Effect of the Wicking Behavior of Multifilament Sutures

JOURNAL OF ENDODONTICS Copyright © 2004 by The American Association of Endodontists

Printed in U.S.A. VOL. 30, NO. 9, SEPTEMBER 2004

Effect of the Wicking Behavior of Multifilament Sutures T. R. Grigg, DMD, F. R. Liewehr, DDS, MS, W. R. Patton, DDS, MS, T. B. Buxton, PhD, and J. C. McPherson, PhD

Homsy et al. (5) reported similar differential tissue response but suggested that the difference was caused by physical features of the suture, such as increased surface area and swelling when hydrated. Wallace et al. (6) reported favorable tissue response to an absorbable braided polyglycolic acid (PGA) MUS in clinical trials on human subjects. Lilly et al. (7) found that PGA MUS produced a tissue response similar to the nylon MOS and did not observe bacteria within the interstices of the MUS in histological specimens. In a follow-on study, the authors concluded that the PGA sutures inhibit bacterial penetration, accounting for its milder tissue response (8). The results of recent studies have confirmed the superiority of monofilament over braided sutures. Scher et al. (9) found significantly fewer bacteria adhered to the polypropylene MOS than either of the braided polyester MUS. Katz et al. (10) found monofilament nylon to bind 1⁄5 to 1⁄8 the number of bacteria bound by MUS. Durdley and Bucknall (11) found that braided materials produced a more prolonged tissue response and harbored more bacteria than MOS. Many clinicians, however, prefer MUS, because MOS are more difficult to manipulate, have sharp ends that irritate oral tissues, and exhibit poor knot security. Its use has been challenged by studies suggesting that it can “wick” bacteria into oral tissues causing severe inflammation (1, 3, 4, 7). The purpose of this study was to study the wicking behavior of currently available MUS.

The purpose of this study was to compare the wicking propensity of multifilament sutures. Dexon II, Vicryl, and black silk suture (BSS) were dipped in saline or soaked for 48 h, then suspended on a microscope slide. Fluorescein isothiocyanate-dextran (FITC-D) was placed at the suture mid points, and its movement was observed using fluorescence microscopy. The experiment was repeated, replacing the FITC-D with mixture of S. salivarius and saline, incubating the suture specimens in culture medium, and evaluating microbial growth. Dipped sutures showed FITC-D movement in the Dexon II group only. All 48-h soaked sutures demonstrated FITC-D movement with significant (p < 0.005) differences in mean times: BSS 179 ⴞ 42 s; Vicryl 120 ⴞ 26 s; and Dexon II 32 ⴞ 2 s. Dexon II suture demonstrated wicking of S. salivarius, whereas Vicryl and BSS did not (p < 0.05). These results suggest that BSS and Vicryl sutures do not wick as readily as Dexon II does.

In 1968, Lilly (1) observed a reduced inflammatory reaction in intraoral tissues sutured with monofilament suture material (MOS) compared with tissues sutured with multifilament suture (MUS). His results supported Katz and Evans’ (2) thesis that the physical characteristics of suture material may be a major determinant of tissue response. Lilly suggested that bacteria and oral fluids may be transmitted into a wound by a “wicking” action that would be more active in MUS than MOS. Lilly et al. (3) tested the hypothesis that the more severe response to MUS was caused by bacteria colonizing the interstices of braided suture by comparing the effect of systemic antibiotics tissue response to MOS and MUS. Although antibiotic administration reduced overall tissue response, a greater response to MUS was still noted. Lilly et al. (4) interpreted numerous small particles seen within the interstices of the MUS in histological specimens to be bacteria. Similar particles associated with MOS were fewer in number and did not permeate the suture. The authors concluded that bacteria sheltered within the interstices produced the severe tissue response.

MATERIALS AND METHODS Experiment 1 Thirty samples each of dry, 24-h, 48-h, and 72-h sterile, salinesoaked black silk suture (BSS) segments were measured on scanning electron microscope (SEM) images and compared using Student’s t test to determine the time needed to reach maximum saturation and expansion. Statistical significance was set at p ⬍ 0.05. Experiment 2 Three sterile 4-0 sutures, Dexon II PGA MUS (Sherwood Medical, St. Louis, MO), coated Vicryl MUS (Polygalactin 910, Ethicon, Somerville, NJ), and BSS MUS (Henry Schein, Melville, NY), were sectioned into 8-cm lengths. Sutures were prewetted by dipping them in sterile saline or soaked in saline for 48 h to achieve 649

650

Grigg et al.

Journal of Endodontics

maximum saturation. The sectioned pieces were then suspended horizontally between two 7-mm diameter aluminum studs 5.7 cm apart on a 7.6-cm long microscope slide, and their mid points labeled. Two microliters of 0.5% fluorescein isothiocyanate-dextran (FITC-D; FITC-Dextran, Sigma Chemical Co., St. Louis, MO) in saline was placed at the mid point of the suture. Movement of the FITC-D along the suture was observed using fluorescence microscopy. A stopwatch was used to record the time required for the FITC-D to travel 2.85 cm from the mid point suture to the right stud. Samples were observed for 1 h, and the experiment was repeated three times for each suture type. An analysis of variance and Student-Newman-Keuls test were used to compare the results between suture types. Statistical significance was set at p ⬍ 0.05.

TABLE 1. Two-way ANOVA of FITC-D movement Source Suture type Soak vs. dip Suture ⫻ Soak vs. dip

df Mean Square F Value Significance 2 1 2

6745530.9 23901698.0 5911212.7

16488.3 p ⬍ 0.0001 58423.5 p ⬍ 0.0001 14448.9 p ⬍ 0.0001

TABLE 2. S. salivarius movement past 1.3 cm

Dexon II Vicryl Black Silk

Dipped

48-h Soaked

1 of 3 0 of 3 0 of 3

2 of 3 0 of 3 0 of 3

Experiment 3 Experiment 3 Three full-length samples of each suture type (45 cm), dipped or soaked in sterile saline for 48 h, were suspended horizontally between sterilized studs mounted 30-cm apart on a sterile 33-cm ⫻ 25-cm Teflon cutting board. A 2-␮l mixture of viable S. salivarius bacteria (ATCC 7073, 1 ⫻ 107 colony-forming units/ml) and saline was placed at the mid point of each sample. After 8 min, the suture was sectioned into three 1.3-cm segments extending outward from the suture mid point, which were referred to as L1 (closest to the site of bacterial insertion), L2, and L3. Segments were immersed in sterile tubes containing 5 ml of brain-heart infusion (BHI) broth (Difco Laboratories, Detroit, MI) and incubated in 5% CO2 at 37°C for 24 h with uniform mixing. Microbial growth was ascertained by turbidity of the culture medium using a spectrophotometer at a wavelength of 540 nm (Spectronic 20, Rochester, NY). A sign test was performed to determine whether significant growth occurred in segment L2. Statistical significance was set at p ⬍ 0.05. An uncontaminated piece of suture was sectioned from the end of each sample before inoculation, incubated, and tested with the experimental samples to verify the sterility of the suture. Uncontaminated BHI broth and a tube containing broth plus a 2-␮l sample of S. salivarius for each suture tested were analyzed for turbidity at 24 h to verify sterility and viability, respectively.

RESULTS Experiment 1 Expansion of the silk suture reached maximal size by 48 h, which was significantly larger (p ⬍ 0.01) than the dry suture. Experiment 2 Using dipped sutures, the average time FITC-D took to travel 2.9 cm along the Dexon II suture was 45 ⫾ 1 s, whereas it exhibited no movement along the Vicryl or silk sutures during the 1-h observation period. In the 48-h soaked sutures, FITC-D movement occurred statistically faster along Dexon II than the other sutures (p ⬍ 0.005); average times were: Dexon II 32 ⫾ 2 s, which significantly faster (p ⬍ 0.05) than Vicryl 120 ⫾ 26 s, which was significantly faster (p ⬍ 0.05) than BSS 179 ⫾ 42 s (Table 1).

Dexon II suture (dipped or soaked for 48 h) also demonstrated wicking of S. salivarius within 8 min, whereas Vicryl and BSS did not (p ⬍ 0.05; Table 2). No growth was seen for any of the uncontaminated samples of suture or BHI broth, confirming sterility. All samples of S. salivarius were found to be viable. DISCUSSION Sutures placed after surgery are partly embedded in tissue and partly bathed in saliva, with a mean concentration of approximately 750 million bacteria/ml. The bacterial plaque that forms on the suture contains 200 billion cells per gram, similar in density to centrifugally packed cultured cells (12). The inflammation caused by these bacteria produces erythema surrounding the puncture wounds and leads clinicians to suspect that the suture could wick the bacteria into the surgical site itself. Investigators (1– 8) performed a variety of in vitro and in vivo studies attempting to identify characteristics of suture material that produced the differential tissue reaction they saw to MOS and MUS. Lilly (1) suggested that the physical character of the suture might play a role and called the presumed ability of the MUS to transmit bacteria into the depth of a wound wicking. Despite inexact histological criteria and lack of statistics, his data do not seem to be divided neatly between MOS and MUS. Monofilament chromic gut, for example, produced approximately the same percentage of moderate responses (37%) as braided polyester fiber (35%), and nearly as high a percentage of severe responses (21%) as dermal silk (29%). Lilly et al. (3) repeated the experiment after premedicating the animals with antibiotics before surgery and daily until they were killed. No statistics are presented, but the data are once again not clearly divided. BSS produced the same percentage of moderate reactions (59%) as monofilament surgical gut, whereas monofilament steel and chromic gut produced more severe reactions (19%) than BSS and nearly as many as braided polyester (25%). More importantly, if bacteria wicked by MUS are responsible for the difference between suture types, why did the antibiotic administration not eliminate this difference? Because no unmedicated control group was included, the overall role of bacteria and the effectiveness of the antibiotic protocol are unknown. Katz and Evans (2) hypothesized that the roughened exterior of braided suture causes physical damage when placed, which produces the differential reaction. However, it seems unlikely that this minor initial trauma could produce the intense inflammatory reaction seen in biopsy specimens (1, 3, 4, 6). Lilly et al. (4) included

Vol. 30, No. 9, September 2004

FIG 1. SEM image of dry, sterile, black silk suture.

siliconized and teflonized suture materials among their samples to reduce this trauma, but reported that these treatments “did not appear to alter appreciably” the tissue response. Although no statistics are presented, coated sutures produced a low frequency of severe reactions (22% and 25%), like monofilament chromic gut (20%), compared with multifilament materials (39 –57%). Lilly et al. (7) soaked silk suture material in antibiotic agents before placement and compared those sutures to MOS and PGA MUS. They additionally examined the sutures histologically and found tissue reaction to be directly related to the presence or absence of bacteria in the filamental interstices. Although no statistics are presented, the data seem to show nylon producing the least tissue response, silk the most, and PGA MUS, chromic gut, and various antibiotic-impregnated BSS producing an intermediate level of inflammation. Additionally, no bacteria were seen associated with the PGA suture. The authors interpreted their findings as indicating that (a) antibiotic impregnation of silk suture reduces inflammation by reducing bacterial colonization, and (b) PGA suture in some way inhibits bacterial infiltration. To elucidate the mechanism of wicking and to devise strategies to prevent bacterial transmission by suture, it is necessary to consider passive bacterial transmission by fluid movement caused by capillary action between the suture fibers separately from the movement of nonmotile bacteria by colony growth and expansion. To operationally separate these two mechanisms, we first looked at the travel of FITC-D, a fluorescent marker for fluid movement, along the suture, and then at the movement of S. salivarius in a similar fashion. During manufacturing, natural waxes are removed from silk suture to allow a tighter, more compact braid. The strands are then coated with a mixture of waxes or silicone to improve handling and reduce capillary action (13). Homsy et al. (5) found that silk swelled up to 70% during a 14-day implantation, suggesting that these coatings did not prevent the suture from absorbing liquid. This swelling could open spaces between the fibrils, increasing both the capillary action of the suture and interstitial space in which bacteria could lodge. In experiment 1, we plotted the expansion of silk suture over time and found that most of the expansion was complete after 48 h. We used this value to saturate the sutures in experiments 2 and 3. Interestingly, the BSS only swelled a maximum of 33% of its diameter compared with the 70% found in Homsy’s study, possibly because of improved impregnation of the silk in contemporary suture material (Figs. 1 and 2). In experiment 2, we used FITC-D, a fluorescent dye marker, to investigate fluid movement along a suture that had been dipped momentarily in saline, to mimic wetting the dry-packed suture before it is used clinically, and suture that had been soaked in

Suture Wicking

651

FIG 2. SEM image of 48-h, sterile, saline-soaked, black silk suture.

saline for 48 h, to simulate suture that had been in place in the tissue long enough to swell to maximum size. We observed the samples for 1 h, because in pilot studies any movement that was going to occur took place within minutes of placement of the FITC-D. Using dipped suture, fluid movement was only seen along the Dexon II suture and not the silk or Vicryl. Using suture soaked for 48 h, movement was seen along all three types, with Dexon II being faster than Vicryl, which in turn was faster than silk. Lilly et al. (4) attempted to siliconize or teflonize MUS to reduce wicking but reported that this treatment did not appreciably alter the tissue response to the suture. Fluid movement was not studied; the authors assumed that increased tissue inflammation was caused by increased numbers of bacteria, and therefore to wicking. The techniques used may have been ineffective because only the outermost surface of the suture was treated. In contrast, in our study fluid movement did not occur in dipped silk or Vicryl suture, possibly reflecting improvements in processing and impregnating the silk suture in recent years. It also is possible that differences between the original Dexon and contemporary Dexon II have increased its capillary fluid transmission. In soaked suture, all types wicked, but silk took the longest of the three materials to exhibit fluid movement. In experiment 3, we looked at the movement of S salivarius, one of the most regularly occurring organisms of the human mouth (14), along the suture to determine whether the bacteria would follow the same pattern as the liquid. Pilot research indicated that if movement were going to occur to segment L2, it would occur within 8 min. Additionally, it seemed that if movement would occur, the migration would surpass 1.3 cm, but perhaps because of the absorption of the liquid by the suture, movement would not exceed 2.6 mm. Oral bacteria require approximately 20 min to replicate and thus for the colony to expand or move by that mechanism. By keeping the exposure time short, we could ensure that any movement seen was caused by fluid movement and not bacterial growth. Wicking was operationally defined as positive growth from segment L2 as evidenced by turbidity. Results with both the dipped and soaked sutures (Table 1) demonstrated that, in contrast to previous studies (6 – 8), Dexon II wicked bacteria under both conditions, whereas neither silk nor Vicryl wicked at all. The results of our in vitro study suggest that silk does not produce more fluid movement by capillary action than the other two braided sutures, and actually produces less than Dexon II. Additionally, it seems to mechanically transmit bacteria less than Dexon II does. These results are in agreement with Chatterjee (15), who found no difference in tissue reaction to Dexon, collagen, or silk in ophthalmic surgery, and Blomstedt (16), who found no difference in serious infections when silk or PGA MUS were used. Further research is needed to investigate mechanisms to reduce active bacterial migration by colony growth.

652

Grigg et al.

Journal of Endodontics

This study is the work of the United States government and may be reprinted without permission. Opinions expressed herein, unless otherwise specifically indicated, are those of the authors. They do not represent the views of the Department of the Army or any other Department or Agency of the United States government. Dr. Grigg is Captain in the U.S. Army Dental Corps, and former resident, U.S. Army Endodontic Residency Program, Fort Gordon, GA. He is currently a staff endodontist at Ft. Hood, TX. Dr. Liewehr is former program director, and Dr. Patton is former assistant program director, U.S. Army Endodontic Residency Program, Fort Gordon, GA. Dr. Buxton is microbiologist, and Dr. McPherson is biochemist, Department of Clinical Investigation, Dwight D. Eisenhower Army Medical Center, Fort Gordon, GA. Address correspondence to Frederick R. Liewehr, DDS, MS; E-mail: [email protected].

References 1. Lilly GE. Reaction of oral tissues to suture materials. Oral Surg Oral Med Oral Pathol 1968;26:128 –33. 2. Katz AB, Evans HD. Linear polyethylene sutures: an evaluation of tissue reaction. Am J Surg 1962;103:208 –16. 3. Lilly GE, Armstrong JH, Salem JE, Cutcher JL. Reaction of oral tissues to suture materials II. Oral Surg Oral Med Oral Pathol 1968;26:592–9. 4. Lilly GE, Salem JE, Armstrong JH, Cutcher JL. Reaction of oral tissues to suture materials III. Oral Surg Oral Med Oral Pathol 1969;28:432– 8.

5. Homsy CA, McDonald KE, Akers WW, Short C, Freeman BS. Surgical suture-canine tissue interaction for six common suture types. J Biomed Mater Res 1968;2:215–30. 6. Wallace WR, Maxwell GR, Cavalaris CJ. Comparison of polyglycolic acid suture to black silk, chromic, and plain catgut in human oral tissues. J Oral Maxillofac Surg 1970;28:739 – 46. 7. Lilly GE, Cutcher JL, Jones JC, Armstrong JH. Reaction of oral tissues to suture materials IV. Oral Surg Oral Med Oral Pathol 1972;33:152–7. 8. Lilly GE, Osbon DB, Hutchinson RA, Heflich RH. Clinical and bacteriologic aspects of polyglycolic acid sutures. J Oral Mafillofac Surg 1973;31: 103–5. 9. Scher KS, Bernstein JM, Jones CW. Infectivity of vascular sutures. Am Surg 1985;51:577–9. 10. Katz S, Izhar M, Mirelman D. Bacterial adherence to surgical sutures. A possible factor in suture-induced infection. Ann Surg 1981;194:35– 41. 11. Durdey P, Bucknall TE. Assessment of sutures for use in colonic surgery: an experimental study. J R Soc Med 1984;77:472–7. 12. Miller CH. Microbial ecology of the oral cavity. In: Schuster GS, ed. Oral microbiology and infectious disease. Philadelphia: B. C. Decker, 1990: 465. 13. Forrester JC. Suture materials and their use. Br J Hosp Med 1972; 578 –92. 14. Miller CH. Oral microbial flora. In: Schuster GS, ed. Oral microbiology and infectious disease. Philadelphia: B. C. Decker, 1990:447. 15. Chatterjee S. Comparative trial of Dexon (polyglycolic acid), collagen, and silk sutures in ophthalmic surgery. Br J Ophthalmol 1975;59:736 – 40. 16. Blomstedt GC. Infections in neurosurgery: a randomized comparison between silk and polyglycolic acid. Acta Neurochir (Wien) 1985;76: 90 –3.