Article Text

Download PDFPDF

Complement-mediated autoimmune haemolytic anaemia as an initial presentation of Legionnaires’ disease
  1. Uzma Sabahat1,
  2. Niaz Ahmed Shaikh1,
  3. Ali Mohammed Mahmood Alameen2 and
  4. Faryal Ashfaq1
  1. 1Internal Medicine, Rashid Hospital, Dubai, UAE
  2. 2Pathology, Rashid Hospital, Dubai, UAE
  1. Correspondence to Dr Uzma Sabahat; uzmasabahat{at}hotmail.com

Abstract

A 42-year-old diabetic man presented to the hospital with severe sepsis and multiorgan dysfunction. A probable respiratory source of sepsis was suspected because of suggestive clinical and radiological findings. He was critically ill and was therefore admitted to intensive care for further management including ventilatory support and renal replacement therapy. He was also found to have marked anaemia requiring multiple blood transfusions with clinical and laboratory evidence pointing towards severe haemolysis. Further workup for the aetiology of pneumonia established a diagnosis of Legionella by confirmatory tests namely legionella antigen in the urine and exponentially rising serum antibody titres. The cause for the severe haemolysis was found to be complement-mediated autoimmune haemolysis as determined by direct antiglobulin test positive for complement components C3 and negative for IgG. Such clinically significant autoimmune haemolysis as a presenting feature, rather than a late complication, has never before been reported in the literature.

  • haematology (incl blood transfusion)
  • pneumonia (infectious disease)

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Legionnaires’ disease, a pneumonic illness caused by Legionella pneumophila, is well known to cause atypical pneumonia. It is precisely because of its ‘atypical’, non-specific and variable multisystemic involvement that this cause of pneumonia is often missed. We report here a case of sepsis with multiorgan dysfunction and severe haemolysis at presentation which was due to Legionella. Though the patient improved with empirical broad spectrum antibiotics (which included quinolones) and aggressive supportive care in the intensive care unit (ICU), the diagnosis was not suspected till much later when he was transferred to the ward.

Though immune-mediated haemolysis has been commonly reported in literature in association with certain infections, autoimmune and neoplastic conditions, it has rarely been reported in association with legionella pneumonia. Specifically, warm antibodies haemolytic anaemia (wAIHA) has been reported twice and severe cold autoimmune haemolysis (cAIHA) reported only once before. This is only the second reported case of clinically significant severe complement-mediated haemolytic anaemia associated with legionella pneumonia and the first ever case of AIHA as its presenting feature.

This case report underlines the high propensity to miss the diagnosis of legionella pneumonia due to its varied presentations, which in turn may have grave consequences. It also reports a hitherto unreported phenomenon of autoimmune complement-mediated haemolytic anaemia as a presenting feature of legionella pneumonia.

Case presentation

A 42-year-old man was brought to the hospital with recent onset shortness of breath and reduced level of consciousness. History was limited as he was a prison inmate but he was reported to be generally unwell for the past week with poor oral intake, decreased activity and fever. His medical history was significant for diabetes, and he had a history of pulmonary tuberculosis treated in 2015. He was also known to be positive for anti HCV antibody since 2007.

At the emergency, he was in a critical condition. He was hypothermic (35°C), bradycardic, hypotensive and hypoxaemic in severe respiratory distress. He was confused but without any focal neurological deficit. General physical examination revealed significant pallor and icterus. There was no rash, arthritis, orogenital ulcers or lymphadenopathy. Respiratory system examination was suggestive of left lower lobe consolidation. Cardiovascular and abdominal examination did not reveal any significant abnormality. Per rectal examination was negative for melena. A urinary catheter inserted in the emergency room drained cola coloured urine.

Initial laboratory tests (table 1) were suggestive of severe sepsis with liver dysfunction, deranged coagulation and significant anaemia with evidence of haemolysis. Chest radiograph showed left lower zone consolidation.

Table 1

Laboratory profile

He was mechanically ventilated and shifted to the medical ICU as a case of severe sepsis of suspected pulmonary origin. He received full supportive care in medical ICU including ventilatory support, vasopressors and broad-spectrum intravenous antibiotics including levofloxacin. He developed acute renal failure for which he required continuous renal replacement therapy. He received multiple blood transfusions. He was eventually extubated on day 6 of his admission and shifted to the medical ward on day 9. All cultures including blood, urine and tracheal aspirate were negative and no definite aetiology for sepsis was identified. Mycoplasma antibody was negative.

On reviewing his case in the medical unit, we considered the possibility of Legionella pneumonia. It was confirmed by a positive urine antigen test and a significantly rising antibody titre from 1:100 on day 10 to 1:100 000 on day 20.

We further reviewed his labs for the possible cause of the severe anaemia at the presentation. Elevated lactate dehydrogenase (LDH), indirect bilirubin and reticulocyte count, low haptoglobin, blood film and positive direct antiglobulin test (DAT) were suggestive of immune-mediated haemolysis. The presence of cola coloured urine, haemoglobinuria (positive urine blood 5+ by strip without significant microscopic haematuria), significantly raised LDH and fragmented red blood cells (RBCs) on blood film suggested an intravascular component of haemolysis. DAT was repeated for further analysis. It was positive for complement components C3b and C3d and negative for Immunoglobulin G (IgG), usually a feature of cAIHA.

In the ward, he required a few more sessions of haemodialysis till his kidney function normalised. He also had evidence of ongoing haemolysis which was milder, but required two units blood transfusion over 1 week. A haematology consultation was done and they advised to give him steroids 1 mg/kg for the continuing haemolysis. Steroids were stopped after 1 week. His haemoglobin remained stable and he did not require blood transfusion during last week of his hospital course. His clinical condition improved significantly as did his laboratory profile including renal and liver functions. He was discharged in a stable condition almost 4 weeks after his initial critical presentation to the hospital.

Investigations

Blood film (day 1)

Severe normocytic anaemia with spherocytes, occasional fragmented RBCs and nucleated RBCs. Neutrophilic leucocytosis with few polymorphs showing cytoplasmic vacuolation and toxic granulation. Adequate numbers of platelets.

Radiology

Chest X-ray—left lower zone consolidation (figure 1).

Figure 1

Chest radiograph on day 1: left lower zone consolidation (AP- Anteroposterior view)

Ultrasound abdomen: unremarkable.

Differential diagnosis

This patient presented with severe sepsis with multiorgan dysfunction and vere haemolytic anaemia.

Further workup for the cause of sepsis was confirmatory for Legionella pneumonia with positive urine antigen (95%–100% specificity) and exponentially rising titres of legionella antibodies. There was no other additional source of sepsis or coexisting infection as evidenced by negative blood, urine and respiratory cultures (table 2).

Table 2

Specific investigations

The other significant question here is the aetiology of haemolytic anaemia. The positive DAT was suggestive of haemolytic anaemia. Further analysis showed positivity for complement components C3b and C3d and negative for IgG which is usually a feature of cAIHA. But rarely complement activation can occur in wAIHA with non-IgG (IgM of warm antibody class). The main criterion differentiating between complement-mediated cAIHA and non-IgG wAIHA is the cold antibody titre. This was not done in our patient, but given the rarity of IgM-mediated complement activation in wAIHA it is more likely to have been cAIHA.

AIHA can be primary when the cause is unknown or secondary when it is associated with various viral and bacterial infections, autoimmune, malignant or lymphoproliferative processes (table 3).

Table 3

Classification of autoimmune haemolytic anaemia (AIHA)

Absence of any significant drug history and presence of immune haemolysis prior to blood transfusion ruled out drug-induced haemolytic anaemia and transfusion reactions, respectively.

Mycoplasma pneumoniae is well known to be associated with cAIHA. However, serum M. pneumoniae IgM titres were negative. Other specific infectious conditions known to be associated with AIHA such as Ebstein-Bar virus (EBV), Cytomegalovirus (CMV) and Human immunodeficiency virus (HIV) infections were also ruled out with negative results. Negative blood, urine and respiratory cultures ruled out haemolysis secondary to sepsis due to other bacterial infections. Hepatitis C is associated with cryoglobulinaemia. The patient was anti-HCV antibody positive, but his hepatitis C RNA was undetectable by PCR, ruling out active hepatitis C infection.

Non-infectious diseases presenting with AIHA like rheumatoid arthritis, systemic lupus erythematosus and underlying lymphoid malignancy were excluded on the basis of negative history, clinical features and radiological examination.

Thus, we could conclude with a fair amount of certainty that our patient had severe sepsis with multiorgan dysfunction due to legionella and the associated rare and clinically significant complement-mediated haemolytic anaemia was also due to the same organism.

Treatment

Treatment of autoimmune haemolytic anaemia depends on the type of the antibodies and whether it is primary or secondary. It is therefore important to determine the subtype of AIHA.

In wAIHA, the role of glucocorticosteroids, rituximab, immunosuppressive agents and splenectomy is well established. In comparison, cAIHA is more challenging to manage and the above modalities have not been found to be uniformly effective. Glucocorticoids and splenectomy are only effective in select patients with primary cAIHA. Glucocorticoids do downregulate phagocytosis by macrophages in extravascular haemolysis but do not block production of monoclonal antibodies. Splenectomy is ineffective as the liver is the main site of extravascular haemolysis in cAIHA. Rituximab, a monoclonal antibody, alone or in combination with other cytotoxic agents has been found to be useful. All the above therapies are recommended for primary AIHA. Generally, secondary AIHA responds to the treatment of the underlying conditions. In secondary wAIHA, additional medical therapy, for example, steroids is useful. However, in secondary cAIHA none of the above-mentioned therapies has been proven to have a beneficial role. The haemolysis secondary to an infectious aetiology is usually mild and self-limited. Secondary cAIHA usually resolves spontaneously, typically over 2–4 weeks following resolution of infection.

Our patient received intravenous levofloxacin for 5 days for the treatment of his pneumonia. He was given only a short course of prednisolone (1 week). His haemolytic markers continued to improve after his infection started to resolve.

Outcome and follow-up

At discharge, he was back to his baseline status and generally well. His renal and hepatic parameters had normalised and haemoglobin was stable. He did not return for a follow-up visit.

Discussion

The importance of this case of complicated atypical pneumonia due to Legionnaires’ disease is its presentation with severe complement-mediated haemolytic anaemia which has never been reported before in literature.

Legionnaires’ disease, caused by an intracellular bacteria L. pneumophila, is a multisystem illness that involves pulmonary, gastrointestinal, central nervous system and to a lesser extent renal system.1–3 It is predominantly a pneumonic illness that accounts for 2%–9% of cases of community acquired pneumonia4 with a mortality rate of about 1%–10%.5 6

Historically well known for its epidemic outbreaks, this multisystemic illness largely remains underdiagnosed, underinvestigated, underestimated and under-reported.7 8 Termed as an often-invisible problem, it is usually lower on the clinicians' list of differentials. It is generally assumed to be a relatively rare disease characterised by deadly outbreaks, which get reported widely in the media.8 The reality, however, is that community acquired Legionella pneumonia typically occurs sporadically9 and has been reported to represent as much as 65%–82% of the cases in a European study.10 According to WHO, the identified incidence of Legionnaires’ disease varies widely according to the level of surveillance and reporting.11

In wAIHA, complement-mediated haemolysis is seen in a proportion (upto 50%) of patients usually in conjunction with IgG—the most common subtype of warm antibody(figure 2). Warm-type IgM antibodies have been reported, and though their exact incidence is uncertain they are unarguably rare. Thus isolated detection of C3d in the absence of IgG is highly suggestive of cAIHA which is almost exclusively mediated by IgM induced complement activation. However, we do not have cold agglutinin titre to definitely diagnose cAIHA.

Figure 2

Extravascular haemolysis in warm autoimmune haemolytic anaemia. C, complement; RBCs, red blood cells.

Warm AIHA is the more common form of AIHA and often causes significant haemolysis, but usually not through non-IgG-mediated complement activation.12

On the other hand, cAIHA, an entirely complement-mediated process, is the less common form of autoimmune haemolytic anaemia. The haemolysis in cAIHA is often mild. This is because of two main reasons. First the cold agglutinins characteristically agglutinate RBCs at an optimum temperature of 3°C–4°C. However, if their thermal amplitude (defined as the highest temperature at which CA reacts with antigens) exceeds 28°C–30°C, they will be pathogenic and cause RBC agglutination in the circulation, even at mild ambient temperatures.13 14 Second, cold agglutinins usually cause extravascular haemolysis by activation of the classical complement pathway. However, intravascular haemolysis can occur through activation of C5 (complement 5) mediated terminal (lytic) pathway and subsequent formation of membrane attack complex (figure 3). RBC destruction rate in intravascular haemolysis (calculated as up to 200 mL of RBCs in 1 hour) is significantly higher than in extravascular haemolysis and can cause severe anaemia. A significantly raised LDH, red cell fragments on smear, and the presence of haemoglobinuria suggests a predominant intravascular component to the haemolytic process.15–18

Figure 3

Complement-mediated cold autoimmune haemolysis. Solid arrows—(major) classical pathway. Dotted arrows—(minor) terminal pathway. C, complement; MAC, membrane attack complex; RBC, red blood cell.

Cold AIHA has been seen in association with many infectious agents. M. pneumoniae is the most frequent cause of secondary cAIHA (8% of the cases of AIHA) followed by EBV infection (1% of the cases of AIHA). Patients with M. Pneumoniae infection usually produce CA as part of their physiologic immune response. Clinically significant haemolysis occurs occasionally, due to the production of high thermal amplitude CA. It is reported to begin usually around 2 weeks after onset of infection and resolves by 2–3 months. Several series and case reports have described transient secondary AIHA after cytomegalovirus infection. Rarely, CA-mediated haemolysis has been reported in adenovirus infections, influenza A, varicella, rubella, listeriosis and pneumonia caused by Chlamydia species.19–21

Sepsis has been reported to cause haemolytic anaemia through various mechanisms including immune-mediated haemolysis which are mostly of warm AIHA. Cold AIHA secondary to sepsis appears to be a rare entity and has been reported specifically with Escherichia coli and Klebsiella pneumoniae.22 23

Another case report published in 2015, attributes sepsis to cold agglutinin-mediated haemolysis. However, there is no mention of any specific infectious agent and tests for L. pneumoniae and M. pneumoniae were not performed.24

Regarding Legionella, association with wAIHA has been reported twice in literature. In an epidemiologic review of the initial Philadelphia outbreak, a case of haemolytic anaemia was reported. However, a DAT was not done. In the first ever case reporting this association (1983), the patient developed warm AIHA a few days after his initial admission with Legionella pneumonia.25The second case of wAIHA as a delayed complication of severe legionella pneumonia was published recently. IgG-mediated extravascular haemolysis was reported to occur as a late complication. The patient was treated with prednisolone.26

Cold AIHA associated with Legionella pneumonia has only been reported once in literature. In 1980, an 80-year-old male with legionella pneumonia was reported to have developed cold AIHA into the third week of his illness.27

The case of our patient stands out as the first ever case reporting the association of Legionella pneumonia and severe complement-mediated haemolysis as an initial presenting feature of this illness.

Learning points

  • High index of suspicion is required for diagnosis of sporadic cases of Legionnaires’ disease in any patient presenting with pneumonia complicated by multisystemic involvement.

  • Autoimmune haemolytic anaemia is a rare associated feature of Legionnaires’ disease.

  • Severe and clinically significant complement-mediated autoimmune hemolysis can be an initial presenting feature of Legionnaire’s disease.

Ethics statements

Acknowledgments

The completion of this manuscript could not have been possible without the guidance and assistance of dear friend and colleague Dr Liza Mariam Thomas. Her contribution is sincerely appreciated and gratefully acknowledged.

References

Footnotes

  • Contributors US wrote the manuscript and NAS reviewed and critically revised the manuscript. AMMA and FA contributed to the compilation of investigations and drafted tables. All the authors reviewed and agreed on the final manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.