A Laparoscopic Knot-Tying Device for Minimally Invasive Cardiac Surgery

Abstract

Objectives

Intracorporeal suturing and knot tying can complicate, prolong or preclude minimally invasive surgical procedures, reducing their advantages over conventional approaches. An automated knot-tying device has been developed to speed suture fixation during minimally invasive cardiac surgery while retaining the desirable characteristics of conventional hand-tied surgeon's knots: holding strength and visual and haptic feedback. A rotating slotted disk (at the instrument's distal end) automates overhand throws, thereby eliminating the need to manually pass one suture end through a loop in the opposing end. Electronic actuation of this disk produces left or right overhand knots as desired by the operator.

Methods

To evaluate the effectiveness of this technology, 7 surgeons with varying laparoscopic experience tied knots within a simulated minimally invasive setting, using both the automated knot-tying tool and conventional laparoscopic tools. Suture types were 2-0 braided and 4-0 monofilament.

Results

Mean knot-tying times were 246 ±116 seconds and 102 ±46 seconds for conventional and automated methods, respectively, showing an average 56% reduction in time per surgeon (p=0.003, paired t-test). The peak holding strength of each knot (the force required to break the suture or loosen the knot) was measured using tensile testing equipment. These peak holding strengths were normalized by the ultimate tensile strength of each suture type (57.5 N and 22.1 N for 2-0 braided and 4-0 monofilament, respectively). Mean normalized holding strengths for all knots were 68.2% and 71.8% of ultimate tensile strength for conventional and automated methods, respectively (p= 0.914, paired t-test).

Conclusions

Experimental data reveal that the automated suturing device has great potential for advancing minimally invasive surgery: it significantly reduced knot-tying times while providing equivalent or greater holding strength than conventionally tied knots.

Introduction

The challenge of intracorporeal suturing and knot tying, at awkward angles and with limited visibility and depth perception, has long been recognized as a significant barrier to the advancement of Minimally Invasive Surgery (MIS). Sutures remain the “gold standard” of fixation for many surgical fields, as the visual and haptic feedback associated with suture tensioning are critical means by which the surgeon assesses good tissue apposition. The clinical effectiveness of MIS frequently depends on the surgeon's suturing skills, however, and knot-tying with conventional laparoscopic tools can be challenging.

The crux of the problem as it relates to MIS is that knot-tying with traditional suture materials using laparoscopic or robot-assisted tools can lead to increased operation times [1,2]. Operating time can be especially critical for cardiac procedures involving cardiopulmonary bypass (CPB). Increased CPB times lead to higher risks of multiple organ dysfunction syndrome (MODS) [3], and procedures involving CPB are linked to adverse cerebral outcomes, including strokes, transient ischemic attacks and seizures [4]. Patient mortality rates increase with the time during which the heart is arrested, especially with older patients. This leads to the double bind between the “need for speed,” which the direct access and visibility of traditional “open heart” procedures allow, and the requirement that the MIS patient be comparatively “young and healthy” in order to survive the increased CPB times required for MIS [5]. Furthermore, older patients and otherwise “unhealthy” patients make up a majority of the candidates requiring surgical procedures.

To address this need, various laparoscopic instruments have been developed to automate the knot-tying process or deploy “pre-tied” sliding knots. Devices include the Autosuture™ Endo Stitch (Covidien, Mansfield MA) [6], the Suture Assist (Ethicon Endo-Surgery, Cincinnati, OH) [7], and a “Suture Cartridge” [8] as well as others [9-11]. Suture-free alternatives such as staples, clips, and glues have proven effective in certain procedures but can have limited applicability. For example, the nitinol U-clip™ (Medtronic® [12]) has been used effectively to secure annuloplasty bands in minimally invasive mitral valve repair [13,14], but cannot be used for leaflet repair. Both pre-tied knots and suture alternatives generally limit tactile assessment and can also lack the strength (resistance to loosening) of conventional surgical knots [15-17].

The goal of this research is to develop an automated knot-tying tool that enhances suture fixation in MIS while retaining the strength of conventional hand-tied surgeon's knots and preserving visual and haptic feedback.

Materials and Methods

Design Constraints

Based on the recommendations of collaborating cardiac surgeons, several design specifications for the automated knot-tying tool were formulated: 1) utilization of conventional suture materials, 2) deployment of conventional (non-sliding) left and right overhand knots, 3) provision for manual tightening of each knot, and 4) operability through 10-mm or smaller trocards.

Design Concepts and Nomenclature

The key impediment to rapid knot-tying in MIS is the process of passing one suture end through a loop generated in the opposite end. With conventional laparoscopic tools, these steps are usually accomplished by wrapping one suture end around a grasper or needle holder and then pulling the opposite suture end through the loop—a process which requires much experience to master.

In this research, a novel mechanism was created to automate the task of generating a loop in the suture and passing the opposite end through the loop. The key feature of this mechanism is a slotted disk that rotates inside its housing at the distal end of the tool (Fig. 1). To function properly, the center of the disk must be free of any bearing support or drive that is centrally located. The standing end of the suture is fixed to the slotted disk.

Figure 1.

Novel procedure for tying knots: a) passing the suture's needle end into the slotted disk, b) - c) rotating the knot-tying disk, d) removing the loop from the disk, e) - f) tensioning the knot

To tie a knot, the operator first passes the needle and suture through the tissue, places the suture's non-needle (bitter) end on either side of the disk housing, and places the needle end of the suture into the disk slot (Fig. 1a). The disk is then rotated by an external actuator to create the overhand loop essential to the knot (Fig. 1b,c). The surgeon then removes the loop from the disk (Fig. 1d) and simultaneously tensions the working and standing ends (attached to the knot-tying mechanism) to tighten the knot as desired (Fig. 1e,f). Additional knots can be placed over the initial knot by repeating the previous steps. The surgeon may alternate between right and left overhand knots by prescribing the appropriate direction of disk rotation, thereby enabling the production of square knots, surgeon's knots, and a variety of related knots.

Prototype Design

After several design iterations, a practical ergonomic prototype was produced. The prototype used in this study features electronic actuation, a single-use suture cartridge, and cost-effective fabrication. The suture cartridge (Fig. 2) attaches to the distal end of the knot-tying instrument (Fig. 3) and contains the suture, knot-tying disk, gears, and a fitting that couples the cartridge with actuators in the handle. After each fixation, the suture cartridge is removed from the instrument and a new cartridge (with suture attached) coupled to the end of the device.

Figure 2.

Suture cartridge a) placement within the device and b) CAD model showing suture slot and knot-tying disk

Figure 3.

Knot-tying device prototype

The design utilizes rapid prototyping to facilitate the design and fabrication processes. The device handle (Fig. 3) and suture cartridge geometries were fabricated using 3D printing technologies: the Dimension® SST 3D Printer (Stratasys, Inc., Eden Prairie, MN) and the Invision HR 3D Modeler (3D Systems Corporation, Rock Hill, SC). Each cartridge (including housing, bobbin, and gears) was printed as a complete unit, eliminating the need for assembly after fabrication.

Actuation of the disk is achieved using a miniature DC gearmotor (MicroMo Electronics, Inc. Clearwater, FL) activated by two push-button switches mounted on the grip. A Parallax® Propeller™ microcontroller chip (Parallax, Inc., Rocklin, CA) ensures precise indexing of the knot-tying disk. Either right or left overhand throws can be made by pressing the appropriate button, allowing flexibility in knot-tying. For example, a surgeon's knot (a double overhand throw followed by a single overhand throw in the opposite direction) can be tied by pressing the left button two consecutive times, followed by the right button (with subsequent alternating throws for reinforcement).

Surgical Evaluations

Seven surgeons with varying levels of laparoscopic surgical training and experience completed knots within a simulated MIS setting (Fig. 4) using both the automated knot-tying device and conventional laparoscopic tools (a needle holder and laparoscopic forceps). All participants had some prior training in tying knots with laparoscopic instruments. Each surgeon completed a minimum of five knots per device, each consisting of five alternating left and right overhand throws. Suturing times and knot quality were compared for both methods. Since the trials were intended to evaluate knot-tying exclusively, each suture cartridge was pre-loaded into the prototype and the needle was passed through a foam board target before delivery to the surgeon. Suture types varied randomly between 2-0 braided and 4-0 monofilament to analyze their effects on suturing times and resistance to loosening.

Figure 4.

Simulated minimally invasive setting: a) overview, and b) view as displayed on the LCD monitor (automated device at right)

The simulated MIS setting consisted primarily of an artificial human skeletal thorax covered in fabric to prevent direct visualization of the suturing site (Fig. 4a). Laparoscopic tools were introduced through two small ports in the second and fifth intercostal spaces of the thorax. A hollow rubber ball (75 mm diameter, Fig. 4b) with a rectangular window was fastened mediastinally within the thorax (over the spine). A video camera placed between the ports allowed visualization of the suturing site on an LCD monitor (Fig. 4a).

Mechanical testing was conducted to compare the quality of knots generated using conventional tools to those generated using the automated knot-tyer. For testing, each loop was placed between two eye hooks and elongated at a constant rate specified by the user (0.3-0.5 mm/sec (0.012-0.020 inches/sec)), while force and elongation data were recorded. A Sherline 2000 computer numeric controlled (CNC) milling machine (Sherline Products, Inc., Vista, CA) was used for precise motion control, while a MLP-50 load cell (Transducer Techniques, Temecula, CA) measured force values.

Results

Average suture fixation times were 246 ±116 and 102 ±46 seconds, respectively, for conventionally tied and automated knots, showing an average 56% reduction in suture times per surgeon (p=0.003, paired t-test). Comparable reductions in suturing times were seen for both 4-0 monofilament and 2-0 braided sutures (Table 1).

Table 1. Reduction in suturing time with automated knot-tyer as compared to conventional tools.

Surgeon	Mean Suturing Time (sec)		% diff	p
	Conventional	Automated
1	172 ±44	149 ±45	-13.4	0.440
2	181 ±89	58 ±8	-67.8	0.015
3	324 ±106	120 ±23	-62.9	0.003
4	168 ±57	55±22	-67.1	0.003
5	269 ±110	116 ±54	-56.9	0.023
6	347 ±177	71 ±17	-79.6	0.008
7	259 ±76	142 ±14	-45.1	0.010
Overall	246 ±116	102 ±46	-58.6	0.003

Peak holding strengths (the force required to break the suture or loosen the knot) were normalized by the tensile strength of its suture type (57.5 N and 22.1 N for 2-0 braided and 4-0 monofilament, respectively). Mean holding strengths for all knots were 68.2% and 71.8% of ultimate tensile strength for conventional and automated methods, respectively (p= 0.914, paired t-test, Table 2). Holding strengths differed widely for monofilament (60.3% and 94.6% for conventional and automated methods, respectively, p= 0.063, paired t-test) and braided sutures (72.6% and 55.4%, respectively, p= 0.399, paired t-test).

Table 2. Normalized tensile strengths of suture loops.

Surgeon	Conventional		Automated		% difference	p
	n	Strength (%)	n	Strength (%)
1	3	90.8± 28.6	3	88.1± 17.4	-3.0	0.895
2	5	33.6 ±41.4	5	90.6 ±26.9	+169.4	0.032
3	5	62.6 ±50.1	5	98.8 ±7.0	+57.7	0.149
4	5	92.5 ±18.4	4	51.2 ±37.8	-44.6	0.067
5	5	80.7 ±28.2	4	39.5 ±52.2	-51.1	0.171
6	5	48.7 ±31.8	5	86.6 ±40.0	+77.7	0.136
7	5	77.2 ±23.0	5	43.8 ±50.5	-43.2	0.216
Overall	33	68.2 ±36.6	31	71.8 ±40.7	+5.4	0.914

n=number of samples

Discussion

Design Strategy: An Emphasis on Rapid Prototyping

Efficiency of the development process was facilitated by rapid prototyping methodology. A distinct advantage of producing as many prototypes as possible (from the earliest stages of a project to the final design) is that concepts can be continuously tested against a wide array of performance criteria. Using rapid prototyping, the design team adopted a practice of building rough models continuously, as a means of making both mistakes and discoveries.

Simulated Laparoscopic Model

The surgical evaluation was limited in its scope by its simulated rather than in vivo setting. While the simulated minimally invasive setting (Fig. 4) exhibited port access, confinement, and reliance on endoscopy, it lacked biological fluids and compliant tissues, both of which could affect suture handling and knot generation. Knots were generated on a suture looped around a single 5.0-mm thick foam-board; therefore this statistical study lacked ample assessment of the device's ability to appose biological tissues. Nevertheless, preliminary tests on explanted pig anatomy have revealed adequate apposition of cardiac tissues (myocardial and valvular tissue) and fixation of an annuloplasty band using the automated device (Fig. 5).

Figure 5.

Preliminary tests on porcine cardiac tissue: a) before “repair”, and b) after “repair”

“Air Knots”

While peak holding strengths for both methods were comparable and favorable, tensile testing did not assess an occasional and undesirable byproduct of knot-tying: “air knots”. These knots are characterized by high holding strength but little or no tension in the suture, which would result in inadequate tissue apposition. The fact that loose knots occurred with both the prototype and conventional tools shows that the condition may be related to the simulated surgical setting rather than the suturing method (conventional knot-tying methods are obviously known to be acceptable for surgery).

Nevertheless, a feature of the knot-tying device could exacerbate the air knot problem. Before producing a knot with the device, the surgeon must often release tension on the suture to position its bitter end, potentially causing an air knot (Fig. 1a). If deemed necessary after further evaluations, the automated suturing device could be redesigned so that the operator does not have to release suture tension during the knot-tying process.

Learning Curve

The automated knot-tying device has the advantage of a fast learning curve compared with conventional laparoscopic knot-tying tools. Surgeons with varying laparoscopic skills were able to operate the device effectively after just a few minutes of training, proving its intuitive operation. A previous study involving surgical residents yielded similar outcomes [11], showing that inexperienced surgeons can be good indicators of a novel device's potential. In the case of three surgeons with limited recent knot-tying experience, the average times on the first attempt were 378 and 78 seconds, respectively, for the conventional and automated methods. As expressed by the participants (especially those with limited relevant experience), the automated suturing device greatly reduced the complexity and anxiety associated with knot-tying. Accompanying reductions in surgeon fatigue could further enhance the efficacy of long surgical procedures.

Integrated Forceps Design

Extensions of this research have focused on the design and fabrication of prototypes with integrated forceps to aid in needle/suture manipulation (without the need for additional surgical instruments). One stainless steel prototype featured forceps and a manually-actuated knot-tying mechanism (Fig. 6). This prototype was not used for the study since its manual actuation proved cumbersome and it lacked a practical means of introducing new sutures.

Figure 6.

Prototype exhibiting integrated forceps

Potential use

Experimental data reveal that the automated suturing device has great potential for use in minimally invasive surgery. Although originally designed for cardiac procedures, the device's ease of use and low profile allow its utilization in virtually any thoracosopic or laparoscopic procedure. This device could reduce suturing times and therefore overall surgery times as compared to conventional laparoscopic approaches. It gives the surgeon complete control of tissue piercing and suture tensioning, yet allows him/her to generate knots without releasing either end of the suture. Knots generated with the device have equivalent strengths to those generated by conventional laparoscopic tools.

Limitations and future work

This study was designed to be an initial comparison of two different methods of laparoscopic knot tying within a simulated MIS setting. The testing protocol (five knots per device and surgeon) yielded statistically significant differences in knot-tying times. It is likely that increasing the number of knots per device would result in even larger differences, as the surgeon would benefit from a “learning curve” and become more skilled with the automated knot-tyer. Knot pushers were not used during these evaluations, but it is likely that their use would decrease the time necessary to complete knots via the conventional method. Although only two types of suture were evaluated (4-0 monofilament and 2-0 braided suture), we anticipate that the cartridge and rotating disk can accommodate a wide range of suture materials. However, further studies are warranted to fully assess the use of this prototype with other suture materials. Also, in vivo testing with biological tissues is needed to ensure the reliable production of high quality knots, not only in holding strength but also in suture tension, before clinical use of this knot-tying approach can proceed.