Persistent Hypoxemia in a COVID-Positive Patient With Systemic Lupus Erythematosus
A 17-year old girl with systemic lupus erythematosus (SLE) and class IV lupus nephritis presents to the emergency department with lower extremity and periorbital edema, which is concerning for a lupus flare. Medical history also includes chronic leukopenia, microcytic anemia, anxiety, and gastroesophageal reflux disease (GERD). She has been taking mycophenolic acid at home but reports missing most doses of methylprednisolone, dapsone, and ferrous sulfate during the last 2 weeks. She is afebrile with a heart rate of 93 beats/min, blood pressure 128/69 mm Hg, and oxygen saturation (SpO2) 97%. Laboratory studies are notable for white blood cells 4400 μL−1 (4.40 × 109 L−1; reference range 4500–10 000 μL−1), hemoglobin 9.5 g/dL (95 g/L; reference range 12.0–14.6 g/dL), platelets 206 000/μL (206 × 109 L−1; reference range 160 000–360 000/μL), potassium 5.4 mEq/L (5.4 mmol/L; reference range 3.6–5.1 mEq/L), phosphorus 5.7 mg/dL (1.84 mmol/L; reference range 2.4–4.7 mg/dL), albumin 1.1 g/dL (11.0 g/L; reference range 3.5–4.8 g/dL), creatinine 1.19 mg/dL (90.74 μmol/L) from a baseline of 0.5 mg/dL (38.13 μmol/L), and urine protein 300 mg/dL on urinalysis. She tests positive for COVID-19, and she recalls having congestion and a cough 10 days prior. She is admitted for intravenous diuretics, albumin, and methylprednisolone 30 mg twice per day. Her home medications are continued; enoxaparin for thromboprophylaxis and remdesivir are started. Concerned that the patient’s use of immunosuppressive agents increases her risk for severe COVID, we delay administration of a steroid pulse.On the third day of her hospitalization, the patient experiences chest heaviness and pressure. She has no dyspnea or pleuritic chest pain. At the onset of symptoms, she is afebrile and normotensive with a heart rate of 125 beats/min and respiratory rate of 22 breaths/min. SpO2 on room air is 96%. She appears uncomfortable but in no acute distress and with capillary refill less than 2 seconds. Chest auscultation reveals regular heart rhythm and no murmurs or abnormal breath sounds. She has facial fullness, but the periorbital edema has resolved. Bilateral lower extremities have trace pitting edema, and the right calf is slightly larger than the left. An electrocardiogram was obtained, at which time her heart rate had spontaneously decreased to 75 beats/min. It demonstrated normal sinus rhythm and no ST segment changes. Troponin was less than 0.010 ng/mL (<0.010 μ/L; reference range ≤0.028 ng/mL). An echocardiogram shows normal function and no pericardial effusion or right heart strain. Lower extremity venous duplex ultrasonography does not detect any thrombosis. The following day, her chest discomfort and tachycardia partially improve, but her SpO2 decreases to 92% on room air.On hospital day 5, her SpO2 continues to fluctuate between 92% and 96%. She has no dyspnea or chest pain but feels fatigued. A chest radiograph does not show evidence of cardiopulmonary disease. An additional laboratory test is drawn and reveals the diagnosis.Chest discomfort, tachycardia, tachypnea, hypoxemia, and lower extremity swelling all supported the diagnosis of pulmonary embolism (PE). The concern for PE was heightened by the patient’s hypercoagulability risk factors associated with SLE, nephrotic-range proteinuria, COVID-19 infection, and immobilization. However, considering the patient’s poor kidney function and hemodynamic stability, in discussion with Pediatric Cardiology, Nephrology, and Rheumatology, the risk of contrast administration outweighed benefit of a computed tomography. We did not order a d-dimer because a negative result would not have assuaged our concerns, and a positive result could be explained by her multiple comorbidities.1 Normal troponin, chest radiograph, ECG, and echocardiogram provided further evidence against an acute cardiopulmonary issue, such as PE, pleuritis, pneumonia, carditis, myocardial ischemia, or pericardial effusion. Moreover, these considerations seemed less likely because she remained afebrile and without cough or dyspnea. We considered GERD, anxiety, deconditioning with atelectasis, and musculoskeletal pain as possible contributors, but none of these fully explained her symptoms and persistently low SpO2.Suspicion was raised for methemoglobinemia in the setting of dapsone use because dapsone is a common cause of methemoglobinemia, and venous blood gas co-oximetry analysis revealed a methemoglobin (MetHb) level of 15.7% (0.157) and mixed venous oxygen tension (PvO2) 42 mm Hg (5.59 kPa), confirming the diagnosis.Dapsone can induce methemoglobinemia through a series of oxidative reactions, causing respiratory distress and neurologic symptoms. However, less-severe cases may present with subtle signs and symptoms that can pose diagnostic uncertainty. In healthy individuals, most hemoglobin exists with heme iron in the reduced ferrous state (Fe2+). Oxidation to the ferric state (Fe3+) forms MetHb, which has decreased oxygen-carrying capacity, and causes a left shift in the oxygen dissociation curve. As a result, the remaining heme moieties have a higher affinity for oxygen and decreased delivery to tissues. Under normal conditions, a small fraction (<1%) of hemoglobin exists as MetHb because of auto-oxidation. Constitutive actions of reducing enzymes associated with nicotinamide adenine dinucleotide (NADH) and nicotinamide adenine dinucleotide phosphate as well as the nonenzymatic compounds glutathione and ascorbic acid constantly convert MetHb back to hemoglobin, thus maintaining low levels of MetHb. However, these regulatory processes can be overwhelmed in settings of increased oxidative stress.2Medications may cause methemoglobinemia either directly via oxidation of hemoglobin or indirectly through creation of superoxide free radicals. Common MetHb-inducing medications include phenazopyridine, primaquine, benzocaine, and nitrite derivatives.3 Dapsone can produce dapsone hydroxylamine through cytochrome P450, leading to formation of reactive oxygen species that oxidize heme iron from the ferrous to the ferric state, resulting in methemoglobinemia.3–5 Hereditary methemoglobinemia due to mutations causing NADH cytochrome b5 reductase deficiency is less common.6Preexisting anemia, likely coexisting anemia of chronic inflammation and iron deficiency anemia in this case, can worsen one’s symptoms.2 For example, our patient’s measured hemoglobin of 9.4 g/dL (94 g/L) in the setting of 15.7% (0.157) MetHb suggests a concentration of functional hemoglobin of only 7.9 g/dL (79 g/L). Cyanosis is uncommon in methemoglobinemia, particularly if patients are anemic, highlighting the importance of a blood gas test as soon as methemoglobinemia is suspected.7 With higher levels of MetHb, SpO2 is inaccurate because of the light absorption pattern interfering with the calculation of the oxyhemoglobin-to-deoxyhemoglobin ratio. At high levels of MetHb, the SpO2 reading is fixed around 85%.6 Commonly reported symptoms of methemoglobinemia include dyspnea and fatigue.2,8–11 Given that our patient had not been taking her medications consistently, she likely had not been exposed to such significant levels of dapsone prior to hospitalization. However, on admission, we ensured she received dapsone daily.Concomitant COVID-19 infection may have increased her susceptibility to methemoglobinemia. Scholkmann et al discuss dozens of cases in which patients with COVID-19 also had elevated MetHb levels, often when they were given hydroxychloroquine. It is possible that COVID-19 infection predisposes to methemoglobinemia.12 A review of other case reports of adolescents and young adults (ages 15–28 years) who developed methemoglobinemia secondary to dapsone use revealed a variety of presentations. The duration of dapsone use prior to diagnosis ranged from days to months, and patients had a variety of medical conditions (acne, Hansen disease, SLE, leukemia).8–11,13–15 As we continue to treat COVID-positive patients, it is important to keep methemoglobinemia as a differential for those with persistent hypoxemia despite receiving standard care.To our knowledge, this is the first report that discusses a patient with inconsistent dapsone use at home who likely developed methemoglobinemia, in part because of regular dosing while hospitalized. Medication adherence, especially among pediatric patients, can be challenging. A recent study of adolescents with SLE found that 65% were nonadherent with at least 1 component of their treatment.16 Providers should be aware that new side effects and adverse events may arise in admitted patients who receive medications more often and at different doses than they had been taking at home.Patients with medication-induced methemoglobinemia should stop the offending drug. Supportive care, as needed, includes supplemental oxygen and intravenous hydration. Additional treatment, typically reserved for symptomatic patients with MetHb of 20% or more or asymptomatic patients with MetHb of 30% or more, involves administration of a reducing agent. First-line therapy is methylene blue, a rapid-acting medication that facilitates electron transfer to Fe3+ iron. However, toxic side effects occur with repeated dosing.2 Alternative treatment with ascorbic acid is used when methylene blue is unavailable or contraindicated. The water-soluble vitamin directly reduces MetHb, but it may take up to 24 hours to exert its effects.17Dapsone is discontinued immediately and replaced with atovaquone for Pneumocystis jirovecii prophylaxis given her use of immunosuppressive medications and chronic leukopenia. In patients with SLE, dapsone and atovaquone are preferred over trimethoprim-sulfamethoxazole, to which there is a high prevalence of allergies among those with SLE.17,18 Given that the patient is hemodynamically stable and MetHb is less than 20%, she does not receive additional treatment. Overnight, her SpO2 drops to 90%, and supplemental oxygen improves her SpO2 to 95%. Two days after discontinuing dapsone, repeat venous blood gas analysis shows improvement in MetHb to 4.7% (0.047). PvO2 is 54 mm Hg (7.18 kPa), and SpO2 ranges from 94% to 97% with intermittent supplemental oxygen for patient’s comfort. Her intermittent chest pressure continues for several days after resolution of methemoglobinemia, raising suspicion for a multifactorial cause, including contributions from musculoskeletal pain and anxiety. On her eighth day of hospitalization, COVID IgG is reported to be positive, indicative of time passed since acute infection, so we initiate a 3-day course of methylprednisolone 1000 mg per day. She requires continued hospitalization for electrolyte abnormalities and labile renal function. She remains hemodynamically stable for the remainder of her hospital stay without oxygen requirement and with improvement in intermittent chest pain.Providers should be aware of drug-induced methemoglobinemia and maintain suspicion for this adverse event in patients prescribed multiple medications who present with fatigue, dyspnea, and hypoxemia.Patients receiving medications associated with methemoglobinemia who are also diagnosed with COVID-19 may have a higher risk of developing methemoglobinemia.Methemoglobinemia causes inaccurate pulse oximetry readings due to interference with the calculation of the oxyhemoglobin-to-deoxyhemoglobin ratio, necessitating co-oximetry for a diagnosis.
Authors
- Grace Riley
- Sarah Gustafson (ORCID: https://orcid.org/0000-0002-3925-5313)
Institutions
- University of California, Los Angeles (US)
- UCLA Medical Center (US)
- Harbor–UCLA Medical Center (US)
- Charles R. Drew University of Medicine and Science (US)
Publication Details
- Journal
- Pediatrics in Review
- Published
- 2025-05-01
- DOI
- https://doi.org/10.1542/pir.2023-006104
- Primary Topic
- COVID-19 Clinical Research Studies
- Type
- article
- Field-Weighted Citation Impact
- 0.00