Lower bleeding volume contributes to decreasing surgical site infection in radical cystectomy: A propensity score-matched comparison of open versus robot-assisted radical cystectomy

INTRODUCTION

Radical cystectomy with urinary diversion is the standard surgical treatment for muscle-invasive bladder cancer and very-high-risk non-muscle-invasive bladder cancer, for which the previously reported complication rates were relatively high (28–60%).^{1, 2} Surgical site infections (SSIs) negatively impact the quality of life, prolong hospital stays, and increase healthcare-related costs. They are common complications of radical cystectomy and the incidence rate of SSI after ORC has been reported as 2%–33%.^2-6

Robot-assisted radical cystectomy (RARC) has become a standard minimally invasive surgery, expected to reduce perioperative mortality and complications, especially bleeding volume and blood transfusion, with oncological outcomes comparable to ORC.^7-10 Minimally invasive surgeries significantly decrease the incidence of SSIs in various surgeries; hence, RARC is also expected to decrease the incidence of SSI compared to ORC.¹¹ However, several previous studies evaluating the perioperative outcomes of RARC did not distinguish between the incidence of SSIs and urinary tract infections when assessing the prevalence of infectious complications or evaluating the prevalence of SSIs according to the criteria defined in the Centers for Disease Control and Prevention (CDC) guidelines.^12-14 Furthermore, no studies have evaluated in detail the differences in SSI characteristics between RARC and ORC or the etiology related to the differences in incidence.

Therefore, this study aimed to compare the incidence rates of SSIs after RARC and ORC and investigated the risk factors for SSIs after radical cystectomy, focusing on the difference in the surgical approach using propensity score matching (PSM).

METHODS

RESULTS

Analysis after PSM

The patients' characteristics of the matched cohort (ORC: 34 and RARC: 34) are shown in the right half of Table 1. For PSM, all preoperative variables and simultaneous surgical procedures (type of urinary diversion, simultaneous nephroureterectomy, and urethrectomy) listed in Table 1 were matched.

In the univariate analysis after PSM, SSI incidence was significantly associated with ORC, increased bleeding volume, blood transfusion, and delayed postoperative oral feeding as well as the results before PSM (23.5% vs. 5.9%, p = 0.040, 3125 mL vs. 530 mL, p < 0.001, 80.0% vs. 43.1%, p = 0.03, and 7 days vs. 5 days, p = 0.046, respectively, Table 3).

TABLE 3. Univariate analysis of risk factors for SSI following radical cystectomy in the matched cohort.

	SSI event + (n = 10)	SSI event − (n = 58)	p-value
Age, year	63.5 (55–72)	70 (64–75)	0.15
Male	5 (50.0%)	31 (53.5%)	0.84
BMI, kg/m2	24.7 (23.1–25.7)	22.8 (20.8–24.9)	0.09
Current smoker	0 (0%)	4 (6.9%)	0.39
Diabetes mellitus	1 (10.0%)	13 (22.4%)	0.37
Steroid use	1 (10.0%)	2 (3.5%)	0.35
ASA-PS			0.72
≤2	8 (14.0%)	49 (88.0%)
3	2 (18.2%)	9 (81.8%)
Preoperative hemoglobin, g/dL	12.3 (10.7–14.2)	11.9 (10.5–13.1)	0.31
Preoperative serum albumin, g/dL	4.2 (3.8–4.4)	4.0 (3.8–4.2)	0.11
Preoperative pyuria	7 (70%)	36 (62.1%)	0.63
Preoperative bacteriuria	2 (28.6%) (n = 7)	19 (38.8%) (n = 49)	0.60
Clinical T Stage			0.96
≤2	7 (14.6%)	41 (85.4%)
≥3	3 (15.0%)	7 (85.0%)
Neoadjuvant chemotherapy	4 (40.0%)	28 (48.3%)	0.63
Operative method			0.040
ORC	8 (23.5%)	26 (76.5%)
RARC	2 (5.9%)	32 (94.1%)
Urinary diversion			0.42
Using the intestinal tract	8 (17.0%)	39 (83.0%)
Other methods	2 (9.5%)	19 (90.5%)
Simultaneous nephroureterectomy	2 (20.0%)	10 (17.2%)	0.83
Simultaneous urethrectomy	9 (90.0%)	41 (70.7%)	0.20
Total operation time, min	470 (411–517)	476 (386–542)	0.89
Bleeding volume, mL	3125 (1453–4063)	530 (223–1490)	< 0.001
Blood transfusion	8 (80.0%)	25 (43.1%)	0.03
Length of antimicrobial prophylaxis			0.78
≤2 days	2 (12.5%)	14 (87.5%)
≥3 days	8 (87.5%)	44 (84.6%)
Oral feeding (POD)	7 (5.8–7.3)	5 (3–7)	0.046

• Abbreviations: ASA-PS, American Society of Anaesthesiologists Physical Status; BMI, body mass index; ORC, open radical cystectomy; POD, postoperative day; RARC, robot-assisted radical cystectomy; SSI, surgical site infection.

Only bleeding volume also remained as an independent risk factor of SSI by multivariate logistic regression analysis after PSM (OR, 1.13 [per 100 mL increase]; 95% confidence interval:1.02–1.25; p = 0.001), and RARC, delayed postoperative oral feeding or blood transfusion were not remained as significant risk factors for SSIs as well as the results before PSM (p = 0.60, 0.11, and 0.78, respectively; Table 4).

TABLE 4. Multivariable logistic regression analyses of risk factors for SSI following radical cystectomy in the matched cohort.

	Odds ratio	95% confidence interval	p-value
RARC (conf. ORC)	0.39	(0.01–14.5)	0.60
Oral feeding (by 1-day delay)	1.18	(0.96–1.43)	0.11
Bleeding volume (by 100 mL increase)	1.13	(1.02–1.25)	0.001
Blood transfusion (conf. No transfusion)	1.65	(0.05–57.7)	0.78

• Abbreviations: ORC, open radical cystectomy; RARC, robot-assisted radical cystectomy; SSI, surgical site infection.

The ROC curve analysis identified 1630 mL as the cutoff bleeding volume for predicting the SSI occurrence after radical cystectomy, with an AUC, sensitivity, and specificity of 0.773, 0.73, and 0.75, respectively (Figure 1). The cumulative incidences of SSI were compared and stratified by the difference in surgical procedures or more or less than the cutoff bleeding volume. They were significantly lower in RARC group (vs. ORC) and in bleeding volume less than 1630 mL group (vs. higher than 1630 mL) (5.9% vs. 23.5%, p = 0.047, 4.3% vs. 38.1%, p < 0.001, respectively) (Figure 2).

DISCUSSION

This study revealed that SSI incidence was significantly lower in patients who underwent RARC than in those who underwent ORC. Higher bleeding volume, even more than that in the ORC group, remained an independent factor for SSI both before and after PSM. In addition, the accuracy of the cutoff bleeding volume of 1630 mL for predicting SSI occurrence was moderate (AUC: 0.773).

Several previous studies have compared the incidence of perioperative complications between ORC and RARC. However, most of these studies either (i) evaluated urinary tract and surgical site infections simultaneously as infectious complications or wound dehiscence and SSI simultaneously as wound complications or (ii) did not evaluate the details of the complications.^{1, 7, 8, 18} On the contrary, a few studies evaluated the difference in the incidence of SSI in detail. One randomized study comparing the efficacy of ORC and RARC reported that wound infection occurred in 8.3% (13/156) and 3.7% (6/161) of the ORC and RARC groups, respectively.⁹ A previous large cohort database study in the United States reported that RARC was significantly associated with decreased SSI (OR: 0.72).¹⁹ The present study observed a lower SSI incidence rate in the RARC group than in the ORC group, consistent with previous studies. These results suggest that, like colorectal or prostate surgery, robot-assisted, minimally invasive procedures are beneficial for reducing SSIs after radical cystectomy.^{11, 20, 21}

This study observed a more significant decrease in superficial incisional SSIs than in deep incisional or organ/space SSIs was observed in the RARC group (11.7% vs. 2.3%). Differences in the causative bacteria or the prevalence of resistant strains between the ORC and RARC groups were unclear because the number of patients with confirmed causative bacteria in the RARC group was insufficient. A previous randomized trial demonstrated that superficial SSIs occurred in 12% and 7% of the ORC and RARC groups, respectively.¹⁰ This report was partly consistent with the present study's results. However, the differences in the incidence of superficial SSIs between the ORC and RARC groups seemed smaller in the previous study's cohort than those in the present cohort. These differences may be partly associated with the urinary diversion methods. The present study utilized ICUD, whereas extracorporeal urinary diversion (ECUD) was utilized in the previous study. ICUD was associated with a shorter skin incision, a shorter operative time, and less blood loss, which may decrease the risk of bacterial adhesion and growth from the resected intestine to the incised wound.²² Hence, RARC with ICUD would be more beneficial; however, further studies comparing the incidence of SSIs between patients that undergo ICUD and those that undergo ECUD are required.

The multivariate analysis showed that only bleeding volume remained an independent risk factor for SSIs after radical cystectomy both before and after PSM. Our study also found that the difference in the cumulative incidence of SSI was more remarkable for stratification by cutoff bleeding volume than for that by ORC versus RARC. These results were consistent with previous studies and seemed rational because increased blood loss was associated with hypoxia induced by anemia and with the decreased concentrations of antibiotic agents and immune cells in the serum or the tissue. Consequently, this may promote the adhesion and growth of pathogenic bacteria to the wound.^23-25 In addition, several previous studies suggested blood transfusion, strongly associated with high volume of blood loss increased the risk of SSI by immunosuppressive effect associated with transfusion-related immune modulation, even though that was not remained as independent factor in our study.^{26, 27} These results suggest that even ORC could reduce the incidence of SSIs to as low as observed with RARC when well-trained surgeons perform the procedure with little blood loss. In contrast, many studies have revealed that RARC significantly and clearly reduces bleeding volume and perioperative transfusions compared with ORC, indicating that robot-assisted procedures are beneficial for reducing SSIs after radical cystectomy by decreasing bleeding volume.^{1, 8, 10, 19, 28}

This study also revealed that the rational cutoff bleeding volume for predicting SSIs after PSM was 1630 mL with a moderate ability of prediction (AUC 0.773). A previous study suggested that antibiotic concentrations for cefazolin in the tissue were decreased by bleeding volume and that the cutoff volume considering additional doses of antibiotics was 1500 mL.²⁵ Our cutoff bleeding volume may be close to the value unable to maintain effective antimicrobial concentrations in the tissue.

The main limitation of this study was its retrospective, single-center design using historical controls. The patients' backgrounds and treatment policies, including the preoperative managements, may differ between the ORC and RARC groups, even though we performed PSM. These differences must have affected SSI incidence. In addition, the length of antimicrobial prophylaxis was quite varied among patients, which may lead to a strong bias for the present results. Hence, further large-scale prospective studies with uniform perioperative and infection control protocols are required.

In conclusion, SSI prevalence after RARC was significantly lower than that after ORC. Although decreased bleeding volume was an independent and more significant factor for reducing SSIs after radical cystectomy compared to the robot-assisted procedure, RARC would be beneficial for reducing SSI incidence because several studies have revealed that RARC significantly reduces bleeding volume.

AUTHOR CONTRIBUTIONS

Jun Kamei: Data curation; Investigation; Writing—original draft. Kaori Endo: Investigation. Masahiro Yamazaki: Data curation. Toru Sugihara: Formal analysis. Ei-Ichiro Takaoka: Supervision. Satoshi Ando: Funding acquisition; Writing—review & editing. Haruki Kume: Supervision; Writing—review & editing. Tetsuya Fujimura: Conceptualization; Funding acquisition; Writing—review & editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

APPROVAL OF THE RESEARCH PROTOCOL BY AN INSTITUTIONAL REVIEWER BOARD

Approval number: A22-023.

INFORMED CONSENT

Written consent was waived because of the study's retrospective nature, and patients provided consent via an opt-out approach.

REGISTRY AND THE REGISTRATION NO. OF THE STUDY/TRIAL

N/A.

ANIMAL STUDIES

N/A.