Treatment of Fusarium Osteomyelitis in a Diabetic Foot Ulcer Complicated by Antineoplastic Chemotherapy

In the United States, the prevalence of diabetes increases annually.¹ With this increasing prevalence of diabetes, a rise in the number of diabetic complications is also seen. Diabetic foot infections (DFIs) are common and represent the leading cause of hospitalization among diabetic complications.² It is estimated that 19% to 34% of patients with diabetes are likely to have DFIs. Infection is affected by hyperglycemia and decreased vascular supply, and infection may spread into the underlying bone.³

Diabetic foot osteomyelitis (DFO) is defined as an infection of the bone and joint structures of the foot. It has been reported that DFO develops in 20% to 60% of patients with DFIs. Diabetic foot osteomyelitis is difficult to treat; it is associated with the risk of relapsing episodes, hospitalizations, and foot amputations.² Nearly all patients with diabetes and pedal osteomyelitis have contiguous chronic infections, meaning the bone becomes infected by an adjacent soft tissue infection. The most common sites of pedal osteomyelitis are the fifth metatarsal, first metatarsal, calcaneus, and great toe distal phalanx.⁴

In the presence of clinical signs or symptoms of infection, it is recommended that a deep specimen of tissue or bone should be obtained for pathology to confirm the diagnosis of osteomyelitis as well as for culture (aerobic, anaerobic, fungal, and acid-fast bacteria [AFB]) to guide antimicrobial therapy.⁵ Cultures of bone specimens provide more accurate microbiologic data than those of soft tissue specimens from patients with DFO. The main advantage of a bone biopsy is to provide reliable data on the organism responsible for the infection and determine its susceptibility profile to antimicrobial agents.² The microbiology of DFO is usually polymicrobial, and Staphylococcus aureus is the most common pathogen cultured from bone samples. Other gram-positive cocci frequently isolated from bone samples are represented by Staphylococcus epidermidis and other coagulase-negative staphylococci, beta-hemolytic streptococcus, and diphtheroids. Gram-negative bacilli (especially Pseudomonas aeruginosa and Escherichia coli) are frequent in some warm climates. The role of other infectious agents such as fungi or dermatophytes in the setting of DFO appears to be less consistent.²

The main objective of the treatment of DFO is to eliminate the infection, reduce the number of amputations, and maintain functional foot biomechanics. Surgical resection of the infected bone(s) can play an important part in the treatment as it effectively and rapidly reduces the infectious organism load at the infected site and removes necrotic tissue that cannot be managed with antimicrobial agents alone. Medical management is characterized by using antimicrobial agents directed against microbes that have been identified by deep tissue or bone biopsy without any bone resection or as an adjunct to surgical management. The main advantages of medical management are the absence of biomechanical changes that may occur after surgical intervention, decreased morbidity and mortality associated with amputation, and a better cost-effectiveness profile.⁶ Conservative surgery (CS) or bone debridement for DFO consists of the minimal resection of infected bones while avoiding amputation. Conservative surgery has proven to be effective in the setting of DFO with a high rate of limb salvage and could also reduce the period of antimicrobial therapy.⁶ Treatment of DFO can be complicated, even more so when a patient has comorbidities such as cancer.

Multiple modalities are involved in the treatment of patients with cancer, all of which affect the ability to treat wounds. Chemotherapeutic agent regimens are frequently used as monotherapy or in conjunction with surgery and radiation in cancer treatment. These pharmacologic agents target proliferating cells. Although chemotherapeutic agents preferentially target rapidly dividing cells, any tissue can be affected by these treatments; macrophages and fibroblasts involved in wound healing areas are susceptible to these effects as cancer cells.⁷ In addition to the effects of chemotherapeutic agents on a cellular level, the nutritional needs of patients must be considered. An estimated 40% to 80% of patients with cancer are clinically malnourished; these patients are increasingly susceptible to infectious complications and delayed wound healing.⁷

A case of a 77-year-old female with a past medical history of type 2 diabetes with peripheral neuropathy, hyperlipidemia, and relapsed acute myeloid leukemia (AML) currently on FLAG (fludarabine, arabinofuranosyl cytidine, granulocyte colony-stimulating factor) chemotherapy is reported. The type 2 diabetes was well controlled, with a most recent hemoglobin A1c of 6.3%. Metformin and lifestyle modifications were used to manage the diabetes. The patient was preestablished to the podiatry service at the authors’ hospital, as the patient also had a past surgical history of right foot bunionectomy with retained hardware and hammertoe correction of the right second toe.

The patient was admitted to Michigan Medicine, Ann Arbor, Michigan, for antineoplastic chemotherapy for AML. This was the third course of chemotherapy for AML. Podiatry was consulted during this hospitalization for a chronic diabetic foot ulcer (DFU) of the right second toe at the distal tuft of the toe and a new DFU on the plantar aspect of the right hallux (Figure 1). The DFU of the right second toe was present for more than 4 months and had been complicated by cellulitis approximately 1 month earlier, requiring hospitalization. During that hospitalization, the patient received intravenous (IV) vancomycin for 3 days and was discharged on a regimen of oral sulfamethoxazole-trimethoprim twice daily for 7 days. The cellulitis resolved on this antimicrobial regimen, but the DFU remained open. During the current hospitalization, there was a return of cellulitis associated with the right second toe DFU. The patient also had a new, superficial ulcer on the right hallux, which correlated with the placement of a crest pad to take pressure off the second toe DFU. No signs of acute infection were associated with the right hallux ulcer, and it did not probe to the bone.

At presentation, the patient was febrile (39.2°C) and neutropenic, with blood cultures positive for P aeruginosa. The DFU of the right second toe was full thickness and probed to the distal phalanx bone, which was also visible in the wound bed. The DFU measured 1.4 cm × 1.0 cm × 1.5 cm. There was erythema streaking greater than 2 cm from the DFU, approximately to the second metatarsal phalangeal joint level. There was nonpitting edema to the right second toe, creating a sausage digit appearance. There was moderate serosanguinous drainage from the DFU, but no purulent drainage was appreciated. The wound base was a mix of granular and fibrotic tissues, with no necrosis to the soft tissue. Pedal pulses were palpable, and all toes had a brisk capillary refill time bilaterally. A previous ankle-brachial index and toe-brachial index did not suggest peripheral vascular occlusive disease to the lower extremities.

Erythrocyte sedimentation rate (ESR) was elevated to 46 mm per hour, and the C-reactive protein (CRP) level was elevated to 29.2 mg/L. The white blood cell count (WBC) and platelet count were extremely below normal (<0.1 K/uL and 8 K/uL, respectively) and were attributed to the current FLAG chemotherapy treatment. Multiple platelet transfusions were administered throughout the hospitalization; however, the platelet count remained below normal and was attributed to active chemotherapy treatment. A right foot plain radiograph demonstrated cortical irregularity and lucency of the distal phalanx of the second toe (Figure 2). Magnetic resonance imaging (MRI) of the right foot revealed changes consistent with osteomyelitis to the second toe distal phalanx only and no appreciated abscess.

The patient’s medications included oral amoxicillin-clavulanate 875/125 mg twice daily, voriconazole 200 mg twice daily, acyclovir 400 mg once daily, and sulfamethoxazole-trimethoprim 800/160 mg 3 times per week for infection prevention in conjunction with chemotherapy. Due to the P aeruginosa bacteremia from day 1 following inpatient admission, the patient was started on IV cefepime and vancomycin, and the amoxicillin-clavulanate and sulfamethoxazole-trimethoprim were held.

Since worsening osteomyelitis and cellulitis associated with the second toe DFU developed in this patient, the options of partial phalangectomy with bone cultures versus partial toe amputation were presented to the patient. Also discussed was the thrombocytopenia with the possibility of excessive bleeding during the procedure and the postoperative period. Given the risks and benefits of the 2 different procedure options, the patient elected for a bedside wound debridement with partial distal phalangectomy. Antimicrobial therapy was not held for the procedure. A #15 blade, sterile rongeur, and sterile curette were used for the debridement. The bone was sent for anatomic pathology and aerobic, anaerobic, fungal, and AFB cultures. As expected, there was more bleeding than anticipated with the procedure due to the thrombocytopenia. The site was packed with an oxidized cellulose polymer for hemostasis. The postprocedure compression dressing was left intact, and the patient was non-weightbearing, with the foot elevated for 48 hours. Postprocedural plain radiographs were obtained (Figure 3). These revealed greater than 50% removal of the distal phalanx bone. Figure 4 shows clinical photos from status postprocedure day 2. The edema and erythema to the second toe greatly improved after the partial phalangectomy.

The bone culture resulted in coagulase-negative staphylococci and Fusarium species. An infectious disease specialist recommended 14 days of IV cefepime with the transition to oral doxycycline 100 mg twice daily for an additional 6 weeks. The patient also remained on oral voriconazole 200 mg twice daily until the DFU was healed.

The patient was followed up in the outpatient setting with the hospital’s podiatry team every 1 to 2 weeks for additional wound care. The DFU was offloaded with a wound shoe with heel weightbearing to the right foot. At 6 weeks after the debridement with partial distal phalangectomy, the DFU was healed (Figure 5). The patient transitioned to depth inlay shoes with custom, multidensity, accommodative orthotics with offloading to the previous ulceration site. At the patient’s last appointment, the DFU remained healed with no further signs of preulceration skin breakdown for 6 months.

Treating DFO can be especially difficult when patients are severely immunocompromised. Patients with AML may present with leukocytosis, anemia, and thrombocytopenia.⁸ Patients undergoing chemotherapy for AML will have further barriers in the diagnosis and treatment of DFO. On presentation, the current patient had a clinically suspected bone infection. The physical examination showed erythema, edema, drainage, and a positive probe-to-bone test. The ability to probe to bone is associated with high specificity for osteomyelitis, exceeding 90%.³ The inflammatory biomarkers, ESR, and CRP, are routinely ordered for evaluating DFIs. An ESR of 60 mm per hour and a CRP level of 7.9 mg/dL are the evidence-based cutoff points for predicting osteomyelitis.⁹ Although the patient had an elevated ESR, it was subthreshold for DFO. The CRP level was elevated above the cutoff point, and a diagnosis of osteomyelitis with an elevated CRP level has a sensitivity of 49% and specificity of 80%.⁹

Regarding imaging workup of suspected osteomyelitis, it is recommended to begin with plain film radiographs (anteroposterior, oblique, and lateral). The patient’s radiograph displayed cortical irregularity and lucency of the distal phalanx. It typically takes 3 weeks for a cortical erosion to develop and it may not be evident until the late stages of the disease.⁴ Due to the time it may take to see cortical erosions on radiographs, and for surgical planning, the decision was made to obtain an MRI to evaluate the extent of the infection. Magnetic resonance imaging is the most accurate imaging test for the assessment of suspected osteomyelitis, with a sensitivity of 90% and specificity of 79%.⁴ Short T1 inversion recovery (STIR) images are the most sensitive sequence type to evaluate for osteomyelitis.⁴ In the current case, MRI revealed positive changes consistent with osteomyelitis to the distal phalanx without signs of abscess.

The histopathology of the current case showed a bone culture growth of coagulase-negative staphylococci and Fusarium species. Coagulase-negative staphylococci are ubiquitous microbes that recently have become more commonly associated with nosocomial infections. The human integument is the natural reservoir for Staphylococcus epidermidis, the most common coagulase-negative staphylococci. It has the greatest pathogenic potential of all species of coagulase-negative staphylococci and can be challenging to manage.¹⁰Fusarium species are filamentous fungi present as saprophytes in soil and animals.¹¹ The widespread distribution of Fusarium species is attributed to their ability to grow on a wide range of substrates and their efficient mechanisms for dispersal.¹²Fusarium was first reported as a systemic pathogen in 1973 and subsequently has been noted to cause an increasing number of infections, particularly in patients who are severely immunocompromised.¹³Fusarium species have been reported to cause cutaneous infections in patients with diabetes. They can develop after skin breakdown at the infection site.¹¹ Invasive fusariosis has been shown to occur in patients receiving high doses of corticosteroids and those with prolonged and profound neutropenia.¹² It has been shown to respond poorly to medical management by antifungal medication, with only 43% of patients with fusariosis having a complete or partial response.¹¹ Therefore, the treatment of Fusarium osteomyelitis usually requires the use of both antifungal therapy and surgical debridement, which occurred in the current case.¹³

Case reports of Fusarium osteomyelitis in patients who are immunocompromised are extremely rare.¹³ One case report noted that there were 10 cases of Fusarium osteomyelitis reported in the literature from 1972 to 2012.¹⁴ Of the 10 case reports, 7 of these infections occurred in individuals who were immunosuppressed. Of the immunocompromised group, in each of the 4 treated with wide surgical debridement or amputation and antifungal therapy, the infection was successfully eradicated. Conversely, none of the 3 patients treated with simple debridement and antifungal therapy had eradication of infection. No case reports of Fusarium osteomyelitis in an immunocompromised individual have demonstrated definitive long-term treatment with simple debridement and antifungal therapy alone.¹⁴

Due to the rare nature of fungal osteomyelitis, the case study is limited by the small sample size of 1 patient. Unfortunately, the patient died after the 6-month follow-up appointment; as such, the study is also limited by the lack of longer term follow-up.

Fusarium is a rare cause of osteomyelitis. When it presents in a patient who is severely immunocompromised, it can be challenging to determine diagnosis and treatment. The current patient received a diagnosis of DFO while undergoing active antineoplastic chemotherapy treatment. The side effects of the chemotherapy and chronicity of the wound with no response to antimicrobial therapy guided the authors’ treatment decision. The patient was treated with a conservative procedure with a bedside partial phalangectomy and antimicrobial medications postprocedure. The patient responded well to the treatment, and the ulceration ultimately healed. This case report may serve as the first to demonstrate successful remission of Fusarium osteomyelitis in a patient of immunocompromised status with a conservative procedure and adjunct antifungal therapy.