically demanding, with the possible need for longer trocars, decreased range of motion and an increase in the volume of retroperitoneal adipose tissue surrounding the kidney.

Only one study so far focused on the outcomes of RPN in obese patients vs non-obese patients. This comprised 97 patients and used a BMI threshold of 30 kg/m2 to categorize patients into obese and non-obese groups – almost 50% of them were obese and there was increased blood loss in the obese group, as well as a trend towards longer ischaemia and operation durations, although these did not reach statistical significance [15].

In the present study, we categorized the patient population into four groups based on BMI according to the WHO classification of obesity. The groups were equivalent in relation to age, gender distribution and CCI. The ASA score was higher in obese patients and the tumour size was larger in morbidly obese patients than in other groups, something that was also noted by Lawrence et al. [23]. We evaluated the association of increased BMI and the effect of different perioperative outcomes. The only factor that showed a significant association with an increase in BMI was the higher EBL (which was estimated using a digital calculator for a suction canister – at the end of the procedure the EBL is reviewed by both the surgeon and the anaesthetist), which is consistent with findings reported by Naeem et al. [15], and this is also true for laparoscopic partial nephrectomy studies [14]. This can be attributed to an increase in perirenal fat density, limited range of movement and larger tumour size in the group with higher BMI. In the present series, increased BMI was associated with longer operation duration after adjusting for other confounding factors, which can be explained by the increased time required to mobilize the kidney, dissection of perinephric fat before tumour excision and potentially more time used to place ports and close port sites.

Warm ischaemia time trended longer with an increase in BMI, although this did not reach statistical significance. This could be due to the extensive experience with the procedure at our institute.

There was no increase in postoperative complications, postoperative transfusion rate and hospital stay for the different groups. Nevertheless, respiratory complications were more common in the obese groups than in the overweight and normal weight groups.

Anast et al. [24] compared perioperative outcomes in obese and non-obese patients. The obese group had a longer operation duration, more blood loss and greater transfusion rate than the non-obese group; only 44 of 211 patients underwent partial nephrectomy. In a report by Feder et al. [25], analysing patients who underwent laparoscopic partial or radical nephrectomy, there was no significant difference between obese and non-obese groups with regard to EBL, operation duration, hospital stay and number of open conversions or complications. Colombo et al. [26] compared the perioperative outcome of laparoscopic partial nephrectomy in obese and non-obese patients, using a cohort of patients undergoing retroperitoneal or transperitoneal approach. There was no significant difference between groups regarding EBL, operation duration, WIT, conversion rate or hospital stay for the transperitoneal approach group.

The present study is limited by the fact that the data have been retrospectively analysed and that they are from a single centre. It is also limited by the small number of patients in the normal BMI and morbidly obese groups. Moreover, the surgeons' learning curves are included and this might account for additional bias. Lastly, this experience is from a tertiary referral centre with a high volume of partial nephrectomy procedures and therefore the current findings might not apply to other populations in different hospital settings.

In conclusion, in a tertiary referral centre, surgical outcomes of RPN in obese and morbidly obese individuals are similar to those in normal-weight individuals. Thus for patients with BMI up to 60 kg/m2, RPN represents a viable treatment option for renal masses amenable to nephron-sparing surgery.

CONFLICT OF INTEREST

Jihad H. Kaouk is a Consultant for Intuitive Surgical and a Speaker for Ethicon incorp.

Ancillary