FormalPara Key Summary Points

Mycoplasma hominis can cause serious disseminated extragenital infections in patients with hypogammaglobulinemia, also after B-cell-depleting therapies including CAR-T.

In patients with disseminated infection in which no pathogen grows on common culture media, Mycoplasmataceae should be considered and specific testing performed with molecular methods and selective culture media.

Doxycycline, azithromycin, and fluoroquinolone, in monotherapy or in combination, have been successfully used in 13 of 15 cases identified through literature review.

In our two cases, diagnosis was obtained thanks to biopsies cultured on standard media, and PET-CT scan was used to guide the length of therapy.

Introduction

Mycoplasma spp. is a member of the Mycoplasmataceae family, class Mollicutes [1, 2]. M. hominis can be found in the vagina or cervix in approximately 20–50% of sexually mature women [3]. It is part of the normal human urogenital tract microbiome, where it resides on the mucosal surface [4].

In the male and female genital tract, M. hominis can be either a colonizing pathogen or the causal agent of infections, such as urethritis, bacterial vaginosis, cervicitis, and pelvic inflammatory syndrome. M. hominis can also be associated with infertility in both men and women [1]. Given the high rate of colonization, screening in asymptomatic men and women for Mycoplasma spp. is not recommended, and testing should be performed only in cases of symptoms of genital infection [5,6,7].

Although rare, extragenital infections due to M. hominis have been reported, especially in immunocompromised patients. They comprise mediastinitis, septic arthritis, osteomyelitis, prosthetic joint infection following surgery and trauma, abscess after transplant or surgical procedures, central nervous system infections (mainly post-surgical and post-trauma brain abscess or meningitis), infective endocarditis, and bacteremia [1]. The first three reported series of extragenital M. hominis infections by Madoff et al. in 1988 [8], McMahon in 1990 [9], and Meyer et al. in 1993 [10] reviewed 36, 13, and 67 cases, respectively. They found that the main risk factor for extragenital infection was immunosuppression, followed by surgery, trauma, and other conditions. Among 32 patients with immunosuppression, two had hypogammaglobulinemia, both due to congenital deficiency [10]. Other authors confirmed that Mycoplasma spp. could be a frequent cause of infection in patients with hypogammaglobulinemia, but most of the infections were respiratory and species other than M. hominis (mainly M. pneumonia) were detected [11].

In this report, we want to focus on extragenital M. hominis infections in patients with hypogammaglobulinemia, also caused by recently introduced therapeutic options such as CAR-T therapy, reporting two cases from our center and providing a literature review.

Cases Description

Case 1

A 61-year-old woman with diffuse large B cell lymphoma underwent eight cycles of first-line plus three cycles of second-line chemotherapy; both first- and second-line regimens included rituximab. As after chemotherapy she experienced disease progression, she underwent chimeric antigen receptor T cell therapy (CAR-T with Yescarta®) on 11/08/2021 (day 0). Of note, as comorbidity she had severe osteoporosis complicated by multiple vertebral fractures.

On day + 20, the patient complained about lumbar pain, first ascribed to multiple already known fractures. At that time, her immunoglobulin level was low [IgG 2.200 g/L (normal value 8.000–17.000), IgA 0.225 g/L (normal value 0.700–4.000), IgM 0.031 g/L (normal value 0.400–2.300)] as a consequence of prolonged rituximab therapy.

On day + 30, she underwent fluorodeoxyglucose–positron emission tomography (FDG-PET) study (Fig. 1), which showed complete metabolic response in sites of lymph nodal disease but revealed elevated high tracer uptake at the first (L1) and second (L2) lumbar vertebrae [standardized uptake value (SUV) max 16.2]. On day + 42, lumbar magnetic resonance imaging (MRI) was performed, showing structural disruption by pathological tissue at multiple levels and suspected pathological fracture at L1–L2. On day + 49, vertebral biopsy and vertebroplasty were performed. Histological examination revealed inflammation and excluded neoplastic process. Biopsy culture grew M. hominis on sheep blood agar plates and identification was performed with matrix-assisted laser desorption–ionization time of flight (MALDI-TOF) after 2 days of incubation at 37 °C in aerobic conditions. Considering histological characteristics and microbiological results, on day + 56, levofloxacin 750 mg every 24 h plus doxycycline 100 mg every 12 h were prescribed.

Fig. 1
figure 1

Baseline (A) and control (B) PET performed in the first patient. A FDG baseline PET performed before the start of antibiotic therapy; arrows mark tracer hyperaccumulation at the soma of the L1 and L2 vertebrae (SUV max 16.2). B control PET performed after 8 weeks of antibiotic therapy; tracer hyperaccumulation no longer evident at L2 soma level

Full size image

Polymerase chain reaction (PCR) (day + 59) for Mycoplasmataceae showed urinary and vaginal colonization by M. hominis and Ureaplasma urealyticum, not associated with genital symptoms of infection. Since Mycoplasma can cause infections of cell cultures, we reviewed the summary of critical quality attributes of the patient’s CAR-T product and found that it had tested negative by PCR for Mycoplasma spp.

A follow-up FDG-PET performed after 3 and 8 (Fig. 1B) weeks of antibiotic therapy documented first reduction and later almost complete resolution of spondylodiscitis, with FDG hyperaccumulation no longer evident at L2 soma, but the relapse of lymphoma was present. PCR search of Mycoplasmataceae was negative for M. hominis and U. urealyticum on both urine and vaginal swabs (day + 88).

The combination antibiotic therapy for Mycoplasma spondylodiscitis with levofloxacina and doxycycline was well tolerated but interrupted at 10 weeks (instead of the planned 12 weeks), as the patient expired due to lymphoma progression on day + 128 after CAR-T therapy.

Case 2

A 50-year-old woman with congenital hypogammaglobulinemia (previously on replacement therapy) presented at the Emergency Department with a 3-month history of malaise, asthenia, back pain, and ankle pain. She was admitted to a different hospital (day 0, 27/12/2022), where she underwent several radiological assessments that revealed multiple abscesses localized in pelvic and perineal muscles (the largest 5 × 8 × 1.6 cm), pleural empyema, and right ankle arthritis. All the cultures performed (of blood, pleural liquid, glute abscess fluid, hip arthrocentesis fluid, and ankle arthrocentesis fluid) came up negative. At that time, all the immunoglobulin classes were below the lower limit of detection since the patients had not recieved immunoglobulin replacement therapy for several months.

As fever, pain, and inflammation persisted, empirical broad-spectrum antimicrobial therapy was started: first, with daptomycin plus metronidazole and meropenem, and then with ceftaroline, fosfomycin, ceftolozane/tazobactam and caspofungin.

Clinical conditions did not improve and generalized malaise with fever persisted. The patient was therefore transferred to our regional reference center for infectious diseases (day + 42), where antimicrobial therapy was discontinued. Cultures of purulent wounds (localized at the site where previous thoracic and gluteal drainages were placed) were performed: all three grew M. hominis after 3 days of incubation (day + 45); molecular analyses through PCR of thoracic wound sample confirmed the presence of M. hominis (day + 45).

PCR screening for Mycoplasmataceae (day + 52) showed urine colonization by U. urealyticum and rectal colonization by M. hominis, not associated with local symptoms of infection.

Based on microbiological findings, on day + 45, levofloxacin 750 mg every 24 h was started.

A baseline FDG-PET performed on day + 48 (Fig. 2) showed high tracer uptake within the organized pleural effusion (SUV max 8), pelvic abscesses (SUV max 11), and right tibia-calcaneal joint (SUV max 8). As the patient was persistently confused since her admission, a brain MRI was performed (day + 51), showing mild signal hyperintensity in T1 along the profile of supratentorial ventricular system, suspected of micro-abscess. The cerebrospinal fluid chemical-physical analysis (day + 58) was significant only for moderate damage of blood–brain, while culture and PCR for Mycoplasmataceae were negative.

Fig. 2
figure 2

Baseline (A) and control (B) PET performed in the second patient. A Baseline PET performed before the start of antibiotic therapy. High tracer uptake in pelvis (A1), at internal obturator muscles of both sides (A2) and on the left extending to the external obturator muscle and great adductor muscle (A3) (SUV max 11). Concomitant high FDG uptake at the right tibio-calcaneal joint extending to the lateral aspect of the calcaneus and plantar region (A4) (SUV max 8). B Control PET performed after 6 weeks of antibiotic therapy. Marked reduction in size and hypermetabolism compared with previous control and suspicious for proportion of active disease at pelvis (B1). Residual modest tracer hyperaccumulation at the internal obturator muscles (B2) and external obturator muscle and great adductor muscle (B3) (SUV max 5). These findings are reported as borderline regarding diagnostic significance. No more significant accumulation evident at the right tibio-calcaneal joint (B4)

Full size image

The fever resolved after 72 h of levofloxacin therapy, and ankle pain and swelling improved rapidly. Levofloxacin was continued for a total of 6 weeks, after which (day + 84) persistent rectal colonization by M. hominis was documented, while U. urealyticum was no longer detected in urine.

A second FDG-PET performed at the end of antibiotic therapy (day + 90) (Fig. 2) revealed the marked reduction of tracer uptake in the internal, external, and great adductor muscles (sites of pelvic abscesses at baseline, SUV max 5), and complete resolution of uptake in the ankle. During subsequent hospital admissions a year later, there was no relapse of M. hominis infection.

Microbiological culture records from out institution from the years 2014–2023 did not reveal any additional cases of extragenital M. hominis infection.

Literature Review

In Table 1, we summarize our 2 and 14 previously reported clinical cases of extragenital M. hominis infections in patients with hypogammaglobulinemia. We have included clinical cases reporting data on the cause of hypogammaglobulinemia, age, sex, diagnosis, treatment, and outcomes, obtained through a MEDLINE search with the keywords Mycoplasma, hypogammaglobulinemia, CAR-T, HSCT, rituximab, and anti-CD20 was performed. Papers’ references were screened for additional cases.

Table 1 Extragenital Mycoplasma hominis infections in patients with hypogammaglobulinemia
Full size table

There was a similar distribution of iatrogenic and congenital causes of hypogammaglobulinemia (7 vs. 8). There was an osteoarticular involvement in 13 cases and the presence of abscesses in 4 cases. In 7 cases, M. hominis grew on a standard medium and diagnosis was confirmed by PCR or culture in selective medium, while a common medium culture was negative in the other 9, and diagnosis was made by culture on selective media or PCR.

Combination therapy was administered in 8 patients. Improvement or total resolution was obtained in 13 patients after a median length of treatment of 8 weeks, relapse occurred in 1 patient treated with combination therapy, and 2 patients died early.

Discussion

We have described two cases of severe extragenital infections due to M. hominis in patients with hypogammaglobulinemia, and we report the first case of M. hominis osteoarticular infection in a patient after CAR-T cell therapy. Indeed, a single case of lethal hyperammonemia related to Ureaplasma pneumonia in a CAR-T cell recipient has been reported, but no data on immunoglobulin levels were provided and no other cases of severe infections by the Mycoplasmataceae family heve been reported in CAR-T recipients [12].

Hypogammaglobulinemia can be associated with congenital (primary) [13, 14] or acquired (secondary) immunodeficiencies, such as drug-induced. Rituximab, which is a monoclonal antibody targeting the CD20 antigen expressed on B cells, can cause hypogammaglobulinemia usually for more than 6 months after administration [15]. In the 16 reported cases, the distribution of primary and secondary cases was almost equal.

It has been demonstrated that patients with hypogammaglobulinemia are more susceptible to Mycoplasma spp. mucosal colonization, and this could be explained by the lack of protective antibodies on the mucosal surface [13]. The possible higher mucosal colonization rate along with the lack of antibodies may also explain the increased risk of disseminated infection originating from urogenital tract in these patients [13, 16]. Regarding arthritis, in vitro study has suggested how a minor joint trauma can attract neutrophils that, in the absence of specific antibodies, can uptake Mycoplasma spp. into phagocytic vacuoles, which can be released into joint spaces leading to septic arthritis [17,18,19].

Both our patients had an osteoarticular involvement (in the second case, associated with other localizations of infection). This is in line with the literature (Table 1), in which osteoarticular involvement was reported in 81% of cases, confirming Mycoplasma spp. tropism for the joint.

In 7 of 16  identified cases (Table 1), diagnosis was made by culture on common media. This was possible since M. hominis is the only human-pathogenic species of Mycoplasma able to grow on blood or chocolate agar [20]. In our cases, as probably also in many others, the growth of M. hominis on a common culture medium was a stroke of luck, since we did not suspect Mycoplasma to be the possible cause of infection. In the absence of clinical suspicion, the difficulty in growth on common culture medium of other members of the Mycoplasmataceae family may lead to the underdiagnosis of infections due to these microorganisms. Indeed, the diagnosis of most Mycoplasma infections requires specific tests and the use of selective media for isolating Mycoplasma from clinical samples. Mycoplasma isolation in culture is often challenging due to the slow growth of small colonies. These microorganisms should be suspected if the isolate grows on agar medium in pinpoint-sized colonies (diameter, 0.2 mm) after 2–7 days of incubation and the Gram stain is negative (due to the lack of cell walls) [20]. Identification to the species level can obtained using MALDI-TOF [21]. Molecular methods, like PCR, provide an alternative for diagnosis of Mycoplasma spp. infections, both genital and extragenital. PCR allows the diagnosis of the disease by detecting the organism directly from clinical samples, and can also be used to identify suspect colonies [22]. The sensitivity and specificity of culture on selective media are reported to be 70% and 100%, respectively; the sensitivity of culture on common media is even lower. Compared to selective culture methods, PCR has demonstrated a sensitivity of 87–96% and a specificity of 87–100%, depending on the site of sample collection [23, 24]. Other diagnostic tests such as serology and cold agglutinins are no longer used.

Because Mycoplasma lacks a cell wall, it is intrinsically resistant to all antibiotics targeting the cell wall, such as ß-lactams [20]. Guidelines for the treatment of genital infection recommend doxycycline or azithromycin as first-line therapy [25,26,27], while fluoroquinolone and clindamycin are alternative therapeutic options. However, no recommendations on the treatment of extragenital infections are available, in particular on the use of monotherapy versus combination therapy or on the length of treatment in immunocompromised patients or in case of osteo-articular involvement.

In the identified cases (Table 1), combination therapy was preferred in 9 of 15 cases with data available (56%), and only one case of relapse was reported in a patient treated with a combination therapy. Combination therapy might be preferable in case of multi-organ involvement or when susceptibility testing results are not available, due to the possibility of infection by resistant strains. In particular, in studies performed around 2010, 10% of M. hominis strains were found to be resistant to doxycycline, 98% to azithromycin, and 8% to ciprofloxacin. Among fluoroquinolones, moxifloxacin demonstrated better in vitro activity against the Mycoplasmataceae compared to other molecules of the same antibiotic class [28, 29].

The length of antibiotic therapy varied significantly, from 3 weeks to long-term therapy (Table 1), but the median length of therapy in successfully treated patients was 8 weeks. In our cases, PET-CT scan at baseline and during treatment was performed to guide its length, as suggested in spondylodiscitis and other infectious settings [30, 31].

In our cases, we preferred a combination therapy of levofloxacin and doxycycline in the first patient considering severe immunodeficiency after CAR-T and bone involvement, while, in the second, less immunocompromised, patient with soft tissue infection and possible arthritis, monotherapy with levofloxacin was successfully administered. The length of therapy was 10 weeks in the first patient and 6 weeks in the second, with PET-CT scan negative after 8 and 6 weeks of therapy, respectively.

The limitations of our paper include the retrospective identification of the cases and the long observation period (from 1981 to 2023) of identified cases. Moreover, among causes of immunosuppression, we focused only on patients with reported hypogammaglobulinemia, which might also have been present but not reported in other immunocompromised patients. For instance, in recent papers of the importance of M. hominis and Ureaplasma respiratory infections in lung transplant recipients, a high rate of respiratory infections with Mollicutes was reported, but no data on IgG levels were provided [32,33,34].

Conclusion

Although it has been reported since 1980 that patients with hypogammaglobulinemia are at increased risk of infections caused by atypical pathogens like the Mycoplasmataceae, it is surprising that only a few cases of extragenital M. hominis infections have been reported in the last decades, despite an increase in B cell-depleting therapies, including CAR-T. Underdiagnosis might play a significant role as demonstrated by the fact that half of the cases were diagnosed thanks to positive results of cultures performed on standard media. Therefore, it is important that, in patients with hypogammaglobulinemia and persistent infections, particularly if disseminated or with bone/joint involvement, and negative standard culture, Mycoplasma infections are suspected. In such cases, molecular methods should be promptly used for correct diagnostic work-up and successful therapy. Prospective studies with molecular diagnostic protocols might provide much needed insight into the incidence of such infections.