Introduction

Radical prostatectomy and cystectomy are curative treatment options for localized and locally advanced prostate and bladder cancers. In recent years, robot-assisted radical prostatectomy has evolved as the predominant surgical approach for radical prostatectomy due to comparable postoperative recovery and surgical outcomes [1], and lately, the number of robot-assisted radical cystectomy is also rising even though the level of evidence for a comparable surgical outcome has been moderate [2]. However, robot-assisted and laparoscopic techniques not only require positioning the patient at a 45° head-down tilt in either a steep Trendelenburg or lithotomy position, but also accompany peritoneal insufflation for laparoscopic vision of the surgical field.

Several studies have reported the steep Trendelenburg and capnoperitoneum to exert adverse effects on intracranial pressure and cerebral circulation [3, 4]. The gravitational effect of the steep Trendelenburg position is thought to divert blood away from the lower extremities and increase the central blood volume, which then increases the cerebral blood flow and the intracranial pressure by impairing the venous outflow from the brain [5]. With increased blood volume, the hydrostatic pressure within the cerebral vasculature is thought to increase and push the fluid into the extracellular space, which could lead to the development of cerebral perivascular edema [6]. This understanding has led to earlier studies investigating whether alterations in the cerebrovascular function due to the steep Trendelenburg position translates to neurological defects after the surgery [7,8,9]; however, the results have been conflicting: the steep Trendelenburg was associated with postoperative cognitive dysfunction in one study [7] but not in others [8, 9].

Conventionally, factors including advanced age, the American Society of Anesthesiologists (ASA) physical status, cognitive impairment, duration of anesthesia, hypoalbuminemia, and preoperative IGF-1 levels have been reported to be associated with postoperative delirium [10]. Nevertheless, studies on the potential role of preoperative brain MRI features as prognostic markers for postoperative delirium have been limited [11, 12]. Furthermore, in these studies [11, 12], volumetry analysis of the total brain, grey and white matter volumes were mainly analyzed with less attention to other brain MRI features relevant to the small vessel disease burden and cognitive impairment.

Therefore, the primary objective of this study was to explore the association between the steep Trendelenburg position and postoperative delirium. Additionally, we aimed to investigate other potentially significant risk factors for postoperative delirium by analyzing preoperative brain MRI features and clinico-demographic data in patients undergoing major urologic surgery.

Materials and methods

Study patients

This retrospective study was approved by the Severance Institutional Review Board (IRB No.2023–1856), and the need to obtain informed consent was waived owing to its retrospective nature. In total, 187 patients who had received a major urologic surgery defined as one of open, laparoscopic, and robot-assisted radical cystectomy or prostatectomy due to bladder or prostate cancer between January 2005 and July 2021, who had also undergone brain MRI within one year prior to the date of the surgery were identified. Patients whose brain MRI included only diffusion-weighted imaging (n = 2), or MR angiography (n = 3) were excluded, and the remaining 182 patients were included in the final analysis (Fig. 1).

Fig. 1
figure 1

Flow chart of study patients. DWI, diffusion weighted imaging; MRI, magnetic resonance imaging

Full size image

Demographic and clinical characteristics of patients, such as age, sex, body mass index (BMI), comorbidities (e.g., hypertension, type 2 diabetes, hyperlipidemia, atrial fibrillation, coronary artery occlusive disease, and cerebral infarction), current smoking status, surgical factors such as position (e.g., steep Trendelenburg vs. supine), crystalloid and colloid infusion, estimated blood loss, duration of anesthesia and surgery, and the preoperative ASA physical status score recorded by anesthesiologists were extracted from the electronic medical records. The protocol for assessing postoperative delirium, MRI acquisition, and for assessing small vessel disease burden and mesial temporal atrophy on preoperative MRI can also be found in the Supplementary Material.

Study outcome

The study outcome was the incidence of postoperative delirium after a major urologic surgery, and the association of patient’s surgical position, preoperative brain MRI features, and clinico-demographic factors with postoperative delirium.

Statistical analysis

Quantitative variables are presented as means and standard deviations while qualitative variables are presented as absolute and percentage frequencies. All analyses were performed using R version 3.4.3. (R Foundation for Statistical Computing). Patients’ characteristics were compared using Student’s t test for near-normal continuous variables, Mann–Whitney U test for other continuous variables, and the Chi-square test for categorical variables unless more than 20% of cells had expected frequencies < 5 in which case Fisher’s exact test was used. The univariate logistic regression analysis was used to identify potential independent predictors of postoperative delirium. Variables significant in the univariate analysis were further analyzed in the multivariate analysis. For all analyses, P-values were two-sided and P < 0.05 was considered statistically significant.

Results

Study patients

Patients’ characteristics are summarized in Table 1. A total of 182 patients were included in the final analysis. The mean age of the patients was 67.9 ± 8.0 years (range: 21.0–84.0 years) and 174 (96%) patients were men. Among these patients, 145 (80%) patients had one or more comorbidities, and hypertension was the most frequent comorbidity present in 120 (66%) patients. Most patients (163 patients, 90%) had a preoperative ASA physical status of II or III. In addition, 128 (70%) patients had undergone a robot-assisted or laparoscopic radical cystectomy or prostatectomy, and thus, had received surgeries in a steep Trendelenburg position. The overall incidence of postoperative delirium was 6% (11 out of 182 patients).

Table 1 Characteristics of all patients and patients without and with postoperative delirium
Full size table

Comparison of characteristics between patients with and without postoperative delirium

Patients with postoperative delirium showed significantly lower BMI (21.3 vs. 25.0 kg/m2, P = 0.003), had received more radical cystectomy (45% vs. 15%, P < 0.001), had received less colloid infusion (0 vs. 34.2 mL, P = 0.001), had undergone longer duration of anesthesia (362.7 vs. 224.7 min, P < 0.001) and surgery (302.2 vs. 174.5 min, P < 0.001) and had more significant baseline mesial temporal atrophy (MTA) (Scheltens’ scale ≥ 2) (64% vs. 30%, P = 0.046) in the preoperative brain MRI. There were no significant differences in age, sex, underlying comorbidities, preoperative ASA physical status, and MRI features relevant to the baseline small vessel disease burden (all P values > 0.050).

Independent predictors of postoperative delirium

In the univariate analysis, surgical position (steep Trendelenburg over supine) was not associated with the events of postoperative delirium (odds ratio [OR], 0.73; 95% CI 0.21–2.86; P = 0.617) (Table 2). Variables found significant were female sex (OR, 6.11; 95% CI 1.08–34.65; P = 0.041), type of surgery (radical prostatectomy over cystectomy) (OR, 0.22; 95% CI 0.06–0.76; P = 0.017), prolonged duration of anesthesia (≥ 6 h) (OR, 4.57; 95% CI 0.11–16.70; P = 0.023), and presence of a significant baseline MTA (Scheltens’ scale ≥ 2) (OR, 4.12; 95% CI 1.19–16.30; P = 0.029). Among these variables, only the presence of a significant baseline MTA in the preoperative MRI was found to be statistically significant in the multivariate analysis (OR, 3.69; 95% CI 0.99–13.75; P = 0.046). Figure 2 illustrates representative preoperative MRI scans with and without event of postoperative delirium.

Table 2 Univariate and multivariate associations with postoperative delirium as dependent variable for other explanatory variables
Full size table
Fig. 2
figure 2

A 80-yr old male patient who underwent robot-assisted radical prostatectomy with no mesial temporal atrophy in preoperative MRI (coronal T1-weighted image) and had no event of postoperative delirium; B 78-yr old female patient who underwent laparoscopic radical cystectomy with mesial temporal atrophy (Scheltens’ scale 2) in preoperative MRI (coronal T1-weighted image) and had an event of postoperative delirium. MRI, magnetic resonance imaging

Full size image

Discussion

The steep Trendelenburg versus supine position and preoperative brain MRI features related to the small vessel disease burden were not associated with the events of postoperative delirium in patients who received a major urologic surgery. There were differences in BMI, type of surgery (radical prostatectomy vs. cystectomy), amount of colloid infusion, duration of anesthesia and surgery, and proportion of a significant baseline MTA (Scheltens’ scale ≥ 2) between patients with and without events of postoperative delirium. However, in the logistic regression analysis, only the presence of a significant MTA (Scheltens’ scale ≥ 2) in the preoperative MRI was demonstrated to be a significant predictor of postoperative delirium.

The incidence of postoperative delirium in our study was 6%, which was similar to the reported incidence of 5.5–7.8% for urologic surgeries in previous studies [13, 14]. The steep Trendelenburg position was not associated with postoperative delirium in the regression analysis, which contrasts to the result of a previous study [7]. Instead, our result was consistent with those of two observational studies based on the same cohort [8, 9], which reported no difference in the events of postoperative delirium between the two groups. The comparison in these studies [8, 9] was also made between robot-assisted radical prostatectomy and open retropubic radical prostatectomy, but a major difference with our study was that the patients had low perioperative risk, as expressed by the ASA physical status and the number of comorbidities, and this may have contributed to a previous nonsignificant difference between the two groups. However, patients who undergo a major urologic surgery due to malignancy are in general advanced aged with one or more pre-existing comorbidities [15]. In our study, the median age was 68 years and 80% of patients had at least one comorbidity similar to the 88% reported by a previous study [15]. The patients analyzed in this study better reflected the real-world patients undergoing urologic surgeries for malignancy, who are also simultaneously more susceptible to cerebrovascular changes and an increase in the intracranial pressure given their preoperative status when placed in a steep Trendelenburg position. However, even under these conditions, the steep Trendelenburg was not a significant contributor to the postoperative delirium. This suggests that a prior concern over postoperative cognitive complication does not have to be considered in the choice of a surgical approach.

Instead, patients with postoperative delirium were associated with lower BMI, longer duration of anesthesia and surgery, and a significant baseline MTA (Scheltens’ scale ≥ 2). In case of BMI, the inverse correlation was also documented in a previous study, possibly owing to the “obesity paradox” [16]. Prolonged duration of anesthesia was also suggested as a possible contributor to postoperative delirium [9]. However, these clinical and surgical factors were non-significant when the radiologic factor, baseline MTA, was concomitantly considered in the regression analysis. Moreover, small vessel disease burden and cerebral microbleed were not significant risk factors of postoperative delirium similar to previous studies [12, 17]. The presence of a significant MTA in the preoperative MRI was the single most important predictor of postoperative delirium.

MTA is a well-established and validated imaging feature of Alzheimer’s disease-related neurodegeneration [18]. This radiologic feature also has been reported to be present in patients with mild cognitive impairment (MCI), and the presence of MTA in even healthy individuals is reported to associate with an increased risk of future cognitive decline [19]. Likewise, presence of postoperative delirium, even incidental, has been reported to accelerate cognitive decline [20] and has been suggested as the strongest risk factor of long-term development of dementia by a recent study [21]. The link between MTA and postoperative delirium, as first demonstrated by this study, may be that patients with a significant baseline MTA in the preoperative MRI have preclinical Alzheimer’s disease, a disease state in which pathologic and morphologic changes in the brain parenchyma are present prior to the onset of cognitive or functional impairment. Aligned with this finding, patients with preclinical Alzheimer’s disease were found to be at a higher risk for developing severe delirium in a previous study [19]. In addition, cortical thinning in the mesial temporal lobe has been suggested to be predictive of dementia following surgery for ensuring 3 years, especially in patients who have had postoperative delirium [22]. Furthermore, the cutoff score of MTA (i.e., Scheltens’ scale ≥ 2) for postoperative delirium in our study was similar to the cutoff score reported to predict patients with MCI who had later transitioned to Alzheimer’s disease [23]. Thus, the presence of a significant MTA in the preoperative MRI may be a useful radiologic marker for screening patients with preclinical Alzheimer’s disease who are more inclined to develop postoperative delirium following surgery. In addition, the MTA scale is simple and easy to measure, and it has been most extensively studied with a high inter-and intra-reader agreement [24], and thus, has many suitable qualities as a potential prognostic marker.

This study has several limitations. First, this study was based on retrospective, single institutional data, and there may have been a selection bias. Additionally, we have analyzed patients who had undergone a major urologic surgery with a brain MRI taken within 1 year prior to the date of surgery. Considering that the brain MRI is not part of a routine evaluation for a urologic malignancy, one could argue that a large proportion of our study patients may have had pre-existing neurologic conditions that could have affected the baseline cognitive function. However, 95 (52%) of our patients had neither small vessel disease nor mesial temporal atrophy in the baseline MRI, and a considerable number of patients had undergone brain MRI as part of a screening health check-up and not for a neurologic symptom evaluation before knowing or being diagnosed for a urologic malignancy. Second, although cognitive impairment is a known risk factor of postoperative delirium, we did not analyze the patients’ preoperative cognitive status since the administration of a cognitive screening tool such as the Mini-Mental State Examination was not a routine preoperative procedure. Moreover, we did not evaluate other relevant characteristics such as education and socioeconomic status that were inaccessible through the medical records, which could have had a confounding effect on our results. Nevertheless, our goal was to identify factors that could predict postoperative delirium in routine clinical scenarios, making our findings relevant to daily medical practice. Lastly, we have only examined postoperative delirium, and thus, our results may not be generalizable to other etiologies of delirium.

Conclusions

In conclusion, steep Trendelenburg was not associated with postoperative delirium in patients undergoing a major urologic surgery. This suggests that the choice of surgical position in urologic surgery need not be influenced by prior concerns about postoperative cognitive complications. However, significant mesial temporal atrophy (Scheltens’ scale ≥ 2) in preoperative brain MRI was predictive of postoperative delirium and may serve as a useful radiologic prognostic marker.