- Email: [email protected]
Actinomyces meyeri, a Common Agent of Actinomycosis
Author's Accepted Manuscript
Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience Robert Rolfe MD, Lisa L. Steed PhD, Cassandra Salgado MD, MS, J. Michael Kilby MD
www.amjmedsci.com
PII: DOI: Reference:
S0002-9629(15)37988-X http://dx.doi.org/10.1016/j.amjms.2016.03.003 AMJMS128
To appear in:
Am J Med Sci
Cite this article as: Robert Rolfe MD, Lisa L. Steed PhD, Cassandra Salgado MD, MS, J. Michael Kilby MD, Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience, Am J Med Sci, http://dx.doi.org/10.1016/j.amjms.2016.03.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience
Robert Rolfe, MD The University of Alabama of Birmingham Internal Medicine Residency Training Program BDB 327 1720 2nd Ave South Birmingham, AL 35294 Phone: 205-934-2490 Fax: 205-975-6424 Email: [email protected]
Lisa L. Steed, PhD Professor of Pathology Director of Diagnostic Microbiology Medical University of South Carolina Department of Pathology and Laboratory Medicine 171 Ashley Avenue, MSC 908 Charleston, SC 29425 843-792-1108 Email: [email protected]
1
Cassandra Salgado, MD, MS (Corresponding Author) Professor of Medicine Medical University of South Carolina 135 Rutledge Ave, Suite 1204 MSC 752 Charleston, SC, 29425, USA Phone: 843-792-4541 Fax: 843-792-6680 Email: [email protected]
J. Michael Kilby, MD* Professor of Medicine and Microbiology Medical University of South Carolina Charleston, SC, 29425
*J. Michael Kilby, MD, a significant mentor and author for this manuscript, passed away suddenly August 10, 2015. The corresponding author, Cassandra Salgado, MD, MS will be responsible for verifying Dr. Kilby’s conflicts of interest as well as any additional necessary documentation.
Running Title: Clinical Experience with Actinomycosis
2
Abstract Background: Actinomyces, particularly A. israelii, may cause indolent, persistent infections or represent normal mucosal flora, leading to management dilemmas. Methods: Prompted by a refractory A. meyeri infection complicating AIDS, clinical data for all Actinomyces isolates at our hospital laboratory since 1998 were analyzed. Results: 140 cases had positive Actinomyces cultures. Of 130 cases with adequate follow-up, 36 (28%) had end-organ or disseminated disease treated with prolonged antibiotics and/or surgery (Group 1). A. meyeri was more common than A. israelii (33% vs 8%; p<0.05) in Group 1, particularly thoracic infections. Another 56 (43%) were considered local pathogens, treated with drainage only or short-course antibiotics (Group 2). 38 (29%) were deemed commensals (Group 3). Immunosuppression was less frequent in Group 1 vs Group 2 or 3 (p=0.05) and HIV/AIDS was uncommon. Foreign bodies or devices (Group 1 vs Group 2 or 3, p=0.003) and alcoholism (Group 1 vs Group 2 or 3; p=0.03) were associated with actinomycosis. Isolates from the central nervous system and musculoskeletal sites were more often treated as definitive disease; skin, abdominal/pelvic or single blood culture isolates were more likely commensals (all p <0.05). Disease progression or recurrence did not occur in Groups 2 and 3, whereas Group 1 had complex and variable courses, including two deaths. Conclusions. In the absence of disseminated or end-organ disease, avoiding prolonged therapy for Actinomyces isolates was not associated with adverse outcomes. Alcoholism or foreign bodies were associated with actinomycosis. A. meyeri may be a more common cause of actinomycosis than previously recognized. 3
Key Words: Actinomyces meyeri, Actinomycosis, epidemiology, clinical outcomes. Conflicts of Interest and Source of Funding: None of the authors have conflicts of interest relevant to the manuscript and no funding was provided for this study.
INTRODUCTION Actinomycosis is a rare infection caused by a Gram positive anerobic bacterium and associated with mass-like lesions, granulomata, slowly progressive growth across tissue planes, and fistula formation1,2. Actinomyces was first identified in the nineteenth century as the etiology of “lumpy jaw” in cattle, and later as a cause of chronic cervicofacial and lung abscesses in humans3. Actinomyces infections can be challenging to eradicate, often requiring long-term administration of antibiotics with or without surgical intervention. Actinomyces species are also found as normal flora at oropharyngeal, gastrointestinal and genitourinary mucosal surfaces1,4. Thus, when Actinomyces is isolated in the setting of severe illness or complex co-morbidity, it may be challenging to distinguish a harmless commensal from a life-threatening pathogen. Actinomyces is known as a “great mimicker”, in that inflammatory masses in the presence of this organism may be mistaken for (or confused with co-existing) malignancies or manifestations of connective tissue diseases1,5. Most actinomycosis cases reported to date have been due to Actinomyces israelii1,6,7, although small series and case reports describe other species as potential causes including A. viscosus, A. odontolyticus, and A. naeslundii. A. meyeri has been rarely associated with actinomycosis--of 26 cases due to A. meyeri reviewed in the medical literature in 19963 and several subsequent case reports8-10, most infections were reported in males with
4
alcoholism, poor dentition and pulmonary manifestations, who frequently developed A. meyeri bacteremia. Recent medical literature on actinomycosis has pointed to the need for more research on 1) the significance of Actinomyces when isolated with other potential pathogens, 2) any role of underlying immunosuppression, and 3) when to consider observation or local control (ie, incision and drainage) versus brief or prolonged systemic antibiotics1. We recently cared for a patient with AIDS who had refractory cervical and thoracic disease due to A. meyeri, in spite of years of aggressive antibiotic administration. Prompted by this case and with these unanswered questions in mind, we reviewed all cases involving Actinomyces isolates at our medical center in the Southeastern United States, seeking clinical and microbiological correlates of invasive disease, including potentially predisposing conditions such as HIV/AIDS and solid organ transplantation. We also provide long-term follow-up of outcomes categorized by anatomical sites and the species involved. MATERIALS AND METHODS Ethics Review This study was reviewed and approved by the Medical University of South Carolina (MUSC) Institutional Review Board. Microbiological Analysis All isolates identified by the microbiology laboratory at MUSC as definite or probable Actinomyces species from any clinical specimen between January 1998 and August 2014 were included in the analyses. Anaerobic or aerotolerant, non-acid fast, Gram positive, non-spore-
5
forming rods with or without branching that grew on blood-containing media were initially identified by colony morphology, pigmentation, susceptibility to kanamycin and vancomycin and resistance to colistin, and variable 15% catalase. Isolates in pure or predominant culture were further characterized by a commercial biochemical identification panel (Analytab Products AnIdent then Remel RapID Ana II). Species identifications were reported if all available data were in agreement. “Actinomyces species” was reported for uncommon species identifications or if data did not distinguish between two or more species. Isolates were not available for further characterization in this retrospective study. Actinomyces isolates sent from other institutions for identification or obtained from patients whose clinical care was provided outside of MUSC were excluded. All other positive bacterial cultures obtained from patients whose histories involved Actinomyces isolates—from concurrent specimens, from anytime in the same calendar year, and from all years subsequent to the first isolate up until the end of the study period were recorded. Chart Reviews and Clinical Definitions Clinical charts for all cases involving Actinomyces isolates were reviewed, using paper charts for the oldest cases and electronic medical records systems from more recent years. Demographic information and diagnoses of medical conditions concurrent with the first positive cultures (HIV/AIDS, diabetes, solid organ or bone marrow transplant status, connective tissue or autoimmune disease requiring chronic immunosuppressive therapy, chronic lung disease, malignancy) were recorded. Key clinical parameters closest to the date of the positive culture (maximum temperature, weight, white blood cell count, absolute neutrophil count, platelet count, AST, ALT, alkaline phosphatase, total bilirubin, and serum albumin) were tabulated. Antibiotic
6
treatments (agents, duration, intravenous or oral administrations) and any discussion by clinicians of the potential pathogenicity of Actinomyces were noted. All related incision and drainage or other surgical procedures were recorded. Longer-term outcomes were assessed by reviewing all subsequent hospitalizations and clinic visits, all subsequent microbiological studies on the patients, and ultimate dates and causes of death when available. Cases involving Actinomyces isolates were separated into three broad categories. If the treating physician specifically noted the organism, identified it in an appropriate clinical setting as a pathogen, and placed the patient on long-term directed therapy or short-term directed therapy plus definitive surgical treatment, the case was considered actinomycosis (Group 1). On the opposite end of the spectrum, isolates dismissed or never mentioned by clinicians, and never treated with targeted antibiotic therapy, were categorized as commensals (Group 3). All remaining cases were grouped in a middle category of probable or possible pathogens (Group 2), which included local, self-limited infections such as pilonidal cysts or simple dental abscesses. Some of these patients were treated with debridement, incision and drainage, and/or a few days of empiric antibiotics. Disease presentations were subdivided into clinical categories based on those used in previous literature regarding actinomycosis1: Oral/Cervicofacial, Pulmonary/Thoracic, Abdominal/Pelvic, Musculoskeletal, Skin/Soft Tissue, Central Nervous System, or Disseminated/Bacteremic. Statistical Analysis The prevalence of underlying medical conditions (HIV/AIDS, transplant, malignancy, immunosuppressive drugs, etc) for actinomycosis, probable/possible disease, and for individual 7
species was calculated. Correlations were sought for individual species and clinical presentations and outcomes using Chi-square or Fisher’s exact test (p<0.05 was considered significant). RESULTS Index Case A 40 year old HIV-infected female, diagnosed 2 years previously with AIDS (absolute CD4+ T cell count of 6 cells and Pneumocystis jirovecii pneumonitis), presented in 2008 with a midline anterior neck fluctuant mass, which was incised and drained. Mycobacterial infection was suspected, but cultures of surgical drainage were negative and a PPD skin test was negative. A right supraclavicular lymph node mass developed and was excised; scant partially acid fast bacilli were seen but cultures were negative. She briefly received empiric broad spectrum therapy with no definitive clinical response. She began combination HIV therapy and experienced an increase in her absolute CD4+ T cell count to over 400, and the drainage from her anterior neck wound resolved. In 2009, she was first referred to our center, when she was found to be pregnant. During the third trimester, she had recurrence of her neck mass and purulent spontaneous drainage from the midline neck wound. A CT scan of the chest revealed mediastinal necrotic adenitis. Bronchoscopic alveolar lavage specimens demonstrated negative fungal and AFB stains, flow cytometry was negative for malignancy, and anaerobic specimens grew A. meyeri. She received 30 days of intravenous penicillin. A repeat CT scan showed progressive disease with surrounding pneumonitis and a fistulous tract connecting the anterior neck to the mediastinum (see Figure 1). Bronchoscopy with transbronchial biopsies revealed lymphoid and histiocytic 8
granulomatous inflammation, negative fungal and AFB smears, and positive cultures again for A. meyeri plus Veillonella species. Therapy was switched to intravenous ceftriaxone plus oral doxycycline for an additional 6 weeks. Pneumonitis features improved on serial images but adenopathy persisted and gradually calcified. Over the subsequent 5 years, she alternated courses of IV ceftriaxone and oral therapy with ampicillin/sulbactam and doxycycline. Attempts to de-escalate or stop chronic antibiotics resulted in re-enlargement of cervical and thoracic adenopathy, persistence of the fistula to the mediastinum, and recurrent swelling and purulent drainage from the anterior neck wound requiring surgical drainage and debridement. Overall Experience with Actinomyces Isolates Because of this refractory infection due to an infrequently reported species, we pursued a detailed review of the clinical experience with Actinomyces at our institution. Between January 1998 and September 2014, 140 cases had positive cultures for Actinomyces at MUSC. 60 (43%) could not be further speciated; 26 (19%) were A. meyeri; 25 (18%) were A. odontoliticus; 17 (12%) were A. israelii; and the remainder various other species. Seven isolates were excluded from the study because they were forwarded to MUSC as a reference laboratory, and clinical information was not accessible. Three additional cases were excluded because positive cultures were reported after transfer of care to another institution. Of 130 cases with adequate clinical data for analysis, 36 (28%) met the definition of actinomycosis (Group 1); 56 (43%) had possible Actinomyces infections (Group 2); and 38 (29%) were commensals (Group 3).
9
Clinical Characteristics Associated with Actinomycosis Demographic characteristics were not significantly different between the three groups (see Table 1). Immunosuppressive conditions were less common in actinomycosis versus all others (p=0.03). Underlying HIV was uncommon and the index case was the only HIVassociated actinomycosis case. Only one actinomycosis case occurred in the setting of neutropenia. Foreign bodies or devices were associated with 25% of actinomycosis cases, significantly more prevalent than among commensal cases (p=0.05) or among Group 2 or 3 together (p=0.003); only one case was related to the presence of an intrauterine device. Alcoholism was reported in more than 10% of all cases involving Actinomyces isolates, and was significantly more common in actinomycosis cases versus Group 2 or 3 together (p=0.03). Eight of the Group 1 cases (22%), all males, had documentation of alcohol abuse, and the most common species among this group (4 of 8) was A. meyeri. Whereas the frequency of a cancer diagnosis was not different among the groups, four Actinomyces isolates were from specimens obtained at or near a subsequently confirmed primary tumor site (mouth; duodenum, biliary tree, lung) leading to initial clinical uncertainty about diagnosing actinomycosis versus malignancy. There were no physical examination or clinical laboratory findings that clearly distinguished Group 1 cases from Group 2 and 3 cases, such as patterns or degree of fever, hepatic enzyme elevations, or hematologic parameters (data not shown). Patterns of other organisms isolated concurrently with the Actinomyces, or within the same year, also did not help to distinguish between cases approached by providers as definitive actinomycosis, possible infection, or commensals. In 50-70% of the abscess or bronchoscopy specimens from each of the three groups, for example, there were concurrent potential pathogens isolated (particularly mixed oral flora, anaerobic Streptococci, coagulase-negative Staphylococci, and yeast), leading to clinician 10
uncertainty about the necessity of specifically targeting the Actinomyces when there were not classic indications such as granulomatous masses or fistula formation. Disease Severity, Disease Presentations, and Associated Actinomyces Species A. meyeri was the most common species causing actinomycosis, significantly more prevalent than A. israelii (12/36 [33%] vs 3/36 [8%]; p<0.05), whereas 11/36 (31%) could not be further speciated (see Figure 2). Other species identified in these cases, in decreasing order of frequency, were A. odontolyticus (4/36 [11%]), A. turicensis (3/36[8%]), A. viscosus (2/36[6%]), and A. bovis (1/36[3%]). The most common clinical categories for actinomycosis were Oral/Cervicofacial (22%) and Pulmonary/Thoracic (22%). The Oral/Cervicofacial infections were caused by various species but only one was identified as A. israelii and none were A. meyeri. There was no predominant clinical pattern among the cases caused by A. israelii, unspeciated Actinomyces and identified species other than A. meyeri. The majority of the Pulmonary/Thoracic definitive cases (5 of 8) were caused by A. meyeri. The most common pathogens in Group 2 were unspeciated Actinomyces (27/56 [48%]) followed by A. odontolyticus (13/56[23%]). Analogous to Group 1, there were no isolates of A. meyeri among the Oral/Cervicofacial cases. A. meyeri isolates in Group 2 were two pelvic cases and five Skin/Soft Tissue abscesses. Single occurrences of positive blood cultures for Actinomyces were significantly likely to be ignored as commensals. Musculoskeletal or central nervous system isolates were more likely to be designated actinomycosis than categorized in Group 2 or 3 (Table 1).
11
Clinical Outcomes In terms of longer-term outcomes, two actinomycosis cases had deaths directly attributable to the infection (See Table 2 for Group 1 clinical details). Remaining cases had apparent microbiologic cures or required long-term antibiotic suppression, although 6 (14%) had moderate or severe disfigurement or disability (neck and orofacial fistulas, chronic bone infections, multiple hardware surgical revisions, or amputations). Outcomes were more consistently favorable among Group 2 and 3 cases, with no long-term evidence of persistence, major related disability and no infection-related deaths. DISCUSSION To our knowledge, this report represents the largest single-site experience on clinical characteristics and outcomes of cases involving Actinomyces isolates to date. A novel observation is that the most commonly identified species causing actinomycosis was A. meyeri, significantly more common than A. israelii. A. meyeri was predominant among the Pulmonary/Thoracic actinomycosis cases, where isolation of this species was particularly associated with substantial disease. Our experience supports previous observations that A. meyeri may be linked with thoracic disease in males with alcoholism [1, 8]. This phenomenon might be related to poor dentition and aspiration events leading to pulmonary complications. However, A. meyeri was one of the less common isolates from Oral/Cervicofacial specimens (none of the definitive infections and <10% of positive orofacial cultures overall). Perhaps A. meyeri represents a small subgroup of many potential Actinomycetes colonizing oral mucosa, but once aspirated or invading directly into the chest in the right clinical settings is uniquely pathogenic.
12
Our experience with Actinomyces isolates overall, however, did not demonstrate the high male-to-female ratio described in the previous literature11-13; males and females were equally represented in all groups. Whereas a challenging, refractory case of cervical and thoracic A. meyeri infection in a patient with HIV/AIDS prompted this investigation, our review of additional cases did not suggest a common association between actinomycosis and HIV infection. Instead, we found these isolates are frequently identified in the setting of other complex and diverse underlying diseases, sometimes making it difficult to arrive at a definitive diagnosis. We found examples of Actinomyces infections initially falsely suggestive of a malignant tumor in the neck, thorax, or abdomen, for example. There was frequent co-existence of cancer, vasculitis, or other complex inflammatory disease with positive cultures for Actinomyces, making it challenging to conclude whether Actinomyces species represent harmless bystanders or opportunistic pathogens invading already necrotized tissues. Unfortunately, we were not able to identify clear patterns, in terms of characteristic clinical presentations, laboratory or imaging findings, to contribute to a formal disease definition. Positive cultures obtained from granulomatous masses invading across tissue planes and creating fistulas clearly point to classically-described actinomycosis. As others have reported, Actinomyces species were often identified among polymicrobial flora isolated from wounds, abscess drainage, or bronchoalveolar lavage. Our analysis of longer-term outcomes, however, does suggest that clinical judgment was generally accurate in these scenarios, in that cases designated as actinomycosis caused more morbidity and mortality, while those approached as “possible infection” or non-disease based on clinical context were cleared or cured without prolonged antibiotic administration.
13
Our report has several limitations. This is a retrospective study with inherent potential biases and the lack of a comparison group without Actinomyces isolates. The diverse and variable clinical presentations of actinomycosis make a unifying case definition challenging1. As a result, we were compelled to categorize cases into groups of graduated severity based on the documented clinical judgment of individual clinicians. While it is theoretically possible clinicians were inclined to label A. meyeri as a true pathogen while approaching A. israelii as a commensal, based on the medical literature the converse bias would be considerably more likely. Another shortcoming of our report is the lack of speciation for many Actinomyces isolates, which substantially compromised the ability to make statistical conclusions about the precise relative contribution of individual species. Identification of Actinomyces species has been difficult for clinical microbiology labs due to the diversity of the members of this genus and the morphological similarity to other anaerobic Gram positive bacilli. Commercial biochemical panels may not identify newer species and may not be able to accurately distinguish between species14, 15. Other techniques providing better identification, such as whole-cell fatty acid analysis, gas-liquid chromatography, 16S rDNA sequencing and matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) have not been widely available. As MALDI-TOF becomes a more clinically accessible technology, identification will become more accurate. Unfortunately, our culture isolates are no longer available for further microbiologic and susceptibility testing. However, in the highly unlikely “worst case” scenario that every one of the unspeciated Actinomyces isolates in Group 1 were found to be A. israelii, the number of A. meyeri isolates in this clinical experience would still represent a remarkable and unexpected observation. A strength of our series is that it represents a large number of isolates of a relatively rare pathogen isolated over more than a 15 year clinical experience, in the microbiology
14
laboratory of a tertiary care center offering primary care to the surrounding region as well as subspecialized care to extensive cohorts of patients with HIV/AIDS, solid organ transplant, and other chronic diseases. In conclusion, A. meyeri may be a more common agent of actinomycosis than previously recognized, particularly in the setting of pulmonary and thoracic infections in patients with complex underlying diseases. Actinomycosis does not appear to be commonly associated with HIV infection, but may occur in both males and females of all ages in the setting of underlying malignancy, solid organ transplantation, diabetes, chronic lung disease, and particularly in the presence of foreign body implants or alcoholism. Differentiating harmless commensals from pathogenic Actinomyces isolates can be challenging, particularly when these organisms are isolated from sites adjacent to tissues necrotized by concurrent malignancy, vasculitis, or other inflammatory processes. Although randomized controlled trials are lacking, aggressive and prolonged antibiotic therapy is sometimes required to manage this chronic, smoldering infection when classic pathological signs such as enterocutaneous fistula or invasive granulomatous masses are present. However, in the absence of convincing clinical evidence of tissue invasion or repeated isolation from normally sterile sites, it appears reasonable to closely observe patients with likely commensal isolates or to treat uncomplicated Actinomyces local infections with incision and drainage or short course antibiotic therapy alone.
15
References 1. Wong VK, Turmezei TD, Weston VC. Actinomycosis. Brit Med J 2011; 343: 60997006. 2. Acevedo F, Baudrand R, Letelier LM, Gaete P. Actinomycosis: a great pretender. Int J Infect Dis 2008; 12: 358-62. 3. Apotheloz C, Regamey C. Disseminated infection due to Actinomyces meyeri: Case report and review. Clin Infect Dis 1996; 22: 621-5. 4. Smego RA, Foglia G. Actinomycosis. Clin Infect Dis 1998; 26: 1255-61. 5. Valour F, Senechal A, Dupieux C, et al. Infect Drug Resist 2014; 7: 183-97. 6. Pulverer G, Schutt-Gerowitt H, Schaal KP. Clin Infect Dis 2003; 21: 490-7. 7. Schaal KP, Lee HJ. Actinomycete infections in humans—a review. Gene 1992; 115: 201-11. 8. Fazili T, Blair D, Riddell S, et al. Actinomyces meyeri infection: case report and review of the literature. J Infect 2012; 65: 357-61. 9. Attaway A, Flynn T. From lumpy jaw to empyema. Infection 2013; 41: 1025-7. 10. Park HJ, Park AH, Kim SH, et al. Infect Chemother 2014; 46: 269-73. 11. Weese WC, Smith IM. A study of 57 cases of actinomycosis over a 36-year period. Arch Intern Med 1975; 135: 1562-68. 12. Brown J. Human actinomycosis. Hum Pathol 1973; 4: 319-30. 13. Kinnear W, MacFarlane J. A survey of thoracic actinomycosis. Respir Med 1990; 84: 57-59. 14. Clarridge JE, Zhang Q. Genotypic diversity of clinical Actinomyces species: phenotype, source, and disease correlation among genospecies. J Clin Micro 2002; 40: 3442-8.
16
15. Sarkonnen N, Kononen E, Summanen P, et al. Phenotypic identification of Actinomyces and related species isolated from human sources. J Clin Micro 2001; 39: 3955-61.
Figure 1. Index Case. Computerized tomography images of disease due to Actinomyces meyeri. On left, midline neck fistula connecting to necrotic anterior mediastinum. On right, cavitary mediastinal masses, scattered nodules, and pericardial fluid.
Figure 2a-c. Actinomyces species and sources of cultures in Groups 1, 2, and 3
Table 1: Demographic and Clinical Characteristics Associated with Actinomyces Isolates Group 1 (n=36) Actinomyco sis Male (%) Mean age (Range) HIV Infection Diabetes Alcohol Abuse Concurrent Cancer (%) Organ Transplantation Foreign Bodies/Devices Renal Failure (%) Immunosuppression (%) Oral/Cervicofacial Thoracic/Pulmonary Abdominal/Pelvic Musculoskeletal CNS
Group 3(n=38) Commensal s
P value (3 vs 1)
P value
P value
(1 vs 2+3)
(1+2 vs 3)
19(53%) 50 (1-84) 1(3%) 11(36%) 8(22%) 6(17%)
Group 2 (n=56) Possible/Lo cal Infections 31(55%) 47 (1-81) 0 (0%) 15(27%) 1 (2%) 8(14%)
21(55%) 45 (10-77) 2 (4%) 14(37%) 7(18%) 7(18%)
ns ns ns ns ns ns
ns ns ns ns 0.03 ns
ns ns ns ns ns ns
2(6%)
7(12.5%)
3(8%)
ns
ns
ns
9(25%)
3(5%)
3(8%)
0.05
0.003
ns
4(11%) 3(8%)
7(13%) 13(23%)
3(8%) 9(24%)
ns ns
ns 0.05
ns ns
8(22%) 8(22%) 5(14%) 4(11%) 5(14%)
8 (14%) 12(21%) 16(28%) 1 (2%) 0 (0%)
4(11%) 14 (36%) 3 (8%) 1 (3%) 1 (3%)
ns ns ns ns ns
ns ns ns 0.03 0.002
ns ns 0.05 ns ns 17
Disseminated/Blood Skin/Soft Tissue
4(11%) 2(5.5%)
3(5%) 16(28%)
15 (40%) 0 (0%)
0.006 ns
ns ns
<0.001 0.01
Table 2: Clinical Details of Actinomycosis Cases (Group 1) A R S Cultur Disease g a e e e c x Sourc e e A. 4 W F BAL + Pneumoniti mey 1 TBNA s, eri mediastiniti s, neck fistula Spe cies
A. 2 A F CSF mey 2 A eri
CSF Shunt Infection
Underlying Conditions
Concurrent Microbiolog y
Treatment and Outcome
HIV
Veillonella
Acute IV PEN <1 mo Chronic IV CEPH + PO TETRA x6wks -> alternating IV CEPH and PO PEN/TETRA
Congenital hydrocephalus
None
A. 6 A M BAL mey 0 A eri
Pneumoniti s
Pulmonary vasculitis, ESRD; Hepatocellular carcinoma, DM
A. 6 W F Blood mey 9 eri
Disseminat ed with wound infection
Breast carcinoma None S/P mastectomy and chemotherapy
A. 3 W F Bone mey 9
Pelvic osteomyelit
Multiple Sclerosis,
None
None
Recurrent swelling and drainage Acute IV CEPH <1mo Chronic PO PEN x6mo Cure Acute IV CEPH <1mo Chronic LTFU Acute IV PEN <1mo Chronic PO PEN x6mo Cure Acute IV CARB x6wks 18
eri
A. 5 A M Brain mey 0 A biopsy eri
is
Brain abscess, pneumoniti s
urethrovaginal fistula, catheter erosion into bone
Alcoholism
Chronic PO PEN x6mo
Actinobacillu s; mixed anerobic flora
A. 6 W M Pericar Pericarditis, Churg Stauss mey 4 dial pneumoniti vasculitis eri fluid s involving myocardium
None
A. 4 L M Bone + mey 4 brain eri absces s
None
Microaerophi lic streptococcus ; Strep mitis; CNS
Heart Transplant
CNS
Sinusitis, orbital cellulitis, frontal brain abscess A. 5 W M Sternal Deep mey 4 necrosi sternal eri s wound infection
A. 5 A M Aortoil mey 6 A liac eri bypass graft
A. 6 W M Pacem mey 4 aker eri pocket tissue
Endovascul ar and deep tissue infection
Pacemaker pocket infection
Alcoholism, vascular disease
CHF with pacemaker.
None
None
Recurrent osteomyelitis Acute IV CEPH <1mo Chronic PO PEN x5mo LTFU Acute IV CEPH x6 wks Chronic PO PEN x5mo Cure Acute Drainage and IV CEPH 1mo Chronic LTFU Acute IV PEN x1mo Chronic Not known Cure Acute IV PEN x1mo Chronic PO PEN x12mo Cure Acute Pacemaker removal Chronic none Death due to heart disease
19
A. 5 A M Bone + Metatarsal mey 0 A tissue osteomyelit eri is
Hepatitis C, alcoholism, leg trauma
Mixed anaerobes, CNS
A. isra elii
3 A F Brain 5 A biopsy
Brain abscess
ESRD, chronic granulomatous disease, DM
None
A. isra elii
6 W F Bone 6
Mandibular infection
Monoclonal gammopathy, RA lumbar laminectomy
Microaerophi lic streptococcus
A. isra elii
A. odo ntol ytic us
A. odo ntol ytic us
A. odo ntol ytic us
4 A M Lymph Pneumoniti 9 A node s, biopsy mediastinal and hilar adenitis 3 A M Ear A draina ge
4 A M Blood 0 A
4 W F Sublin 7 gual Absces s
Alcoholism and cocaine abuse
Periauricula Chronic otitis r abscess media, bilateral myringotomy tubes, developmental delay Disseminat ed disease with osteomyelit is
Abscess with cutaneous fistula
DM, ESRD after transplant, chronic osteomyelitis
Sputum: Mycobacteri um arupense
Acute CARB x7wks Chronic PO PEN >1yr Lifelong therapy planned Acute IV CARB x7wks Chronic PO TETRA LTFU Acute Debridement IV CARB x8wks Chronic PO TETRA x4wks Cure Acute IV CEPH <1mo Chronic PO PEN
Ongoing Peptostreptoc Acute occus I&D Chronic Serial surgeries and PO PEN
None
Ongoing Acute Foot amputation and IV CARB PO PEN Chronic None
Deficits Gunshot wound to Pseudomonas Acute the face aeruginosa IV CLINDA <1mo Chronic PO PEN <1mo
20
A. turi cen sis
5 A M Perior 7 A bital absces s
Infected Orbital floor orbital plate fracture, recurrent abscesses
Microaerophi llic streptococcus
Recurrent Infections Acute Hardware removal Chronic PO Bactrim x4wks PO PEN x3wks Cure Acute IV METRO x1wk Chronic PO PEN x4wks
A. turi cen sis
2 W F Abdo 2 minal Absces s
Necrotic uterus after cesarean section
Narcotic abuse
Mixed gram positive organisms
A. turi cen sis
8 A F Brain 4 A biopsy + CSF
Cerebellar abscesses
Cerebellar stroke, hydrocephalus
None
A. visc osu s
4 A F Pelvic Ruptured DM 4 A Absces appendicitis s with abscesses
A. visc osu s
7 W M Blood TIPS 7 + TIPS infection shunt
Alcoholic cirrhosis
A. bov is
6 W M Cornea Corneal 1 l ulcer scrapin g
None
None
A. spe cies
6 A M Bone 0 A
Ischium osteomyelit is
Spina bifida, chronic left sacral decubitus ulcer, DM
Streptococcu s mitis, mixed flora
A. spe
7 W M Lung 1 fluid
Empyema
Lung cancer
Eschericia coli
Mixed anaerobic flora; Microaerophi lic streptococcus None
Cure Acute IV PEN x6wks Chronic PO PEN x12mo Death Acute IV PEN x6wks Chronic PO PEN x2wks LTFU Acute IV ABX (Poor documentation) Chronic LTFU Acute AG solution <1mo Chronic TETRA x3 wks Cure Acute IV CEPH x6wks Chronic PO PEN 1yr Cure Acute Decortication, IV 21
CEPH x >1mo Chronic None
cies
A. spe cies
A. spe cies
3 W F IUD 4
8 A F Blood 1 A
Pelvic inflammati on
Blood
A. spe cies
3 A M Arm 8 A tissue
A. spe cies
4 W F Liver Liver 9 Absces abscess s
A. spe cies
A. spe cies
A. spe cies
Hardware infection
2 A F Pelvic Pelvic 4 A Absces abscess s
1 W F Pharyn geal absces s+ blood 4 W F Woun 8 d draina ge
None
Cholangiocarcinoma, DM, CAD
None
Gunshot wounds, multiple surgeries
Proprionibact erium acnes
Duodenal carcinoma, DM
None
Obesity, hyperglycemia
Pharyngeal None abscess, prevertebral extension
Orofacial wound infection
Microaerophi lic streptococcus
Head and Neck Cancer, Obesity, DM
Mixed anaerobic flora
Usual oral flora
None
LTFU Acute TETRA <1mo, IUD removal Chronic None Cure Acute IV PEN x2wks Chronic PO pen x 6mo LTFU Acute IV CEPH x6wks Chronic PO PEN LTFU Acute CARB x8wks Chronic PO PEN x12mo Cure Acute IV PEN x6wks Chronic PO PEN x6wks Cure Acute IV PEN <1mo Chronic PO PEN x4wks Cure Acute TETRA x unspecified weeks Chronic LTFU 22
A. spe cies
A. spe cies
5 A M Neck 3 A mass + Facial Absces s 7 A F Pleural 3 A fluid
Soft tissue mass and cellulitis
Alcoholism, Hepatitis B, Hepatitis C
CNS
Empyema
Acute renal failure, CHF, DM
None
Acute I&D PO PEN x6wks Chronic LTFU Acute Drainage IV PEN x1wk Chronic PO PEN 1mo
Cure Abbreviations: W, white; AA, African American; F, female; M, male; BAL, bronchoalveolar lavage; TBNA, transbronchial needle aspirate; CSF, cerebrospinal fluid; TIPS, transjugular intrahepatic porto-systemic shunt; HIV, human immunodeficiency virus; ESRD, end-stage renal disease; DM, diabetes mellitus; RA, rheumatoid arthritis; CAD, coronary artery disease; CHF, congestive heart failure; CNS, coagulase-negative Staphylococcus; IV, intravenous; I&D, incision and drainage; PO, oral; LTFU, lost to follow-up; PEN, penicillin or ampicillin; CEPH, cephalosporin; CARB, carbapenem; CLINDA, clindamycin; METRO, metronidazole; ABX, antibiotics; AG, aminoglycoside; TETRA, tetracyclines.
Figure 1. Index Case. Computerized tomography images of disease due to Actinomyces meyeri. On left, midline neck fistula connecting to necrotic anterior mediastinum. On right, cavitary mediastinal masses, scattered nodules, and pericardial fluid.
Figure 2a-c. Actinomyces species and sources of cultures in Groups 1, 2, and 3
23
Figure 1. Computerized tomography images of disease due to Actinomyces meyeri
24
Percent
Figure 2a. Group 1: Actinomycosis 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
25
Percent
Figure 2b. Group 2: Possible or Local Infections 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
Percent
Figure 2c. Group 3: Commensals 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
26
Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience Robert Rolfe MD, Lisa L. Steed PhD, Cassandra Salgado MD, MS, J. Michael Kilby MD
www.amjmedsci.com
PII: DOI: Reference:
S0002-9629(15)37988-X http://dx.doi.org/10.1016/j.amjms.2016.03.003 AMJMS128
To appear in:
Am J Med Sci
Cite this article as: Robert Rolfe MD, Lisa L. Steed PhD, Cassandra Salgado MD, MS, J. Michael Kilby MD, Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience, Am J Med Sci, http://dx.doi.org/10.1016/j.amjms.2016.03.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Actinomyces meyeri, a Common Agent of Actinomycosis: Case Report and 16-Year Clinical Experience
Robert Rolfe, MD The University of Alabama of Birmingham Internal Medicine Residency Training Program BDB 327 1720 2nd Ave South Birmingham, AL 35294 Phone: 205-934-2490 Fax: 205-975-6424 Email: [email protected]
Lisa L. Steed, PhD Professor of Pathology Director of Diagnostic Microbiology Medical University of South Carolina Department of Pathology and Laboratory Medicine 171 Ashley Avenue, MSC 908 Charleston, SC 29425 843-792-1108 Email: [email protected]
1
Cassandra Salgado, MD, MS (Corresponding Author) Professor of Medicine Medical University of South Carolina 135 Rutledge Ave, Suite 1204 MSC 752 Charleston, SC, 29425, USA Phone: 843-792-4541 Fax: 843-792-6680 Email: [email protected]
J. Michael Kilby, MD* Professor of Medicine and Microbiology Medical University of South Carolina Charleston, SC, 29425
*J. Michael Kilby, MD, a significant mentor and author for this manuscript, passed away suddenly August 10, 2015. The corresponding author, Cassandra Salgado, MD, MS will be responsible for verifying Dr. Kilby’s conflicts of interest as well as any additional necessary documentation.
Running Title: Clinical Experience with Actinomycosis
2
Abstract Background: Actinomyces, particularly A. israelii, may cause indolent, persistent infections or represent normal mucosal flora, leading to management dilemmas. Methods: Prompted by a refractory A. meyeri infection complicating AIDS, clinical data for all Actinomyces isolates at our hospital laboratory since 1998 were analyzed. Results: 140 cases had positive Actinomyces cultures. Of 130 cases with adequate follow-up, 36 (28%) had end-organ or disseminated disease treated with prolonged antibiotics and/or surgery (Group 1). A. meyeri was more common than A. israelii (33% vs 8%; p<0.05) in Group 1, particularly thoracic infections. Another 56 (43%) were considered local pathogens, treated with drainage only or short-course antibiotics (Group 2). 38 (29%) were deemed commensals (Group 3). Immunosuppression was less frequent in Group 1 vs Group 2 or 3 (p=0.05) and HIV/AIDS was uncommon. Foreign bodies or devices (Group 1 vs Group 2 or 3, p=0.003) and alcoholism (Group 1 vs Group 2 or 3; p=0.03) were associated with actinomycosis. Isolates from the central nervous system and musculoskeletal sites were more often treated as definitive disease; skin, abdominal/pelvic or single blood culture isolates were more likely commensals (all p <0.05). Disease progression or recurrence did not occur in Groups 2 and 3, whereas Group 1 had complex and variable courses, including two deaths. Conclusions. In the absence of disseminated or end-organ disease, avoiding prolonged therapy for Actinomyces isolates was not associated with adverse outcomes. Alcoholism or foreign bodies were associated with actinomycosis. A. meyeri may be a more common cause of actinomycosis than previously recognized. 3
Key Words: Actinomyces meyeri, Actinomycosis, epidemiology, clinical outcomes. Conflicts of Interest and Source of Funding: None of the authors have conflicts of interest relevant to the manuscript and no funding was provided for this study.
INTRODUCTION Actinomycosis is a rare infection caused by a Gram positive anerobic bacterium and associated with mass-like lesions, granulomata, slowly progressive growth across tissue planes, and fistula formation1,2. Actinomyces was first identified in the nineteenth century as the etiology of “lumpy jaw” in cattle, and later as a cause of chronic cervicofacial and lung abscesses in humans3. Actinomyces infections can be challenging to eradicate, often requiring long-term administration of antibiotics with or without surgical intervention. Actinomyces species are also found as normal flora at oropharyngeal, gastrointestinal and genitourinary mucosal surfaces1,4. Thus, when Actinomyces is isolated in the setting of severe illness or complex co-morbidity, it may be challenging to distinguish a harmless commensal from a life-threatening pathogen. Actinomyces is known as a “great mimicker”, in that inflammatory masses in the presence of this organism may be mistaken for (or confused with co-existing) malignancies or manifestations of connective tissue diseases1,5. Most actinomycosis cases reported to date have been due to Actinomyces israelii1,6,7, although small series and case reports describe other species as potential causes including A. viscosus, A. odontolyticus, and A. naeslundii. A. meyeri has been rarely associated with actinomycosis--of 26 cases due to A. meyeri reviewed in the medical literature in 19963 and several subsequent case reports8-10, most infections were reported in males with
4
alcoholism, poor dentition and pulmonary manifestations, who frequently developed A. meyeri bacteremia. Recent medical literature on actinomycosis has pointed to the need for more research on 1) the significance of Actinomyces when isolated with other potential pathogens, 2) any role of underlying immunosuppression, and 3) when to consider observation or local control (ie, incision and drainage) versus brief or prolonged systemic antibiotics1. We recently cared for a patient with AIDS who had refractory cervical and thoracic disease due to A. meyeri, in spite of years of aggressive antibiotic administration. Prompted by this case and with these unanswered questions in mind, we reviewed all cases involving Actinomyces isolates at our medical center in the Southeastern United States, seeking clinical and microbiological correlates of invasive disease, including potentially predisposing conditions such as HIV/AIDS and solid organ transplantation. We also provide long-term follow-up of outcomes categorized by anatomical sites and the species involved. MATERIALS AND METHODS Ethics Review This study was reviewed and approved by the Medical University of South Carolina (MUSC) Institutional Review Board. Microbiological Analysis All isolates identified by the microbiology laboratory at MUSC as definite or probable Actinomyces species from any clinical specimen between January 1998 and August 2014 were included in the analyses. Anaerobic or aerotolerant, non-acid fast, Gram positive, non-spore-
5
forming rods with or without branching that grew on blood-containing media were initially identified by colony morphology, pigmentation, susceptibility to kanamycin and vancomycin and resistance to colistin, and variable 15% catalase. Isolates in pure or predominant culture were further characterized by a commercial biochemical identification panel (Analytab Products AnIdent then Remel RapID Ana II). Species identifications were reported if all available data were in agreement. “Actinomyces species” was reported for uncommon species identifications or if data did not distinguish between two or more species. Isolates were not available for further characterization in this retrospective study. Actinomyces isolates sent from other institutions for identification or obtained from patients whose clinical care was provided outside of MUSC were excluded. All other positive bacterial cultures obtained from patients whose histories involved Actinomyces isolates—from concurrent specimens, from anytime in the same calendar year, and from all years subsequent to the first isolate up until the end of the study period were recorded. Chart Reviews and Clinical Definitions Clinical charts for all cases involving Actinomyces isolates were reviewed, using paper charts for the oldest cases and electronic medical records systems from more recent years. Demographic information and diagnoses of medical conditions concurrent with the first positive cultures (HIV/AIDS, diabetes, solid organ or bone marrow transplant status, connective tissue or autoimmune disease requiring chronic immunosuppressive therapy, chronic lung disease, malignancy) were recorded. Key clinical parameters closest to the date of the positive culture (maximum temperature, weight, white blood cell count, absolute neutrophil count, platelet count, AST, ALT, alkaline phosphatase, total bilirubin, and serum albumin) were tabulated. Antibiotic
6
treatments (agents, duration, intravenous or oral administrations) and any discussion by clinicians of the potential pathogenicity of Actinomyces were noted. All related incision and drainage or other surgical procedures were recorded. Longer-term outcomes were assessed by reviewing all subsequent hospitalizations and clinic visits, all subsequent microbiological studies on the patients, and ultimate dates and causes of death when available. Cases involving Actinomyces isolates were separated into three broad categories. If the treating physician specifically noted the organism, identified it in an appropriate clinical setting as a pathogen, and placed the patient on long-term directed therapy or short-term directed therapy plus definitive surgical treatment, the case was considered actinomycosis (Group 1). On the opposite end of the spectrum, isolates dismissed or never mentioned by clinicians, and never treated with targeted antibiotic therapy, were categorized as commensals (Group 3). All remaining cases were grouped in a middle category of probable or possible pathogens (Group 2), which included local, self-limited infections such as pilonidal cysts or simple dental abscesses. Some of these patients were treated with debridement, incision and drainage, and/or a few days of empiric antibiotics. Disease presentations were subdivided into clinical categories based on those used in previous literature regarding actinomycosis1: Oral/Cervicofacial, Pulmonary/Thoracic, Abdominal/Pelvic, Musculoskeletal, Skin/Soft Tissue, Central Nervous System, or Disseminated/Bacteremic. Statistical Analysis The prevalence of underlying medical conditions (HIV/AIDS, transplant, malignancy, immunosuppressive drugs, etc) for actinomycosis, probable/possible disease, and for individual 7
species was calculated. Correlations were sought for individual species and clinical presentations and outcomes using Chi-square or Fisher’s exact test (p<0.05 was considered significant). RESULTS Index Case A 40 year old HIV-infected female, diagnosed 2 years previously with AIDS (absolute CD4+ T cell count of 6 cells and Pneumocystis jirovecii pneumonitis), presented in 2008 with a midline anterior neck fluctuant mass, which was incised and drained. Mycobacterial infection was suspected, but cultures of surgical drainage were negative and a PPD skin test was negative. A right supraclavicular lymph node mass developed and was excised; scant partially acid fast bacilli were seen but cultures were negative. She briefly received empiric broad spectrum therapy with no definitive clinical response. She began combination HIV therapy and experienced an increase in her absolute CD4+ T cell count to over 400, and the drainage from her anterior neck wound resolved. In 2009, she was first referred to our center, when she was found to be pregnant. During the third trimester, she had recurrence of her neck mass and purulent spontaneous drainage from the midline neck wound. A CT scan of the chest revealed mediastinal necrotic adenitis. Bronchoscopic alveolar lavage specimens demonstrated negative fungal and AFB stains, flow cytometry was negative for malignancy, and anaerobic specimens grew A. meyeri. She received 30 days of intravenous penicillin. A repeat CT scan showed progressive disease with surrounding pneumonitis and a fistulous tract connecting the anterior neck to the mediastinum (see Figure 1). Bronchoscopy with transbronchial biopsies revealed lymphoid and histiocytic 8
granulomatous inflammation, negative fungal and AFB smears, and positive cultures again for A. meyeri plus Veillonella species. Therapy was switched to intravenous ceftriaxone plus oral doxycycline for an additional 6 weeks. Pneumonitis features improved on serial images but adenopathy persisted and gradually calcified. Over the subsequent 5 years, she alternated courses of IV ceftriaxone and oral therapy with ampicillin/sulbactam and doxycycline. Attempts to de-escalate or stop chronic antibiotics resulted in re-enlargement of cervical and thoracic adenopathy, persistence of the fistula to the mediastinum, and recurrent swelling and purulent drainage from the anterior neck wound requiring surgical drainage and debridement. Overall Experience with Actinomyces Isolates Because of this refractory infection due to an infrequently reported species, we pursued a detailed review of the clinical experience with Actinomyces at our institution. Between January 1998 and September 2014, 140 cases had positive cultures for Actinomyces at MUSC. 60 (43%) could not be further speciated; 26 (19%) were A. meyeri; 25 (18%) were A. odontoliticus; 17 (12%) were A. israelii; and the remainder various other species. Seven isolates were excluded from the study because they were forwarded to MUSC as a reference laboratory, and clinical information was not accessible. Three additional cases were excluded because positive cultures were reported after transfer of care to another institution. Of 130 cases with adequate clinical data for analysis, 36 (28%) met the definition of actinomycosis (Group 1); 56 (43%) had possible Actinomyces infections (Group 2); and 38 (29%) were commensals (Group 3).
9
Clinical Characteristics Associated with Actinomycosis Demographic characteristics were not significantly different between the three groups (see Table 1). Immunosuppressive conditions were less common in actinomycosis versus all others (p=0.03). Underlying HIV was uncommon and the index case was the only HIVassociated actinomycosis case. Only one actinomycosis case occurred in the setting of neutropenia. Foreign bodies or devices were associated with 25% of actinomycosis cases, significantly more prevalent than among commensal cases (p=0.05) or among Group 2 or 3 together (p=0.003); only one case was related to the presence of an intrauterine device. Alcoholism was reported in more than 10% of all cases involving Actinomyces isolates, and was significantly more common in actinomycosis cases versus Group 2 or 3 together (p=0.03). Eight of the Group 1 cases (22%), all males, had documentation of alcohol abuse, and the most common species among this group (4 of 8) was A. meyeri. Whereas the frequency of a cancer diagnosis was not different among the groups, four Actinomyces isolates were from specimens obtained at or near a subsequently confirmed primary tumor site (mouth; duodenum, biliary tree, lung) leading to initial clinical uncertainty about diagnosing actinomycosis versus malignancy. There were no physical examination or clinical laboratory findings that clearly distinguished Group 1 cases from Group 2 and 3 cases, such as patterns or degree of fever, hepatic enzyme elevations, or hematologic parameters (data not shown). Patterns of other organisms isolated concurrently with the Actinomyces, or within the same year, also did not help to distinguish between cases approached by providers as definitive actinomycosis, possible infection, or commensals. In 50-70% of the abscess or bronchoscopy specimens from each of the three groups, for example, there were concurrent potential pathogens isolated (particularly mixed oral flora, anaerobic Streptococci, coagulase-negative Staphylococci, and yeast), leading to clinician 10
uncertainty about the necessity of specifically targeting the Actinomyces when there were not classic indications such as granulomatous masses or fistula formation. Disease Severity, Disease Presentations, and Associated Actinomyces Species A. meyeri was the most common species causing actinomycosis, significantly more prevalent than A. israelii (12/36 [33%] vs 3/36 [8%]; p<0.05), whereas 11/36 (31%) could not be further speciated (see Figure 2). Other species identified in these cases, in decreasing order of frequency, were A. odontolyticus (4/36 [11%]), A. turicensis (3/36[8%]), A. viscosus (2/36[6%]), and A. bovis (1/36[3%]). The most common clinical categories for actinomycosis were Oral/Cervicofacial (22%) and Pulmonary/Thoracic (22%). The Oral/Cervicofacial infections were caused by various species but only one was identified as A. israelii and none were A. meyeri. There was no predominant clinical pattern among the cases caused by A. israelii, unspeciated Actinomyces and identified species other than A. meyeri. The majority of the Pulmonary/Thoracic definitive cases (5 of 8) were caused by A. meyeri. The most common pathogens in Group 2 were unspeciated Actinomyces (27/56 [48%]) followed by A. odontolyticus (13/56[23%]). Analogous to Group 1, there were no isolates of A. meyeri among the Oral/Cervicofacial cases. A. meyeri isolates in Group 2 were two pelvic cases and five Skin/Soft Tissue abscesses. Single occurrences of positive blood cultures for Actinomyces were significantly likely to be ignored as commensals. Musculoskeletal or central nervous system isolates were more likely to be designated actinomycosis than categorized in Group 2 or 3 (Table 1).
11
Clinical Outcomes In terms of longer-term outcomes, two actinomycosis cases had deaths directly attributable to the infection (See Table 2 for Group 1 clinical details). Remaining cases had apparent microbiologic cures or required long-term antibiotic suppression, although 6 (14%) had moderate or severe disfigurement or disability (neck and orofacial fistulas, chronic bone infections, multiple hardware surgical revisions, or amputations). Outcomes were more consistently favorable among Group 2 and 3 cases, with no long-term evidence of persistence, major related disability and no infection-related deaths. DISCUSSION To our knowledge, this report represents the largest single-site experience on clinical characteristics and outcomes of cases involving Actinomyces isolates to date. A novel observation is that the most commonly identified species causing actinomycosis was A. meyeri, significantly more common than A. israelii. A. meyeri was predominant among the Pulmonary/Thoracic actinomycosis cases, where isolation of this species was particularly associated with substantial disease. Our experience supports previous observations that A. meyeri may be linked with thoracic disease in males with alcoholism [1, 8]. This phenomenon might be related to poor dentition and aspiration events leading to pulmonary complications. However, A. meyeri was one of the less common isolates from Oral/Cervicofacial specimens (none of the definitive infections and <10% of positive orofacial cultures overall). Perhaps A. meyeri represents a small subgroup of many potential Actinomycetes colonizing oral mucosa, but once aspirated or invading directly into the chest in the right clinical settings is uniquely pathogenic.
12
Our experience with Actinomyces isolates overall, however, did not demonstrate the high male-to-female ratio described in the previous literature11-13; males and females were equally represented in all groups. Whereas a challenging, refractory case of cervical and thoracic A. meyeri infection in a patient with HIV/AIDS prompted this investigation, our review of additional cases did not suggest a common association between actinomycosis and HIV infection. Instead, we found these isolates are frequently identified in the setting of other complex and diverse underlying diseases, sometimes making it difficult to arrive at a definitive diagnosis. We found examples of Actinomyces infections initially falsely suggestive of a malignant tumor in the neck, thorax, or abdomen, for example. There was frequent co-existence of cancer, vasculitis, or other complex inflammatory disease with positive cultures for Actinomyces, making it challenging to conclude whether Actinomyces species represent harmless bystanders or opportunistic pathogens invading already necrotized tissues. Unfortunately, we were not able to identify clear patterns, in terms of characteristic clinical presentations, laboratory or imaging findings, to contribute to a formal disease definition. Positive cultures obtained from granulomatous masses invading across tissue planes and creating fistulas clearly point to classically-described actinomycosis. As others have reported, Actinomyces species were often identified among polymicrobial flora isolated from wounds, abscess drainage, or bronchoalveolar lavage. Our analysis of longer-term outcomes, however, does suggest that clinical judgment was generally accurate in these scenarios, in that cases designated as actinomycosis caused more morbidity and mortality, while those approached as “possible infection” or non-disease based on clinical context were cleared or cured without prolonged antibiotic administration.
13
Our report has several limitations. This is a retrospective study with inherent potential biases and the lack of a comparison group without Actinomyces isolates. The diverse and variable clinical presentations of actinomycosis make a unifying case definition challenging1. As a result, we were compelled to categorize cases into groups of graduated severity based on the documented clinical judgment of individual clinicians. While it is theoretically possible clinicians were inclined to label A. meyeri as a true pathogen while approaching A. israelii as a commensal, based on the medical literature the converse bias would be considerably more likely. Another shortcoming of our report is the lack of speciation for many Actinomyces isolates, which substantially compromised the ability to make statistical conclusions about the precise relative contribution of individual species. Identification of Actinomyces species has been difficult for clinical microbiology labs due to the diversity of the members of this genus and the morphological similarity to other anaerobic Gram positive bacilli. Commercial biochemical panels may not identify newer species and may not be able to accurately distinguish between species14, 15. Other techniques providing better identification, such as whole-cell fatty acid analysis, gas-liquid chromatography, 16S rDNA sequencing and matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) have not been widely available. As MALDI-TOF becomes a more clinically accessible technology, identification will become more accurate. Unfortunately, our culture isolates are no longer available for further microbiologic and susceptibility testing. However, in the highly unlikely “worst case” scenario that every one of the unspeciated Actinomyces isolates in Group 1 were found to be A. israelii, the number of A. meyeri isolates in this clinical experience would still represent a remarkable and unexpected observation. A strength of our series is that it represents a large number of isolates of a relatively rare pathogen isolated over more than a 15 year clinical experience, in the microbiology
14
laboratory of a tertiary care center offering primary care to the surrounding region as well as subspecialized care to extensive cohorts of patients with HIV/AIDS, solid organ transplant, and other chronic diseases. In conclusion, A. meyeri may be a more common agent of actinomycosis than previously recognized, particularly in the setting of pulmonary and thoracic infections in patients with complex underlying diseases. Actinomycosis does not appear to be commonly associated with HIV infection, but may occur in both males and females of all ages in the setting of underlying malignancy, solid organ transplantation, diabetes, chronic lung disease, and particularly in the presence of foreign body implants or alcoholism. Differentiating harmless commensals from pathogenic Actinomyces isolates can be challenging, particularly when these organisms are isolated from sites adjacent to tissues necrotized by concurrent malignancy, vasculitis, or other inflammatory processes. Although randomized controlled trials are lacking, aggressive and prolonged antibiotic therapy is sometimes required to manage this chronic, smoldering infection when classic pathological signs such as enterocutaneous fistula or invasive granulomatous masses are present. However, in the absence of convincing clinical evidence of tissue invasion or repeated isolation from normally sterile sites, it appears reasonable to closely observe patients with likely commensal isolates or to treat uncomplicated Actinomyces local infections with incision and drainage or short course antibiotic therapy alone.
15
References 1. Wong VK, Turmezei TD, Weston VC. Actinomycosis. Brit Med J 2011; 343: 60997006. 2. Acevedo F, Baudrand R, Letelier LM, Gaete P. Actinomycosis: a great pretender. Int J Infect Dis 2008; 12: 358-62. 3. Apotheloz C, Regamey C. Disseminated infection due to Actinomyces meyeri: Case report and review. Clin Infect Dis 1996; 22: 621-5. 4. Smego RA, Foglia G. Actinomycosis. Clin Infect Dis 1998; 26: 1255-61. 5. Valour F, Senechal A, Dupieux C, et al. Infect Drug Resist 2014; 7: 183-97. 6. Pulverer G, Schutt-Gerowitt H, Schaal KP. Clin Infect Dis 2003; 21: 490-7. 7. Schaal KP, Lee HJ. Actinomycete infections in humans—a review. Gene 1992; 115: 201-11. 8. Fazili T, Blair D, Riddell S, et al. Actinomyces meyeri infection: case report and review of the literature. J Infect 2012; 65: 357-61. 9. Attaway A, Flynn T. From lumpy jaw to empyema. Infection 2013; 41: 1025-7. 10. Park HJ, Park AH, Kim SH, et al. Infect Chemother 2014; 46: 269-73. 11. Weese WC, Smith IM. A study of 57 cases of actinomycosis over a 36-year period. Arch Intern Med 1975; 135: 1562-68. 12. Brown J. Human actinomycosis. Hum Pathol 1973; 4: 319-30. 13. Kinnear W, MacFarlane J. A survey of thoracic actinomycosis. Respir Med 1990; 84: 57-59. 14. Clarridge JE, Zhang Q. Genotypic diversity of clinical Actinomyces species: phenotype, source, and disease correlation among genospecies. J Clin Micro 2002; 40: 3442-8.
16
15. Sarkonnen N, Kononen E, Summanen P, et al. Phenotypic identification of Actinomyces and related species isolated from human sources. J Clin Micro 2001; 39: 3955-61.
Figure 1. Index Case. Computerized tomography images of disease due to Actinomyces meyeri. On left, midline neck fistula connecting to necrotic anterior mediastinum. On right, cavitary mediastinal masses, scattered nodules, and pericardial fluid.
Figure 2a-c. Actinomyces species and sources of cultures in Groups 1, 2, and 3
Table 1: Demographic and Clinical Characteristics Associated with Actinomyces Isolates Group 1 (n=36) Actinomyco sis Male (%) Mean age (Range) HIV Infection Diabetes Alcohol Abuse Concurrent Cancer (%) Organ Transplantation Foreign Bodies/Devices Renal Failure (%) Immunosuppression (%) Oral/Cervicofacial Thoracic/Pulmonary Abdominal/Pelvic Musculoskeletal CNS
Group 3(n=38) Commensal s
P value (3 vs 1)
P value
P value
(1 vs 2+3)
(1+2 vs 3)
19(53%) 50 (1-84) 1(3%) 11(36%) 8(22%) 6(17%)
Group 2 (n=56) Possible/Lo cal Infections 31(55%) 47 (1-81) 0 (0%) 15(27%) 1 (2%) 8(14%)
21(55%) 45 (10-77) 2 (4%) 14(37%) 7(18%) 7(18%)
ns ns ns ns ns ns
ns ns ns ns 0.03 ns
ns ns ns ns ns ns
2(6%)
7(12.5%)
3(8%)
ns
ns
ns
9(25%)
3(5%)
3(8%)
0.05
0.003
ns
4(11%) 3(8%)
7(13%) 13(23%)
3(8%) 9(24%)
ns ns
ns 0.05
ns ns
8(22%) 8(22%) 5(14%) 4(11%) 5(14%)
8 (14%) 12(21%) 16(28%) 1 (2%) 0 (0%)
4(11%) 14 (36%) 3 (8%) 1 (3%) 1 (3%)
ns ns ns ns ns
ns ns ns 0.03 0.002
ns ns 0.05 ns ns 17
Disseminated/Blood Skin/Soft Tissue
4(11%) 2(5.5%)
3(5%) 16(28%)
15 (40%) 0 (0%)
0.006 ns
ns ns
<0.001 0.01
Table 2: Clinical Details of Actinomycosis Cases (Group 1) A R S Cultur Disease g a e e e c x Sourc e e A. 4 W F BAL + Pneumoniti mey 1 TBNA s, eri mediastiniti s, neck fistula Spe cies
A. 2 A F CSF mey 2 A eri
CSF Shunt Infection
Underlying Conditions
Concurrent Microbiolog y
Treatment and Outcome
HIV
Veillonella
Acute IV PEN <1 mo Chronic IV CEPH + PO TETRA x6wks -> alternating IV CEPH and PO PEN/TETRA
Congenital hydrocephalus
None
A. 6 A M BAL mey 0 A eri
Pneumoniti s
Pulmonary vasculitis, ESRD; Hepatocellular carcinoma, DM
A. 6 W F Blood mey 9 eri
Disseminat ed with wound infection
Breast carcinoma None S/P mastectomy and chemotherapy
A. 3 W F Bone mey 9
Pelvic osteomyelit
Multiple Sclerosis,
None
None
Recurrent swelling and drainage Acute IV CEPH <1mo Chronic PO PEN x6mo Cure Acute IV CEPH <1mo Chronic LTFU Acute IV PEN <1mo Chronic PO PEN x6mo Cure Acute IV CARB x6wks 18
eri
A. 5 A M Brain mey 0 A biopsy eri
is
Brain abscess, pneumoniti s
urethrovaginal fistula, catheter erosion into bone
Alcoholism
Chronic PO PEN x6mo
Actinobacillu s; mixed anerobic flora
A. 6 W M Pericar Pericarditis, Churg Stauss mey 4 dial pneumoniti vasculitis eri fluid s involving myocardium
None
A. 4 L M Bone + mey 4 brain eri absces s
None
Microaerophi lic streptococcus ; Strep mitis; CNS
Heart Transplant
CNS
Sinusitis, orbital cellulitis, frontal brain abscess A. 5 W M Sternal Deep mey 4 necrosi sternal eri s wound infection
A. 5 A M Aortoil mey 6 A liac eri bypass graft
A. 6 W M Pacem mey 4 aker eri pocket tissue
Endovascul ar and deep tissue infection
Pacemaker pocket infection
Alcoholism, vascular disease
CHF with pacemaker.
None
None
Recurrent osteomyelitis Acute IV CEPH <1mo Chronic PO PEN x5mo LTFU Acute IV CEPH x6 wks Chronic PO PEN x5mo Cure Acute Drainage and IV CEPH 1mo Chronic LTFU Acute IV PEN x1mo Chronic Not known Cure Acute IV PEN x1mo Chronic PO PEN x12mo Cure Acute Pacemaker removal Chronic none Death due to heart disease
19
A. 5 A M Bone + Metatarsal mey 0 A tissue osteomyelit eri is
Hepatitis C, alcoholism, leg trauma
Mixed anaerobes, CNS
A. isra elii
3 A F Brain 5 A biopsy
Brain abscess
ESRD, chronic granulomatous disease, DM
None
A. isra elii
6 W F Bone 6
Mandibular infection
Monoclonal gammopathy, RA lumbar laminectomy
Microaerophi lic streptococcus
A. isra elii
A. odo ntol ytic us
A. odo ntol ytic us
A. odo ntol ytic us
4 A M Lymph Pneumoniti 9 A node s, biopsy mediastinal and hilar adenitis 3 A M Ear A draina ge
4 A M Blood 0 A
4 W F Sublin 7 gual Absces s
Alcoholism and cocaine abuse
Periauricula Chronic otitis r abscess media, bilateral myringotomy tubes, developmental delay Disseminat ed disease with osteomyelit is
Abscess with cutaneous fistula
DM, ESRD after transplant, chronic osteomyelitis
Sputum: Mycobacteri um arupense
Acute CARB x7wks Chronic PO PEN >1yr Lifelong therapy planned Acute IV CARB x7wks Chronic PO TETRA LTFU Acute Debridement IV CARB x8wks Chronic PO TETRA x4wks Cure Acute IV CEPH <1mo Chronic PO PEN
Ongoing Peptostreptoc Acute occus I&D Chronic Serial surgeries and PO PEN
None
Ongoing Acute Foot amputation and IV CARB PO PEN Chronic None
Deficits Gunshot wound to Pseudomonas Acute the face aeruginosa IV CLINDA <1mo Chronic PO PEN <1mo
20
A. turi cen sis
5 A M Perior 7 A bital absces s
Infected Orbital floor orbital plate fracture, recurrent abscesses
Microaerophi llic streptococcus
Recurrent Infections Acute Hardware removal Chronic PO Bactrim x4wks PO PEN x3wks Cure Acute IV METRO x1wk Chronic PO PEN x4wks
A. turi cen sis
2 W F Abdo 2 minal Absces s
Necrotic uterus after cesarean section
Narcotic abuse
Mixed gram positive organisms
A. turi cen sis
8 A F Brain 4 A biopsy + CSF
Cerebellar abscesses
Cerebellar stroke, hydrocephalus
None
A. visc osu s
4 A F Pelvic Ruptured DM 4 A Absces appendicitis s with abscesses
A. visc osu s
7 W M Blood TIPS 7 + TIPS infection shunt
Alcoholic cirrhosis
A. bov is
6 W M Cornea Corneal 1 l ulcer scrapin g
None
None
A. spe cies
6 A M Bone 0 A
Ischium osteomyelit is
Spina bifida, chronic left sacral decubitus ulcer, DM
Streptococcu s mitis, mixed flora
A. spe
7 W M Lung 1 fluid
Empyema
Lung cancer
Eschericia coli
Mixed anaerobic flora; Microaerophi lic streptococcus None
Cure Acute IV PEN x6wks Chronic PO PEN x12mo Death Acute IV PEN x6wks Chronic PO PEN x2wks LTFU Acute IV ABX (Poor documentation) Chronic LTFU Acute AG solution <1mo Chronic TETRA x3 wks Cure Acute IV CEPH x6wks Chronic PO PEN 1yr Cure Acute Decortication, IV 21
CEPH x >1mo Chronic None
cies
A. spe cies
A. spe cies
3 W F IUD 4
8 A F Blood 1 A
Pelvic inflammati on
Blood
A. spe cies
3 A M Arm 8 A tissue
A. spe cies
4 W F Liver Liver 9 Absces abscess s
A. spe cies
A. spe cies
A. spe cies
Hardware infection
2 A F Pelvic Pelvic 4 A Absces abscess s
1 W F Pharyn geal absces s+ blood 4 W F Woun 8 d draina ge
None
Cholangiocarcinoma, DM, CAD
None
Gunshot wounds, multiple surgeries
Proprionibact erium acnes
Duodenal carcinoma, DM
None
Obesity, hyperglycemia
Pharyngeal None abscess, prevertebral extension
Orofacial wound infection
Microaerophi lic streptococcus
Head and Neck Cancer, Obesity, DM
Mixed anaerobic flora
Usual oral flora
None
LTFU Acute TETRA <1mo, IUD removal Chronic None Cure Acute IV PEN x2wks Chronic PO pen x 6mo LTFU Acute IV CEPH x6wks Chronic PO PEN LTFU Acute CARB x8wks Chronic PO PEN x12mo Cure Acute IV PEN x6wks Chronic PO PEN x6wks Cure Acute IV PEN <1mo Chronic PO PEN x4wks Cure Acute TETRA x unspecified weeks Chronic LTFU 22
A. spe cies
A. spe cies
5 A M Neck 3 A mass + Facial Absces s 7 A F Pleural 3 A fluid
Soft tissue mass and cellulitis
Alcoholism, Hepatitis B, Hepatitis C
CNS
Empyema
Acute renal failure, CHF, DM
None
Acute I&D PO PEN x6wks Chronic LTFU Acute Drainage IV PEN x1wk Chronic PO PEN 1mo
Cure Abbreviations: W, white; AA, African American; F, female; M, male; BAL, bronchoalveolar lavage; TBNA, transbronchial needle aspirate; CSF, cerebrospinal fluid; TIPS, transjugular intrahepatic porto-systemic shunt; HIV, human immunodeficiency virus; ESRD, end-stage renal disease; DM, diabetes mellitus; RA, rheumatoid arthritis; CAD, coronary artery disease; CHF, congestive heart failure; CNS, coagulase-negative Staphylococcus; IV, intravenous; I&D, incision and drainage; PO, oral; LTFU, lost to follow-up; PEN, penicillin or ampicillin; CEPH, cephalosporin; CARB, carbapenem; CLINDA, clindamycin; METRO, metronidazole; ABX, antibiotics; AG, aminoglycoside; TETRA, tetracyclines.
Figure 1. Index Case. Computerized tomography images of disease due to Actinomyces meyeri. On left, midline neck fistula connecting to necrotic anterior mediastinum. On right, cavitary mediastinal masses, scattered nodules, and pericardial fluid.
Figure 2a-c. Actinomyces species and sources of cultures in Groups 1, 2, and 3
23
Figure 1. Computerized tomography images of disease due to Actinomyces meyeri
24
Percent
Figure 2a. Group 1: Actinomycosis 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
25
Percent
Figure 2b. Group 2: Possible or Local Infections 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
Percent
Figure 2c. Group 3: Commensals 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
Skin/Soft Tissue Disseminated CNS Musculoskeletal Abdomen/Pelvic Thoracic/Pulmonary Orofacial
Species
26