Rare causes of haemoptysis in suspected pulmonary embolism

When a patient presents with haemoptysis and pleuritic chest pain, a pulmonary embolism is an important and common diagnosis to consider. There is a tendency in busy medical admissions units to start treatment of certain conditions without thorough investigation, with the intention of reducing delays in starting treatment. Patients are usually treated for suspected pulmonary embolism with heparin early to reduce mortality and morbidity. It is important, however, to remember other less common causes of haemoptysis.

Case report

A 59 year old woman was admitted to our medical assessment unit with chest pain and haemoptysis. She had experienced pleuritic left sided, chest wall pain intermittently for the previous week, with gradually increasing intensity. On the day of admission she had also produced about a cupful of bright red blood while coughing. She had no medical history of note, except that she was a smoker.

On admission the patient was in distress but not objectively dyspnoeic—her respiratory rate was not raised and her oxygen saturations on air were 97%. She did not show any signs of shock; she had no tachycardia and her systolic blood pressure remained around 120 mm Hg throughout. Chest examination showed some decreased air entry at her left lung base with a loud pleural rub. She had no cardiac murmurs, and no clinical evidence was found of a deep venous thrombosis in her legs or pelvis.

The admitting doctor's differential diagnosis included pulmonary embolism, pneumonia, and carcinoma of the lung; a plain chest radiograph and routine blood tests were ordered including a d-dimer assay.

The chest x ray film (fig 1) showed moderate left basal shadowing with blunting of the right costophrenic angle. There was no cardiomegaly, and, importantly, the mediastinum appeared normal.

Fig 1.

Chest x ray film showing moderate left basal shadowing with blunting of the right costophrenic angle

Blood tests indicated a mild anaemia (haemoglobin 96 g/l), a raised white cell count (28.5, neutrophils 25.9) and a normal prothrombin time. A d-dimer assay was greatly raised at 3893 (normal < 250) ng/ml. Inflammatory markers were raised, with an erythrocyte sedimentation rate of 73 seconds and a C reactive protein of 411 mg/l. Her alkaline phosphatase was mildly raised, though corrected calcium and the rest of her liver enzymes were normal. Renal function was normal and blood cultures were subsequently reported as “no growth.”

Based on the chest radiograph and inflammatory markers, treatment was initially begun for a left basal pneumonia with oral co-amoxiclavs (amoxicillin 500 mg, clavulanic acid 125 mg). She was not started on low molecular weight heparin as is the case normally for suspected pulmonary embolism. She was admitted overnight and seen in the morning by the on-call consultant. She had made no improvement and needed opioids to control her pain.

The consultant on call was concerned that this was an atypical presentation for a pulmonary embolus and arranged urgent spiral computed tomography of the patient's thorax to elucidate the cause of the chest pain. The tomogram showed an aneurysm of the lower part of the descending thoracic aorta with surrounding haematoma (fig 2). She was immediately transferred to the local cardiothoracic unit for surgery.

Fig 2.

Spiral computed tomogram showing aneurysm of lower part of descending thoracic aorta with surrounding haematoma

Through a posterolateral thoracotomy and with partial cardiopulmonary bypass, the aneursymal segment of the aorta was replaced with a Dacron tube graft. There were no postoperative complications. A mixed growth of Gram positive and Gram negative cocci was cultured from the contents of the aneurysm, and so although the patient was clinically well, she was treated with a course of intravenous antibiotics.

Discussion

Thoracic aneurysms are rare, estimated at 6 per 100 000 a year.¹ In contrast, pulmonary embolism is more common (60 to 70 cases per 100 000 a year).² We know that “common things occur commonly,” and since pulmonary emboli have a considerable mortality associated with them, there is a great temptation to treat before the diagnosis is confirmed. The British Thoracic Society recommends initial investigation with d-dimer assays and, if combined with a high clinical suspicion, treatment with low molecular weight heparins until imaging is available. In this case, however, to have given any kind of anticoagulation might have proved fatal.

The society's guidelines also state that if a pulmonary embolus is ruled out (or thought unlikely) then high resolution or multislice computed tomography is necessary to identify the true nature of the pain.

The clues that pointed this case towards an “atypical” chest pain were the high white cell count; the large level of haemoptysis; no significant hypoxia; no evidence of deep venous thrombosis on clinical examination; and a pain much more severe than would be expected for pulmonary embolism.

This mode of presentation is also somewhat unusual for a thoracic aortic aneurysm in terms of the nature of the pain and the presence of haemoptysis. The location of the aneurysm was such that it was not apparent on the plain chest radiograph. The identification of micro-organisms would suggest an infective aetiology.

We advise to always consider alternative diagnoses of pulmonary embolism if atypical features are present.