Published ahead of print on March 1, 2007, doi:10.1164/rccm.200609-1362OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200609-1362OC
Cryptogenic HemoptysisFrom a Benign to a Life-threatening Pathologic Vascular Condition1 Service de Pneumologie et Unité de Réanimation Respiratoire, 2 Service de Radiologie, and 3 Service d'Anatomie Pathologique, Hôpital Tenon, Assistance Publique–Hôpitaux de Paris and Université Pierre et Marie Curie, Paris, France Correspondence and requests for reprints should be addressed to Muriel Fartoukh, M.D., Hôpital Tenon, Assistance Publique–Hôpitaux de Paris, 4 rue de la Chine, 75020 Paris, France. E-mail: muriel.fartoukh{at}tnn.aphp.fr
Rationale: Data on hemoptysis of unknown origin (i.e., "cryptogenic") are scarce and the outcome of patients affected is controversial. Objectives: To describe the clinical spectrum and course of patients with cryptogenic hemoptysis, as well as pathologic findings when surgery is performed. Methods: A cohort of 81 patients referred for cryptogenic hemoptysis after clinical evaluation, chest radiography, fiberoptic bronchoscopy, and computed tomographic scan to a respiratory intermediate care and intensive care unit, from December 1995 to August 2004, with a prospective follow-up by visit or telephone interview. Measurements and Main Results: The 81 patients (69 males) had a mean cumulative volume of hemoptysis averaging 190 ml on admission. First-line conservative measures and bronchial artery embolization controlled hemoptysis in 73 patients (90%). Emergency surgery was performed in six patients (7%) because of failure of bronchial artery embolization, and secondary surgery was scheduled in a seventh patient. A total of 73 patients were followed for a mean of 47 (± 35) months. No lung cancer developed. Hemoptysis recurred in 10 patients (4 within the first year; 6 between 1 and 8 yr later), 2 of whom underwent surgery. A specific bronchial vascular involvement (Dieulafoy disease) was demonstrated in five of the nine patients who had undergone surgery, especially in those with high amounts of bleeding. Conclusions: Cryptogenic hemoptysis may be a life-threatening condition. Nonsurgical approaches provide immediate control of bleeding in most patients with cryptogenic hemoptysis, with few recurrences in both short and long terms. Dieulafoy disease of the bronchus, unsuspected after routine imaging investigations, may be involved in a subset of patients yet to be determined.
Key Words: bronchial artery embolization • cryptogenic hemoptysis • Dieulafoy disease of the bronchus • management • outcome
Hemoptysis refers to a wide clinical spectrum, ranging from a nonalarming bloody expectoration to a life-threatening condition associated with an immediate risk of airway obstruction and death. The challenge for the physician caring for patients with hemoptysis is to assess the severity and to identify the underlying cause of bleeding, which may influence both management and outcome. The diagnosis of hemoptysis is usually based on the combination of physical examination, chest X-ray, fiberoptic bronchoscopy, and computed tomography (CT) scan. Despite the currently available diagnostic tools, some 7 to 25% cases of hemoptysis remain without identified etiology (i.e., "cryptogenic" hemoptysis), whatever the severity of the bleeding (1–6). Recently, Herth and colleagues (2) reported a substantial incidence of bronchial carcinoma complicating the course of cryptogenic hemoptysis, whereas older studies performed before the era of CT scanning suggested a favorable outcome (7–9). The aims of our study were to describe the clinical spectrum, the management, and both the short- and long-term outcome of a large cohort of consecutive patients diagnosed with cryptogenic hemoptysis, after history, physical exam, chest X-ray, fiberoptic bronchoscopy, and CT scan had excluded any specific lesion of the lung parenchyma that could have been identified as the source of bleeding. The study was conducted in accordance with French law, which does not require approval of an institutional review board or the consent of patients for such retrospective analysis of medical records.
Selection of Patients The study was conducted between December 1995 and August 2004 in Tenon hospital, a tertiary university hospital and referral center for hemoptysis. All consecutive patients admitted to the respiratory intermediate care ward and intensive care unit (6 and 10 beds, respectively) for hemoptysis were prospectively screened. Definite causes of hemoptysis were bronchiectasis, tuberculosis (active and sequels), cancer, and mycetoma; probable causes were pulmonary embolism, infectious pneumonia, and emphysema. Only patients diagnosed with cryptogenic hemoptysis were considered in this study. Patients with recurrent hemoptysis were included in the first episode. A fiberoptic bronchoscopy was performed in all patients within the first 24 hours of admission to assess the presence and location of bronchial blood, active bleeding, and clot. To exclude any specific anatomic lesion as the source of bleeding, chest X-rays and CT scans were reviewed by a radiologist (A.K.) unaware of the patients' clinical condition. Alveolar infiltrate(s) on chest X-ray were considered as abnormalities locating the bleeding. Although all patients had a CT scan performed, only 60 were available for review by the radiologist and were analyzed, considering ground-glass, alveolar, and nodular opacities to be reflections of the filling of the alveolar lumen with blood.
Management of Hemoptysis
Analysis of Outcome
Statistical Analysis
Demographic and Clinical Presentation on Admission During the study period, 84 of 653 patients (13%) referred to our institution for hemoptysis were diagnosed as having cryptogenic hemoptysis, and 3 of them were secondarily excluded because of missing information. The remaining 81 patients (69 males) were aged 47.9 (± 12.3) years. The time from bleeding onset to hospitalization or referral to our unit was 1.75 days (± 3 d; median, 1 d) and 2.7 days (± 4.2 d; median, 1.5 d), respectively. A smoking history was recorded for 64 patients (79%), 54 of whom were former smokers (Table 1). The cumulative volume of blood was 190 (± 195) ml on admission (range, 10 to > 1,000 ml; median, 100 ml) (Figure 1).
Chest X-ray, Fiberoptic Bronchoscopy, and CT Scan A total of 16 patients (20%) had a single lobar alveolar infiltrate on chest X-ray. The fiberoptic bronchoscopy showed bronchial blood in 75 patients (93%), 32 of whom had localized active bleeding. The CT scan evidenced ground-glass opacities in most patients (n = 47; 78%). A repeated CT scan was performed several months later in 29 patients, demonstrating a complete resolution of the initial abnormalities related to the bleeding.
Management
Overall, the initial therapeutic approach based on conservative measures alone or completed BAE led to an immediate control of bleeding and to hospital discharge in 31 of 31 (100%) and 39 of 43 (91%) patients, respectively. The mean hospital length of stay of patients initially managed conservatively or receiving BAE differed significantly (3 ± 1.5 vs. 6.2 ± 4.9 d; p < 0.001).
Surgery
Follow-up Follow-up data were available for 73 patients (90%), for a mean duration of 47.3 (± 35) months (range, 0.4–133 mo) after initial referral. No lung cancer developed. There were 10 episodes of recurrence of bleeding, which occurred mostly after 1 month (median, 20 mo) (Table 2). Hemoptysis recurred at a similar rate in patients managed conservatively and in those in whom the initial BAE was completed (n = 6/34 vs. 3/43; p = 0.17). The overall cumulative rate of control of bleeding without recurrence was 98.5% at 1 month, 93.6% at 1 year, and 86.3% at both 2 and 4 years.
Our study aimed at characterizing the clinical spectrum and the course of cryptogenic hemoptysis observed at a single referral center over a 9-year period. One-third of our patients had a cumulative volume greater than 200 ml on admission, emphasizing that cryptogenic hemoptysis may be life-threatening. The first-line management was similar to that currently proposed in hemoptysis associated with a specific cause, favoring conservative measures and interventional radiology over emergency surgery. There were relatively few episodes of recurrence. Surgery was mainly performed in cases of failure of initial BAE and recurrences; unexpectedly, a specific bronchial vascular involvement (Dieulafoy disease) was frequently evidenced on pathologic examination. Recent studies focusing on hemoptysis of various severities have reported a 7 to 25% incidence rate of cryptogenic hemoptysis, even with the use of CT scan (1–5, 10). We found a comparable rate of 13% of cryptogenic hemoptysis, after performing complete clinical evaluation, chest X-ray, fiberoptic bronchoscopy, and CT scan. The clinical presentation of cryptogenic hemoptysis usually refers to minor and chronic bleeding with a benign course (7–9), although recent studies have reported life-threatening cases (3–5). Our study suggests that cryptogenic hemoptysis may commonly present as a life-threatening condition, as one-third of our patients had a volume greater than 200 ml on admission, and surgery was needed to control the bleeding in a few of these. The management of hemoptysis has evolved during the past decade, favoring a least-invasive therapeutic approach (i.e., conservative measures and interventional radiology over emergency surgery). The efficacy and safety of BAE have been established for controlling hemoptysis of various etiologies and severities, both in the short and long term, except for that related to aspergilloma and lung cancer (4, 5, 11–13). However, there is no available information for cryptogenic hemoptysis. In our study, BAE was first attempted in 50 patients (62%) and, when completed, was associated with an immediate control of bleeding in 90% of patients. Altogether, failure of BAE or noncontrol of hemoptysis despite a completed BAE occurred in 11 patients (22%) in whom BAE was first attempted, 6 of whom underwent emergency surgery. Finally, the initial "nonsurgical" approach led to an immediate control of bleeding and to hospital discharge in 73 patients (90%) (Figure 2). There were 10 episodes of recurrence (12%) after discharge, occurring between 1 month and 8 years. No lung cancer occurred during the follow-up. Previous studies have suggested a favorable prognosis for patients with cryptogenic hemoptysis (7–9). Adelman and colleagues (7) reported a spontaneous but more progressive resolution in 67 patients with unexplained minor and chronic hemoptysis, although hemoptysis had resolved completely in only 43% of patients at hospital discharge. In a more recent study of 135 patients with presumed cryptogenic hemoptysis, Herth and colleagues (2) reported that lung cancer was diagnosed in 6% within 3 years after the initial bleeding episode. However, the subset of patients initially studied with CT scan was not specified. Nevertheless, these data suggest that exploration of the underlying lung parenchyma at the time of hemoptysis and during follow-up to exclude small lung carcinoma is warranted. Using multidetector row CT should be helpful in this regard (14–17). In our study, no specific lung parenchyma abnormality was evidenced on CT scan, either on initial evaluation or during follow-up, when performed. The diagnosis of cryptogenic hemoptysis is usually based on the normality of physical examination and endoscopic and radiologic exploration. Our study is the first to provide pathologic findings in this setting. An abnormal superficial vessel contiguous to the epithelium of the bronchial mucosa was demonstrated at pathologic examination of pulmonary resection in more than half of the patients undergoing surgery, mostly in those with high amounts of bleeding. This description corresponds to Dieulafoy's disease of the bronchus, a vascular condition first described in the stomach and regarded as being responsible for 1 to 2% of cases of severe upper gastrointestinal bleeding (18). Dieulafoy disease of the bronchus has been noted in a few case reports of massive hemoptysis (19–25). In our study, the total amount of bleeding (defined as the amount of bleeding on admission plus the subsequent amount of bleeding during hospitalization) was greater than 500 ml in 13 patients (16%), 5 of whom underwent surgery. Dieulafoy disease was demonstrated in four of those five patients who had undergone surgery. None of the preoperative information provided by fiberoptic bronchoscopy, chest X-ray, or CT scan could have suggested such pathologic findings. The incidence of Dieulafoy disease of the bronchus among patients diagnosed with cryptogenic hemoptysis, after a complete clinical exam and imaging investigations, including CT scan, remains unknown. Our results suggest that this vascular condition may be underestimated among patients with life-threatening hemoptysis presumed to be cryptogenic. The mechanisms underlying Dieulafoy disease of the bronchus remains unknown. One may hypothesize that chronic bronchial inflammation might participate in the pathogenesis of the disease, as most patients were smokers. In summary, the clinical spectrum of cryptogenic hemoptysis ranges from a benign to a life-threatening condition. The outcome appears to be rather good, both in the short and long term. A substantial subset of cryptogenic hemoptysis with high amounts of bleeding may reveal a specific bronchial vascular involvement that is not illuminated by the routine methods, including CT scan. Unfortunately, we could not determine any preoperative clinical or radiologic criteria suggesting such pathologic lesions.
Originally Published in Press as DOI: 10.1164/rccm.200609-1362OC on March 15, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form September 25, 2006; accepted in final form February 28, 2007
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
Failure of arteriography occurred in seven patients, four of whom underwent emergency surgery due to the amount of bleeding (range, 200–600 ml). 