Page 14/15

Studies of wound healing indicate that tissue reaction begins immediately after any break in the integument. Body fluids and bleeding from capillaries form a coagulum of a hydrophilic fibrin- fibronectin gel and cellular debris. Various cytokines coordinate the subsequent entry of inflammatory cells, fibroblasts, and the formation of new blood vessels (11). Polymorphonuclear leukocytes phagocytize local bacteria and participate in the formation of a scab, which is desiccated coagulum. Healing of wounds starts with transformation of the coagulum into granulation tissue, composed mostly of new vessels and fibroblasts. New epithelial cells spread upon this tissue. To the best of our knowledge, the time sequence of this process in humans, in the presence of a foreign body, has not been studied.

Optimal Healing Exit: We considered fast-healing exits as normal or optimally healing. At one week slight tenderness was present in about 30% of exits, a scab was visible in almost all exits, and epidermis surrounding the exit orifice was pale pink or pink. A small amount of serosanguineous, bloody, or serous drainage was visible around the exit in about half of patients; only one patient had a larger amount of serous drainage. There was no swelling. Drainage inside the sinus was almost always visible and was similar in character to that seen outside. There was no epithelium visible in the sinus; the sinus was lined with a white tissue, which appeared hard and resembled aponeurosis. It is most likely that this lining is composed of collagen fibers, since granulation tissue was not yet formed at week 1.

External drainage abated by week 2 and was absent by week 3. Scab size diminished by week 3 and scabs were not seen after week 4. Exit color remained pale pink or pink throughout the 6-week evaluation. Drainage in the sinus diminished, and most sinuses were dry at week 6. The sinus lining remained flat, but was gradually transformed to plain granulation tissue. Sometimes as early as week 2 the vessels became visible and/or the surface appeared mottled (partly white and partly pink). The color gradually changed to pink, but sometimes remained white as late as 6 weeks. Epithelium started entering the sinus by week 2 or 3, progressed steadily, and covered at least half the visible sinus tract by 5 weeks after implantation. At times it was difficult to differentiate between granulation tissue and epithelium. If the surface wrinkled on pressure, it was identified as epithelium. Epithelium was fragile and pale pink, or occasionally white. The white epithelium (Figure K) may indicate that the epithelium progressed over collagen fibers before granulation tissue developed. This observation challenges the role of collagen in inhibiting epithelial cell spreading.

Slow-healing exits appeared similar to the fast-healing exits, but the progression of epithelium into the sinus was delayed by approximately 2 - 4 weeks. Drainage lasted longer, and the amount of sinus drainage was somewhat larger.

Interrupted healing exits looked similar to fast-healing exits during the first 2 weeks. Between 3 and 5 weeks the progression of epithelium was halted. Signs of impending infection developed. In five of nine sinuses the epithelium regressed; in the other four progression ceased. In all exits the drainage increased and became purulent, and granulation tissue became soft or fleshy. The signs were visible only in the sinus in most patients. These signs resemble those of equivocal exits according to the classification of healed exits (6). Bloody and/or purulent external drainage was seen in only two patients at weeks 5 or 6. In one patient the external color became deep pink. Overt infection (purulent external drainage, proud flesh) developed within 4 weeks in more than half the patients, and all but one had an exit infection within 18 weeks postimplantation (Figure 3).

Early infected exits did not show signs of healing (epithelium entering the sinus or progressing, decreasing drainage amount). Instead, drainage became purulent, the sinus lining became composed of granulation tissue at week 1 or 2, and the tissue became slightly or frankly exuberant. In two exits signs of infection [purulent drainage in one, necrotic rim (Figure L) and inflammation in one] were present as early as at one week. Early infection was diagnosed using similar criteria as for acute infection of healed exits (6); however, the definition was modified to include the presence of purulent drainage and/or exuberant granulation tissue in the sinus only when associated with no signs of healing. A healed exit with the same signs would be classified as equivocal (6). Early infected exits did not begin to heal until appropriate treatment was implemented.

Washout Cytology: We anticipated that the sinus washout cytology would be of value in the categorization of the healing pattern. Unfortunately, the number of cells depended on the method by which the sample was taken. This method could not be sufficiently standardized for comparison purposes, even though only one person took all samples. The depth of the sinus dictated the amount of saline that could be injected into the sinus, and manipulation of the catheter during photography prior to sample-taking influenced the amount of drainage that could be washed out from the sinus. Ultimately, the cytology was found inferior to simple exit and sinus inspection with the use of a magnifier.

Systemic Factors Interfering with Healing: Impaired nutrition, diabetes mellitus, uremia, and corticosteroids are all known factors decreasing wound healing by decreasing fibrosis (12). In our study these factors played a minor role, if any (Table 1). Judging by serum albumin concentration, the four groups were similar regarding nutrition. Obesity and diagnosis of diabetes tended to be higher in early infected groups. This is consistent with the surgical literature.

Exit Colonization: Colonization of the exit by bacteria seems to be the most important risk factor for early infection. Seventy-five percent of early infected exits had a positive periexit smear by week 1, and all were colonized by the second week. Results from the sinus washout were similar; however, positive cultures from washouts in early infected exits tended to occur later than those from periexit smears. This may indicate that bacterial invasion was coming from outside in the majority of cases.

Survival to the first peritonitis episode was not different among the four classification groups (Figure 6); however, patients with the first positive culture from the periexit smear at 5 weeks or later developed the first peritonitis episode later than other groups (Figure 7). The first positive culture from washout was not significantly related to survival to the first peritonitis episode, although there was a trend for later peritonitis in the groups with the first positive culture at 3 weeks or later (Figure 8). The early infected group had a higher peritonitis rate than the fast- healing exit group; peritonitis rates among the groups were significantly different (Table 7). A definitive trend for lower peritonitis rates in groups having exits colonized later is an interesting and previously unreported observation (Tables 8 and 9).

Early colonization seems to be an important risk factor for catheter removal for exit/tunnel infection. Although the organisms first cultured from the periexit smears and sinus washouts were usually different from those cultured from the exit exudate and peritoneal dialysate at the time of catheter removal (Table 12), we believe that early colonization and subsequent early infection interferes with normal healing. One may speculate that early infection disrupts the ingrowth of tissue into the cuff, leaving a defect; therefore, the cuff does not constitute an effective barrier to bacterial penetration into the tunnel. In study patient No. 25 [Table 12 and case reported in the accompanying paper (6)] the mechanism might be different. It seems that in this case early infection led to a partially suppressed cuff infection, which ultimately caused tunnel and deep cuff infection leading to peritonitis and catheter demise.

Sinus epithelialization is an excellent indicator of the quality of healing. In well- (fast-) healing exits epithelium entered into the sinus by week 2 or 3, progressed steadily, and covered at least half the visible sinus tract by 5 weeks after implantation. In slow-healing exits without signs of infection sinus epithelialization was slower. These exits were less resistant to infection compared to fast-healing exits. Although no slow-healing exit developed infection during the 6- week study period, there was more than a 50% chance of developing exit infection by 16 weeks. Regression of epithelium was an early sign of impending infection. At the same time granulation tissue became exuberant. Whereas epithelium may spread over plain granulation tissue, exuberant granulation tissue not only inhibits progression of epithelium, but almost always causes regression of epithelium. It looked as if exuberant granulation tissue was "pushing" epithelium out of the sinus.

Bacterial Colonization and the Healing Process: Infection is the major cause of impaired wound healing (13). It has been well established in the surgical literature that wound infection is the result of major disturbance in the balance between host defense and bacteria (13). The number of bacteria as a critical factor in wound infection was already recognized in World War I (14). Elek (15) demonstrated that it requires 7.5 - 106 staphylococcal organisms to produce a pustule in normal human skin, but the number of bacteria necessary to cause infection was reduced 10 000- fold in the presence of a single suture. Bacterial virulence is also important; Staphylococcus aureus or Pseudomonas aeruginosa is more likely to induce an inflammatory response than is Staphylococcus epidermidis.

Our study clearly demonstrates the importance of delayed colonization for optimal healing, reduced exit infection and peritonitis rates, and catheter survival. Whereas Staphylococcus coagulase-negative prevailed in the first positive cultures in all groups, the infected groups were more likely to have Gram-negative bacteria and Staphylococcus aureus in the first positive culture. Late colonization, after the healing process is completed, is inevitable and mostly harmless, provided that the defense mechanisms are intact.

This study supports the rationale for the Moncrief-Popovich technique of peritoneal catheter insertion. This technique allows tunnel healing without exposure to the skin surface area and ipso facto eliminates bacterial interference with tissue ingrowth into the cuff material. Unlike any other technique the distal (external segment) of the catheter is completely buried and remains in the subcutaneous tunnel until exteriorized after 3 - 8 weeks post catheter insertion (16). Using Swan neck catheters with this technique, Moncrief and co-workers reported a significant reduction in peritonitis incidence (16). Healing of the sinus after exterioration is subjected to bacterial interference; hence the exit infection rates are not decreased.

Staphylococcus aureus Nasal Carriage: The importance of Staphylococcus aureus as an etiologic agent of peritoneal catheter exit-site infection has been well established (17,18). Nasal carriage status of is reported to be common in patients undergoing hemodialysis (19) and peritoneal dialysis (20,21). A recent multicenter study found an increased incidence of exit-site infections in nasal carriers of Staphylococcus aureus; in 85% of these infections the strain from the nares and the strain causing the infection were similar in phage type and antibiotic profile (22). However, in previously published preliminary data we reported that even though Staphylococcus aureus was more likely to be found in nares of patients with exit infections, by antibiotic profile the strain causing exit infection and the strain cultured from nares were different (23). Judging from those preliminary data, there is an increased probability of exit infection in patients who carry Staphylococcus aureus in nares, but the strain is usually different.

According to the expanded data in this report, organisms and/or strains, regardless of species, cultured from nares, are usually different than those cultured from periexit smears and sinus washouts. The organisms cultured from periexit smears and sinus washouts are more likely to be identical (or culture may be negative in both places) than the organisms cultured from nares are to be identical to either sinus washouts or periexit smears. It is worth remembering that whereas cultures of the skin surrounding the exit and of the nares were taken by smear with a sterile, rayon-tipped swab, the cultures from the sinuses were taken by saline washouts and the initial culture media were different. Thus it seems that the close proximity between the culture sources rather than identical culture techniques is more likely to yield identical strains.

We cannot explain why our patients usually showed colonization of nares and the exits by different strains, whereas in other series the strains were more uniform. Hands are probably the major means of spreading bacteria to distant parts of the integument. Differences in habits, hygiene, exit-site care, and/or surgical practices are among the possible explanations.

Thus our results suggest that intranasal antimicrobial agents would be less efficacious in preventing exit infections than topical application at the exit or systemic use of antimicrobial agents.

Exit Care Principles: Based on the present study several principles emerge as important elements of preimplantation patient preparation, surgical technique, and postimplantation care. This study indicates that the exits are colonized predominantly by Gram-positive bacteria; however, in three cases the exits were colonized by Gram-negative bacteria, and these cases developed early infection. Therefore, coverage for both Gram-positive and negative organisms is desirable for prophylaxis. The surgical preparation of the integument should be similar to that used for major surgery. Perfect hemostasis during surgery and wound irrigation before closure are of utmost importance for preventing hematoma formation.

Because of the large amount of drainage in the postimplantation period, local antibiotics have a lower chance of effectiveness compared to the systemic route of delivery. It seems logical that systemic antibiotic therapy covering both Gram-positive and Gram-negative organisms would decrease the incidence of early infection and interrupted healing, as well as long-term infectious complications.

The exit should be covered with several layers of high absorbency gauze dressings and secured with good adhesive to immobilize the catheter. An occlusive or semipermeable dressing is not indicated, because it keeps drainage trapped around the exit. Such drainage is an excellent medium for bacterial growth. The dressing can be left in place for a week. There are two reasons for infrequent dressing changes: first, each dressing change may introduce bacteria into the exit even though a sterile procedure is used. Second, the less manipulation of the catheter, the lower chance of exit trauma. In cases with excessive bleeding or large quantity of drainage from the incision or exit, the dressing should be changed earlier and more frequently.

To minimize chances of external contamination, dressing changes should be done using sterile procedure. The skin surrounding the exit should be cleansed with a nonirritating agent to decrease the number of bacteria. Strong oxidants are cytotoxic to mammalian cells in bacteriocidal concentrations (24) and are harmful to granulation tissue if allowed to enter the sinus. Nonionic surfactant that cleanses the exit and is not harmful to the granulation tissue seems to be preferable to strong disinfectants. The quality of healing should be evaluated until the exit is healed. Weekly dressing changes may be continued throughout the 6-week healing period if drainage is minimal or absent and epithelium progresses steadily.

If healing is not progressing as desired, it is likely that the exit is already colonized and a clinical culture should be taken. The frequency of dressing changes should be increased to every other day, because the major rationale for infrequent dressing changes (avoidance of exit colonization) no longer exists, and more frequent cleansing of the exit will decrease the number of bacteria. If the amount of drainage is minimal, local antibiotics according to sensitivity may be sufficient. With large drainage and overt infection, systemic antibiotics are necessary. Antibiotics should be adjusted according to the sensitivity results.

Peritoneal Dialysis International, Vol. __, pp. ___ Printed in Canada. All rights reserved. Copyright © 1996 International Society for Peritoneal Dialysis ____-____/96 $3.00 + .00