Pulmonary haemorrhage is a common necropsy finding in acute leukaemia, however, it is rarely diagnosed during life. A man with acute myeloid leukaemia is reported who presented with disseminated fungal infection, anaemia, thrombocytopenia, and subconjuctival and petechial haemorrhages. During the course of the patient's illness, the chest infection was complicated with bilateral pulmonary haemorrhage. The diagnosis of pulmonary haemorrhage was based on characteristic clinical and radiological findings. The patient improved on treatment.
- pulmonary infiltrate
Statistics from Altmetric.com
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.
A 52 year old man presented with the complaints of fever, cough, weakness, oral ulcers, painful swallowing, hoarseness of voice, and painful defecation of two weeks' duration. On examination, he was febrile and had severe pallor; his general condition was poor. Severe thrush was seen over the soft palate and oropharynx. There was generalised petechial rash on the skin along with right subconjunctival haemorrhage and proctitis. There was no organomegaly or peripheral lymphadenopathy.
On investigation, the blood haemoglobin concentration was 53 g/l, total leucocyte count 16.6 × 109/l, and platelet count 28.0 × 109/l. Peripheral blood smear revealed 90% myeloblasts. Cytochemistry was suggestive of acute myeloid leukemia (AML M2). Indirect laryngoscopy showed sloughing ulcers over the epiglottis. A chest radiograph was obtained (fig 1) and showed the presence of diffuse, poorly defined nodular interstitial opacities in both lungs. The transverse fissure was thickened. Based on these features, the diagnosis of acute myeloid leukaemia with disseminated fungal infection was made.
He was put on low dose cytosine arabinoside along with amphotericin B, ceftazidine, amikacin, and metronidazole. Daily transfusion of 200 ml of concentrated platelets was given for three days in view of thrombocytopenia and presence of petechial and subconjunctival haemorrhages. He received two units of blood transfusion in addition. There was mild improvement in his general condition over a few days and he became afebrile. The oral thrush decreased and he was able to swallow semisolid food.
A further chest radiograph was obtained on the 19th day (fig 2) and this showed bilateral, segmental, asymmetric airspace consolidation. It was located in bilateral mid and lower zones and was more pronounced in the perihilar region. The periphery of the lungs were spared and a thick band of the aerated lung was seen separating the area of consolidation from cardiac sillhouette. There was no dyspnoea or haemoptysis. The platelet count was 35.0 × 109/l at this stage. Coagulation testing was not performed. Bronchoalveolar lavage was not performed because of his poor general condition and the presence of severe oral and pharyngeal thrush.
A third chest radiograph was obtained five days later (fig 3) and showed significant clearance of bilateral airspace consolidation. Chemotherapy and antimicrobial therapy were continued. The general condition and oral thrush improved further. Bone marrow biopsy performed on 37th day of admission confirmed complete remission. The haemogram became normal. He was subsequently discharged with an appointment for second consolidation chemotherapy but was lost to follow up.
The radiographic differential diagnosis included diffuse pulmonary haemorrhage (DPH) and chest infection. In a patient with acute leukaemia and oral thrush, the diffuse abnormalities seen on the first chest radiograph should suggest the diagnosis of candida pneumonia.
In a patient with acute leukaemia and a chest infection, the sudden appearance of bilateral consolidation with characteristic features as seen in the second chest radiograph should suggest acute pulmonary haemorrhage as the most likely cause of consolidation. Clinical evidence of petechial haemorrhages, laboratory evidence of thrombocytopenia, and rapid resolution the consolidation as seen on the third radiograph further substantiate this diagnosis. Segmental distribution of consolidation as well as its sudden appearance and rapid clearance make fungal pneumonia an unlikely cause of consolidation.
DPH is a syndrome with few defining clinical and radiological features.1 It is characterised by widespread haemorrhages from the microvasculature of the lung into the alveolar spaces. While uncommon, it can occur in a wide variety of clinical disorders. The usual causes are Goodpasture's syndrome, connective tissue disorders with vasculitis, and idiopathic pulmonary haemosiderosis among others (box 1).2 It can also occur in variety of bleeding disorders including thromocytopenia in the context of leukaemia.1 3 4 A triad of anaemia, haemoptysis, and diffuse alveolar consolidation on chest radiography strongly suggests DPH.1 4 5 As haemorrhages occur distal to the mucociliary escalator mechanism, many patients with DPH do not have haemoptysis.4 5 Even in absence of haemoptysis, DPH should be suspected on the basis of chest radiographs if the patient has one of the diseases known to be associated with pulmonary haemorrhage.5 Radiographic changes of the acute alveolar haemorrhage are the same regardless of the aetiology and consist of airspace consolidation.1 2
Box 1: Causes of diffuse pulmonary haemorrhage2
Antiglomerular basement membrane disease (Goodpasture's syndrome)
Connective tissue disorders with systemic vasculitis: systemic lupus erythematosus, Wegener's granulomatosis, Henoch-Schönlein disease, rheumatoid disease
Idiopathic pulmonary haemosiderosis
Rare causes: blood dyscrasias, drugs, gases, tumours, trauma, necrotising pneumonia
Bone marrow transplant
Chest radiographs in DPH typically show the sudden appearance of diffuse alveolar filling pattern that ranges from patchy alveolar shadows to widespread confluent consolidation with air bronchgram.1 2 4 5 Initially the opacities may have an interstitial pattern but often progress to airspace consolidation as seen in our patient. The consolidation can be widespread or show a perihilar or mid to lower zone predominance and tends to be more pronounced centrally.2 5 It is usually bilateral and asymmetric but may be unilateral in distribution. 1 3 4 The lung periphery, apices, and costophrenic angles are typically spared.4 5As the blood is being absorbed into the interstitium, the consolidation clears rapidly, often within two or three days, either completely or partially, to leave a reticular pattern or ground glass haziness.2 4 Unless bleeding recurs, this interstitial opacities are also transient and clear completely within two weeks from the beginning of the episode.2 4 Computed tomography may be helpful and is indicated in patients with haemoptysis in whom a focal cause of haemorrhage such as bronchiectasis or tumour is suspected but it plays a limited part in patients with DPH.1 The predominant findings on computed tomography are diffuse nodular opacities or ground glass haze.
The radiographic abnormalities are not specific for DPH since diffuse bilateral opacities can be caused by any substance filling the alveoli, for example oedema fluid (pulmonary oedema) or inflammatory exudate (pneumonia). Fortunately, the traditional clinical and radiographic criteria are present in most cases of DPH and the diagnosis can be made with some degree of confidence.5 When doubt exists or when treatment options would differ, transbronchial or open lung biopsy may be required to make the definitive diagnosis.1 5
PULMONARY INFILTRATES IN LEUKAEMIA
The chest radiographs of patients with leukaemia can be difficult to interpret. These patients are often febrile and have impaired immunological and microbicidal defences as well as bleeding tendencies. The usual causes of pulmonary infiltrates in patients with leukaemia are infection and haemorrhage. Other conditions which need to be included in the differential diagnosis are leukaemic involvement, alveolar proteinosis, and adverse drug reactions (box 2).6Although pulmonary leukaemic infiltrates are commonly found at necropsy, symptomatic pulmonary disease is uncommon. Alveolar proteinosis is uncommon but a well known cause of alveolar opacities in leukaemia and is thought to be a result of impaired phagocytosis by defective macrophages resulting in accumulation of cellular debris and phospholipids in alveoli.6 Rarely, pulmonary infiltrates can also occur as an adverse reactions to certain anticancer drugs such as cytosine arabinoside, busulfan, and methotrexate. A diagnosis of drug induced pulmonary disease requires exclusion of other causes and demonstration of histological features specific to the drug effect.6
Box 2: Causes of pulmonary infiltrates seen on chest radiography in patients with leukaemia6
Opportunistic tumours (Richter's transformation)
Infection, particularly by opportunistic pathogens, is usually the first and most important consideration after the appearance of infiltrates on chest radiographs of a leukaemic patient. Whereas common bacterial pathogens are responsible for most of the focal infiltrates, opportunistic organisms tend to produce more diffuse abnormalities.7 Bilateral and relatively uniform opacities are usually caused by pneumocystis, viruses, and disseminated fungal or mycobacterial infections. Pulmonary candidiasis may be suspected in presence of oral thrush.8 The usual radiographic appearance of candida pneumonia is of multiple nodular or patchy infiltrates in both lungs. These are usually poorly defined and do not follow a segmental distribution. Multiple nodular abscesses may be present occasionally.9 A diffuse, miliary, nodular pattern may also be seen.1 In most patients, however, there are no specific features, partly because of high frequency of other coexisting infections, oedema and haemorrhages. Hence early institution of antifungal therapy based primarily on clinical grounds is recommended.9 Candida pneumonia is usually a part of disseminated candidiasis and has a poor prognosis.
The radiographic findings in aspergillus and mucoraceae pneumonia are also non-specific. They commonly produce solitary or multiple areas of pneumonia that slowly increase and may ultimately produce pulmonary infarction.1 4 Pulmonary infarction may be seen as a peripheral wedge shaped, pleural based area of consolidation. Viral infections produce interstitial infiltrates or bilateral extensive and diffuse airspace consolidation.
DPH IN LEUKAEMIA
Pulmonary haemorrhage is the most common cause of non-infectious pulmonary infiltrates in patients with leukaemia.7 Thrombocytopenia is universal in acute myeloid leukaemia and platelets also show abnormalities of morphology and function. Alveolar haemorrhage is often found at necropsy in patients with leukaemia.6 7 However, it is usually not suspected or diagnosed clinically before death because haemoptysis is absent in most patients and chest radiographs often have a non-specific appearance.6 10 Although DPH associated with thrombocytopenia alone does occur, it is more commonly a complication of diffuse alveolar damage caused by coexistent sepsis or thrombosis of blood vessels by invasive fungal infection.1 4
The radiographic abnormalities of pulmonary haemorrhage in patients with leukaemia are usually bilateral. Initially the opacities may have interstitial pattern but often progress to bilateral airspace consolidation, which tends to clear rapidly.7 These typical radiographic findings are clearly exemplified in the present case. Radiographic changes of uncomplicated acute DPH in patients with leukaemia are the same as those from other causes.2However, most patients have non-specific appearance because of coexisting infection.1 2 The risk of uncontrolled bleeding makes invasive diagnostic procedures such as lung biopsy and bronchial brushing hazardous.10 The presence of increased haemosiderin content in the macrophages obtained by bronchoalvelar lavage may provide indirect evidence of intrapulmonary haemorrhage.10 However this procedure may not be possible in patients with severe oropharyngeal thrush and the dignosis may remain presumptive as in our patient. Although DPH occurring in an immunocompromised patient is often terminal,1 prompt institution of treatment in our patient resulted in marked clinical and radiographic improvment.