Background

Pressure sores in patients with spinal cord injury (SCI) are sometimes refractory to treatment. Chronic gastrointestinal symptoms are also frequently seen in patients with SCI [1, 2], and malnutrition caused by decreased gastrointestinal motility in SCI patients is a major cause and exacerbating factor of pressure sores. Evaluation of nutritional status in patients with pressure sores is essential [3], as nutritional intervention can be a valuable treatment option for pressure sores. However, small intestinal bacterial overgrowth (SIBO) is rarely considered in the evaluation of malnutrition in SCI patients with pressure sores. SIBO is defined as the presence of more than 1 × 105 colony forming units (CFU)/mL of bacteria or any amount of E. coli in the proximal small bowel content [4]. Here, we report the case of an SCI patient with a refractory sacral pressure sore that healed after starting treatment for SIBO. To the best of our knowledge, this is the first report of an association between a pressure sore and SIBO.

Case presentation

A 66-year-old woman visited our hospital for the purpose of treating her sacral pressure sore (day 0), which she developed 8 months prior due to bed rest during treatment of a left humeral fracture in another hospital. She had paraplegia as well as bladder and rectal disturbance due to SCI at the fourth lumbar level (L4) caused by a suicidal jump in response to paranoid delusions at 52 years of age. Spinal fusion surgery and cystostomy were performed early after SCI. Otherwise, she had a history of hysterectomy due to uterine cancer at 31 years of age, lymphaticovenular anastomosis as a treatment for post-hysterectomy lymphedema in the bilateral lower extremities at 50 years of age, and cholecystectomy at 60 years of age.

When she visited our hospital, she was taking the following oral medicines: propiverine hydrochloride, vitamin B12, etizolam, flunitrazepam, sodium bicarbonate anhydrous monobasic sodium phosphate mixture, Clostridium butyricum tablets, sodium risedronate hydrate, rebamipide, sodium ferrous citrate, fursultiamine hydrochloride, alfacalcidol, and potassium L-asparate. She did not take proton pump inhibitors (PPI). Her vital signs were as follows: body temperature of 36.9 °C, low blood pressure of 83/50 mmHg, pulse rate of 82 bpm, and respiratory rate of 12 per min. Physical examination showed abdominal distension, emaciation with a body mass index of 15, and a sacral pressure sore 10.0 × 6.5 cm in size including a pocket entrance of 6.0 × 3.5 cm (Fig. 1a). Most of the surface of the pressure sore was covered by granulation. Our evaluation of the pressure sore with DESIGN-R [5] was D3 e3 s8 i0 g3 N3 P24, with a total score of 44 (Table 1).

Fig. 1
figure 1

Patient images. a, b, c, and d Sacral pressure sore. a Day 0: sore 10.0 × 6.5 cm in size with an entrance of 6.0 × 3.5 cm; DESIGN-R score was D3 e3 s8 i0 g3 N3 P24 with a total score of 44. b Day 405 (i.e., 285 days after starting SDD for treating SIBO): reduced size of sore; DESIGN-R score was D3 e3 s3 i0 g1 n0 p0 with a total score of 10. c Day 742 (i.e., 617 days after staring SDD): healed sore; DESIGN-R score was d0 e0 s0 i0 g0 n0 p0 with a total score of 0. d Day 1057 (i.e., 932 days after staring SDD): no recurrence of the sore. e and f Osteoporosis and multiple fractures. e X-ray showing left tibial fracture. f Tc-99 m bone scan showing accumulation in multiple ribs, vertebrae, and right ulna. g, h, and i Endoscopic examination and results of bacterial culture of the upper digestive tract. All stomach, duodenum, and proximal jejunum samples were positive for E. coli. g Stomach. Food residue can be seen. Acid level was decreased to pH 7.0. h Duodenum. Food residue is evident. i Proximal jejunum. Flat villi and a jejunal ulcer are observed

Full size image
Table 1 DESIGN-R assessment tool for pressure sore. (Reprinted with permission from John Wiley and Sons. In: Matsui et al., Development of the DESIGN-R with an observational study: an absolute evaluation tool for monitoring pressure ulcer wound healing. Wound Repair Regen 2011, 19(3):309–315)
Full size table

Bacterial culture examination of the pressure sore showed Corynebacterium striatum and methicillin-resistant Staphylococcus aureus. Laboratory data showed an elevated serum alkaline phosphatase level of 1260 U/L and low serum levels of hemoglobin 11.9 g/dL, albumin 2.8 g/dL, calcium 7.8 mg/dL, and zinc 51 μg/dL. On day 22, we observed a sudden decrease of hemoglobin to 6.4 g/dL with a positive fecal occult blood test, bilateral pleural effusion on chest x-ray, and serum albumin level of 2.3 g/dL. Upper gastrointestinal endoscopy showed a gastric ulcer at H2 stage.

As a result of searching for the cause of alkaline phosphatase elevation, bilateral tibial fracture, multiple rib fracture, and osteoporosis were found (Fig. 1e and f). Femoral bone density was 17% of the young adult mean. A low serum inorganic phosphorous level was found (Table 2), along with a low serum level of 25-hydroxy vitamin D3 (25(OH)VitD3) below the detection limit and elevated level of parathyroid hormone (Table 3). Levels of other fat-soluble vitamins were also low: vitamin A < 5 μIU/dL, vitamin K1 < 0.05 ng/dL, and vitamin E 0.21 mg/dL. Examination using ultrasound and computed tomography showed normal thyroid and parathyroid glands. Basing on these finding, we diagnosed osteomalacia with vitamin D deficiency.

Table 2 Laboratory data before starting supplementation with vitamin D
Full size table
Table 3 Vitamins and bone metabolism markers before starting supplementation of vitamin D
Full size table

On day 61, oral supplementation of calcium, phosphorous, and vitamin D was started. Despite supplementation, serum levels of calcium, phosphorous, and 25(OH)VitD3 on day 89 showed poor improvement: calcium 7.5 mg/dL, phosphorous 2.4 mg/dL, and 25(OH)VitD3 below the detection limit.

On day 117, we performed bacterial examination of collected gastric and duodenal fluid with suspicion of a disturbance in absorption, which showed elevated numbers of bacteria in gastric content (104 E. coli, 105 Streptococcus species, and 105 Neisseria species) and duodenal content (106 E. coli, 104 Candida glabrata) (Fig. 1g, h, and i). Therefore, we diagnosed SIBO. On day 125, we started selective decontamination of the digestive tract (SDD) using oral administration of polymyxin B sulfate, 2.5 million units, daily and oral administration of amphotericin B, 300 mg, daily. Before starting SDD, the pressure sore was refractory to multiple methods of wound treatment, including depressurization, irrigation, debridement, ointment, basic fibroblast growth factor, and negative pressure wound therapy. After starting SDD, the pressure sore began to heal. On day 410 (i.e., 285 days after starting SDD), the pressure sore DESIGN-R score was D3 e3 s3 i0 g1 n0 p0 with a total score of 10 (Fig. 1b). Regarding the nutritional status, the serum albumin level increased from 1.8 g/dL, just before starting SDD, to 3.2 g/dL at 285 days after starting SDD. Also, the hemoglobin level increased to 12.1 g/dL and the serum zinc level increased to 67 μg/dL. Serum levels of calcium, phosphorous, vitamin D, and other fat-soluble vitamins also gradually increased (Fig. 2 and Table 4). A repeat diagnostic bacterial examination of the upper digestive tract contents, for SIBO, was not performed because of obvious improvements in most of the laboratory data. There were no adverse effects of SDD, such as antibiotic-associated diarrhea. In contrast, the patient presented with constipation that was noted before starting SDD. The sacral pressure sore was completely healed on day 742 (i.e., 617 days after starting SDD), with a DESIGN-R score of d0 e0 s0 i0 g0 n0 p0 and total score of 0 (Fig. 1c). In addition, the patient showed improved nutritional status, and had a serum albumin level of 3.4 g/dL. We successfully reduced the dose of polymyxin B from 2.5 to 1.5 million units daily, similarly the dose of amphotericin B was reduced from 300 to 200 mg daily on day 904 (i.e., 779 days after starting SDD) without any signs of SIBO recurrence. There was no recurrence of the sacral pressure sore with a serum albumin level of 3.5 g/dL on day 1057 (i.e., 932 days after staring SDD) (Fig. 1d). On day 1493, we successfully ended the use of amphotericin B; however, the use of polymyxin B at 1.5 million units per day continued. On day 1815 (i.e., 1690 days after starting SDD), while still using polymyxin B at 1.5 million per day, the serum albumin level was 3.9 g/dL, the hemoglobin level was 13.9 g/dL, and the serum zinc level was 114 μg/dL. There were no signs of SIBO recurrence or the sacral pressure sore.

Fig. 2
figure 2

Bone metabolism markers. After starting SDD, levels of bone metabolism markers gradually improved

Full size image
Table 4 Vitamins and trace elements before and after supplementation and selective digestive decontamination (SDD)
Full size table

Discussion and conclusions

We report the case of a patient whose sacral pressure sore and osteoporosis were improved by treatment for SIBO. Although nutrition status is known to be important for the healing of pressure sores, SIBO is rarely checked as a cause of malnutrition in patients with pressure sores. However, SIBO is a potential cause of malnutrition in patients with SCI due to decreased intestinal motility resulting from autonomic disturbances and reduced physical activity [6]. SCI is also a risk factor for pressure sores [7]. However, to the best of our knowledge, there are no previous reports of an association between pressure sores and SIBO. Thus, our case draws attention to the fact that SIBO can be an overlooked cause of poor wound healing during the treatment of pressure sores.

SIBO was first reported by Vantrappen et al. as an increased concentration of 14CO2 in a bile acid breath test for patients with an absent interdigestive motor complex [8]. Today, consensus diagnostic criteria for SIBO are the presence of more than 1 × 105 CFU/mL of bacteria or any amount of E. coli in the proximal small bowel content [4]. Relatively little is known about commensals inhabiting the small intestine, mainly due to the limited accessibility of this environment for microbiological analysis [9]. In the healthy state, the numbers of intestinal bacteria range from 102 to 105 CFU/mL and mainly include gram-negative and gram-positive aerobes, such as Streptococcus, Lactobacillus, and Bacteroides species [10]. Regarding the amount of bacteria in proximal jejunal aspiration, Khoshini et al. report that normal subjects rarely exceed 1 × 103 CFU/mL and therefore proposed more than 1 × 103 CFU/mL coliform bacteria as the threshold for SIBO [11]. In our case, 106 CFU/mL E. coli existed in duodenal content, and 104 CFU/mL E. coli existed in gastric content, which met the traditional diagnostic criteria of SIBO.

Other diagnostic methods for SIBO are breath tests using hydrogen or hydrogen methane with lactulose or glucose [10]. Breath tests have clinical utility for diagnosing SIBO because they are less invasive than obtaining proximal small bowel content. However, there are no standardized criteria for diagnosing SIBO using breath tests [10]. We did not perform a breath test in our case study.

Despite no previous reports of an association between unhealed pressure sores and SIBO, nutritional status is known to be important for the healing of pressure sores [3]. In our case, the sacral pressure sore, which was initially refractory, began to heal after starting treatment for SIBO. Among intrinsic factors related to the healing of pressure sores, blood levels of hemoglobin, albumin, and zinc are especially important [12]. Our patient had anemia, hypoalbuminemia, and a low zinc concentration, which gradually improved after starting treatment for SIBO.

SIBO is caused by multiple factors, including disturbances in defense mechanisms of the digestive tract, anatomical abnormalities, surgical interventions, and disturbed gastrointestinal motility [6, 10]. Bures et al. described several endogenous defense mechanisms that prevent bacterial overgrowth [4], including secretion of gastric acid, intestinal motility, a properly functioning ileocecal valve, production of secretory immunoglobulins on the surface of the gastrointestinal mucous membrane, and the bacteriostatic properties of pancreatic juice and bile. In our case, the patient’s history of cholecystectomy and hysterectomy were possible causes or exacerbating factors of SIBO. Disturbed gastrointestinal motility caused by paraplegia below the L4 level due to SCI is another possible cause of SIBO in our case, as well as decreased physical activity due to fracture of the left humerus and bilateral tibia.

Chronic gastrointestinal involvement is seen in 27–62% of patients with SCI [1, 13, 14]. SCI patients lack central nervous system control over the gastrointestinal system [15]. Liu et al. report that bowel problems in SCI patients are related to high levels of cord injury, completeness of cord injury, and post-injury durations of 10 years or more [2]. Moderate or severe grade depressive status is also associated with neurologic bowel dysfunction in SCI patients. Of these risk factors, our patient had complete cord injury that had occurred more than 10 years ago. Also, many bowel symptoms appear in patients with SCI (e.g., constipation, distension, incontinence, abdominal pain, bowel accidents, nausea, diarrhea, straining, rectal bleeding, hemorrhoids, autonomic hyperreflexia, headaches or sweat relieved by a bowel movement [2, 14, 16,17,18,19,20,21]). However, our patient showed no appetite loss, a sufficient amount of food intake, and non-severe bowel symptoms. Thus, the presence of malnutrition despite adequate food intake and low levels of lipid-soluble vitamins that were unresponsive to supplementation led us to suspect SIBO.

Although gastrointestinal symptoms are frequently observed in patients with SCI, there are few reports of SIBO in SCI patients. Cheng et al. report that 39% (145 of 377) of SCI patients were diagnosed with SIBO based on the glucose hydrogen/methane breath test [22]. However, the prevalence of SIBO among SCI patients as confirmed by the consensus diagnostic criteria of the presence of more than 105 CFU/mL bacteria or any amount of E. coli in upper digestive tract content is unknown. In patients with SCI, absent central nervous system innervation of the digestive tract can change the inhabiting environment of bacteria. Gungor et al. report differences in gut microbial patterns between SCI patients and control individuals as measured by bacterial genome sequencing [15]. Specifically, they found that butyrate-producing bacteria were specifically reduced in SCI patients. Thus, it is possible that SIBO is overlooked in patients with SCI. In our case, it is unclear when SIBO occurred relative to the time of SCI, but we suspect that it arose due to gastrointestinal motility disorder caused by autonomic disturbances.

Disturbances in fat absorption and deficiency in fat-soluble vitamins (i.e., vitamins A, K, E, and D3) are observed in patients with SIBO [10]. Excess bacteria in the small intestine promotes a change from conjugated bile acid into deconjugated bile acid, which decreases the micellar solubilization of dietary fat. Bacterial fermented short chain fatty acid causes osmotic water movement to the intestinal lumen, which results in diarrhea and malabsorption [10]. Intestinal epithelial damage in SIBO also interferes with fat absorption. Mucosal damage is caused by metabolites of aerobic bacteria, endotoxins of anaerobic bacteria, and lithocholic acid, which is a bacterial degradation product of unconjugated bile acid [23,24,25]. Our patient, however, showed constipation rather than diarrhea in spite of SIBO. Whether or not diarrhea occurs in patients with SIBO is determined by multiple factors. Constipation frequently occurs in patients with SCI due to decreased physical activity and autonomic dysfunction. De Looze et al. reported that the rate of constipation in the patients with SCI is 58% [18]. A certain proportion of the patients with SCI show constipation in spite of the coexisting SIBO. Cheng et al. reported that in patients with both SCI and SIBO, 22% showed constipation [22]. We believe that the, factors leading to constipation in our patient were stronger than those leading to diarrhea. Vitamin D deficiency in SIBO causes osteomalacia. Our patient showed multiple fractures and osteoporosis with serum vitamin D3 levels below the detection limit and refractory to supplementation.

There is no consensus on the choice, dose, or duration of antibiotics for treating SIBO [6]. In principle, antibiotics should be chosen based on the results of an antimicrobial susceptibility test, but this approach cannot address the great diversity in microbiota of the digestive tract [26, 27]. Metronidazole is a first-line choice for SIBO [28], with other choices being rifaximin, ciprofloxacin, norfloxacin, amoxicillin/clavulanate, trimethoprim, sulfamethoxazole, cephalexin, or their combination [6]. However, these antibodies are selected based on custom rather than scientific evidence [27]. In our case, we used oral polymyxin B and amphotericin B in accordance with SDD, which was first reported as a method of preventing ventilation-associated pneumonia and microbial translocation of gram-negative rod bacteria and fungi in critically ill patients treated in the intensive care unit [29,30,31]. Polymyxin B administered to the digestive tract is non-absorbent into the human body and has strong bactericidal power against gram-negative rod bacteria except for naturally polymyxin-resistant bacteria such as Proteus, Providencia, Morganella, Burkholderia, and Serratia [32]. Amphoteric B is an antifungal drug that is also non-absorbent into the human body when administered to the digestive tract. In our case, after starting SDD, fat-soluble vitamins were increased, and osteoporosis was improved. No obvious adverse effects of SDD, such as antibiotic-associated diarrhea, were observed in our case.

When and how to stop antibiotherapy for the treatment of patients with SIBO are difficult problems. Few reports are available for the method and timing for making a decision to stop antibiotherapy in SIBO. Lauritano et al. reported that the recurrence rate at 9 months after stopping antibiotherapy in SIBO patients is 44% [33]. They also showed that an older age, history of appendectomy, and chronic use of PPIs are associated with SIBO recurrence. Bures et al. reported that cyclical gastrointestinal selective antibiotics are needed for SIBO treatment [4]. These reports indicate that in many patients with SIBO, it is actually impossible to stop antibiotherapy because of the underlying conditions that lead to SIBO. Similarly, in our case, it was difficult to ameliorate the underlying condition of decreased motility of the digestive tract due to SCI. We were compelled to continue SDD for a long duration. We did, however, succeed in gradually reducing the dose of polymyxin B and end the use of amphotericin B without signs of SIBO recurrence. With careful consideration, it may be possible and feasible to stop SDD completely.

Probiotics are also a treatment approach for SIBO, as some species of bacteria are thought to protect against high numbers of E. coli and fungi in the digestive tract [34]. However, the role and effects of probiotics are still unclear. The digestive tract microbiome has both pathogenic potential and a protective role in maintaining health. However, metagenomic analysis reveals that 20–60% of microorganisms in the digestive tract cannot be cultured under laboratory conditions [10]. The effects of SDD and probiotics on the digestive tract microbiome in patients with SIBO should be investigated to further understand the pathogenesis of the disease.

In conclusion, we treated a patient with a sacral pressure sore who also had SCI, multiple fractures with osteoporosis, and malabsorption, especially of fat-soluble vitamins. Based on culture of upper digestive tract content, we diagnosed the patient with SIBO and started SDD using polymyxin B and amphotericin B, which effectively ameliorated the absorbency disturbance and allowed healing of the pressure sore. In light of several common risk factors between pressure sores and SIBO, such as decreased physical activity, our case provides additional information on the associations among pressure sores, malnutrition, and SIBO.