Subtle case of dapsone-induced methaemoglobinaemia

Abstract

Methaemoglobinaemia is a rare disease that is typically caused by a medication or other exogenous agent, with dapsone being the most common. It occurs when the concentration of methaemoglobin rises via ferrous haeme irons becoming oxidised to the ferric state, which shifts the oxygen dissociation curve to the left. The net result of an elevated methaemoglobin concentration is functional anaemia and impaired oxygen delivery to tissues. At lower blood levels, this can cause symptoms such as cyanosis, lethargy, headache and fatigue, whereas at higher levels it can be fatal. Here we discuss a subtle case of dapsone-induced methaemoglobinaemia presenting as subacute mental status changes and apparent hypoxia, thus highlighting the association between methaemoglobinaemia and dapsone. This case demonstrates the importance of thorough medication reconciliation and maintaining a broad differential diagnosis, while also recognising the significance of conflicting data and their implications for the workup.

Background

Methaemoglobin is a form of haemoglobin in which the irons of haeme, typically ferrous (Fe⁺⁺), become oxidised to the ferric state (Fe⁺⁺⁺). These ferric haeme molecules cannot reversibly bind oxygen and have an increased oxygen affinity, with the net result being a shift in the oxygen dissociation curve to the left (figure 1).¹ Auto-oxidation of haemoglobin to methaemoglobin normally occurs slowly and results in a physiological methaemoglobin concentration of less than 1% of total haemoglobin.² However, if the concentration of methaemoglobin rises for any reason, then this results in functional anaemia and impaired oxygen delivery to tissues. Cyanosis typically develops at methaemoglobin levels of 10%–20%, whereas lightheadedness, headache, tachycardia, anxiety and confusion occur at 30%–40%. As levels approach 70%, coma, shock, seizure and death can ensue.^2–4 Methaemoglobinaemia can be congenital or acquired. The majority of cases are secondary to various medications or exogenous agents, most commonly dapsone (table 1). Antimalarial agents such as chloroquine have also been implicated, which have seen increasing utilisation during the 2019–2020 coronavirus disease (COVID-19) pandemic. Here we discuss a subtle case of dapsone-induced methaemoglobinaemia presenting as subacute mental status changes and apparent hypoxia, which emphasises the importance of thorough medication reconciliation and maintaining a broad differential diagnosis.

Figure 1.

Oxygen–haemoglobin dissociation curve. Reproduced with slight modification from a public domain image.¹¹

Table 1.

Exogenous agents associated with methaemoglobinaemia3 5
Drug class	Agent
Antibiotics	Dapsone, sulfamethoxazole
Antimalarials	Chloroquine, primaquine
Antineoplastic agents	Cyclophosphamide, flutamide
Industrial products	Aniline dyes, naphthalene, aminophenols
Local anaesthetics	Benzocaine, lidocaine, prilocaine, procaine
Nitrite/nitrates	Amyl nitrate, sodium nitrite, nitroglycerin, nitroprusside, nitric oxide
Other	Metoclopramide, Phenazopyridine

Case presentation

A 69-year-old generally healthy woman presented to the emergency department with 10 days of confusion, memory loss and generalised weakness with unsteady gait resulting in multiple falls. She had been diagnosed with giant cell arteritis (GCA) 2 months prior that had been complicated by vertebral artery stroke with residual ataxia and visual deficits; this was being treated with high-dose steroids and tocilizumab, along with dapsone for pneumocystis pneumonia (PCP) prophylaxis.

Vitals were notable for a heart rate in the low 100s and oxygen saturation of 85%, which improved to 91%–92% on 3–5 L oxygen by nasal cannula. The examination showed a pale woman in mild respiratory distress with dry mucus membranes. She was alert and oriented, but intermittently confused during the conversation. Her cardiopulmonary examination was unremarkable and she had no focal neurological deficits. Laboratory values demonstrated a leucocytosis of 13.2×10⁹/L with a left shift, bicarbonate of 16 mmol/L with an anion gap of 17 mmol/L and lactate 4.5 mmol/L. Haemoglobin was in the 90s from a baseline of 120–130 g/L. Total bilirubin was elevated (1.9 mg/dL), lactate dehydrogenase was at the upper limit of normal (247 U/L) and haptoglobin at the lower limit of normal (49 mg/dL); all of which were suggestive of low-grade haemolytic anaemia. Erythrocyte sedimentation rate (ESR) was 1 mm/hour (reference range 0–30) and C-reactive protein (CRP) was <2.9 mg/L. Urinalysis was unremarkable.

Investigations

Arterial blood gas (ABG) showed a partial pressure of oxygen of 91% and an oxygen saturation of 95%, whereas peripheral oxygen saturation remained in the mid-80s. The chest radiograph showed clear lungs with no significant findings. CT of the abdomen pelvis did not show any abnormalities, and the CT scan of the head was also unremarkable. She was started on intravenous fluids, after which her anion gap metabolic acidosis, lactate and leucocytosis began to improve.

Given the saturation gap between the pulse oximeter and the ABG in the setting of her overall presentation, a methaemoglobin level was sent and came back at 15.2%, confirming a diagnosis of methaemoglobinaemia. The serum dapsone level was not tested.

Differential diagnosis

A urinary tract infection (UTI) was definitely possible, given her non-specific symptoms (including confusion) and leucocytosis. Often, elderly patients present in such a manner without endorsing lower UTI symptoms. But her urine was completely benign, thus ruling this diagnosis out.

In addition, given her recent diagnosis of GCA, confusion and memory loss, a GCA flare was also on the differential. But the patient did not endorse other symptoms, such as vision changes (although she had residual visual deficits) or headache, and her ESR and CRP were not elevated, all of which made this diagnosis less likely.

A recurrent stroke was also possible, but was made less likely by her normal CT scan of the head. If the methaemoglobin level was within normal limits, then it is likely that an MRI brain would have been done for further workup.

Pneumonia was also on the differential in the setting of her immunosuppressive medications. It could have been that her altered mental status and leucocytosis were caused by pulmonary pathology, with a lack of clinical findings such as cough and fevers secondary to immunosuppression. But this diagnosis was made less likely by the lack of chest X-ray findings.

Furthermore, a slow retroperitoneal bleed was also possible, given her confusion, anaemia and multiple falls that could have led to a traumatic haemorrhage. But this was ruled out based on the CT scan of the abdomen pelvis, which was unremarkable.

Treatment

Following consultation with the toxicology department, she was treated with supportive therapy consisting of nasal oxygen and intravenous fluids. Methylene blue was not started because although she was symptomatic, her methaemoglobin level was not above 20%. Dapsone was discontinued and replaced with pentamidine for PCP prophylaxis. The patient gradually improved from a clinical standpoint, with methaemoglobin levels decreasing (15.2%>10.9%>7.8%) and haemolysis laboratory values normalising.

Outcome and follow-up

She was discharged 3 days later with close follow-up and has been doing well on her current therapy, including pentamidine, at 6 months postdischarge. Of note, an MRI brain was done closely after discharge to rule out any neurological pathology and it was negative.

Discussion

The patient described here displayed mild symptoms of methaemoglobinaemia including confusion, unsteadiness, haemolytic anaemia and hypoxia. These symptoms are very non-specific, especially given her recent vertebral artery stroke. Thus, a thorough medication reconciliation, and appreciating that she had recently started dapsone, was essential to the diagnosis. Dapsone is the most common cause of acquired methaemoglobinaemia. In a retrospective review of 138 patients with acquired methaemoglobinaemia, dapsone was the causal agent in 42% of cases.⁵ That study found a mean methaemoglobin concentration of 7.6% among patients taking dapsone at therapeutic doses, so it is postulated to cause chronic, low-grade methaemoglobinaemia. These patients are then susceptible to symptomatic hypoxia when a ‘second-hit’ occurs that further decreases oxygen tissue delivery, with the most common ‘second-hit’ being anaemia^{2 5}; our patient’s haemolytic anaemia was in line with this finding. Dapsone works by inhibiting the bacterial synthesis and is metabolised by the liver via oxidation reactions of N-acetylation and N-hydroxylation, which produces metabolites that are hypothesised to cause its adverse haematological effects.⁴ In this case, recognising the association between dapsone and methaemoglobinaemia early allowed for an accurate diagnosis and prompt initiation of treatment. This avoided potentially unnecessary diagnostic studies, as well as more serious complications that may have occurred had dapsone been continued.

It would have been easy to misattribute her neurological symptoms and weakness to delirium, or manifestations of her recent stroke, which is why maintaining a broad differential diagnosis and recognising discordant data were key in this case. Her cardiopulmonary examination was benign and her chest radiograph was clear, so the reason for her apparent hypoxia was uncertain. This discrepancy was secondary to the ‘saturation gap’ that is characteristic of methaemoglobinaemia. Standard pulse oximetry measures light absorbance at distinct wavelengths. Methaemoglobin distorts this light absorbance and lowers the measured saturation level.^{3 6} On the other hand, the oxygen saturation level reported by ABG is calculated from the partial pressure of oxygen dissolved in the blood, and this value is independent of the methaemoglobin level.⁶ Thus, in patients with elevated methaemoglobinaemia, there will be discordance between the pulse oximeter and the ABG, a so-called ‘saturation gap’. The patient presented here had a saturation gap of 10%, which is typical in methaemoglobinaemia. Keeping a broad differential and considering other pathologies led to a timely ABG in order to elicit her true oxygen saturation level, which guided the treatment team to the saturation gap and ultimately the diagnosis.

Although this patient had classic symptoms and diagnostic findings of methaemoglobinaemia, her lack of cyanosis on examination was unique. Nearly every case of methaemoglobinaemia in the literature reports cyanosis (or at least dusky skin) on examination. One study contended that peripheral and central cyanosis is almost always present when methaemoglobin is 15% or more in the blood.⁴ But even though this patient’s methaemoglobin level was over 15%, she was never overtly cyanotic. It is unclear why she did not display cyanosis, but this further highlights the importance of keeping a broad differential.

Dapsone’s role in this case has implications for its dosing regimen, as well as for which alternative prophylaxis agents to use. There is controversy in the literature with regards to dapsone dosing and methaemoglobinaemia, with one study concluding that methaemoglobinaemia risk was increased (HR of 6.25) with dapsone dosing >20% above the target range, whereas another study argued that dapsone was not dose dependent.^{7 8} Although there may be a dose-related risk, one cannot just decrease the dose of dapsone, as this leads to increased rates of PCP infection. PCP rates were shown to increase from 2.6% to 11.3% when the dapsone dose was decreased from 50 mg two times per day to 50 mg four times per day.⁹ This is significant, as the majority of cases of depose-induced methaemoglobinaemia occur at therapeutic doses of dapsone, rather than overdoses.² Our patient was on 100 mg daily of dapsone prior to admission. Although dapsone levels were not tested, given that methaemoglobinaemia often occurs at therapeutic dapsone doses, the serum dapsone level may have been in the normal range and would not have changed management. Moreover, our patient’s dapsone was stopped and she was discharged with monthly pentamidine, but pentamidine has been shown to be an inferior PCP prophylaxis, as PCP breakthrough rates are higher with pentamidine than with either dapsone or trimethoprim–sulfamethoxazole.¹⁰ Further studies are needed to better define this dose–response relationship between dapsone and methaemoglobinaemia, as well as to determine optimal alternative prophylaxis regimens for these patients.

All patients with methaemoglobinaemia should receive supportive therapy along with the identification and removal of any causative agents; our patient steadily improved with such management.^{3 4 6} Select patients are also treated with methylene blue, which serves as a cofactor to reduce methaemoglobin to its unoxidised state. Methylene blue should be given intravenously at a dose of 1–2 mg/kg for 5 min, with the maximum effect occurring around 30 min; additional doses may be given after 1 hour if the response is inadequate.⁶ The literature varies regarding whom to treat with methylene blue, but the overall recommendations are to reserve treatment for symptomatic patients with methaemoglobin levels >20%, or anyone with levels >30%.^{3 6} Given the variation in the literature, further studies that better define treatment guidelines would be of great value, as methylene blue administration does not come without risk; it can cause haemolytic anaemia and paradoxical worsening of methaemoglobin levels, especially in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency.⁶

Methaemoglobinaemia is a rare disease that is typically caused by a medication or other exogenous agents. This patient’s recent initiation of dapsone therapy for PCP prophylaxis resulted in non-specific symptoms and discordant objective findings that were ultimately secondary to methaemoglobinaemia. This case demonstrates the importance of thorough medication reconciliation and maintaining a broad differential diagnosis while also recognising the significance of conflicting data and their implications for the workup. More studies are needed with regards to dapsone dosing and alternative PCP prophylaxis agents, as well as to better define guidelines for treating patients with methylene blue.

Patient’s perspective.

Feeling confused and needing oxygen to help me breathe was stressful, especially because at first we couldn’t figure out what was causing this. I didn’t know what was wrong with me or why, and it was scary! First the stroke and now this, it was a lot to deal with. It was hard when I found out that the dapsone I was prescribed had caused all of this to happen, but the doctors did a good job explaining why I needed that treatment in the first place. Overall, I was happy with the care I received.

Learning points.

• Methaemoglobinaemia occurs when haeme irons become oxidised to the ferric state, leading to functional anaemia and impaired oxygen delivery. Symptoms are wide ranging and can be non-specific.

• A ‘saturation gap’ between pulse oximetry and arterial blood gas is a characteristic finding in methaemoglobinaemia.

• Dapsone is the most common cause of acquired methaemoglobinaemia.

• A thorough medication reconciliation, broad differential diagnosis and recognition of discordant data are essential, especially among patients presenting with non-specific symptoms.