J Robotic Surg DOI 10.1007/s11701-016-0604-x
ORIGINAL ARTICLE
Vaginal cuff dehiscence: a randomized trial comparing robotic vaginal cuff closure methods Laurie B. Landeen1,2 • Elizabeth M. Hultgren2 • Taylor M. Kapsch2 Paul W. Mallory2
•
Received: 4 February 2016 / Accepted: 18 May 2016 Ó Springer-Verlag London 2016
Abstract The purpose of this study is to compare the rate of vaginal cuff dehiscence between two different methods of closure in patients undergoing robotic-assisted total laparoscopic hysterectomy and explore variables related to postoperative breakdown. This was a prospective, randomized controlled study with two arms. The control group (Arm 1) underwent single-layer continuous closure while the study group (Arm 2) had three additional imbricating figure-of-X sutures placed in addition to the standard protocol. Of the 263 patients who completed the study, 4 patients (1.49 %) experienced dehiscence of the vaginal cuff. Three of the four patients with dehiscence received the standard single vaginal cuff closure (Arm 1) and the one remaining case of dehiscence underwent the protocol with additional sutures (Arm 2). All patients who experienced dehiscence were current smokers. Our study suggests that there may be benefit in adding additional sutures to the standard single-layer vaginal cuff closure procedure. Physicians should evaluate smoking status before deciding on a vaginal cuff closure method. Keywords Vaginal cuff dehiscence Robotic hysterectomy Smoking Benign gynecology
& Laurie B. Landeen
[email protected] 1
Sanford USD Medical Center, 1400 W 22nd St, Sioux Falls, SD 57105, USA
2
Sanford School of Medicine of The University of South Dakota, Sioux Falls, SD, USA
Introduction Nearly 600,000 hysterectomies are performed each year in the United States, with approximately one-third of American women having undergone a hysterectomy by the age of 60 [1]. Vaginal cuff dehiscence (VCD) is a rare but serious complication involving separation of the vaginal cuff previously closed intra-operatively during a hysterectomy. Clinical manifestations of VCD include: pelvic or abdominal pain, vaginal bleeding, vaginal discharge or gush of fluid, vaginal pressure or mass [2]. Dehiscence of the vaginal cuff can cause evisceration, where pelvic or abdominal contents are expelled through the vaginal cuff opening, requiring prompt medical intervention to reduce the risk of bowel injury, necrosis, peritonitis, or sepsis [3]. In comparing incidence rates of VCD in total laparoscopic hysterectomies (0.64 %), abdominal hysterectomies (0.20 %), vaginal hysterectomies (0.13 %), and roboticassisted total laparoscopic hysterectomies (RATLH) (0.46–1.5 %), the incidence of VCD was significantly higher in RATLH [2, 4]. The relative risk of VCD following laparoscopic hysterectomy is 21 times greater than vaginal surgery, and 53 times greater than abdominal surgery [5, 6]. The unusual occurrence of VCD increases the difficultly in identifying predisposing risk factors; however, menopausal status is theorized to play a role in this complication. Premenopausal females are most likely to dehisce post-coitus whereas postmenopausal dehiscence is often due to pelvic prolapse [5]. Other proposed contributing risk factors include vaginal cuff cellulitis or abscess, vaginal trauma, cigarette smoking, vaginal hematoma, pelvic radiation therapy, chronic conditions that increase intra-abdominal pressure, impaired wound healing, and the mode of surgery [7].
123
J Robotic Surg
As patient demographics and anatomical changes of the pelvis are not easily modified, changes in surgical technique have been suggested to decrease rates of dehiscence [8]. Suggested methods include limiting the use of electrosurgery, the use of a two-layer closure as well as the use of barbed suture. However, there remains no standard method of closure proven to decrease VCD [9, 10]. The purpose of this study was to explore VCD rates in a Midwestern University Medical Center following a single-layer or double-layer closure of the vaginal cuff following RATLH. Our hypothesis was that the addition of a second layer of suture would decrease the incidence of VCD.
Materials and methods Study design This study was a two arm, prospective, randomized, controlled trial at a Midwestern University Medical Center. All participating surgeons were proficient in the field of robotics, having at least 2 years of experience. The control arm (Arm 1) consisted of the current standard closure of the vagina, which involved using a single continuous 0-Maxon suture. The study arm (Arm 2) consisted of standard closure of the vagina with a single continuous 0-Maxon suture plus three additional imbricating figure-ofX sutures of 0-Polysorb, over the first layer. The primary outcome of the study was VCD, defined as the partial or total separation of the edges of the vaginal cuff with or without bowel evisceration. Wound dehiscence as well as other secondary complications and outcomes were evaluated, and noted by the operating surgeon. Secondary outcomes of the study included surgical time from skin incision to skin closure, blood loss as estimated by the operating surgeon, post-operative complications, and length of stay in the hospital post-operatively.
Fig. 1 Image of vaginal cuff using simulation on the robot. The thin, long arrows highlight the green 0-Maxon suture used in the continuous closure of the vaginal cuff. The bold arrows highlight the additional three figure-of-X purple 0-Polysorb sutures placed over the continuous closure to complete the double-layer closure
123
Population All women scheduled for RATLH with or without the removal of the ovaries and fallopian tubes for benign gynecologic indications were invited to participate in the study. During the study period (March 2008–December 2013), 2221 patients underwent RATLH at the participating institution. Of this subject group, 272 patients undergoing RATLH were enrolled in the study. All subjects received and signed an informed consent for participation in the trial at the time of surgery scheduling, as per IRB approval (Sanford Health IRB 617). Patients were informed of the risks involved with each arm of the study, including infection, bleeding, injury to adjacent organs, risk of VCD, and possible increased operative time. The subjects were randomized and assigned to one of the two study arms upon entrance into the operating room on the day of their procedure. All other aspects of the surgery and follow-up care were similar. Patients were excluded from the study if they had a final diagnosis of malignancy on pathology, if the procedure was converted from a robotic to an abdominal- or vaginal-assisted approach, or if the patient was lost to follow-up. These exclusion criteria allowed the study of a group of patients with benign gynecological conditions who successfully underwent RATLH (Fig. 1). Randomization Patients were randomized to either arm of the study immediately before the operation. A biostatistician developed a random number table using a computerized random number generator. The results were placed in sequential, opaque, sealed envelopes. At the time of surgery, the surgeon performing the case opened the next envelope containing a number between 1 and 1000 to determine allocation into Arm 1 or Arm 2 of the study. Those patients
J Robotic Surg
whose assigned envelope contained a number between 1 and 500 received the standard treatment of single-layer closure (Arm 1) while those assigned a number between 501 and 1000 received the study treatment of a doublelayer closure (Arm 2).
272 Paents Enrolled in Study
9 Paents Excluded
Surgical method Nine surgeons participated in the study, each performing the robotic hysterectomy with the use of bipolar electrosurgery, monopolar scissors, and the harmonic scalpel. Two surgeons chose to use the monopolar spatula to open the vagina, while seven surgeons used the harmonic scalpel to open the vagina. Vaginal cuff closure was performed by a robotic-assisted laparoscopic approach. All surgeons used a single continuous 0-Maxon suture, as stated above, but the study arm included three additional imbricating figureof-X sutures with 0-Polysorb over the first layer (Fig. 1). All patients received prophylactic antibiotic therapy with cefazolin or gentamicin as well as clindamycin, if a significant penicillin allergy was present. Evidence of postoperative complications was evaluated at time of discharge from the hospital, at 4 weeks, and again at 6 months. Statistical analysis Statistics were analyzed using GraphPad Prism software analysis. Numerical data were compared using t tests to determine the presence of statistical heterogeneity between the two study arms. Categorical data were compared using f test analysis to determine variance between the study arms. ANOVA was used for further comparison of subsets of the population. Statistical significance was achieved when p was less than 0.05.
Results During the study period, 272 patients underwent a RATLH and were enrolled into study. Nine patients were excluded from the study after randomization; one for final diagnosis of malignancy on pathology, one converted from a robotic to a vaginal approach and seven were lost to follow-up (Fig. 2). Neither patient excluded for reasons outside of follow-up care experienced VCD. Demographic information for the two patient groups was similar following analysis of confounding variables. There were no significant differences in age or BMI between the two arms of the study (Table 1). Smoking incidence was statistically different between the two arms (p \ 0.02), with first and second arm smoking rates of 13.98 and 23.53 %, respectively.
263 Study Parcipants
Arm 1: 144 Paents
Arm 2: 119 Paents
Fig. 2 Patient enrollment. Flow diagram of patient enrollment and final participant numbers. Nine patients were excluded from the trial. Exclusion criteria included having a final diagnosis of malignancy, losing patients to follow up, or converting from a robotic to either an abdominal or a vaginal assisted approach. The control arm (Arm 1) was closed with single continuous suture. The study arm (Arm 2) consisted of closure of the vagina in standard fashion plus three additional imbricating figure-of-X sutures over the first layer
Four of the 263 patients (1.49 %) who underwent RATLH experienced dehiscence of the vaginal cuff. Coitus was the triggering event for all four cases of VCD. No surgeon involved in the study had more than one patient experience VCD during the timeframe of this study. There was a 2.08 % dehiscence rate found with Arm 1 compared to a 0.804 % rate for those in Arm 2 of the trial (p \ 0.001) (Table 1). Further risk factor analysis revealed that all patients who experienced VCD were smokers; 15 % of smokers who underwent single-layer cuff closure and 3.75 % of smokers who underwent double-layer closure of the vaginal cuff experienced VCD (p \ 0.05) (Fig. 3). Secondary outcomes measured were not statistically different between the two study arms. No difference was seen in the length of hospital stay (average 1.13 days), urinary tract infections, dyspareunia, abdominal pain, or post-coital bleeding between the two groups. Both arms of the study showed similar intra-operative blood loss with an average of 97 cc as recorded by anesthesia (range 25–1150 cc). Average surgical time was 114 min (range 41–282 min). No significant increase in surgical time could be determined with the addition of the second layer of sutures. This is a consequence of the wide range of operative times in each arm affected by varying levels of case difficulty. Furthermore, there was no difference in the rate of urinary tract injury between the two surgical closure methods. The relative risk of VCD in smokers in Arm 1 of the trial was 4.23 times higher than VCD in smokers in Arm 2 of the trial. Attributable risk analysis reveals that risk of
123
J Robotic Surg Table 1 Patient demographics
Total population avg. (n = 263)
Arm 1 (n = 144) 43 ± 0.7114
Arm 2 (n = 119)
p value
Age
43
43 ± 0.8104
0.532
BMI
31
31.16 ± 0.67773
30.82 ± 0.7481
0.734
Smokinga
18 %
13.89 ± 2.89 %
23.53 ± 3.91 %
0.020
VCDa
1.47 %
2.08 ± 1.19 %
0.804 ± 0.048 %
0.001
The mean age and body mass index (BMI) of each arm of the study was calculated. T test analysis was used to determine if the two groups studied were significantly different (p \ 0.05). Standard deviations were calculated. Smoking status and vaginal cuff dehiscence (VCD) were determined and represented as a percentage of the total number of patients in each study arm. F test analysis (a) was used to determine variance between the two arms of the trial. The number of individuals in each group is represented by the letter ‘‘n’’
Arm 2: n=119
Arm 1: n=144
NS
S
NS
S
n=124 0% VCD
n=20 15% VCD
n=88 0% VCD
n=28 3.57% VCD
Fig. 3 Cuff dehiscence rates for each arm of the trial subcategorized by smoking status. The number of non-smokers (NS) and the number of smokers (S) in each group was subcategorized in each arm of the trial. The percentage of vaginal cuff dehiscence (VCD) was calculated for each subcategory. ANOVA was used to determine that the groups were significantly different (p \ 0.001)
dehiscence from smoking was reduced by 11.5 % in Arm 2, when a double-layer closure of the cuff was performed compared to a single-layer closure of the cuff in Arm 1. The population attributable risk of the study, when narrowed to the smoking population, was 0.45. Subsequently, the population attributable risk percentage shows that 56 % of dehiscence in the study population of smokers is attributable to single versus double-layer closures.
Discussion RATLH utilization has drastically increased for surgical treatment of benign conditions over the past decade. Despite advantages of laparoscopic surgery such as reduction in blood loss, need for blood transfusion, and shorter hospital stay, laparoscopic hysterectomies continue to have higher rates of VCD compared to traditional vaginal or abdominal surgical methods [5]. Our study of 263 subjects, demonstrated a significant reduction in VCD with a double-layer closure compared to a single-layer closure. Our findings suggest that smokers who undergo RATLH are at an increased risk of VCD compared to non-smokers. Interestingly, there were no VCD in the non-smoking
123
participants in our study. This supports the widely accepted hypothesis that tobacco has a negative impact on wound healing. Nicotine causes impaired healing and ischemia of tissue due to a reduction in blood flow to the tissue secondary to nicotine’s vasoconstrictive properties. Nicotine also increases platelet adhesion, which leads to increased microvascular thrombotic occlusions and further tissue ischemia. Additionally, carbon monoxide decreases the transport and metabolism of oxygen, limiting its use by peripheral tissues. At the cellular level, the transport and metabolism of oxygen is inhibited by hydrogen cyanide. Cigarette ingredients thus delay wound healing in tobacco users, playing an intricate role in the development of VCD following RATLH [11]. In addition, tobacco use can lead to chronic cough that causes an increase in intra-abdominal pressure and may result in an increased risk of VCD. The addition of a second layer in closing the vaginal cuff may decrease the risk of VCD in smokers without further increasing the patient’s intra-operative blood loss, surgical time, or post-operative complications. Bladder injuries are a common risk of laparoscopic and robotic hysterectomies with recent studies suggesting occurrence in 4–5 % of cases [12]. Our study showed no difference in the rate of bladder injury with the single- or double-layer closure. Overall, no negative outcomes were noted in this study while decreasing VCD in our patients. Our study had a high completion rate with only nine excluded patients. Limitations of the study were the small sample size (n = 263) and the few women with VCD (n = 4). Strength of the study is that it was a randomized, controlled study in which the surgeons were blinded until entering the operating room. The patients were also blinded to their treatment arm assignment before the beginning of the case, but were informed of the treatment arm postoperatively, if they inquired. As the patients were able to inquire as to the study arm immediately post-operation, this may have influenced their activities and behaviors after the procedure. Another limitation to the study is that there may have been decreased VCD in the study arm due to the different suture material and not entirely in response to the
J Robotic Surg
double-layer closure itself. To enhance this study in the future, a single suture material could be studied for all layers of closure of the vaginal cuff. In conclusion, by assessing tobacco use pre-operatively, surgeons have the opportunity to consider minimizing the VCD risk for women undergoing hysterectomy by adding a second layer to the vaginal cuff closure, significantly reducing the post-operative dehiscence rate. The demonstrated greater than fourfold reduction in VCD with a double-layered closure is an important point of consideration in patient care, eliminating the need to re-operate and reducing prolonged restrictions on physical and sexual activities. Furthermore, it highlights the importance of counseling patients in smoking cessation before surgery. Through further study and evaluation of vaginal cuff closure methods, we aim to progress the field of robotics in gynecological surgeries and improve patient outcomes. Acknowledgments We thank the surgeons and support staff at Sanford Medical Center for their participation in this study. We also want to express our gratitude to Dr. Peter Van Eerden for his computer-generated envelopes and to Stacy Wempe for her assistance in data collection. Compliance with ethical standards Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Conflict of interest Laurie B. Landeen, M. D. declares that she owns stock with Intuitive Surgical (ISRG); Elizabeth M. Hultgren, Taylor
M. Kapsch and Paul W. Mallory declare that they have no conflict of interest.
References 1. Whiteman MK et al (2008) Inpatient hysterectomy surveillance in the United States, 2000–2004. Am J Obstet Gynecol 198(1):34 (e1–e7) 2. Hur HC et al (2011) Vaginal cuff dehiscence after different modes of hysterectomy. Obstet Gynecol 118(4):794–801 3. Kashani S et al (2012) Vaginal cuff dehiscence in robotic-assisted total hysterectomy. JSLS 16(4):530–536 4. Stevens A et al (2013) Vaginal cuff dehiscence in a series of 12,398 hysterectomies: effect of different types of colpotomy and vaginal closure. Obstet Gynecol 121(1):189–190 5. Hur HC et al (2007) Incidence and patient characteristics of vaginal cuff dehiscence after different modes of hysterectomies. J Minim Invasive Gynecol 14(3):311–317 6. Kho RM et al (2007) Robotic hysterectomy: technique and initial outcomes. Am J Obstet Gynecol 197(1):113 (e1–e4) 7. Weizman NF et al. (2015) Vaginal cuff dehiscence: risk factors and associated morbidities. JSLS 19(2) 8. Koo YJ et al (2013) Vaginal cuff dehiscence after hysterectomy. Int J Gynaecol Obstet 122(3):248–252 9. Kho RM et al (2009) Incidence and characteristics of patients with vaginal cuff dehiscence after robotic procedures. Obstet Gynecol 114(2 Pt 1):231–235 10. Siedhoff MT, Yunker AC, Steege JF (2011) Decreased incidence of vaginal cuff dehiscence after laparoscopic closure with bidirectional barbed suture. J Minim Invasive Gynecol 18(2):218–223 11. Silverstein P (1992) Smoking and wound healing. Am J Med 93(1A):22S–24S 12. Vakili B et al (2005) The incidence of urinary tract injury during hysterectomy: a prospective analysis based on universal cystoscopy. Am J Obstet Gynecol 192(5):1599–1604
123