Introduction
Obesity is a major public health issue in Western countries. The worldwide prevalence of overweight and obesity in the population has nearly tripled since 1975 and now represents 39% of the world population [1]. Obesity is a well-known cause of multiple comorbidities; the most significant is insulin resistance, which has been recognized as an integral feature of metabolic syndrome and cardiovascular diseases [2]. Moreover, obesity is presumably associated with an increased risk of complications after surgery. Although numerous studies have investigated the effects of obesity on surgical outcomes, there is still no consensus on the topic [3, 4]. While some studies have shown that obesity is a definitive risk factor for surgical complications, others have challenged this notion by demonstrating that obesity alone may not necessarily increase the risk [4–6].
Laparoscopic liver resection (LLR) may be considered a safe alternative to traditional open liver surgery, and these procedures have been increasing worldwide [7]. Three consensus conferences acknowledged the benefits of this approach when experts applied it to selected patients and encouraged its dissemination [8–10]. With the increasing application of minimally invasive surgery for benign and malignant liver lesions, the results of LLR in obese patients are of great interest. In a recent systematic review, Kwan et al. studied the impact of body mass index (BMI) on surgical outcomes in LLR [11]. Most studies did not show associations between BMI groups and intra- and postoperative complications. The authors concluded that current evidence shows that LLR in obese patients is safe; however, further studies are still needed [11].
In our everyday practice, we have been observing an increase in the proportion of obese patients undergoing LLR for various indications. Intrinsically, we feel that surgery on these patients poses a greater challenge for the surgeon during the operation and the patients on their path to recovery.
Aim
This study aimed to evaluate the short-term operative outcomes after LLR in obese patients and compare them to patients with normal weight and overweight.
Material and methods
Patients
A retrospective review of a prospectively maintained database of all consecutive patients who underwent LLR in a single tertiary referral centre was performed. Data were collected from the first LLR in this centre (April 2008) until October 2023. The surgeon responsible for implementing LLR in this centre (AI) operated the patients or supervised the procedures. All patients were discussed in multidisciplinary team meetings to assess the feasibility of the laparoscopic approach, dependent on disease characteristics and expertise. With time, a higher level on the learning curve has been reached, and difficulty scores have become available to guide the selection of more complex cases [12, 13]. Surgical techniques were applied as reported previously, and laparoscopic intraoperative ultrasound of the liver was mandatory [14–17].
Patients gave their consent that anonymous data could be used for research at the time of the surgery. The patient’s medical records were anonymized and de-identified before analysis. The local Ethics Committee waived the requirement for approval due to the retrospective nature of this study.
Inclusion and exclusion criteria
The inclusion criterion for the study was patients who underwent elective LLR for benign or malignant lesions. Only pure LLR was performed. The exclusion criteria were patients who underwent hybrid/hand-assisted resections, cyst fenestrations, or liver biopsies [18].
Definitions and clinical variables
Based on World Health Organization definitions and preoperative BMI scores, all patients were categorized into 3 groups: a normal weight group (BMI = 18.5–24.9 kg/m2), an overweight group (BMI = 25–29.9 kg/m2), and an obese group (BMI ≥ 30 kg/m2) [1]. The groups were compared regarding preoperative data and intra- and postoperative outcomes.
Patients’ demographics, preoperative clinical parameters, operative details, histopathological data, and postoperative outcomes were analysed from the database. Diagnoses were based on final pathology.
Only pure LLR was done. Absolute contraindications for the laparoscopic approach included the need for en bloc multi-visceral resection, vascular resection and reconstruction, hepato-jejunostomy, and resections for hilar cholangiocarcinoma.
The nomenclature for describing liver segments and resections is based on The Brisbane 2000 Terminology of Liver Anatomy and Resections [19]. The classification of resections (minor versus major) is based upon the largest volume of liver resection and reflects the proposal from the Louisville Statement 2008 [8]. In addition to this classification, technical major resections, which do not meet the criteria of anatomical major resections but include the technically demanding segments 1, 4a, 7, and 8, have been included [20]. Concurrent procedures are minor LLRs, which were done along with main operations (colorectal resections, lymphadenectomies, nephrectomies, adrenalectomies, and microwave or radiofrequency ablations).
The histological surgical margins for malignant lesions were defined as microscopically positive (< 1 mm, R1) or negative (R0). R0 resection was defined as the complete removal of the tumours with clear microscopic margins.
The conversion was defined as the requirement for laparotomy at any time of the procedure, except for the extraction of the resected specimen.
Postoperative 90-day morbidity and mortality were graded according to the Clavien-Dindo classification and based on the most severe complication; grades 1–2 representing minor complications that only required medical therapies as treatment, and grades 3a–5 representing complications that required radiological or surgical intervention, the use of organ support, and fatality [21].
Length of hospital stay was defined as the time from completion of the operation to discharge. Patients were discharged when the following criteria were met: to be able to tolerate oral fluids and a solid diet in order that intravenous fluid supplementation was not required, pain controlled with oral analgesics, mobility at the level to take care of themselves, and normalized or improving liver function tests.
The difficulty of the liver resection was assessed by the Iwate scoring model and the Southampton difficulty score, which are used to predict the risk of intraoperative and postoperative complications [9, 22].
Outcomes
The primary endpoint of the study was an intraoperative complication. It was described as an objective marker of a complex operation, and its key markers were blood loss of over 775 ml, unintentional damage to the surrounding structures, and conversion to an open approach [22].
Established indicators of technical difficulties, such as blood loss, blood transfusion requirements, operative time, the need for the hepatic pedicle clamping, its duration, and R0 resection, were used as surrogate endpoints [17].
The secondary endpoint was a postoperative complication, defined as morbidity and mortality according to the Clavien-Dindo classification [21]. Surrogate markers were the length of hospital stay and readmission rate.
Statistical analysis
Statistical analysis was performed using SPSS (version 29.0, IBM Corp., Armonk, NY, USA). The p-value < 0.05 was considered statistically significant.
Categorical variables were displayed as a number and percentage. The differences between categorical variables were tested using the Fisher-Freeman-Halton test. Continuous variables were expressed as median (interquartile range (IQR)) and analysed with the Kruskal-Wallis test because the distribution of data was non-normal [23]. Univariate binary logistic regression was performed [24]. The results are shown as an odds ratio (Exp [B]) with a confidence interval (CI).
Results
Preoperative characteristics and surgery-related characteristics
A total of 235 subjects were enrolled in this study. There were no underweight patients (BMI < 18.5 kg/m2). Our study population also consisted of 11 (4.9%) patients with obesity class 2 (BMI ≥ 35 kg/m2) and 2 (0.9%) patients with obesity class 3 (BMI ≥ 40 kg/m2) [25]. Due to their low frequency, they were assigned to the obese group. The preoperative characteristics of patients in the 3 groups are given in Table I, and surgery-related characteristics are given in Table II.
Table I
Preoperative characteristics of 235 patients in the 3 groups
Table II
Surgery-related characteristics of 235 patients in the 3 groups
| Variable | Normal weight (n = 72; 30.6%) | Overweight (n = 119; 50.6%) | Obese (n = 44; 18.7%) | P-value |
|---|---|---|---|---|
| Previous abdominal surgery | 29 (40.3%) | 41 (34.5%) | 21 (47.7%) | 0.287 |
| Previous liver surgery | 5 (6.9%) | 5 (4.2%) | 2 (4.5%) | 0.679 |
| Malignant liver tumour | 51 (70.8%) | 89 (74.8%) | 32 (72.7%) | 0.891 |
| Bilateral liver tumours | 2 (2.8%) | 8 (6.7%) | 4 (9.1%) | 0.785 |
| Size of the largest tumour [mm] | 40 (IQR 23–63) | 40 (IQR 27–55) | 41 (25–55) | 0.998 |
| Multiple liver tumours | 12 (16.7%) | 25 (21.0%) | 12 (27.3%) | 0.386 |
| Posterosuperior location | 22 (31.9%) | 34 (29.6%) | 11(26.8%) | 0.837 |
| Deep location | 24 (34.8%) | 31 (27.0%) | 10 (24.4%) | 0.412 |
| Proximity to the inferior vena cava | 13 (18.8%) | 21 (18.3%) | 8 (19.5%) | 0.974 |
| Major liver resection | 25 (34.7%) | 38 (31.9%) | 11 (25.0%) | 0.558 |
| Concurrent procedures | 23 (31.9%) | 53 (44.5%) | 21 (47.7%) | 0.146 |
| Median Southampton difficulty score | 3 (IQR 3–6) | 4 (IQR 3–6) | 4 (IQR 2–6) | 0.880 |
| Median Iwate score | 5 (IQR 3–9) | 5 (IQR 4–8) | 5 (IQR 3–7) | 0.705 |
| Cirrhosisa | 11 (15.3%) | 22 (18.5%) | 12 (27.3%) | 0.280 |
| Steatosisa | 7 (9.7%) | 23 (19.3%) | 19 (43.2%) | < 0.001 |
Intraoperative outcomes
Intraoperative outcomes are given in Table III. The rate of intraoperative complications among groups was statistically insignificant (p = 0.619). The greatest share of the sum was contributed by conversion and the smallest by damage to the surrounding structures. Among surrogate outcomes, transfusion was required most frequent in the normal-weight group (p = 0.036).
Table III
Intraoperative outcomes in the 3 groups
Postoperative outcomes
Postoperative outcomes are given in Table IV. The morbidity rates were insignificant. Regarding 90-day mortality, 1 patient in the normal-weight group died from ischaemic colon perforation and multiorgan failure. Two patients in the overweight group died, one of post hepatectomy liver failure and the other because of pneumonia. There was no 90-day mortality in the obese group.
Table IV
Postoperative outcomes in the 3 groups
Univariate logistic regression for intraoperative complication
Variables from Tables I and II were tested in univariate binary logistic regression for the outcome of intraoperative complications. Statistically significant variables along with BMI and the group with obesity are given in Table V.
Table V
Univariate binary logistic regression analysis for intraoperative complications
Discussion
This study evaluated the impact of obesity on the short-term outcomes after LLR and refuted our clinical supposition that LLR on patients with obesity was associated with more complications. The results showed that short-term outcomes were similar among the normal weight, the overweight, and the obesity group.
A few decades ago, obesity was generally considered a contraindication for laparoscopic surgery due to attributed technical difficulties. However, recent studies have shown that laparoscopic surgery can be considered a standard procedure in patients with obesity [11, 26]. To date, several studies have examined the relationship between obesity and perioperative outcomes in LLR, but there is still a significant degree of heterogeneity, and applicability to the Western population is questionable [11]. Overall, surgical outcomes do not seem to be overwhelmingly worse in patients with obesity, but some newer studies bring the negative impact of obesity on LLR back into question [27, 28].
As expected, the obese group in our study had statistically higher rates of comorbidities and, specifically, higher rates of diabetes mellitus. It has been shown that this disease directly influences short-term postoperative outcomes with an increased risk of morbidity [29]. Our study did not confirm this because the obese group had similar overall and major morbidity rates as the normal weight group. These results are reflected by the similar length of hospital stay in all 3 BMI groups and similar readmission rates, which were lowest in the obese group. This could be partly because of the smaller incisions and faster wound healing in the laparoscopic approach, a well-known benefit that might be even more pronounced in patients with obesity with several comorbidities. The benefits of reduced abdominal wall trauma in LLR translate into earlier postoperative rehabilitation and may facilitate cardiopulmonary recovery in patients with obesity [28].
Patients with higher BMI have an increased risk of developing non-alcoholic fatty liver disease, a spectrum of diseases ranging from liver steatosis to cirrhosis [30]. In our study, the rate of steatosis was significantly higher (p < 0.001) in the obese group (43.2%), while it was low in the normal-weight group (9.7%) and overweight group (19.3%). However, the presence of cirrhosis did not significantly differ amongst different BMI groups. This could be due to selection bias because patients with liver cirrhosis might be less eligible to undergo laparoscopic or open liver resection. It has been shown that liver cirrhosis affects intraoperative technical difficulty and postoperative outcomes [30, 31].
Many difficulty scoring systems (e.g. the Iwate score and the Southampton difficulty score) have been implemented to predict the risk of intra- and postoperative complications after LLR [12, 13]. Increased difficulty is assumed to be associated with estimated blood loss, prolonged operative time, as well as morbidity and mortality [11]. The median Iwate score in our study did not differ among different BMI groups, meaning the difficulty of performing liver resections was similarly distributed. Consequently, we found similar rates of intraoperative complications as well as postoperative morbidity and mortality rates irrespective of the BMI. Furthermore, when we univariably tested BMI and obesity in predicting intraoperative complications, they were insignificant, and factors known from difficulty scoring systems again proved their significance (Table V) [9, 22, 32, 33].
Only 2 studies demonstrated longer operative time in the obese group than in the normal-weight group [28, 34]. Furthermore, Lee et al. only showed longer operative time in the overweight group [35]. More commonly, the results showed superiority of the laparoscopic approach over open liver surgery in terms of shorter operative time [26, 36]. It can be seen as an important benefit of laparoscopic surgery in this group of patients. Laparoscopy provides a magnified view and clearly shows small structures, such as bile ducts and blood vessels. There is better visualization of deep vascular structures and better exposure of the right hepatic lobe. These are all benefits in an obese patient with a wide, deep abdomen and a thick layer of subcutaneous fat, which can be difficult to retract for adequate exposure in the open method [11, 34].
Previously, poor surgical field exposure due to obesity, difficulty in controlling bleeding, and difficulty in developing a surgical field due to adhesions were shown to be a cause of conversion to open surgery [37]. However, our results did not show higher conversion rates in the obese group, which is accordant with current literature [27, 28, 34, 38, 39]. Conversely, Yu et al. showed a much higher conversion rate in obese patients, the main indication being unclear exposure [40]. It should be noted that this is a study based on an Asian population, where differences in the distribution of BMI among the population exist compared to Western nations. For the same BMI, there is a higher body fat percentage in Asian compared to Western patients, with a higher ratio of intraperitoneal rather than subcutaneous fat [11], which might contribute to a higher conversion rate in the Asian study of the population with obesity and other discrepancies when comparing Western and Asian studies.
Another major intraoperative complication is excessive blood loss. Only Chua et al. were able to show higher rates of intraoperative blood loss in obese patients [27]. The authors noted that their negative impact of obesity compared to other studies might be because 28% of patients underwent a major hepatectomy in their population, and more than half were in technically challenging locations, as opposed to the study by Yu et al. in which most patients underwent a minor hepatectomy for mostly benign lesions [40]. In our study, 72.7% of patients in the obese group had a malignant lesion, and the rate of major hepatectomy was 25.0%, comparable to the population in the study of Chua et al. [15, 16, 27].
Our results show that there were no higher rates of estimated bleeding intraoperatively in the obese group. Hepatic pedicle clamping is the traditional technique of controlling bleeding in liver resection, and it was not used more commonly when performing LLR on our group of obese patients [17]. In laparoscopy, pneumoperitoneum reduces bleeding from exposed vessels at the transected surface of the liver, making this another benefit of laparoscopy in obese patients with a higher number of comorbidities that lessen their cardiovascular reserve [41].
BMI has a non-specific role as an anthropometric measure of obesity. Therefore, it might explain why the distinct negative association between BMI and perioperative outcomes cannot be shown. A high BMI may be a result of increased fat-free mass rather than true visceral or peripheral adiposity [27, 42]. This questions whether other measures for body fat composition are more accurate at determining outcomes after LLR, for example, body shape index, which measures waist circumference adjusted for height and weight. Ratti et al. postulate that this index may be more useful than BMI in determining the difficulty of LLR because it more accurately assesses central obesity [43].
Apart from intraoperative complication, a textbook outcome (absence of intraoperative adverse events of grade II or higher, postoperative bile leak grade B or C, severe postoperative complications, readmission within 30 days after discharge, in-hospital mortality, and the presence of R0 resection margin) and benchmarking have been emerging as desired treatment outcomes and study endpoints that could be a future perspective of our work [44–46].
This study has some limitations. Due to its retrospective nature, there is bound to be some selection bias, especially in the context of selecting obese patients with many comorbidities for surgery. The rate of extremely obese patients was too low to create a separate group. What is more, this is a single-institution study with a modest sample size, which can have some effect on the calculation of regression analyses and the interpretation of our results.
Conclusions
Surgeons will encounter more and more overweight and obese patients. Therefore, it is important to understand the effect of elevated BMI on the outcomes of these patients. Our results suggest that obesity is not a significant, strong risk factor for worse short-term outcomes, and LLR may be considered also in patients with overweight and obesity.
