A Rare Case of Hypoxia and Cyanosis Secondary to Multifactorial Medication-Induced Sulfhemoglobinemia

Introduction

Hypoxia is a common clinical presentation with a broad differential. It often results from ventilation/perfusion mismatch due to conditions such as pneumonia, pulmonary emboli, right-to-left shunt, pleural effusions, or pulmonary edema, where impaired oxygen exchange leads to decreased arterial oxygenation. Less commonly, hypoxia can be caused by acquired hemoglobinopathies, which affect oxygen transportation and release. While methemoglobinemia is a well-recognized cause, sulfhemoglobinemia remains underdiagnosed despite its potential to impair oxygen delivery by reducing hemoglobin’s ability to bind and release oxygen effectively. The presentations of the acquired hemoglobinopathies are remarkably similar: fatigue, weakness, cyanosis, hypoxia refractory to oxygen supplementation, oxygen saturation gap, metabolic acidosis, seizure, coma, and arrhythmia in extreme cases. Most spectrophotometers used in arterial blood gas analysis have difficulty differentiating between methemoglobinemia and sulfhemoglobinemia due to their similar range of light absorption. This requires the clinician to have a high index of suspicion to make the correct diagnosis. The similarity in presentation is in part due to similar pathogenesis and disruption of oxygen-carrying capacity of hemoglobin. While methemoglobinemia can be reversed, sulfhemoglobinemia lasts for the lifespan of the erythrocyte – approximately 90 days – and as such can have chronic indolent presentations.

Presentation can be variable due to the percentage of methemoglobin/sulfhemoglobin, rate of accumulation, duration, and degree of exposure to the oxidizing agent, and concurrent comorbidities of individuals with underlying lung disease. Due to this, it is common for patients to have chronic sulfhemoglobinemia due to long-term exposure to oxidizing agents. For sulfhemoglobinemia, treatment modalities are limited due to the irreversible oxidation of hemoglobin. The backbone of treatment remains discontinuation of the offending agent, providing supplemental oxygenation, red blood cell transfusions, and exchange transfusions in severe cases.

Here, we report the case of a 37-year-old woman who presented with fatigue and dyspnea, who was soon found to have hypoxia refractory to oxygen supplementation, cyanosis, and hemolysis.

Case Report

Our patient was a 37-year-old woman who was a former smoker with a past medical history of provoked left lower-extremity deep venous thrombosis and bilateral segmental pulmonary embolism, both diagnosed 2 months ago, on apixaban, depression on fluoxetine, and recently-diagnosed cyclical Cushing’s syndrome, on osilodrostat 1 mg twice a day initiated 10 days prior to presentation, presented to the Emergency Room (ER) with fatigue, dyspnea, dysuria, nausea, and emesis of 1-week duration. Her primary care physician had prescribed trimethoprim-sulfamethoxazole (TMP-SMX) 800/160 mg twice a day for possible urinary tract infection (UTI), which she took for the first time, completing a total of 5 doses before presentation. The patient also reported first-time use of over-the-counter (OTC) phenazopyridine hydrochloride 97.5 mg tablet, taking 3–4 doses per day for the preceding 5 days. Additionally, 4 days prior to presentation, she underwent a dental procedure with local anesthesia using lidocaine, which she had been exposed to previously. The patient reported that her fatigue and dyspnea started after the initiation of osilodrostat but got significantly worse within the hour following her dental procedure.

In the ER, her vitals included temperature 36.9°C, heart rate 120 bpm, blood pressure 101/69 mmHg, respiratory rate 13 breaths per minute, and oxygen saturation 83% on room air, which improved to 91% on 6 liters per minute supplementary oxygen by nasal cannula. On the physical exam, she was pale, anicteric, and in no acute distress. Cardiac examination revealed tachycardia with regular rhythm, with no murmur, rub, or gallop. Her lung auscultation was without rales or wheezing. Her abdomen was non-distended, with mild suprapubic tenderness to palpation and without guarding or rebound. There was no lower-extremity edema. Her neurological exam revealed no abnormalities.

Her initial laboratory workup was remarkable for hemoglobin 9.9 g/dL (baseline 11.3 g/dL 1 month before this admission), white blood cell count 7.9×109/L (3.4–9.6×109/L), creatinine 0.94 mg/dL (0.59–1.04 mg/dL), urinalysis with positive ketones, positive bilirubin, white blood cells 5/hpf (0–10/hpf), RBC 155/hpf (0–2/hpf), moderate hemoglobin, leukocyte esterase negative, and no bacteria (Table 1). Intravenous (IV) ceftriaxone 1 g every 24 hours was initiated for possible UTI. A hemolysis blood workup was ordered due to a drop in hemoglobin and the presence of bilirubin and hemoglobin on urinalysis. Additional imaging was ordered to rule out other causes of hypoxia. A CT chest angiogram showed peripheral filling defects within bilateral lower lobes compatible with chronic pulmonary emboli, no signs of pneumonia, no pulmonary edema, and no acute pulmonary emboli.

Initially, there was also a concern for possible adrenal insufficiency in the setting of recent initiation of osilodrostat. Due to this, osilodrostat was held, and the patient received 50 mg of IV hydrocortisone followed by 25 mg IV hydrocortisone twice daily the following day, with a plan to continue 20 mg oral hydrocortisone twice daily thereafter. Urine culture showed <10 000 CFU/ml, so UTI was ruled out and IV ceftriaxone was discontinued. Hemolysis workup showed LDH 204 U/L (122–222 U/L), haptoglobin <14 (30–200 mg/dL), and a peripheral blood smear with few schistocytes (Table 1). Hemolysis related to TMP-SMX was considered in the differential. Additional etiologies for hypoxia were considered, but the patient reported compliance with her home anticoagulation, her imaging results were not compatible with worsening or acute clot burden, and there were no acute lung parenchyma changes. Her hypoxia was deemed to be related to possible hemolysis.

In the first 6 hours after admission, she developed worsening hypoxia, requiring high-flow nasal cannula support with FiO2 100%. Despite this, her oxygen saturation remained low at 88%. She was noted to have a generalized slate-gray cyanotic appearance. An arterial blood gas (ABG) with co-oximetry showed respiratory alkalosis (pH 7.54, PCO2 22.1) and oxygen saturation of 77.3% which was compatible with a saturation gap. Curiously, her ABG reported a “?” mark in her methemoglobin level. The core lab was contacted regarding the significance of this “?” mark, which was considered either to be due to a very high level of methemoglobin or to interference with any other additional substance that has the same range of light absorption as methemoglobin. A methemoglobin/sulfhemoglobin panel test was ordered.

Methemoglobinemia was suspected given her recent medication history and new findings. Methylene blue therapy was deferred due to risk of serotoninergic syndrome in a patient taking fluoxetine at home. High-dose vitamin C (ascorbic acid) with a total of 10 g in divided doses over 24 hours was initiated. Exchange transfusion was considered, so the case was discussed with hematology.

The patient’s hemoglobin was monitored, additional hemolytic workup results were reviewed, and serial ABG with co-oximetry showed a persistent saturation gap (Table 1). A methemoglobin/sulfhemoglobin panel showed a sulfhemoglobin level elevated up to 1.3% (normal range: 0.0–0.4%) and a normal methemoglobin level at 0.8% (normal range: 0.0–1.5%).

The patient continued to receive supportive treatment. High-flow oxygen was weaned gradually as tolerated. During her hospital stay, she received 3 units total packed red blood cells as recommended by Hematology to improve her hypoxia. She was started on folic acid and vitamin B12 supplements to assist with cell production in the setting of ongoing hemolysis.

Her hypoxemia resolved prior to discharge. Oxygen saturation was 98% on room air on day 7 of hospital admission. She was discharged home with instructions for follow-up as an outpatient with new blood work. She was instructed to avoid potential triggers, including TMP-SMX, phenazopyridine, and local anesthetics.

The patient was followed up in the outpatient clinic 1 month after discharge. She reported minimal dyspnea on exertion. Her hemoglobin level had improved to 12.1 g/dL. Follow-up in the hematology clinic included genetic testing for methemoglobinemia and sulfhemoglobinemia to assess for M-Hemoglobins (M-Hbs) and cyb5r activity, which was inconclusive, failing to identify any pathogenic variants associated with inherited forms. Given these results and after consultation with Hemopathology, the etiology was deemed acquired. This conclusion was supported by the patient’s exposure history, absence of a family history, and resolution of symptoms after discontinuation of the offending agents.

Discussion

Erythrocytes are continuously exposed to oxidant stress from oxygen, chemicals, drugs, and infections, and it is protected against oxidation-induced hemolysis mainly via reduced glutathione, catalases, and ascorbate [1]. Oxidant stress can cause structural hemoglobin changes and/or hemolysis through various mechanisms. Dyshemoglobinemias alter the hemoglobin structure, impairing oxygen transport and causing functional anemia. Oxidation of hemoglobin can lead to Heinz body hemolytic anemia, methemoglobinemia, or sulfhemoglobinemia, depending on the affected component [2]. Clinicians should recognize the wide range of agents that can cause these conditions, as they may coexist.

Methemoglobin continuously forms within erythrocytes and it is naturally present in small amounts in blood, constituting 1–2% of total hemoglobin, with higher levels indicating methemoglobinemia [3,4].

Sulfhemoglobin forms when a sulfur atom is added to hemoglobin’s porphyrin ring. In the past, the sulfur source was thought to be the offending agent [1]. However, many offending agents do not contain sulfur, and a study suggested that bacterial metabolism in the gastrointestinal tract can be the source of the sulfur found in sulfhemoglobin [5]. Sulfhemoglobin cannot transport oxygen and persists for the life of the erythrocyte. Small concentrations of only 0.5 g/dL can cause slate-gray cyanosis, as seen in our case [2]. The presence of anemia can also influence the severity of symptoms [6]. Previous cases have documented sulfhemoglobin levels as high as 22.1% (equivalent to 3.17 g/dL in a patient with hemoglobin of 14.4 g4L) [7], while symptoms have been observed at much lower levels, such as 0.2 g/dL in a case of chronic daily sumatriptan use [8] and 1.3 g/dL in a patient taking sulfizoxasole and phenazopyridine [9]. Our patient developed symptoms with sulfhemoglobin levels of 1.3%, more than 3 times the upper normal limit (0.0–0.4%), corresponding to 0.13 g/dL with a hemoglobin of 9.9 g/dL. Although this is not an exceptionally high level, it was sufficient to cause clinical compromise.

Our patient was exposed to multiple agents within 1 week, including lidocaine, phenazopyridine, and TMP-SMX. Table 2 summarizes the most common substances associated with sulfhemoglobinemia [1,8,10]. A PubMed search using the keywords “phenazopyridine” and “sulfhemoglobinemia” identified only 5 cases of phenazopyridine-induced sulfhemoglobinemia in the past 35 years [5,6,11–13]. While lidocaine and TMP-SMX have been linked to methemoglobinemia, there are no documented cases linking them to sulfhemoglobinemia. However, TMP-SMX carries a theoretical risk, as sulfhemoglobinemia has been reported with other sulfonamides, particularly sulfanilamide [10], and a case in 1970 where a patient taking sulfisoxazole and phenazopyridine developed sulfhemoglobinemia [9]. In our case, with normal methemoglobin and G6PD levels, phenazopyridine was the most likely cause of sulfhemoglobinemia, although a contribution from TMP-SMX cannot be completely ruled out.

Pulse oximetry is unreliable in cases of methemoglobinemia or sulfhemoglobinemia, as it can provide false low or false high values due to interference of dyshemoglobins with the absorption ratios of the pulse oximeter readings [12,14]. Many blood gas devices do not measure percent saturation, but rather report it based on a calculated value from the oxygen tension and pH, assuming an absence of abnormal levels of hemoglobin pigments [1,15]. Hence, multiple-wave-length co-oximetry in arterial blood gas, as in our case, aids to support the diagnosis measuring true percentage saturation. The difference between the oxygen saturation measured by pulse oximeter and that measured by co-oximetry is called the saturation gap, which should be less than 3–5% in arterial blood. Higher saturation gaps are suggestive of methemoglobinemia, carboxyhemoglobinemia, or, rarely, sulfhemoglobinemia [16]. In our case, hemolysis was suspected based on the initial workup, a saturation gap was noted in ABG with co-oximetry by the time our patient had worsening hypoxia, and all this in the setting of recent medication exposure led to consideration of dyshemoglobinemia in the differential diagnosis.

Co-oximetry measures percentages of oxyhemoglobin, deoxyhemoglobin, carboxyhemoglobin, and methemoglobin [3]. Co-oximeters do not differentiate between methemoglobin and sulfhemoglobin and thus report elevated methemoglobin fractions in the presence of sulfhemoglobinemia [2]. More recent devices do not report a false elevated methemoglobin in the case of isolated sulfhemoglobinemia. In our case, a “?” mark was reported in the methemoglobin level by our device on initial ABGs with co-oximetry, which was found to be related to interference with sulfhemoglobin, which has a similar range of light absorption as methemoglobin [4,11,14]. A similar situation of co-oximetry interference with a “?’ mark result was reported in a case report of sulfhemoglobinemia secondary to drug–drug interaction [17]. It is important to know your co-oximeter device’s specifications and always use a reference laboratory such as spectrophotometry analysis or gas chromatography to measure sulfhemoglobin if there is any possibility of this diagnosis. Hemoglobin electrophoresis and cyb5r activity are also important in evaluating hereditary causes of cyanosis [18]. In our case, hemoglobin electrophoresis did not reveal abnormal variants, and genetic testing did not conclusively show inherited forms.

Ongoing biomedical in silico research to develop inexpensive noninvasive optical methods is exploring the use of spectral analysis techniques to identify unique absorption patterns to differentiate methemoglobinemia from sulfhemoglobinemia [19,20]. However, further validation of these experimental methods against in vivo conditions will be required.

Most patients with sulfhemoglobinemia are asymptomatic, except for slate-gray cyanosis, unless other abnormal hemoglobin such as methemoglobin or hemolytic anemia is present. The slate-gray discoloration can persist for weeks or months due to the irreversible nature of sulfhemoglobin. Rarely, sulfhemoglobinemia is severe enough to cause tachycardia, tachypnea, dyspnea, and altered mentation [2].

Methylene blue, one of the preferred antidotes used in methemoglobinemia, acts as a cofactor of the enzyme NADPH methemoglobin reductase, which increases its activity and reduces methemoglobin levels, producing deoxyhemoglobin. Other therapies such as ascorbate or vitamin C are not as effective as methylene blue due to its slow action in reducing methemoglobin [3]. There is no specific antidote for sulfhemoglobinemia, and methylene blue is useless [2,3]. Treatment of dyshemoglobinemias is mostly supportive and includes oxygen therapy to maximize the oxygen-carrying capacity of the normal hemoglobin, serial monitoring for signs of hemolysis, serial co-oximetry, and blood transfusions if needed, especially in case of sulfhemoglobinemia, as an attempt to increase the total hemoglobin concentration and decrease the sulfhemoglobin fraction and concentration [3]. Patients who develop cyanosis without other symptoms may not require specific treatment. Once exposure to the oxidizing agent ends, methemoglobin levels return to normal, usually within 36 hours, unlike sulfhemoglobin, which lasts the lifetime of the erythrocyte [1,3].

Glucose-6-phosphate dehydrogenase (G6PD) deficiency was ruled out in our patient. Methylene blue should never be administered to a patient with known G6PD deficiency, and it is not totally contraindicated in a symptomatic patient with unknown G6PD deficiency status. G6PD-deficient patients have low NADPH concentrations, which limits the action of methylene blue as a cofactor for NADPH methemoglobin reductase, and even more important, methylene blue triggers hemolysis in G6PD-deficient patients [2]. Patients who do not respond to methylene blue may have profound toxicity from the offending agent, an unrecognized G6PD deficiency, NADPH methemoglobin reductase deficiency, or sulfhemoglobinemia. In our case, methylene blue was not initially considered due to the risk of serotoninergic syndrome in a patient taking fluoxetine. Other optional treatments are blood transfusions and exchange transfusions, which can be lifesaving. Hyperbaric oxygen is only a temporizing measure in severe cases while a patient is prepared for a blood transfusion.

Further research is needed to identify the risk factors that lead some patients to develop dyshemoglobinemia, and to discover medications that can break the bond between the sulfur atom and the beta-pyrrole ring in the heme moiety without causing heme degradation and thus reverse sulfhemoglobinemia.

Conclusions

This case report highlights the diagnostic challenges and clinical complexities associated with dyshemoglobinemia, particularly between methemoglobinemia and sulfhemoglobinemia. While methemoglobinemia was initially suspected based on the patient’s clinical symptoms and recent exposure to oxidizing agents, the failure to confirm elevated methemoglobin levels on ABG and the presence of a significant saturation gap led to further investigation. Ultimately, sulfhemoglobinemia was diagnosed, likely induced by the combination of medications, including TMP-SMX and phenazopyridine. This case underscores the importance of considering sulfhemoglobinemia in patients with unexplained hypoxia, especially when common diagnostic tests for methemoglobinemia fail to provide definitive answers. Although the management of sulfhemoglobinemia remains supportive, research is needed to better understand its pathophysiology and potential treatment options. This case also highlights the importance of a comprehensive medication history and vigilant monitoring in identifying rare but critical causes of persistent cyanosis and hypoxemia.