|Home | About | Journals | Submit | Contact Us | Français|
Background.Major advances in combat casualty care have led to increased survival of patients with complex extremity trauma. Invasive fungal wound infections (IFIs) are an uncommon, but increasingly recognized, complication following trauma that require greater understanding of risk factors and clinical findings to reduce morbidity.
Methods. The patient population includes US military personnel injured during combat from June 2009 through December 2010. Case definition required wound necrosis on successive debridements with IFI evidence by histopathology and/or microbiology (Candida spp excluded). Case finding and data collected through the Trauma Infectious Disease Outcomes Study utilized trauma registry, hospital records or operative reports, and pathologist review of histopathology specimens.
Results.A total of 37 cases were identified: proven (angioinvasion, n = 20), probable (nonvascular tissue invasion, n = 4), and possible (positive fungal culture without histopathological evidence, n = 13). In the last quarter surveyed, rates reached 3.5% of trauma admissions. Common findings include blast injury (100%) during foot patrol (92%) occurring in southern Afghanistan (94%) with lower extremity amputation (80%) and large volume blood transfusion (97.2%). Mold isolates were recovered in 83% of cases (order Mucorales, n = 16; Aspergillus spp, n = 16; Fusarium spp, n = 9), commonly with multiple mold species among infected wounds (28%). Clinical outcomes included 3 related deaths (8.1%), frequent debridements (median, 11 cases), and amputation revisions (58%).
Conclusions.IFIs are an emerging trauma-related infection leading to significant morbidity. Early identification, using common characteristics of patient injury profile and tissue-based diagnosis, should be accompanied by aggressive surgical and antifungal therapy (liposomal amphotericin B and a broad-spectrum triazole pending mycology results) among patients with suspicious wounds.
Invasive fungal wound infections (IFIs) of skin and soft tissue occur in both immunocompromised and immunocompetent hosts . Among immunocompetent individuals, traumatic injury is the most common risk factor for IFI, occurring in settings such as agricultural, motor vehicle, and blunt crush injuries, as well as during natural disasters as recently reported among tornado victims [2–6]. Local invasive infection often requires repeated and extensive surgical debridement and systemic antifungal therapy complicated by amputations (31%) and mortality (as high as 25%) .
Major advances in combat casualty care have led to increased survival of patients with complex extremity trauma [7–9]. Blast traumas, secondary to those caused by improvised explosive devices, are among the most severe injuries [7, 10]. Gross contamination of these wounds with organic material may introduce environmental filamentous fungi (molds), among other pathogens [2, 4]. This report describes IFI epidemiology, clinical findings, and laboratory findings among American combat casualties in Afghanistan from June 2009 through December 2010.
The population includes US military personnel who were injured in combat in Afghanistan, then medically evacuated through Landstuhl Regional Medical Center (LRMC) in Germany and admitted to 1 of 3 US military hospitals: Walter Reed Army Medical Center (WRAMC) in Washington, DC; National Naval Medical Center (NNMC) in Bethesda, Maryland; and Brooke Army Medical Center in San Antonio, Texas.
Case definitions were adapted from the Mycosis Study Group . Diagnostic criteria were as follows: (1) traumatic wounds, (2) after ≥1 irrigation and debridement, (3) tissue necrosis on ≥2 consecutive debridements, and (4) IFI evidence. Cases were classified as either proven IFI (fungal hyphae angioinvasion of viable tissue), probable IFI (viable tissue invasion, but no angioinvasion), or possible IFI (mold grown in culture but either histology not performed or no fungal elements seen). Candida spp were specifically excluded because their role in pathogenesis in wound infections is unclear.
The investigation was undertaken through the Trauma Infectious Disease Outcomes Study (TIDOS) . Trauma history, International Classification of Diseases (Ninth Revision) injury codes, and surgical management history were obtained through the US Department of Defense Joint Theater Trauma Registry (JTTR) . IFI-specific data were collected through a supplemental JTTR infectious disease module. All histopathology specimens were reviewed by 2 surgical pathologists. The study was approved by the institutional review board of the Uniformed Services University of the Health Sciences.
Cases were identified through review of the TIDOS database for all positive fungal wound cultures and histology during the investigation period. To assure complete case capture, histopathology and clinical microbiology reports were reviewed, as were case records from infectious disease and trauma surgery services.
IFI diagnosis date was the earliest specimen collection yielding a positive result from either culture or pathology. Time to clean wound was defined as the number of days from IFI diagnosis to first documented clean wound (operative report with no evidence of infection and no subsequent histopathology or culture evidence). High-level amputations were defined as hip disarticulation or hemipelvectomy.
Fisher exact and χ2 tests were used to test the association of levels of categorical variables with IFI classes and clinical mycology groups. The Kruskal-Wallis test was implemented to compare medians among these groupings.
The investigation included the period from 1 June 2009 (TIDOS start date) through 31 December 2010. Trauma-patient evacuations to LRMC from Afghanistan (n = 2413) are displayed in Figure 1 by calendar quarter against patients diagnosed with IFI (n = 37; proven, 20 patients [54.1%]; probable, 4 patients [10.8%]; possible, 13 patients [35.1%]). No IFI cases were identified from Iraq during the study period. The IFI rate gradually increased to the highest level of 3.5% in the fourth quarter of 2010. The progressive increase was also observed among trauma patient admissions (not restricted to Afghanistan evacuations) to the LRMC intensive care unit (ICU), reaching a high of 8% in the fourth quarter of 2010.
All cases were male enlisted personnel (25 Marines and 12 Army) with median age 23 years (interquartile range [IQR], 21–27). All injuries were secondary to explosive blasts, 34 (91.9%) were dismounted (ie, on foot patrol) at the time of injury (more common in proven and probable groups), and 35 (94.6%) received initial care in a military facility in southern Afghanistan (Helmand or Kandahar provinces; Table 1). All had extremity injuries with 29 (78.4%) sustaining an amputation at time of injury or first surgery as exemplified in one patient's infected wound (Figure 2). Multiple limb amputations were common with 25 (67.6%) bilateral and 6 (16.2%) involving three limbs. Other common findings included perineal and genital wounds (67.6%), vascular repair (51.4%), and penetrating abdominal trauma (32.4%).
Characteristics of patients’ clinical findings and early trauma care are detailed in Table 1 by IFI classification. Overall, most patients presented with borderline hypotension and tachycardia, with significant base deficit and required massive blood transfusion, approximately 30 units on average of both packed red blood cells and fresh frozen plasma, in the first 24 hours (proven and probable cases had higher transfusion requirement than did possible cases). The majority of patients were febrile (temperature, >38°C) with leukocytosis (white blood cell count, >10 × 109 cells/L) within 1 week of IFI diagnosis irrespective of IFI classification.
Among the 37 patients, a total of 54 distinct wounds with IFI were identified, 27 (50.0%) classified as proven, 8 (14.8%) as probable, and 19 as (35.2%) possible (12 of which did not have specimens sent for histopathology).These wounds were characterized by myonecrosis (81%), liponecrosis (52%), eschar (14%), purulence (43%), and fibrinous exudates (18%) observed on surgical examination.
Among the wounds with pathological examination (n = 41), hyphae were seen, from any staining technique, in 34 (82.9%) wounds. Hyphal element staining was more commonly observed with Gomori methenamine silver staining (72.4%) than with periodic acid Schiff (PAS) staining (51.4%). Of critical importance in specimens with aseptate hyphae, visualization with routine hematoxylin-eosin (H&E) staining was occasionally the only method of identification (19.5%). Although an intra-operative frozen section was not commonly obtained, among the 15 wounds for which frozen sections were evaluated, there were 6 positive frozen sections (all also found positive for hyphae on standard histopathology) and 9 negative frozen sections (6 later found to be positive on definitive staining).
A total of 31 (83.8%) patients had at least 1 mold isolated (Table 3). In 16 patients, a mold from the order Mucorales was identified (multiple species isolated in some cases). Pathogens from the order Mucorales were isolated from 62.5%, 100%, and 38.5% of proven, probable, and possible cases, respectively. Evidence of order Mucorales, from culture and/or aseptate hyphae on histopathology, was observed in 25 (47.2%) infected wounds: 10 (40%) from culture and aseptate morphology, 6 (24%) from culture only, and 9 (36%) from pathology only. In addition, there were 16 patients with Aspergillus spp and 9 with Fusarium spp. Two or more mold isolates, with one from Mucorales and another from a non-Mucorales species, were recovered in 28.3% of infected wounds. There were no statistically significant clinical differences based on fungal pathogen isolation. No mold isolates were recovered from sites distant from the affected extremity or blood with the exception of abdominal cavity involvement contiguous with extensive perineal and hip invasive disease in 2 of the fatal cases as well as respiratory secretion cultures (commonly sputum and tracheal aspirates) which were not felt to represent invasive pulmonary disease.
All 54 wounds were also either colonized or infected with a bacteria or yeast by the date of IFI diagnosis, with the most commonly identified organisms being Enterococcus faecium (n = 11; 20.4%) and Acinetobacter calcoaceticus-baumannii complex (n = 9; 17%). Candida albicans grew in cultures from 5 wounds (9.3%).
The overall median duration from the time of injury to IFI diagnosis was 10 days (IQR, 7–14 days). The time from IFI diagnosis to discharge from the US facility was 57.5 days (IQR, 35–68 days). The median number of surgeries at the IFI wound site after medical evacuation from the operational theater was 11 surgeries (IQR, 7–16 surgeries) consisting of frequent debridements and amputation revisions. Time from IFI diagnosis to clean wounds, documented in 20 cases, was a median of 21 days (IQR, 18–35 days). Ten patients (27%) ultimately required a high-level lower extremity amputation.
Thirty-three patients (89.2%) received antifungal therapy, most commonly lipid formulations of amphotericin B (LFAB) in 91% of patients, voriconazole in 85% of patients, and posaconazole in 49% of patients. Eight (21.6%) patients received monotherapy, whereas the most common combination therapy (≥3 days of overlap) was LFAB plus a triazole (voriconazole or posaconazole) in 65% of patients. Overall, patients with possible cases received less total antifungal therapy than did patients with proven or probable cases as reflected by overall receipt (69% vs 100%), shorter treatment duration, and less common combination therapy (46% vs 79%). The 4 patients who did not receive antifungal therapy were all classified as having possible IFI, with 1 diagnosis of IFI on the day of death. The other 3 patients underwent frequent debridements (range, 8–11 debridements) leading to infection resolution. Several topical agents were used, including 0.025% sodium hypochlorite solution (modified Dakin solution) applied through a negative pressure therapy device in 12 (32.4%) patients, amphotericin B beads in 3 (8.1%) patients, and amphotericin peritoneal irrigation in 1 patient. The median time between the IFI diagnosis date (when specimen yielding first positive result was collected) and antifungal therapy initiation was 4 days (IQR, 1.5–7 days). The median duration of antifungal therapy was 26.5 days (IQR, 14.5–42.5 days).
There were no significant differences for duration of hospitalization, ICU stay, new amputations, or revised amputations by IFI classification (Table 2). Five patients with IFI died. IFI contributed to death in 3 (8.1%) of these cases. In each of these 3 cases, new-onset necrotic wounds presented 8–12 days after injury with widespread progression of necrosis despite intensified debridements over a period of 2–7 days. In each of these fatal cases the necrosis extended beyond affected extremities into truncal areas as well as through the abdominal wall with involvement of intra-abdominal tissues. All fatal cases were in the proven group except 1 possible case (no tissue specimen sent for histopathology). Three patients who died with positive cultures near time of death were all from the group with Mucorales infection. One patient tested positive for an Alternaria species at the infected wound site 1 week prior to patient death; however, wound culture was negative with documented angioinvasion the day prior to death (primarily attributed to severe head injury and not IFI). The fifth patient had aseptate angioinvasive fungal elements observed but no culture obtained (Table 3).
To our knowledge, this is the largest case series of trauma-related IFI described to date. Invasive fungal infections have emerged as an important cause of morbidity and mortality among US military personnel who have suffered combat-related traumatic injuries. The most common features among these patients are lower extremity amputation with perineal or pelvic injury and receipt of massive blood transfusions following blast injuries incurred while on foot patrol in southern Afghanistan. This profile is consistent with a previous British military report in the “green zone” of Helmand province in Afghanistan . This region, with large agricultural areas, may have a more dense concentration of environmental molds than more arid regions of Afghanistan or in Iraq; however, the primary risk may be more related to the mechanism of injury in an individual on foot patrol than to a regional endemic threat given the ubiquitous nature of these molds. The increasing number of IFI cases follows the upward trend in lower extremity amputations among injured personnel returning from Afghanistan .
Previous case reports of US combat-related IFI describe wound infections with rare mold species [15–17], one of which was a fatal case included in this investigation . Cases have also been reported from a military burn center noting higher numbers of attributable deaths related to invasive mold infection in combat-related compared with civilian burn patients at the center . In addition, there were 6 patients with IFI, all injured in Iraq, from March 2002 through July 2008 among US military personnel sustaining traumatic war wounds who wer hospitalized at WRAMC, yielding estimates of 0.4 cases per 1000 admissions, much lower than observed during this investigation .
Blast extremity wounds typically undergo meticulous low-pressure irrigation and debridement within 1–2 hours of injury and are commonly dressed using vacuum-assisted wound closure. Serial debridement occurs approximately every 48 hours with wound closure often first considered after injury day 8 if evidence of wound contraction and healthy noninfected-appearing granulation tissue. Techniques of wound closure are as follows, in order of preference: delayed primary closure, split thickness skin graft, rotational flap, and free flap. The average injury severity score (ISS) from one case series of mucormycosis after traumatic injury was 24 (range, 9–41), similar to the average ISS observed in this series . The high fatality rate exceeds observed mortality among wound infections secondary to bacterial infections , although causes of death in this series were multifactorial in nature and 2 deaths were not related to IFI diagnosis. IFI case fatality is lowest among trauma-associated etiologies that typically occur in previously immunocompetent individuals; however, mortality of 25% exceeded our observed rate despite the more extensive trauma in our series . A recent large series of cutaneous mucormycosis among persons injured in the tornado that struck Joplin, Missouri, in 2011 reported an even higher fatality rate of 38.5% . The 13 confirmed case patients in the Joplin series had a higher median age with a wide range (as high as 76 years) and diabetes in 2 cases but none were immunocompromised.
All but 1 patient in this series were transfused with ≥10 units of blood in theater (a commonly used threshold delineating massive transfusion requirement)  with an initial 24-hour requirement almost 3 times this threshold. The receipt of blood products can have immunosuppressive effects and thereby increase the risk for infection [22, 23]. An additional consideration pertaining to IFI is the iron burden that is received with blood transfusions, given that these molds are known to utilize iron as a nutrition source, which may be further exacerbated with acidosis during the immediate posttrauma period . Iron burden was not assessed in these patients.
In recent years, rates of IFI have increased, most commonly among immunocompromised patients with mucormycosis . Mucor spp, Saksenaea vasiformis, and Apophysomyces elegans were the most common species isolated from the order Mucorales in this case series. The invasive potential of S. vasiformis and A. elegans observed in this series is consistent with previous reports and a recent review [15, 25, 26]. Of note, there were no isolates of Rhizopus spp, the most common identified mucormycosis etiology; although this is primarily due to the predominance of immunocompromised hosts among cases . Other environmental molds such as Aspergillus spp, Fusarium spp, and Scedosporium spp have been reported to cause wound infections and were isolated in this series [2, 4, 27]. Only the Mucorales group and Aspergillus spp were observed in proven cases as solitary pathogens. It is likely that many of the diverse mold species isolated, particularly given the common finding of multiple isolates per wound, represent wound contamination.
A favorable prognosis is dependent upon early identification and aggressive surgical debridement combined with antifungal treatment [3, 4]. IFI diagnosis relies on histopathological confirmation of tissue invasion, with fungal culture providing species identification . Mold cultures often require several days to grow and may be insensitive (20% of wounds in this series had negative cultures) ; therefore, surgical tissue samples for histopathology are essential for accurate diagnosis particularly since fungal colonization alone is not indicative of an invasive infection [29, 30]. Standard H&E staining is reported to be insensitive, prompting recommendations for special staining ; however, in this series more wounds with hyphae were identified with H&E staining than with PAS staining, particularly aseptate hyphae. Real-time frozen sections demonstrated poor negative predictive value, limiting clinical utility, although the number of cases examined was small and no prospective evaluation was undertaken. Calcofluor white, a fluorescent stain used for rapid detection of fungal elements in the microbiology laboratory [31–33], was positive in only 49% of 33 mold culture positive specimens at one of the hospitals. The 4-day average “delay” between IFI diagnosis and start of antifungal therapy reflects the time required for tissue specimen collection, laboratory work-up, reporting results, and starting therapy. This delay was most evident among cases reliant upon microbiological findings to fulfill IFI criteria, highlighting the need for improved rapid and sensitive diagnostic methods, as well as use of empiric antifungal therapy pending laboratory results. Encouraging reports applying rapid diagnostic molecular methods warrant further investigation [34, 35].
Debridement of necrotic material is paramount, as the thrombosed blood vessels from the IFI results in poor tissue penetration of antifungal agents [1, 36]. Frequent debridement is necessary (potentially daily); one series had an average of 10 surgeries, comparable with the average number in this series . The majority (89%) of the patients in this series received antifungal therapy, with those managed with surgery alone classified as possible IFI. No well-controlled studies are available on which to base antifungal therapy in trauma-associated IFI. A lipid polyene, amphotericin B (LFAB), at a dose of 5 mg/kg per day, is the preferred agent for mucormycosis without controlled trial evidence for added benefit when combined with other antifungal agents . Voriconazole is not active against mucormycosis, but it is primary therapy for invasive aspergillosis including molds such as Aspergillus terreus which are resistant to amphotericin B . Empiric regimens should include LFAB and voriconazole pending results of serial wound cultures and histopathology findings to better target treatment on probable pathogens rather than transient colonizing species.
Topical therapy was commonly used, including specific antifungal agents or, more commonly, the disinfectant sodium hypochlorite (0.025% Dakin solution instilled in wound vacuum device) . Anecdotally, Dakin solution was felt to be effective, leading to common use in the latter half of the investigation period; however, no outcome difference was apparent. Potential adjunctive therapy options for invasive mucormycosis have been reported including hyperbaric oxygen , with encouraging results from case series but no controlled studies, and deferasirox iron chelation, recently reported to have higher mortality among patients receiving adjunctive treatment in a small randomized placebo-controlled trial .
In summary, IFI is an increasingly important cause of morbidity and mortality among US military personnel who suffer combat-related traumatic injury. Awareness of the common features (blast injury while on foot patrol, extensive lower extremity traumatic amputations with associated perineal injury, and massive blood transfusion) in the setting of necrotic wounds requires aggressive and extensive surgical debridement and empiric systemic antifungal therapy. In addition, clinicians should also recognize that the median time from injury to diagnosis is 10 days (IQR, 7–14 days). Increased awareness and early diagnostic efforts utilizing tissue histopathology and clinical mycology may shorten this interval, lessening the requirement for extensive debridement and ultimately minimizing long-term disability due to high-level amputations. Future strategies for IFI prevention, diagnosis, and treatment require preclinical and clinical research efforts to support ongoing refinement of clinical practice guidance.
Acknowledgments.We are indebted to the Infectious Disease Clinical Research Program (IDCRP) Trauma Infectious Diseases Outcomes Study team of clinical investigators, clinical coordinators, microbiology technicians, data managers, clinical site managers, and administrative support personnel to ensure the success of this project. We also wish to thank John Bennett, National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH), for his critical review of the manuscript.
The content of this publication is the sole responsibility of the authors and does not necessarily reflect the views or policies of the NIH or the Department of Health and Human Services, the Department of Defense (DoD), or the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the US Government.
This work was supported by the IDCRP, a DoD program executed through the Uniformed Services University of the Health Sciences. This project (IDCRP-024) has been funded by the National Institute of Allergy and Infectious Diseases, NIH, under Inter-Agency Agreement Y1-AI-5072 and the Department of the Navy under the Wounded, Ill, and Injured Program. D. R. T. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Potential conflicts of interest.All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.