Noninfectious Skin Fistulization as a Delayed Complication of Lumbar Interlaminar Device Implantation: A Report of Two Cases

The main cause of delayed wound healing is infection. Infection rates after initial instrumented spine surgery reportedly range from 0.7% to 11.9%.^1-4 Delayed infectious skin fistulization after spinal instrumentation has been rarely reported. As of this writing, and to the knowledge of the authors of the current case reports, NIF after spinal instrumentation has never been reported.

This report presents 2 patients with NIF, with 1 case occurring 11 months and the other occurring 2 years after implantation of a lumbar interlaminar device (IntraSPINE [Cousin Surgery]).  In 1 patient, the device was implanted at 3 levels via a posterior approach for the management of kissing spine disease, and in the other patient the device was implanted for the management of recurrence of single-level lumbar disc herniation. The lumbar interlaminar device is used in the management of several conditions, including kissing spine syndrome, lumbar spinal canal stenosis, lumbar disc herniation, lumbar facet syndrome, and topping off fusion of the lumbar spine.^1,5-7 This device is made of soft silicone covered with a layer of polyester.

The 2 cases of NIF reported herein presented with a cutaneous lesion. Infection and granuloma were considered in the differential diagnosis. This report documents the management of this delayed, unusual complication of lumbar interlaminar device implantation.

Two patients were readmitted owing to the presence of a cutaneous lesion near a surgical incision scar. One patient (case 1) had undergone surgery 11 months before readmission, and the other patient (case 2) had undergone surgery 2 years before readmission. Contrast-enhanced MRI; measurement of serum procalcitonin CRP, erythrocyte sedimentation rate, and WBC counts; and microbiological sampling of the skin were performed for both patients. Both patients underwent surgery to manage the fistulization. Written informed consent was obtained from both patients.

Case 1

A 59-year-old female was readmitted to the neurosurgery department at Centre Clinical de Soyaux with a granuloma-like skin lesion at the lower part of the incision scar from previous spine surgery. Eleven months before readmission, she underwent implantation of lumbar interlaminar devices at the L2-3, L3-4, and L4-5 levels for the management of Baastrup syndrome (ie, kissing spine disease). The result of the initial surgery was satisfactory, with no complications. The patient returned home on the second postoperative day and returned to normal activity 1 month postoperatively. Postoperative wound healing occurred by primary intention, and the incision cicatrix appeared normal. The stitches were removed 14 days postoperatively. At follow-up examination 1.5 months and 3 months postoperatively, the scar was clean and dry, and the patient had no pain or other signs or symptoms.

On readmission 11 months after the initial surgery, the patient presented with a non-oozing, granuloma-like skin lesion that had appeared 2 weeks earlier at the lower part of the scar. The referring physician had advised the patient to apply silver nitrate solution to the skin, but doing so resulted in no improvement. The patient had no signs or symptoms. Contamination-free removal of the skin lesion was performed at the readmission visit. The result of microbiological culture was negative. The CRP level was 3 mg/L, procalcitonin level was less than 0.2 ng/L, and WBC count was 6.400/µL. Contrast-enhanced lumbar spine MRI demonstrated enhancement around the posterior spinous processes of L3, L4, and L5 lumbar vertebrae and around 3 interlaminar devices, particularly the device at the level of L4-5, as well as a fistulization of the device from the L5-S1 space to the skin (Figure).

After informed consent was obtained from the patient, debridement and rinsing of the subcutaneous aspect of the wound was performed. No purulent tissue was found. Macroscopically, the tissue had a consistency that was soft, non-exuding, and tearing, but it was neither liquid nor bloody. Anatomopathological analysis was positive for the presence of interstitial fluid. Several samples were obtained, and all bacterial cultures (aerobes, anaerobes) remained sterile. Subcutaneous tissues and skin were hermetically closed.

The patient returned to the neurosurgery department 3 weeks later. The lesion had reappeared exactly as before. Revision surgery was performed to remove the interlaminar devices. Several bacteriological samples were obtained intraoperatively, all of which were sterile. As of the time of this writing, the lesion had not recurred after 2-year follow-up.

Case 2

A 62-year-old female was readmitted to the neurosurgery department with a granuloma-like skin lesion at the lower part of the incision scar from previous spine surgery. The lesion had appeared 1 month earlier. Two years before this readmission, the patient underwent diskectomy and implantation of the interlaminar device at the L4-5 level for disc herniation at the same institution. The result of the initial surgery was satisfactory, with no complications and with a return to home on the first postoperative day and a return to normal activity 1 month postoperatively. Wound healing was by primary intention, and the incision cicatrix looked normal. The stitches were removed 12 days postoperatively. At follow-up examinations performed 1.5 months and 3 months postoperatively, the scar was clean and dry, and the patient had no pain or other symptoms or signs.

The referring physician performed microbiological samplings of the granuloma-like lesion, all of which were sterile. The CRP level was 2 mg/L, and the procalcitonin value was less than 0.2 ng/L. Contrast-enhanced lumbar spine MRI demonstrated enhancement around the posterior spinous process at the L5-S1 level as well as a fistulization of the device from the L5-S1 space to skin. However, the patient had no signs or symptoms. At the cutaneous level, the tissue had a consistency that was soft, non-exuding, and tearing, but it was neither liquid nor bloody.

Given the experience with case 1, the authors of this report proceeded to device removal. No suspicious purulent collection or inflammatory tissue was found intraoperatively. Several bacteriological samples (aerobes, anaerobes) were obtained. As in case 1, all samples remained sterile. Anatomopathological tests were not conducted. As of the time of this writing, the lesion had not recurred after 2-year follow-up.

Cutaneous opening of scar tissue after spine surgery is always concerning. It suggests first and foremost a postoperative surgical site infection, which, as a complication of spine surgery, is the primary cause of unplanned early readmission and sometimes requires revision implant procedures.^2-4,8-11 The incidence of infection after spine surgery is reportedly between 0% and 18%.^12,13 Most risk factors are well established, such as diabetes, obesity, smoking, malnutrition, and timing and duration of surgery.^14-24 Surgical site infection manifests with wound redness, swelling, pain, tenderness, purulent drainage, fever, headache, chills, and palpitation. CRP and procalcitonin levels are high.

MRI with gadolinium contrast is the best method for detecting postoperative spine infection. MRI findings associated with infection include fluid collections with rim enhancement and ascending epidural collections.^24,25 The 2 cases discussed in the current report exhibited no clinical or paraclinical findings of infection. Fistulization between the skin and device was clearly shown on MRI. The granuloma-like skin lesions in these 2 cases were suggestive of fistulization such as can occur after abdominal surgery.

Fistulization is a rare complication of visceral or obstetric and gynecologic surgery and is defined by abnormal communication between the digestive tract or genital organs and the skin. This communication gives rise to externalization of digestive or genital fluid by the newly formed fistular pathway.^25-27 In the vast majority of cases, surgery is performed.²⁸ The authors of the current report have not seen such a case following spine surgery, and to date, no cases have been documented in the medical literature. Only 1 article reported a case of delayed fistulization formation after elective spinal instrumentation,²⁹ but that case involved infectious fistulization formation rather than NIF. Some cases of postoperative skin lesions associated with device-related allergies have been reported, however.^30,31 These cases were associated with increased concentration of IgG protein. In addition, they were accompanied by other signs of allergy, such as pain, itching, and hives.

The primary limitation of this report is that it consists of only 2 cases. A lack of sufficient cases and insufficient data do not permit the authors to pinpoint causes and consequences of this fistulization.

To the knowledge of the authors of the current report, the 2 cases presented herein are the first documented cases of noninfectious fistulization formation after lumbar device implantation. Despite the rarity of their presentation, these cases serve as a possible complication for surgeons to consider when investigating the cause of delayed wound healing after spinal surgery.