Severe falciparum malaria complicated by prolonged haemolysis and rhinomaxillary mucormycosis after parasite clearance: a case report

The mechanism of the haemolytic component of anaemia following parasite clearance in falciparum malaria is poorly understood and challenging to differentiate. This case does not completely conform to one of the three patterns of post-treatment malaria anaemia as it appeared bimodal within a short time period, where the initial episode of haemolysis occurring while parasitaemic resolved by 30 h after artesunate treatment, and a second episode of massive haemolysis began on day 3 with 75 % drop in haemoglobin that persisted for 10 days requiring multiple transfusions. This case highlights that drug-induced haemolysis may contribute. Several antimalarial drugs can cause haemolysis by differing mechanisms including quinine, mefloquine, halofantrine, and primaquine. Two possible culprit drugs in our case include quinine and ceftriaxone. Quinine was given prior to admission, and blackwater fever (intravascular haemolysis with haemoglobinuria) was observed on his initial presentation with P. falciparum parasitaemia. Quinine-induced haemolytic anaemia is considered drug-dependent and proposed to occur through an immune-complex mechanism [30, 31], however since the acute haemolysis began on day 3, by which time around 95 % of the drug would have been eliminated [32], and the DAT was negative, quinine seems less likely to be the cause. Drug-induced immune haemolytic anaemia (DIIHA) is rare but the commonly cited incidence of 1 per million people is likely an underestimate due to under-diagnosis [33]. In patients with DIIHA, ceftriaxone is the second most common antimicrobial cause after cefotetan [30, 33]. The onset of ceftriaxone-induced DIIHA ranges from 30 min to 34 days following exposure [30]. In this case, the negative DAT and positive IAT suggest the presence of an alloantibody or free autoantibody. The DAT is typically positive in most cases of drug-induced haemolytic anaemia but a negative DAT is possible if massive haemolysis occurs before testing [33, 34]. A diagnosis of ceftriaxone-induced DIIHA is supported in our case because the haemoglobinuria stopped two days after discontinuing ceftriaxone, however the haemoglobin was still recovering to baseline 2 weeks post-ceftriaxone discontinuation. Steroids do not improve DIIHA, but were administered in this case for potential autoimmune haemolytic anaemia [33]. Since steroid was started at the same time that ceftriaxone was stopped, establishing the cause of haemolysis is confounded. The diagnosis of ceftriaxone-induced haemolytic anaemia could not be confirmed because specific ceftrixone- antibody testing was not available.

The differential diagnosis for acute onset haemolytic anaemia with positive IAT and negative DAT following parasite clearance includes alloantibody mediated from previous transfusion, drug-induced haemolytic anaemia, microangiopathic haemolytic anaemia, an intrinsic RBC abnormality, or may be a non-specific finding [30, 33]. Delayed haemolytic transfusion reactions are generally mild and the onset of haemolysis is usually 2 to 11 days after subsequent transfusion. This diagnosis was unlikely as there was no history of previous blood transfusion and haemolysis started before the transfusions. Reports of post artesunate delayed haemolysis typically occur in hyperparasitaemic patients and manifest 7–14 days following the first dose of artesunate [19, 20]. This diagnosis is unlikely since this patient did not have a high parasitaemia and haemolysis started three days after the start of artesunate. The available blood results do not support the diagnosis of a microangiopathic haemolytic anaemia, intrinsic RBC abnormality or functional asplenia since there were no obvious red cell abnormalities or Howell-Jolly bodies on peripheral blood smear.

Fever in severe malaria usually resolves by day 4 to 5 [35]. Thus, recurrent or prolonged fever after severe malaria should prompt investigation into alternative causes. We identified a necrotic oral ulcer, from which a sample showed Mucor morphology on slide microscopy (Fig. 2). While our diagnosis was not confirmed by culture, this is not a prerequisite for the diagnosis since these organisms are difficult to grow and frequently cultures yield no growth. While mucormycosis is frequently fatal in immunocompromised hosts, such as those with haematological malignancies and diabetes mellitus, cases in patients without any immune impairment have been reported [28, 36]. These infections manifest as localized cutaneous or disseminated deep infections and are characterised by necrotic lesions due to the angioinvasive behavior of these organisms. Rhinocerebral mucormycosis is the most common clinical presentation of mucormycosis, representing around 60 % of all cases [37]. This category is further divided into 2 subtypes: the highly aggressive, often fatal rhino-ocular-cerebral form and the less fatal rhino-maxillary form. The latter subtype manifests as necrosis of the palate due to angioinvasion and thrombosis of the sphenopalatine and greater palatine arteries [38, 39].

Invasive fungal infections are rarely reported in association with severe malaria [22]. A recent review summarized nine cases of invasive fungal infections complicating falciparum malaria; 5 Aspergillus, 1 hyalohyphomycosis, 2 Cryptococcus, 1 Absidia, and 1 disseminated candidiasis [21–27]. The immune suppression due to malaria infection itself may predispose individuals to invasive fungal infection [3, 22]. Parasitic digestive vacuoles containing malaria pigment phagocytosed rapidly by polymorphonuclear granulocytes can cause functional exhaustion, which blunts microbicidal activity of granulocytes [40]. Neutrophil dysfunction might result from haemolysis-induced haem oxygenase-1 induction [41]. High-dose steroids may also have inhibited the polymorphonuclear leukocyte response and decreased granulocyte phagocytosis. In this case, the iron overload state secondary to massive haemolysis and multiple transfusions may have promoted fungal growth [37, 42, 43]. All reported Aspergillus and Absidia infections complicating severe malaria have occurred in patients with marked haemolysis [3, 23, 24, 26].

In this case it was difficult to definitively exclude a predisposing immunocompromising condition. However, bone marrow biopsy did not suggest an underlying haematologic malignancy, HIV serology was negative, and random glucose was normal in the setting of acute illness, although haemoglobin A1C testing was unavailable to definitely rule out diabetes. It is likely that malaria related immune suppression augmented by the administration of steroids and multiple blood transfusions with massive haemolysis (iron overload) contributed. The initial low parasitaemia is consistent with post-treatment presentation and the severity criteria present on admission are consistent with a delayed time to recovery in severe malaria patients from low transmission settings [1]. Concomitant bacterial infection was not definitively excluded since admission blood culture results were not available, however the day 0 white blood cell count was normal. The kidney and liver dysfunction that developed when mucormycosis was diagnosed has several explanations but is unlikely related to disseminated Mucor. More plausibly, the kidney injury was due to high cell-free haemoglobin generated during massive haemolysis, ischaemia, or nephrotoxic drugs (i.e. amikacin). The liver dysfunction was likely due to ischaemia during hypotensive episodes.

In addition to urgent surgical debridement, the recommended first line anti-fungal therapy for the treatment for mucormycosis is amphotericin B deoxycholate or liposomal amphotericin B. Unfortunately, the availability of these medications is often inadequate in resource-limited settings. In this case, surgical debridement was successful but itraconazole was the only available anti-fungal with activity against Mucorales. Although the patient follow up was truncated there was a favorable in-hospital clinical response to itraconazole, which is in keeping with other reports of successful treatment with itraconazole monotherapy [44, 45].

The oral ulceration differential diagnosis includes squamous cell carcinoma, deep fungal infection, tertiary syphilis and acute necrotizing ulcerative gingivitis [46–49]. In our case, the acute course of disease goes against a diagnosis of squamous cell cancer. Negative non-treponemal syphilis serology and an atypical presentation exclude tertiary syphilis. The sinus involvement and presence of fungal hyphae makes acute necrotizing gingivitis unlikely. Most cases of rhinocerebral mucormycosis involve both sides of the palate, however hemi-palatal involvement has been reported [38, 39]. This patient presented with typical symptoms of rhinomaxillary mucormycosis including fever, headache, lethargy, pain and swelling of the maxillary sinus. Microscopy differentiated Mucor from aspergillus as broad, aseptate hyphae at 90° angles were observed.

This case illustrates that both prolonged haemolysis and recurrent fever occurring after parasite clearance in falciparum malaria are challenging to diagnose. A systematic approach to the differential diagnoses is vital as management can be critically different. The potential diagnosis of ceftriaxone-induced haemolytic anaemia is important as this drug is widely used empirically. Although the incidence of ceftriaxone-induced haemolytic anaemia is low, this diagnosis should be considered and ceftriaxone discontinued to prevent unnecessary blood transfusions, steroid treatment and mortality [30, 33]. Steroids are not recommended as treatment for severe malaria because steroids increase gastrointestinal bleeding, convulsions, and prolong coma [50, 51]. Investigating fever in resource-limited setting is challenging, as microbiologic diagnoses are frequently not possible. However, delayed evaluation of recurrent fever after parasite clearance may lead to unnecessary empirical anti-malarial and anti-bacterial use, which has important consequences on the individual and global level. Although invasive fungal infections rarely complicate severe malaria, early diagnosis and treatment with surgical debridement plus anti-fungals are essential to improve patient outcomes. Empiric overuse of antimicrobials contribute to anti-microbial resistant organisms, which is a global public health concern in this era of reduced research and development into novel antimicrobial therapeutics. To combat this pressing issue, improved microbial diagnostics are urgently needed in resource-limited settings.