Idiopathic hyperammonemia is a rare complication with a high mortality rate that occurs in persons with hematologic malignancies or hematopoietic stem cell or solid organ transplant. Patients present with encephalopathy and hyperammonemia in the absence of liver disease or inborn errors of metabolism. Several etiologies have been proposed, including chemotherapeutic agents, medications, and a catabolic state with an elevated nitrogen load in the setting of acute illness. Recently, cases of hyperammonemia in adult lung transplant recipients have been attributed to infection from Ureaplasma parvum or U urealyticum. Herein, we report a 12-year-old girl with acute myeloid leukemia and neutropenic fever who developed acute encephalopathy. Laboratory testing revealed severe hyperammonemia (blood ammonia level >1609 μmol/L) with normal liver function studies. U parvum was detected in blood, urine, and respiratory specimens by polymerase chain reaction testing. After antibiotic therapy directed against U parvum, blood ammonia levels normalized, the infection was eradicated, and the patient recovered. We propose that clinicians should test for invasive infection from Ureaplasma species in immunocompromised children with unexplained hyperammonemia.
Hyperammonemia is a rare complication with high mortality described in adults and children with hematologic malignancies or hematopoietic stem cell or solid organ transplant.1,–4 Patients present with acute hyperammonemic encephalopathy in the absence of liver disease or inborn errors of metabolism. This syndrome, frequently called “idiopathic hyperammonemia” (IHA), was initially described by Watson et al5 in 1985. The etiology of IHA is unclear and multiple etiologies have been proposed, including chemotherapeutic agents (eg, asparaginase, 5-fluorouracil, or cytarabine), medications (eg, valproic acid, barbiturates), and an increased catabolic state and/or increased nitrogen load during an acute illness.1,2 Recently, systemic infections from Ureaplasma parvum or U urealyticum have been determined to cause hyperammonemia in adult lung transplant recipients.6 Ureaplasma species, commensal bacteria of the urogenital tract that produce ammonia when hydrolyzing urea for energy production, can cause invasive infections in immunocompromised hosts.7
We describe the case of a 12-year-old girl with leukemia who developed hyperammonemic encephalopathy due to a disseminated infection from U parvum. To the best of our knowledge, this is the first published pediatric case of hyperammonemia due to Ureaplasma infection.
In November 2017, a 12-year-old girl with T-cell acute lymphoblastic leukemia, in remission and undergoing maintenance chemotherapy, was found to have treatment-associated acute myeloid leukemia. She was hospitalized to receive induction chemotherapy consisting of clofarabine and high-dose cytarabine. On hospital day 36, she was started on a second round of induction chemotherapy because of persistent disease; the regimen consisted of high-dose cytarabine, mitoxantrone, and gemtuzumab ozogamicin. On hospital day 64, bone marrow evaluation revealed a failure to induce remission. During this time of profound cytopenia, she experienced multiple brief episodes of fever, prompting the administration of short courses of empirical broad-spectrum antibiotics. Additionally, because of chemotherapy-induced hyperemesis, she was started on parenteral nutrition.
On hospital day 62, the patient developed a new fever and was empirically started on cefepime. Because of daily fevers, computed tomography of the chest and abdomen was performed on hospital day 66, which revealed new pulmonary nodules, prompting the initiation of empirical voriconazole. On hospital day 68, the patient had an acute change in her mental status. Initially, she appeared confused and had trouble answering questions. This progressed to excessive sleepiness, and then she became increasingly difficult to arouse. Emergent computed tomography of the head was normal. An ammonia level was obtained revealing a level that exceeded the laboratory’s upper level of detection (>1609 μmol/L; reference range: 18–72 μmol/L). Liver function tests were found to be normal. Plasma amino acids and urine organic acids were sent and later proved normal.
On admission to the PICU, she had a Glasgow coma scale score of 9. She was sedated and intubated. Given her critical state and because the etiology of the severe hyperammonemia was unclear, we pursued multiple strategies to reduce blood ammonia levels rapidly. She was placed on high-dose continuous venovenous hemodiafiltration to remove excess ammonia. Sodium benzoate, arginine, and carnitine were started. To decrease endogenous ammonia production, we initiated rifaximin and enteral lactulose. Parenteral nutrition was stopped, and we began administration of dextrose-containing fluids. We also used hypertonic saline and implemented therapeutic hypothermia.
On hospital day 69, we sent blood and urine samples for polymerase chain reaction (PCR) testing for Ureaplasma species (U parvum and U urealyticum) and Mycoplasma species (Mycoplasma hominis and M genitalium). Doxycycline was then added for empirical coverage. During the next 48 hours, the patient’s ammonia levels decreased to 785 μmol/L before quickly rebounding to previous levels (>1609 μmol/L) (Fig 1).
On hospital day 71, we were notified that U parvum was detected in specimens from the patient’s urine (Associated Regional and University Pathologists Laboratories, Salt Lake City, UT) and blood (Mayo Clinic Laboratories, Rochester, MN). Azithromycin and levofloxacin were added to broaden the antibiotic regimen. Over the subsequent 48 hours, the patient’s blood ammonia levels decreased to as low as 265 μmol/L. When ammonia levels began to rise again, the patient was switched to a different fluoroquinolone because available susceptibility data suggest that moxifloxacin may have a lower rate of resistance.8,9 Eventually, the patient’s ammonia levels declined to normal levels by hospital day 77.
U parvum was also detected by PCR testing in a tracheal aspirate specimen (Mayo Clinic Laboratories). However, attempts to culture the organism from urine and respiratory specimens were unsuccessful. Serial PCR testing revealed microbiologic clearance of urine and blood on hospital day 72 and 85, respectively. After a therapeutic response was noted, doxycycline and azithromycin were stopped. Moxifloxacin was continued during the remainder of the patient’s hospital stay.
By hospital day 76, the patient was extubated, had a Glasgow coma scale of 15, and was conversing appropriately. The patient was hospitalized for another 3 weeks and was ultimately discharged with treatment-refractory acute myeloid leukemia to receive outpatient palliative chemotherapy. After discharge, she lived an additional 2 months, ultimately dying from complications of leukemia.
IHA is a rare complication that occurs with hematologic malignancy or after hematopoietic stem cell or organ transplant. In published series, authors describe a mortality rate of >65%.1,4,10 Recent evidence supports a causal relationship between disseminated infection from Ureaplasma species and hyperammonemia in immunocompromised persons. Bharat et al6 detected U parvum or U urealyticum in 6 adult lung transplant recipients with hyperammonemic encephalopathy. The organisms were discovered in clinical specimens from blood and bronchoalveolar lavage (BAL) fluid. Antibiotic therapy directed against Ureaplasma species was administered to 3 patients and resulted in clinical improvement and normalization of blood ammonia levels. One patient experienced a subsequent episode of hyperammonemia concurrent with a relapse of Ureaplasma bacteremia. In postmortem studies in 3 patients, evidence of disseminated infection was found, identifying Ureaplasma species in the lung, liver, bladder, and spleen. When the investigators tested BAL fluid and serum in 20 lung transplant patients without hyperammonemia, they did not encounter any indication of Ureaplasma infection. Further support of a causal link is found in mouse models in which immunocompromised mice challenged with U parvum or U urealyticum develop hyperammonemia.11,12
Ureaplasma species are small pleomorphic bacteria that lack a cell wall. They produce adenosine triphosphate from the hydrolysis of urea, which results in the formation of ammonia and carbon dioxide. The urogenital tract is the primary reservoir. Colonization has been reported in 40% to 80% of healthy women.13 Ureaplasma species has been associated with adverse pregnancy outcomes, including pregnancy loss and preterm birth.14 Additionally, perinatal Ureaplasma exposure may contribute to chronic lung disease of prematurity.13 Invasive disease, including pericarditis, mediastinitis, septic arthritis, meningitis, brain abscess, and renal abscess, have been reported in immunocompromised children and adults.15,–17
Ureaplasma species are fastidious organisms that require specialized, urea-containing media and are particularly susceptible to dryness and temperature changes during processing. Because they lack a cell well, they are not visible by Gram-stain. Given the challenges in growing Ureaplasma species by culture, rapid and sensitive real-time PCR assays were developed.18
There are few therapeutic options for Ureaplasma infections. Antibiotics active against Ureaplasma species include tetracyclines, fluoroquinolones, and azithromycin. Additionally, there are limited data on antimicrobial susceptibility. Studies to date suggest variable rates of resistance to doxycycline and fluoroquinolones, whereas macrolide resistance is uncommon.19,20 It is not clear why doxycycline monotherapy was ineffective for our patient. One possibility is the infection was due to a doxycycline-resistant strain. Alternatively, tetracyclines may be less effective for invasive Ureaplasma infection. The optimal therapy for invasive Ureaplasma infection in immunocompromised patients is uncertain. Authors of previous cases describe therapeutic success with azithromycin monotherapy, levofloxacin and doxycycline, and levofloxacin and azithromycin.6,16,21 Notably, it is reported that Ureaplasma infection and the accompanying hyperammonemia can relapse after stopping antibiotics.6,21
The etiology of IHA has eluded clinicians. Multiple factors have been proposed, including chemotherapeutic agents, steroids, sepsis, mucositis, gastrointestinal bleeding, dehydration, and increased nitrogen load from parenteral nutrition.10 Although our patient had some of these features (ie, exposure to chemotherapeutics and parenteral nutrition), we submit that she developed hyperammonemia because of the Ureaplasma infection. Because 2 commercial laboratories independently detected U parvum from multiple sites, we were able to corroborate the microbiologic results and establish that the patient had an invasive, widespread infection. Additionally, after starting antibiotics directed against U parvum, there was a notable decline in blood ammonia levels that paralleled clearance of the pathogen from urine and blood specimens. When combined with the aforementioned studies revealing a causal relationship between hyperammonemia and Ureaplasma species, there is compelling evidence that the U parvum infection caused the elevated ammonia levels in our patient.
M hominis, also a urogenital pathogen, metabolizes arginine to ammonia. Authors of a recent report described an adult lung transplant recipient with hyperammonemia and disseminated M hominis infection.22 Subsequently, after the patient’s death, residual blood and BAL fluid were tested and were positive for both M hominis and U parvum.6 Although the relationship between M hominis and hyperammonemia is less clear, clinicians should consider testing for M hominis when evaluating a patient for unexplained hyperammonemia.
We propose that invasive Ureaplasma infection is a cause of IHA in immunocompromised children. Because the organism cannot grow using routine microbiologic cultures, it may go overlooked by clinicians during the evaluation of hyperammonemia. Our case supports testing for Ureaplasma infection in immunocompromised children who develop unexplained hyperammonemia. Antibiotics directed against Ureaplasma species should be initiated while microbiologic tests are pending.
Drs Smith and Crews drafted the initial manuscript, collected data, and reviewed and revised the manuscript; Drs Cheek and Srivastava reviewed and revised the manuscript; Dr Appachi supervised data collection and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspect of the work.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.