Abstract
Mycoplasma hominis is a rare cause of infective endocarditis, typically reported in immunocompetent patients following valve replacement. We report the first case of post-renal transplant prosthetic valve infective endocarditis and concurrent endophthalmitis caused by M. hominis. A 47-year-old woman with prior aortic valve replacement and renal transplantation presented with fever & atrial fibrillation. She was diagnosed with culture-negative endocarditis complicated by cerebral septic emboli and visual symptoms. Plasma cell-free DNA metagenomic next-generation sequencing identified M. hominis, which was confirmed by culture of aortic abscess tissue. Management included valve replacement surgery and antibiotic therapy with doxycycline and levofloxacin. This case highlights the diagnostic challenges of M. hominis infections, the utility of advanced molecular diagnostics, and the importance of considering M. hominis in immunocompromised patients with culture-negative endocarditis. Donor and recipient screening for M. hominis in recipients with prosthetic heart valves may help prevent infection.
Introduction
Mycoplasma hominis is a cell-wall-deficient bacterium commonly found in the human oropharynx and urogenital tract. It belongs to the class Mollicutes and the family Mycoplasmataceae, which additionally includes M. genitalium and Ureaplasma urealyticum. 1 These organisms are often commensals, but can cause disease, particularly in immunocompromised patients.2,3 To the best of our knowledge, this case is the first reported instance of M. hominis causing both infective endocarditis and endophthalmitis in a renal transplant recipient.
Case
A 47-year-old woman presented to the emergency department with several days of fever and palpitations preceded by 3 weeks of diarrhea.
Her medical history included autosomal polycystic kidney disease complicated by Type A aortic dissection and end-stage renal disease. Four years back, she underwent aortic root replacement with a 23 mm On-X mechanical conduit, along with ascending aorta and hemiarch replacement using a 26 mm Hemashield graft. Three months before presentation, she received a renal transplant from a living related donor (see timeline, Figure 1). Her immunosuppressive regimen included induction with thymoglobulin and maintenance with tacrolimus, mycophenolate sodium, and prednisone. Trimethoprim-sulfamethoxazole was prescribed for Pneumocystis prophylaxis, and valganciclovir for cytomegalovirus prophylaxis per our institutional post-transplant protocol. Her post-transplant course was complicated by poor wound healing and serosanguinous drainage at the renal transplant site. Ultrasound revealed a small perinephric fluid collection that was not amenable to drainage. The wound improved with topical care and resolved completely. Given low clinical suspicion for infection, cultures were not obtained.
Timeline of events.
At the time of presentation to the emergency department, she had no fever or leukocytosis. Physical examination was notable for atrial fibrillation and a grade IV systolic murmur, loudest in the right upper sternal border. There were no other peripheral stigmata of infective endocarditis.
At admission, blood cultures were drawn, and a transthoracic echocardiogram was performed, which showed a mild paravalvular leak along the anterior aspect of the mechanical aortic valve with a filamentous hypermobile echodensity adherent to the ventricular side of the valve. Three additional sets of blood cultures were collected prior to empiric antibiotic initiation with vancomycin and cefepime. All blood cultures remained without growth.
Transesophageal echocardiogram showed normal left ventricular ejection fraction and confirmed the presence of a vegetation at the base of the mechanical valve measuring 1.5 × 0.3 cm with increased thickening of the aortic root suggestive of an evolving abscess. Screening MRI of the brain on hospital day 8 demonstrated diffuse punctate infarcts throughout the supratentorial and infratentorial parenchyma consistent with septic emboli.
No neurologic sequelae of embolization were noted until day 10 of admission, when the patient developed acute central vision loss in the left eye with central scotoma. Fundoscopic exam revealed central and perifoveal infectious chorioretinal infiltrates with associated subretinal hemorrhage, suggestive of infectious chorioretinitis. The right eye was unaffected. Intravitreal injections of vancomycin (1 mg), ceftazidime (2 mg), and voriconazole (100 mcg) were performed. A vitreous tap was performed, and the specimen was sent to the University of Washington for broad-range bacterial polymerase chain reaction (PCR). However, the nucleic acid extraction was below the limit of detection, so neither broad-range PCR nor sequencing was able to be performed on the vitreous fluid.
Additional microbiological tests for Brucella, Bartonella, Coxiella, Chlamydia, Tropheryma whipplei, Candida, Legionella, Aspergillus, Cryptococcus, and Histoplasma were negative. A plasma sample was sent for metagenomic next-generation sequencing (mNGS) of cell-free DNA (cfDNA, Karius) on day 11 of admission.
On day 13 of admission, the patient underwent aortic root replacement with bioprosthetic aortic valve replacement and hemiarch replacement. Purulent material was observed in the graft. The aortic root was noted to be grossly destroyed with a circumferential abscess cavity underneath the aortic valve, and the aorto-mitral curtain and interventricular septum were observed to be completely destroyed. Most of the infected graft material was removed except for a few centimeters retained at the distal aortic anastomosis. She also had a dual-chamber permanent pacemaker placed for complete heart block.
Tissue samples were sent from the aortic graft and the aortic root abscess for routine, fungal, anaerobic, and acid-fast bacillus stains and cultures. The mNGS, which had been sent preoperatively, returned positive on hospital day 15 for M. hominis, although the result was not quantifiable. Doxycycline 100 mg twice daily was initiated. Intravitreal injection of clindamycin was also performed at this time for Mycoplasma endophthalmitis. In view of the mNGS result, Mycoplasma and Ureaplasma PCR testing, as well as specialized cultures on the aortic abscess tissue, were ordered. Samples were inoculated in A8 agar and 10B broth (Remel Inc., San Diego, CA, USA). The broth was incubated aerobically at 35°C, while the agar was incubated at 35°C in 5% CO2 per the manufacturer’s packet insert. Cultures were read daily for the detection of turbidity or color change in the broth or the presence of colonies on the agar. After 5 days of incubation, cultures from the aortic root abscess also returned positive for M. hominis (see Figure 2(a) and (b)). Mycoplasma PCR of aortic tissue was also positive for M. hominis.
Mycoplasma culture results for this case: (a) Fried egg appearance of Mycoplasma colonies when viewed under a microscope. (b) Urogenital Mycoplasma growth is indicated when the liquid medium turns from yellow to red (alkaline).
Based on prior case reports of M. hominis endocarditis, levofloxacin 500 mg daily was added to her ongoing doxycycline 100 mg twice daily regimen. She was planned for 6 weeks of dual therapy followed by indefinite suppressive doxycycline due to her immunosuppression and the presence of unresectable graft material. Susceptibilities were tested at the University of Alabama, Birmingham Diagnostic Mycoplasma Laboratory on an isolate from the aortic tissue, which was found to be susceptible to clindamycin, tetracycline, and moxifloxacin.
At the 3-month post-hospital discharge follow-up, she was doing well with resolution of ocular symptoms with a small juxtafoveal subretinal scar in the left eye.
Discussion
The prevalence of M. hominis colonization is reported to be 21%–53%. 1 While most patients with genitourinary carriage do not have symptoms, this organism can cause severe disease in the immunocompromised. There are several cases of both genitourinary (pyelonephritis, graft infection, perinephric abscess) and extra-genital (septic arthritis, pneumonia, meningitis) infections in solid organ transplant recipients.2–5 A notable presentation is hyperammonemia, which is predominantly observed in lung transplant recipients infected with M. hominis, though it is not exclusive to this group. 6 The source of disseminated M. hominis infections in solid organ transplant recipients is suspected to be translocation of endogenous flora or donor-derived infection.1,7 There are at least 13 cases in the literature of endocarditis attributed to M. hominis, typically in immunocompetent patients who have undergone valve replacement (Table 1).8–20 The onset is often within 1 year of valve replacement, suggesting perioperative infection.
Cases of M. hominis infective endocarditis reported in the literature.
ICD, implantable cardioverter-defibrillator; mNGS, metagenomic next-generation sequencing; PCR, polymerase chain reaction.
Our case is unique as it is the first case of endocarditis described following a solid organ transplant. The etiology of M. hominis infection in our patient was unclear, with both donor-derived and endogenous sources considered. The delayed surgical site healing raises the possibility of an undiagnosed perinephric M. hominis infection that subsequently disseminated, making donor-derived transmission the most likely explanation. However, we were unable to test the kidney donor (the patient’s sister) for M. hominis carriage to confirm this hypothesis.
To our knowledge, this is the first reported case of M. hominis causing an ocular infection. Although broad-range PCR of the vitreous fluid did not confirm M. hominis microbiologically, the simultaneous occurrence of cerebral septic emboli from valvular vegetations strongly suggests it caused the chorioretinal infiltrates. Furthermore, the retinal infiltrates improved after starting clindamycin, supporting M. hominis as the cause.
This case also demonstrates the potential benefit of plasma mNGS for the diagnosis of M. hominis infection. Newer techniques, such as mNGS, offer a more agnostic diagnostic approach than culture or PCR. mNGS has been used to diagnose M. hominis infection by the testing of body fluid, tissue, and cerebrospinal fluid, and some data suggest that mNGS of plasma cfDNA could be helpful in eliciting the cause of chorioamnionitis.14,21–23 Our case is the first reported instance where plasma cfDNA was used to diagnose an M. hominis infection. This result prompted us to pursue confirmation and susceptibility testing with Mycoplasma/Ureaplasma-specific culture and PCR.
Correct diagnosis of M. hominis is challenging but important to guide therapy. As it lacks a cell wall, this organism is intrinsically resistant to beta-lactams and other cell-wall active agents. Tetracyclines are effective in most instances; however, resistance has been documented. 1 Fluoroquinolones, clindamycin, and macrolides typically have activity as well. In milder infections, monotherapy with doxycycline or azithromycin is recommended, but in more severe infections, combination therapy with at least two agents is usually initiated. 1 In this case, valve function was restored with surgical revision. However, heart transplantation has been performed as a therapeutic measure in two reported cases of M. hominis endocarditis.8,18
Culturing Mycoplasma hominis is challenging for several reasons. First, mollicutes are highly susceptible to adverse environmental conditions, which can impair recovery if transport is delayed or suboptimal media are used. 24 Recommended transport media include UTM, UVT, M4, M5, and M6. Second, culture requires specialized media, as M. hominis does not grow in standard blood culture bottles due to the presence of sodium polyanethol sulfonate, which inhibits its growth. Instead, it requires arginine-containing media such as Shepard’s A8 and 10B, SP4, or PPLO. 25 These media also include antibiotics to suppress other microbes—for example, A8 and 10B contain cefoperazone, with A8 additionally including penicillin and amphotericin B. Since few clinical laboratories perform these specialized cultures, they are usually sent to reference labs, prolonging transport time and further delaying diagnosis and reducing recovery likelihood. Third, mollicutes grow slowly, which also delays diagnosis. Fourth, they lack a cell wall, so identification cannot rely on Gram staining. In our case, culture detection was based on microscopic observation of colonies with the classic “fried egg” appearance on A8 agar and a colorimetric change in 10B medium due to arginine hydrolysis by Mycoplasma spp.
PCR offers several advantages in diagnosing Mycoplasma hominis infection, including higher sensitivity and specificity. 26 However, its utility is limited by the lack of commercial NAAT assays, with most PCRs being laboratory-developed tests available only at reference or specialized laboratories—mirroring the limited accessibility of mollicute-specific cultures. Furthermore, testing must be specifically requested based on clinical suspicion, increasing the risk of missed diagnoses.
In conclusion, M. hominis infection should be considered in cases of culture-negative endocarditis. While a prior study of pre-transplant screening for mollicutes did not show a clear benefit, 27 it is possible that screening and treatment of M. hominis in solid organ transplant donors and recipients may be of benefit if the recipient has a prosthetic heart valve. This case also illustrates the utility of plasma cfDNA sequencing in the diagnosis of challenging infections.
