Mouth in Foot Disease

Abstract

Toothpicks are commonly used household items that rarely cause serious injury or infection. Toothpick-related injuries often occur due to ingestion with subsequent trauma/infection at distal sites within the gastrointestinal tract; however, cardiovascular, pleural, and soft tissue infections have been reported. Eikenella corrodens is a gram-negative, facultative anaerobic bacillus found in oral flora associated with bite wound infections. A few case reports describe E. corrodens osteomyelitis from toothpick puncture wounds. We report a case of foot cellulitis and abscess in an elderly diabetic after toothpick puncture injury that was unresponsive to empiric antibiotics. Wound cultures grew E. corrodens and rare Peptostreptococcus species. E. corrodens is resistant to first-generation cephalosporins, macrolides, aminoglycosides, clindamycin, and metronidazole. This case highlights the insidious nature of E. corrodens infections and the need to tailor empiric antibiotics for skin and soft tissue infections based on the mechanism of injury. In addition, this case stresses the importance of protective footwear in diabetics and serves as a cautionary tale regarding the use of seemingly innocuous toothpicks.

Introduction

Toothpicks are frequently used household items that can be a potential source of trauma leading to serious infection. Although injuries from toothpicks are relatively common, serious injuries are thought to be rare. Often, toothpick-related injuries result from ingestion leading to subsequent trauma/infection at distal sites within the gastrointestinal tract; however, cardiovascular, pleural, and soft tissue infections have been reported in rare instances in the medical literature. We report a case of Eikenella corrodens foot cellulitis and abscess in an elderly diabetic patient after a toothpick puncture injury.

Background

Between 1979 to 1982, an estimated 8,176 toothpick-related injuries occurred annually.^1,2 According to the US Consumer Product Safety Commission's National Electronic Injury Surveillance System (NEISS), which tracks emergency room visits related to consumer products, the annual incidence of toothpick-related injuries remains relatively unchanged since the early 1980s.² The estimated number of persons hospitalized annually due to toothpick-related injuries ranged from 105 to 184, with the highest incidence among the pediatric population (less than 15 years of age). Additionally, the most common site of injury involved the extremities and/or trunk within all age groups except for those less than 5 years of age.¹

Few case reports describe extra-oral E. corrodens infections, such as skin and soft tissue infections/abscesses, intra-abdominal abscesses, and osteomyelitis.³ It is most commonly recovered as part of a polymicrobial infection involving human bite wounds and less commonly as part of the HACEK organisms causing endocarditis.^4,5 Even fewer report E. corrodens infections from toothpick-related injuries.^5,6 Eikenella corrodens is a gram-negative, facultative anaerobic bacillus found within the normal flora of the human mouth, upper respiratory tract, gastrointestinal tract, and genitourinary tracts. E. corrodens grows slowly on blood and chocolate agar. The colonies are small, emit a bleach-like odor, and yield a characteristic pitting of the agar for which the organism was given its name.⁴

Bite wounds are typically treated with broad spectrum beta-lactam antibiotics (eg, ampicillin-sulbactam, piperacillin-tazobactam, meropenem, etc.) and coverage for E. corrodens should be considered in injuries involving contamination with oral flora. E. corrodens is resistant to first-generation cephalosporins, macrolides, aminoglycosides, and antimicrobials traditionally used against anaerobes, such as clindamycin and metronidazole.⁷ For penicillin-allergic patients, clindamycin plus a fluoroquinolone is often substituted as empiric therapy in bite wounds. In penicillin-allergic patients, therapy for E. corrodens proves more difficult. One case report showed successful treatment of penicillin-allergic patients with doxycycline and in vitro data report susceptibility to newer fluoroquinolones.^7,8

Case Report

A 79-year-old Japanese-American woman with a history of type 2 diabetes mellitus, hypertension, hyperlipidemia, and left eye prosthesis was evaluated in the Acute Care Clinic for left foot pain. She endorsed walking in her home barefoot and sustaining a toothpick impalement injury to the bottom of her left foot approximately 1 week prior. The toothpick was removed intact. The patient was uncertain if the toothpick had been used as she denied personal use however, her vision-impaired husband used toothpicks frequently. The patient first noticed warmth, erythema, and tenderness at the site of injury several days following the incident. She attempted home management with bacitracin ointment and over-the-counter analgesics, but sought medical attention once the pain affected her ability to ambulate. At her initial appointment, the patient denied symptoms of systemic infection or drainage from the injury site. The patient had no prior history of peripheral vascular disease or peripheral neuropathy and her most recent hemoglobin A1C was 7%. Her outpatient medication regimen included metformin, atorvastatin, lisinopril, and alendronate. On physical examination the patient was afebrile with normal vital signs. Her left mid-foot revealed a small puncture wound to the medial plantar aspect with circumferential erythema and warmth. The affected area was tender to palpation however, no fluctuance was appreciated. Bedside superficial wound exploration did not identify a retained foreign body. A small amount of exudate was expressed after incision but no cultures were obtained. The patient was treated with clindamycin to cover potential methicillin-resistant Staphylococcus aureus and a tetanus booster was given.

Two days following her initial appointment, the wound looked improved with decreased erythema and edema but a wound check three days later revealed a new area of fluctuance with increased edema, erythema, and tenderness. The patient remained well-appearing and afebrile. Laboratory analysis showed no leukocytosis and normal inflammatory markers. Plain films of the affected foot did not show retained foreign material or bony involvement.

Approximately 2 weeks after the initial injury, the patient reported increasing pain and difficulty ambulating. Bedside sonography revealed a small fluid collection to the plantar aspect of her left foot. She was admitted to the hospital to start empiric intravenous antibiotics with vancomycin and piperacillin/tazobactam. Surgical debridement showed no involvement of bone or tendon. Inflammatory markers and white blood cell counts remained normal and blood cultures were negative. She was transitioned to amoxicillin/clavulanate for empiric coverage of skin and oral flora given her recent history of toothpick injury. Six days after hospital admission, wound cultures grew Eikenella corrodens and rare Peptostreptococcus species. The patient experienced a full recovery from the injury/infection.

On a subsequent follow up home visit, it was noted the patient observed the custom of removing footwear before entering the home. Shag carpeting was present in all rooms except the kitchen and bathroom. Several round, pointy toothpicks were spotted within the carpet of the living room and on the kitchen floor. The couple lived alone and the patient was the primary caretaker for her ailing husband. She stated that her husband frequently chewed on toothpicks as a means to curb his desire to smoke. The patient was counseled on the importance of wearing protective footwear even while in the house to prevent foot trauma. It was recommended the household switch from round to flat toothpicks if possible, given the latter has more blunted ends that are less likely to cause serious injury. In addition to education on diabetic foot care and injury prevention, homecare and cleaning services were requested for the elderly couple.

Discussion

Toothpicks are found in 96% of American homes.⁹ A major toothpick manufacturer in the United States estimates annual sales at $60 million, producing between 4–20 billion toothpicks a year.¹⁰ Toothpicks are ubiquitous in today's society as their use expands far beyond the initial intention of dental cleaning. They are often seen garnishing fancy cocktails, assisting with finger-food consumption, and securing food items. Toothpicks used by bartenders and in the food industry should be clearly labeled to help prevent accidental ingestion by patrons. The common utilization of toothpicks precludes them from being seen as potentially injurious household items. Budnick, et al's, article highlights the danger of toothpick use in specific populations including the intellectually disabled, denture-wearers, alcohol consumers, and those with dulled palatal sensoria due to increased risk of toothpick ingestion.¹ Cautious use of toothpicks should be expanded to elderly persons, the visually impaired, and diabetics as evidenced in this case. In the United States, data on the incidence of toothpick-related injuries/deaths, such as those reported by the NEISS, are likely underestimated as they only track patients cared for in an emergency room setting.

Susceptibility testing is not routinely performed in standard clinical microbiology laboratories for anaerobic organisms and treatment is often based on those reported in the literature. This patient's isolate was consistent with susceptibilities previously reported.¹¹ We reviewed microbiological data at Tripler Army Medical Center (TAMC) from 2009 to 2014 and found Eikenella species were isolated from 56 patients (data not published). These isolates were often polymicrobial in nature and took several days for adequate growth (2–16 days). The vast majority of TAMC's Eikenella species were isolated from extra-oral sites of infection.

E. corrodens infections are insidious in nature with a predilection for deep tissue infection. It may take more than one week from the time of injury to manifest clinical disease with few systemic symptoms or signs of inflammation on laboratory testing. This case underscores the importance of tailoring empiric antibiotics for skin and soft tissue infections based on the mechanism of injury in combination with surgical debridement. Cultures should be performed to aid in diagnosis and implementation of appropriate antibiotic therapy, especially in those failing to respond. If oral flora is suspected as a causative agent, the laboratory should be notified to optimize isolation of key organisms.

Conflict of Interest

None of the authors identify a conflict of interest.