- Email: [email protected]
Major and Massive Hemoptysis: Reassessment of Conservative Management
Major and Massive Hemoptysis: Reassessment of Conservative Management BY RALPH COREY, MO,
KHIN MAE HLA, MO
ABSTRACT: The etiologic factors of major (~200 ml/24 hr) and massive (~1,000 ml/24 hr) hemoptysis may well affect the outcome and, therefore, the treatment of this often life· threatening problem. The decline in the inci· dence of tuberculosis (TB) and bronchiectasis, along with the increase in bronchitis and neo· plasia, have led to a strong institutional bias against operating on patients with major and massive hemoptysis. A retrospective case study and an extensive literature review were under· taken to critically evaluate this policy. Fifty· nine consecutive patients with major hemopty· sis, 26 of whom had massive hemoptysis, were identified from 887 patients seen in our institution over a 10.year period. Only four of these 59 patients underwent surgery, while 55 were managed conservatively. Etiologic factors, operability, and bleeding rate all appeared to playa major role in outcome. No patients with bronchitis, bronchiectasis, tuberculosis, or who were on anticoagulation therapy died com· pared to a mortality rate of 59% in patients with carcinoma (CA) of the lung and 71% in patients with leukemia. Eleven percent of operable pa· tients treated conservatively died compared to a 46% mortality rate for nonoperable patients. And, 9% of patients with bleeding rates <1,000 m1l24 hr died compared to 58% of those with ~1,000 ml/24 hr. Conservative management appears to have a low mortality in patients with non·tuberculosis· related major hemoptysis as well as in many patients with massive hemoptysis, especially those patients who are operable and those with· out neoplastic disease. KEY INDEXING TERMS: Hemoptysis; Bronchitis; Carcinoma Of The Lung. [Am J Med Sci 1987; 294(5):301-309.]
From the Department of Medicine, Duke University Medical Center, Durham, North Carolina. The authors would like to acknowledge Dr. Patrick McKee for his help on this manuscript. Reprint requests: Ralph Corey, MD, Box 3038, Department of Medicine, Duke University Medical Center, Durham, NC 27710. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
C
onservative management of patients with major and massive hemoptysis has been associated with increased mortality in several studies over the last 20 years. 1- 9 This led to the recommendation of an aggressive, early surgical approach for this problem by most investigators, although more recently, advocates of conservative management have reappeared,10-12 as have innovators of the important new techniques of bronchial artery embolization,13,14 endobronchial tamponade,15 and endobronchial lavage. 2 In most reported series of major and massive hemoptysis, tuberculosis and bronchiectasis represent the majority of cases studied. 1,3-5,8,9,13 The decline in the incidence of TB16 and bronchiectasis,17 combined with an increase in the recognition of chronic bronchitis1? in both the general population as well as in the underlying etiologic factors of hemoptysis (Table 1), may make previous observations and recommendations concerning major and massive hemoptysis invalid. In an attempt to evaluate this hypothesis we reviewed all hemoptysis cases at our institutions over a 10-year period to determine (1) the current etiologic characteristics of major and massive hemoptysis; (2) the clinical and historical factors that have prognostic significance; and (3) the hospital course and outcome of patients treated primarily in a conservative manner for major and massive hemoptysis. A review of the largest series of patients with major and massive hemoptysis including their etiologic factors, management, and outcome is provided for comparison to the present report. PatIents and Methods
An extensive retrospective chart review was performed of all patients admitted to the Durham Veterans Administration and Duke University Medical Centers from January 1972 through December 1981 with a coded primary or secondary admission or discharge diagnosis of hemoptysis. Eight hundred eighty-seven patients with a diagnosis of hemoptysis were identified by the computer coding system at both institutions. The overall chart retrieval rate was 94.3%. Fifty-nine patients with major hemoptysis (~200 ml/24 hr), 26 of whom had massive hemoptysis (~1,000 m1l24 hr), were identified. Cross 301
Major and Massive Hemoptysis: Reassessment of Conservative Management
TABLE 1 Incidence and Cause of Hemoptysis Heller* Abbottt Levltt* Souders§ Moerschll 194818 1951 19 195220 195221 194617 (683) (105) (200) (416) (497) Bronchogenic Carcinoma Chronic Bronchitis Bronchiectasis Tuberculosis Other
Johnston~
195622 (324)
Boucot** Purseltt soll** Corey§§ 195923 1961 24 197825 1984 (395) (105) (200) (120)
2
21
12
3
24
4
8
19
13
13
15 7 40 26
2 21 22 34
3 15 50 20
12 29 2 54
9 27 6 30
17 13 10 59
19 4 7 64
5 24 13 40
55 0 3 29
50 2 3 32
* All patients with hemoptysis attending a hospital based chest clinic.
t Patients referred to a thoracic surgery seNice for hemoptysis. :j: Hemoptysis In patients admitted to
a tuberculosis sanatorium and to a general hospital.
§ All patients admitted to a referral hospital with a primary complaint of hemoptysis.
II All patients with hemoptysis who underwent bronchoscopy at a referral center. ~ All patients
with hemoptysis attending a hospital-based chest clinic.
* * Unselected males in the Philadelphia Pulmonary Neoplasm Research Project.
tt Consecutive patients with hemoptysis admitted to a community hospital. :j::j: Consecutive admissions for hemoptysis to
a metropolitan hospital.
§§ Random patients admitted to Duke University Medical Center with hemoptysis.
checks using major underlying etiologic characteristics such as bronchogenic carcinoma, tuberculosis, bronchitis, and lung abscess ensured completeness of case detection. Their charts were reviewed in detail by one of the investigators. Standardized data abstraction forms were used to record patients' demographic, clinical, laboratory, and radiological characteristics; underlying etiologic factors of hemoptysis; quantity and rate of hemoptysis; transfusion requirements; operability status; therapy instituted; and the patients' outcomes at discharge or follow-up. Quantification of bleeding amounts and rates was attempted only for those patients in whom the hemoptysis was observed directly. All patients in whom the amount of hemoptysis had not been specifically quantitated in the medical record were excluded. The relative frequencies of the underlying causes of hemoptysis were determined in the 59 patients with major or massive hemoptysis. The patients were divided into two groups: survivors and nonsurvivors. Age, sex, race, cause of hemoptysis, presenting symptoms, clinical and radiological characteristics, quantity and rate of bleeding, operability, bronchoscopic localization of bleeding site, and different therapies instituted were compared between the survivors and nonsurvivors. In-hospital medical
302
and surgical mortality rates also were determined in relation to underlying disease, rate of bleeding, operability, and surgery versus no surgery. Surgical mortality was defined as death within seven days of operative intervention. The criteria were inoperable carcinoma as evidenced by distant metastases; involvement of the mediastinum or greater vessels or both; severe advanced cardiac disease; severe chronic obstructive pulmonary disease with pulmonary function impairment characterized by a forced expiratory volume (FEV I) <35% of predicted normal value; and advanced, terminal, hematologic or other malignancies. Further, 120 random patients with hemoptysis were reviewed to determine etiologic factors alone (Table 1). Demographics were not further characterized on these patients. Statistical Analysis
Statistical analysis of the data was done with either the chi-square or the exact test, with p < 0.05 being considered significant. IS Results Deaths. Eighteen of the 59 patients who fulfilled
the study criteria of major and massive hemoptysis died from hemoptysis. These patients were compared November 1987 Volume 294 Number 5
Corey an? Hla
to the 41 patients who lived (Table 2). There was little difference in the ages or the predominance of males in both groups. There was a lower whitelblack ratio in the patients who died when compared to the survivors (1.6 vs. 4.1); however, it was not statistically significant. Previous episodes of hemoptysis were found less frequently in the patients who died (26% vs. 46%). Unfortunately, data were unavailable in 12 of the 59 patients, and the difference between the remaining patients again did not achieve a significant level. Hemoptysis as a presenting symptom was significantly less frequent in nonsurvivors (56% vs. 85%; p< 0.05). Hypotension was more frequent in the patients who died (33% vs. 12%; P < 0.05), but tachypnea was seen more commonly in the survivors (54% vs. 17%; p < 0.01). Again, the database was incomplete on 15 patients. Of the 18 patients who died, 13 (72%) had neoplastic disease, one of which was complicated by aspergillus infection (Table 3). Lung cancer was responsible for seven of the deaths, and leukemia was responsible for five others. Other causes of fatal hemoptysis included severe cystic fibrosis (CF) in two patients, complicated by sepsis in one and pneumonia in the other; hyper-IgE syndrome and Aspergillus pneumonia in one patient; sarcoid and sudden
TABLE 2 Clinical Characteristics of 59 Patients with Major/Massive Hemoptysis Number Age-mean (range In years) Sex (M/F) Race (W/B) Previous episodes of hemoptysis Hemoptysis as presenting symptom Hypotension (BP, 100/70) Respiratory rate of 20 Evidence of aspiration onCXR
Yes No Yes No Yes No Yes No Possible No
Lived
Died
41
18
51 (15-75) 46 (18-73) 4.1: 1 2.6 :1 1.6: 1 4.1: 1 19 (46%) 5(28%) 14 9 35 (85%) 10 (56%) 6 8 5 (12%) 34 22 (54%) 10 9(22%) 32
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
6 (33%) 11 3 (17%) 12 10 (56%) 4
TABLE 3 Comparison of Underlying Etiologic Factors in Survivors and Nonsurvivors with Major/Massive Hemoptysis
Bronchitis/Bronchiectasis Tuberculosis Neoplasia CA of lung Leukemia Metastatic CA Cystic fibrosis Anticoagulation Unknown Miscellaneous
Living (41)
Died (18)
Total (59)
42 7 20 12 5 2 5 10 7 10
0 0 72 39 28 6 11 0 6 11
28 5 36 20 12 3 7 7 7 10
All values given in percentages.
exsanguinating hemoptysis in one patient; and an unknown cause of hemoptysis in one patient. Interestingly, because of the endstage nature of the illnesses, only three of the nonsurvivors underwent bronchoscopy (Table 4). 'l\vo had positive cytologies and one had extrinsic compression of a bron-
TABLE 4 Diagnostic and Therapeutic Procedures Lived (41) Operable Surgery Transfusions Antibiotics Bronchoscopy Diagnosis on bronchoscopy
Localization of bleed on bronchoscopy
Yes No ND* Yes No ND
Died (18)
63 5 27 63 78
28 11 72 67 17
51 27 22 39 39 22
0 17 83 0 17 83
All values given as percentages. • ND = not done.
303
Major and Massive Hemoptysis: Reassessment ot Conservative Management
TABLE 5 Analysis of Mortality According to Rate of Bleeding and Underlying Disease Survivors (41 Patients)
Nonsurvivors (18 Patients)
Bleeding Anti- Lung Met Bran- BronchiCA Met Grand Rate (m1/24 hr) chitis ectasis TB coag CA CA Leuk CF Unk Misc Total Lung CA Leuk CF Misc Unk Total Total 200-499 500-999 1.000 Total
9 5
1 1
15
2
1 3
3 4
1 3 1 5
3 2' 2
1 2
1 3
4
17 13 11 41
1 6 7
5 5
1 22 15 18
2 2
2
18 15 26 59
'Both had pulmonary Aspergillosis: chronic lymphocytic leukemia (CLLj-aspergi/loma; acute myelogenous leukemia (AMLj-aspergi/lus pneumonia,
chus, but location of bleeding was not identified in any of the three. Conservative management with bed rest, sedatives, cough suppressants, and airway control was considered standard therapy and was not quantitated. Among the nonsurvivors, 67% received antibiotics and 72% received transfusions. More non survivors received transfusions than survivors (72% vs. 27%; p < 0.01). Of the five not receiving transfusions, three died suddenly and two had terminal illnesses. As can be seen from Table 5, the amount of bleeding in the patients who died was always greater than 1,000 m1l24 hr, except in the two patients with CF who had "complicated" hemoptysis where the cause of death was multifactorial, and in one patient with nonresectable bronchogenic carcinoma. Only five (28%) patients in this group of nonsurvivors were considered surgical candidates, and two of these underwent a thoractomy (Table 6). Three patients, however, died without the possible benefit of surgery. '!\vo of these had "late but not terminal" chronic leukemia and one died suddenly and unexpectedly with Stage IV sarcoid lung disease and hemoptysis. The two patients who were operated on died during or shortly after surgery with hemoptysis secondary to hyper-IgE syndrome complicated by Aspergillus pneumonia and unresectable CA of the lung, respectively. Survivors. Forty-one patients survived their episode of major or massive. hemoptysis. Seventeen (41 %) of these had bronchitis or bronchiectasis; eight (20%) had neoplasia; four (10%) were receiving anticoagulants (one of whom had had a documented pulmonary embolus); three (7%) had tuberculosis; two had cystic fibrosis (CF); two had aspiration pneumonia (one with foreign body removed at bronchoscopy); one had congestive heart failure (CHF); and one had an arteriovenous malformation (AVM). No diagnosis was made in three (Table 3). In this group, 32 (78%) of the 41 patients under304
went bronchoscopy and a bleeding site was seen in 50% of these (Table 4). Antibiotics were given to 26 (63%), but only 11 (29%) received transfusions. The severity of the bleeding is seen in Table 5 with 17 patients having 200 to 499 m1l24 hr, 13 having 500 to 999 ml/24 hr, and 11 having more than 1,000 ml/24 hr. Although 26 (63%) of these patients were considered operable, only two patients underwent a surgical procedure. One patient with AML lived, only to
TABLE 6 Mortality in the Operable and Nonoperable Groups of Patients with Major/Massive Hemoptysis Operable Surgery
No Surgery
Nonoperable
Patients Died Patients Died Patients Died Bronchitis Bronchiectasis CAof lung Metastatic CA Leukemia Tuberculosis Anticoagulation Cystic fibrosis Unknown Miscellaneous Total 4
11 1
3 3 2
4
2
11 2
6
3
3
2
2
4
2
1
1
5
2 (50%)
27
3 (11%)
28
13 (46%)
November 1987 Volume 294 Number 5
Corey and Hie
TABLE 7 Mortality According to Underlying Cause and Rate of Bleeding Mortality Bronchitls/BronchiectasisfTBI Anticoagulation CA of lung Leukemia Neoplasia <1.000 ml 2::1.000 ml 2::1.000 ml and neoplasia Overall
o 59 71
59 9
58 80 31
All values given in percentages.
die of other causes two months later during the same hospitalization. A second patient left the hospital with the diagnosis still unknown. Factors Influencing Mortality. The overall mortality rate was 31% (18/59). The presence of carcinoma of the lung or other neoplasia increased the mortality rate to 59% (Table 7). Also, if the bleeding was ;:::1,000 ml/24 hr, the mortality rate was 58%, while it was only 9% if bleeding was <1,000 ml/24 hr. If the patient had both neoplasia and bleeding ;:::1,000 ml/24 hr, the mortality rate rose to 80%. On the other hand, if the patient had bronchitis, bronchiectasis, tuberculosis, or was anticoagulated, the mortality rate was 0% (0/24). As expected, operable patients had a higher survival rate than nonoperable patients (84% vs. 54%; P < 0.05). Operative intervention was undertaken in only four patients despite the fact that 31/59 (53%) were surgical candidates (Table 6). Of these patients, one lived to leave the hospital (diagnosis unknown), one died at surgery (hyper-IgE with Aspergillus pneumonia), one died shortly after surgery (CA of lung), and one died from AML two months later while still in the hospital. Five (16%) of 31 of the operable patients died, three without surgery (see above). Discussion
The importance of hemoptysis in significant intrathoracic disease has been known since Hippocrates. 19 The frequency of hemoptysis as a symptom of intrathoracic disease has ranged from 11% to 38%,20.21 and in patients evaluated for hemoptysis as a presenting symptom, an underlying lesion will be found in 55% to 96% of the cases. 20 -28 In reviewing the causes of hemoptysis in several series from 1946 THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
to the present,20-28 it becomes apparent that there is significant variation in the underlying pathologic features (Table 1). In recent years, bronchitis has been diagnosed much more frequently, perhaps as a result offiberoptic bronchoscopy. Bronchiectasis and tuberculosis, on the other hand, have decreased markedly in frequency, while carcinoma of the lung has remained fairly stable. Death from hemoptysis is not common and is felt to occur primarily in those patients who are bleeding briskly. This form of hemoptysis has been referred to in the literature as major,2 massive,9 significant,t or life-threatening8 hemoptysis. The treatment of major and massive hemoptysis has been a much debated issue in the medical literature for the past 20 years. Despite several reports showing a high mortality rate with conservative treatment,t-9 the Surgery Department at our institution steadfastly refused to operate on these patients, basing the decisionon personal experience. In an attempt to critically evaluate this policy, we undertook this retrospective assessment of patients with major and massive hemoptysis. In reviewing the literature, it is apparent that there is no generally accepted definition of major or massive hemoptysis. Amounts have ranged from 100 ml/24 hr2 to 600 ml/16 hr.4 By defining major hemoptysis as 200 ml or greater in a 24-hr period, we hoped to eliminate a large group of patients with significant but clearly not life-threatening hemoptysis. This was, of necessity, an arbitrary decision. Fifty-nine patients were identified with bleeding rates of ;:::200 ml/24 hr. Twenty-six were hemoptysizing at a rate of 1,000 ml/24 hr or more, a rate we deemed to be massive hemoptysis. In this case series, because of the extremely low incidence of surgery (four cases), we observed the natural history of conservative treatment of hemoptysis in 55 medical patients. Unfortunately, the uniformity of this group of patients is limited because of the 10-year span of this case series, during which time significant improvements were made in bronchoscopy as well as supporting care on intensive case units for these patients. However, since new therapeutic modalities such as bronchial or pulmonary artery embolization,13.14 which are now coming into more widespread use, were not used during this 10-year period, our study is not further complicated by multiple therapeutic options. In trying to predict outcome, we were unable to determine any differences between the two groups in age or sex, although being of black race tended to be an adverse predictive factor . As expected, previous episodes of hemoptysis were more common in the surviving group of patients, suggesting chronic, more benign underlying causes in this group. Hypotension, on the other hand, was more common in the patients who died, indicating serious compromise of cardio305
Major and Massive Hemoptysis: Reassessment of Conservative Management
v~scular
stability, whether from blood loss or hypqxia. Surprisingly, an unexplained increase in the incidence of tachypnea was seen in the survivors. Because of the difficulties in delineating clear-cut e\iidence of aspiration in our patients, we do not feel tll.at any strong conclusions could be drawn from the fact that 56% of the 19 patients with possible aspiration died. This does, however, support the findings of other seriesl.3.5.10 that asphyxiation may be more important than exsanguination as a cause of death in patients with major and massive hemoptysis. In contrast to the rather weak associations above, etiologic factors and operability were excellent indicators of outcome. No patients with bronchitis or bronchiectasis died. However, the mortality rate of hemoptysis in patients with lung cancer was almost 60%, and that of patients with leukemia was over 70%. Similarly, in operable patients, the overall mortality was 16% (11 % in patients who did not have surgery, while patients who were nonoperable had a 46% mortality rate. Only five (28%) of the 18 patients who died were operable, and two of these succumbed despite aggressive surgical management. Furthermore, quantity of hemoptysis did appear to playa role in outcome as it had in previous series. In patients with ;:;:1,000 m1l24 hr of hemoptysis, the mortality rate was more than six times the rate in patients who bled
306
surgery, however, is a major advantage of the present series because it allowed us to look at the conservative management of major/massive hemoptysis. As can be seen from Table 8, the surgical mortality rates in these series ranges from 0% to 50%, but realistically is probably between 9% and 23% (our 50% mortality rate reflects only four operations). The medical mortality rate on the other hand varies from 0% to 100%, and dividing patients into operable/nonoperable/undefined groups fails to bring uniformity to this disparity. However, when one looks at only the two most-recent series, more order is apparent. Mortality rates in the medicallymanaged-but-operable groups are 2% (>100 m1l24 hr), 11% (>200 m1l24 hr), and 30% (> 1,000 m1l24 hr), increasing with rate of bleeding; similarly, the mortality rates in the medically-managed-butnonoperable groups are 29%, 46%, and 65%, again increasing with bleeding rate. Thus, it is apparent that in the operable patients the surgical and medical mortality rates are quite comparable. Furthermore, it is suggested that at lower bleeding rates (100-1,000 mll24 hr), medical therapy may well be slightly superior, while at higher bleeding rates (;:;:1,000 m1l24 hr) surgical intervention may improve survival. Nonoperable patients, on the other hand, have a consistently higher mortality rate, which may well have skewed the medical mortality rate considerably in the previous studies where mortality in the medically treated patients was often not stratified according to operability. The remarkably high mortality rate in medically treated yet operable patients in the two series by Crocco and Sehhat remains unexplained, although it is interesting to note that the operable patients managed medically by Crocco did worse than the nonoperable patients. This may have resulted from the higher rate of bleeding in the operable patients. When one compares the etiologic factors of he moptysis in previous studies with those of the present study, the largest discrepancies are noted (Table 9). In most previous series, tuberculosis was the overwhelmingly predominant cause of hemoptysis, accounting for 59% of all cases, with bronchiectasis providing an additional 24%. In contrast, in the present series, the incidence of tuberculosis was 5%; bronchiectasis, 3%; bronchitis, 25%; and neoplasia, 36%. This marked difference in cause may well help explain the differences in mortality with nonsurgical therapy. The source of bleeding in pulmonary tuberculosis is most commonly a ruptured bronchial artery aneurysm or bronchiopulmonary anastomosis in the wall of a tuberculous cavity.29-32 Less commonly, cavernoliths, broncholiths, and aspergillomas cause rupture of the same bronchial arteries. All of these lead to a relatively high pressure November 1987 Volume 294 Number 5
Corey and Hla
TABLE 8 Operability and Mortality Rates in 11 Series of Major/Massive Hemoptysis
Name Yeoh (1967) Crocco (1968) (Subsets) Amlrana (1968) Gaurin/ Garzon (1974) Garzonl Gourln (1977) McCollum (1975) Sehhat (1978) Yang (1978) Conlan (1983) Bobrowltz (1983) Corey (major) (massive)
Definition of Major/Massive Hemoptysis
Number Number of Operable ("10) Patients
Mortality
Number Undergoing Operation ("10)
Surgical
Medical Operable
Nonoperable
Undefined
200 ml/24 hr
56
NA
23"10
15"10
600 ml/48 hr 600 mll16 hr 600 m1l16-48 hr
67 46 21
61"10 65"10 52"10
48"10 46"10 52"10
19"10 23"10 9"10
100 ml/24 hr
17 45
NA NA
0 33"10
0"10
600 ml/24 hI'
62
92"10
92"10
18"10
80"10
600 ml/24 hr
75
91"10
91"10
18"10
71"10
150 ml/24 hr
15
100"10
87"10
23"10
600 ml/48 hr
146
91"10
73"10
0.9"10
200 ml/24 hr
20
85"10
15"10
0"10
18"10
600 ml/24 hr
100*
NA
34"10
18"10
32"10
100 ml/24 hr
113
65"10
27"10
13"10
2"10
29"10
200 ml/24 hr 1000 ml/24 hr
59 26
53"10 41"10
7"10 8"10
50"10 50"10
11 "10 30"10
46"10 65"10
23"10 78"10 78"10
46"10 75"10 0"10 29"10 0"10
50"10 85"10
69"10
100"10
• Twenty·three patients were treated with endobronchial iced saline lavage, with a 4% mortality rate.
hemorrhage which may not stop without surgical intervention. In bronchitis and carcinoma of the lung, on the other hand, hemorrhage is usually from venous sources that are more likely to clot with withdrawal of irritating stimuli such as cough, smoke, or infection. Importantly, Bobrowitz et apo recently reported a series of 113 patients, most of whom had tuberculosis or bronchiectasis, and concluded that conservative management was the treatment of choice. Unfortunately the rate of bleeding used in this series was :::=100 m1l24 hr, which may explain the low (2%) mortality rate seen in the operable patients. This low bleeding rate also makes it impossible to apply THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
the results of that study to the patient with lifethreatening hemoptysis. In summary, we feel that our surgical colleagues may well have been correct in refusing to operate on many of our patients with major and massive hemoptysis. Our data suggest that a conservative approach is reasonable and that, indeed, patients with non-tuberculosis causes of hemoptysis often do quite well without surgery. There are two subsets of patients, however, where the mortality of conservative management is quite high: (1) in patients with bleeding rates> 1,000 mll 24 hI'; and (2) in patients with active or inactive tuberculosis. Further studies will be needed to deter307
Major and Massive Hemoptysis: Reassessment of Conservative Management
TABLE 9 Incidence of Underlying Pulmonary Disease in Different Series of Major/Massive Hemoptysis Yeoh (1967) Crocco (1968) Amirana (1968) Geurin/Garzon (1974) Garzon/Gourln (1977) McCollum (1975) Sehhat (1978) Yang (1978) Conlan (1983) Bobrowltz (1983) Corey (1984) (Major) (Massive)
Tuberculosis
Bronchiectasis
100 73 100 73 71 33 20 50 38 71
0 10 0 11 12 20 63 15 30 18
5 4
3 4
mine the role of surgery in these subgroups and to look at the efficacy of local measures such as bronchial or pulmonary artery occlusion, endobronchial tamponade, and endobronchial iced saline lavage. References 1. Amirana M, Frater R, Tirschwell P, Janis M, Bloomberg A, State 0: An aggressive surgical approach to significant hemoptysis in patients with pulmonary tuberculosis. Am Rev Respir Dis 97:187-192, 1968. 2. Conlan AA, Hurwitz SS, Krige L, Nicolaou N, Pool R: Massive hemoptysis: Review of 123 cases. J Thorae Cardiovasc Surg 85:120-124, 1983. 3. Crocco JA, Rooney JJ, Fankushen OS, DiBenedetto RJ, Lyons HA: Massive hemoptysis. Arch Intern Med 121: 495-498, 1968. 4. Garzon AA, Gourin A: Surgical management of massive hemoptysis. Ann Surg 137:267-271, 1977. 5. Gourin A, Garzon AA: Operative treatment of massive hemoptysis. Ann Thorac Surg 18:52-60,1974. 6. McCollum WB, Mattox KL, Guinn GA, Beall AC: Immediate operative treatment for massive hemoptysis. Chest 67: 152-155, 1975. 7. Sehhat S, Oreizie M, Moinedine K: Massive pulmonary hemorrhage: Surgical approach as choice of treatment. Ann Thorac Surg 25:12-15, 1978. 8. Yang CT, Berger HW: Conservative management of lifethreatening hemoptysis. Mount Sinai J Med 45:329-333, 1978. 9. Yeoh CB, Hubaytar RT, Ford JM, Wylie RH: Treatment of massive hemorrhage in pulmonary tuberculosis. J Thorae Cardiovase Surg 54:503-510,1967. 10. Bobrowitz ID, Ramakrishna S, Shim Y-S: Comparison of medical vs. surgical treatment of major hemoptysis. Arch Intern Med 143:1343-1346,1983. 11. Porter OK, Van Every MJ, Anthracite RF, MackJW: Massive
308
Bronchitis
Neoplasia
Other
0 0 0 0 0 0 0 0 0 2
0 8 0 3 4 7 12 15 5 3
0 10 0 13 13 40 5 20 27 6
25 4
36 58
31 30
hemoptysis in cystic fibrosis. Arch Intern Med 143:287-290, 1983 12. Stern RC, Wood RE, Boat TF, Matthews LW, Tucker AS, Doershuk CF: Treatment and prognosis of massive hemoptysis in cystic fibrosis. Am Rev Respir Dis 117:825, 1978. 13. Remy J, Arnaud A, Fardou H, Giraud R, Voisin C: Treatment of hemoptysis by embolization of bronchial arteries. Radiology 122:33-37, 1977. 14. Ferris EJ: Pulmonary hemorrhage. Vascular evaluation and interventional therapy. Chest 80:710-714,1981. 15. Feloney JP: Repeated massive hemoptysis. Successful control using multiple balloon-tipped catheters for endobronchial tamponade. Chest 683-685, 1978. 16. Centennial: Koch's discovery of the tubercle bacillus. MMWR. US Public Health Service 31(10):121-123, 1982. 17. Epidemiology of Respiratory Diseases-Task Force Report. US Department of Health and Human Services, NIH #81-2019, October, 1980. 18. Armitage P: Statistical Methods in Medical Research. London, Blackwell Scientific Publications, 1971. 19. The Aphorisms of Hippocrates, Special Edition. Divis.ion of Gryphon Editions, Ltd., Birmingham, AL, 1982, P 230. 20. Heller R: The significance of hemoptysis. Tubercle 27:70, 1946. 21. Abbot OA: The clinical significance of pulmonary hemorrhage: A study of 1316 patients with chest disease. Dis Chest 14:824, 1948. 22. Levitt N: Clinical significance of hemoptysis. JMSMS June:606-610, 1951. 23. Souders CR, Smith AT: The clinical significance of hemoptysis. N Engl J Med 247:790-793,1952. 24. Moersch HJ: Clinical significance of hemoptysis. JAMA 148:1461, 1952. 25. Johnston RN, Lockhart W, Ritchie RT: Haemoptysis. Br Med J 1:592, 1960. 26. Boucot KR, Cooper DA, Weiss W: Hemoptysis in older men. Geriatrics 14:67, 1959. 27. Pursel SE, Lindskog GE: Hemoptysis: A clinical evaluation of
November 1987 Volume 294 Number 5
Corey end Hie
105 patients examined consecutively on a thoracic surgical service. Am Rev Respir Dis 84:329, 1961. 28. Soll B, Selecky PA, Chang R, Cleary MG, Casaburi R: The use of the flberoptic bronchoscope in the evaluation of hemoptysis. AARD 15(2):165,1978. 29. Rasmussen V: On hemoptysis, especially when fatal in its anatomical and clinical aspects. Edinburgh Med J 14:384, 1868.
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
30. Thompson JR: Mechanisms of fatal pulmonary hemorrhage in tuberculosis. Am J Surg 80:637-644, 1955. 31. Liebow AA, Hales MR, Lindskog GE: Enlargement of the bronchial arteries and then anastamosis with the pulmonary arteries in bronchiectasis. Am J Pa/hol 25:211-220, 1949. 32. Wood DA, Miller M: The role of the dual pulmonary circulation in various pathologic conditions of the lungs. J Thoracic Surg 7:649-670, 1937-38.
309
KHIN MAE HLA, MO
ABSTRACT: The etiologic factors of major (~200 ml/24 hr) and massive (~1,000 ml/24 hr) hemoptysis may well affect the outcome and, therefore, the treatment of this often life· threatening problem. The decline in the inci· dence of tuberculosis (TB) and bronchiectasis, along with the increase in bronchitis and neo· plasia, have led to a strong institutional bias against operating on patients with major and massive hemoptysis. A retrospective case study and an extensive literature review were under· taken to critically evaluate this policy. Fifty· nine consecutive patients with major hemopty· sis, 26 of whom had massive hemoptysis, were identified from 887 patients seen in our institution over a 10.year period. Only four of these 59 patients underwent surgery, while 55 were managed conservatively. Etiologic factors, operability, and bleeding rate all appeared to playa major role in outcome. No patients with bronchitis, bronchiectasis, tuberculosis, or who were on anticoagulation therapy died com· pared to a mortality rate of 59% in patients with carcinoma (CA) of the lung and 71% in patients with leukemia. Eleven percent of operable pa· tients treated conservatively died compared to a 46% mortality rate for nonoperable patients. And, 9% of patients with bleeding rates <1,000 m1l24 hr died compared to 58% of those with ~1,000 ml/24 hr. Conservative management appears to have a low mortality in patients with non·tuberculosis· related major hemoptysis as well as in many patients with massive hemoptysis, especially those patients who are operable and those with· out neoplastic disease. KEY INDEXING TERMS: Hemoptysis; Bronchitis; Carcinoma Of The Lung. [Am J Med Sci 1987; 294(5):301-309.]
From the Department of Medicine, Duke University Medical Center, Durham, North Carolina. The authors would like to acknowledge Dr. Patrick McKee for his help on this manuscript. Reprint requests: Ralph Corey, MD, Box 3038, Department of Medicine, Duke University Medical Center, Durham, NC 27710. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
C
onservative management of patients with major and massive hemoptysis has been associated with increased mortality in several studies over the last 20 years. 1- 9 This led to the recommendation of an aggressive, early surgical approach for this problem by most investigators, although more recently, advocates of conservative management have reappeared,10-12 as have innovators of the important new techniques of bronchial artery embolization,13,14 endobronchial tamponade,15 and endobronchial lavage. 2 In most reported series of major and massive hemoptysis, tuberculosis and bronchiectasis represent the majority of cases studied. 1,3-5,8,9,13 The decline in the incidence of TB16 and bronchiectasis,17 combined with an increase in the recognition of chronic bronchitis1? in both the general population as well as in the underlying etiologic factors of hemoptysis (Table 1), may make previous observations and recommendations concerning major and massive hemoptysis invalid. In an attempt to evaluate this hypothesis we reviewed all hemoptysis cases at our institutions over a 10-year period to determine (1) the current etiologic characteristics of major and massive hemoptysis; (2) the clinical and historical factors that have prognostic significance; and (3) the hospital course and outcome of patients treated primarily in a conservative manner for major and massive hemoptysis. A review of the largest series of patients with major and massive hemoptysis including their etiologic factors, management, and outcome is provided for comparison to the present report. PatIents and Methods
An extensive retrospective chart review was performed of all patients admitted to the Durham Veterans Administration and Duke University Medical Centers from January 1972 through December 1981 with a coded primary or secondary admission or discharge diagnosis of hemoptysis. Eight hundred eighty-seven patients with a diagnosis of hemoptysis were identified by the computer coding system at both institutions. The overall chart retrieval rate was 94.3%. Fifty-nine patients with major hemoptysis (~200 ml/24 hr), 26 of whom had massive hemoptysis (~1,000 m1l24 hr), were identified. Cross 301
Major and Massive Hemoptysis: Reassessment of Conservative Management
TABLE 1 Incidence and Cause of Hemoptysis Heller* Abbottt Levltt* Souders§ Moerschll 194818 1951 19 195220 195221 194617 (683) (105) (200) (416) (497) Bronchogenic Carcinoma Chronic Bronchitis Bronchiectasis Tuberculosis Other
Johnston~
195622 (324)
Boucot** Purseltt soll** Corey§§ 195923 1961 24 197825 1984 (395) (105) (200) (120)
2
21
12
3
24
4
8
19
13
13
15 7 40 26
2 21 22 34
3 15 50 20
12 29 2 54
9 27 6 30
17 13 10 59
19 4 7 64
5 24 13 40
55 0 3 29
50 2 3 32
* All patients with hemoptysis attending a hospital based chest clinic.
t Patients referred to a thoracic surgery seNice for hemoptysis. :j: Hemoptysis In patients admitted to
a tuberculosis sanatorium and to a general hospital.
§ All patients admitted to a referral hospital with a primary complaint of hemoptysis.
II All patients with hemoptysis who underwent bronchoscopy at a referral center. ~ All patients
with hemoptysis attending a hospital-based chest clinic.
* * Unselected males in the Philadelphia Pulmonary Neoplasm Research Project.
tt Consecutive patients with hemoptysis admitted to a community hospital. :j::j: Consecutive admissions for hemoptysis to
a metropolitan hospital.
§§ Random patients admitted to Duke University Medical Center with hemoptysis.
checks using major underlying etiologic characteristics such as bronchogenic carcinoma, tuberculosis, bronchitis, and lung abscess ensured completeness of case detection. Their charts were reviewed in detail by one of the investigators. Standardized data abstraction forms were used to record patients' demographic, clinical, laboratory, and radiological characteristics; underlying etiologic factors of hemoptysis; quantity and rate of hemoptysis; transfusion requirements; operability status; therapy instituted; and the patients' outcomes at discharge or follow-up. Quantification of bleeding amounts and rates was attempted only for those patients in whom the hemoptysis was observed directly. All patients in whom the amount of hemoptysis had not been specifically quantitated in the medical record were excluded. The relative frequencies of the underlying causes of hemoptysis were determined in the 59 patients with major or massive hemoptysis. The patients were divided into two groups: survivors and nonsurvivors. Age, sex, race, cause of hemoptysis, presenting symptoms, clinical and radiological characteristics, quantity and rate of bleeding, operability, bronchoscopic localization of bleeding site, and different therapies instituted were compared between the survivors and nonsurvivors. In-hospital medical
302
and surgical mortality rates also were determined in relation to underlying disease, rate of bleeding, operability, and surgery versus no surgery. Surgical mortality was defined as death within seven days of operative intervention. The criteria were inoperable carcinoma as evidenced by distant metastases; involvement of the mediastinum or greater vessels or both; severe advanced cardiac disease; severe chronic obstructive pulmonary disease with pulmonary function impairment characterized by a forced expiratory volume (FEV I) <35% of predicted normal value; and advanced, terminal, hematologic or other malignancies. Further, 120 random patients with hemoptysis were reviewed to determine etiologic factors alone (Table 1). Demographics were not further characterized on these patients. Statistical Analysis
Statistical analysis of the data was done with either the chi-square or the exact test, with p < 0.05 being considered significant. IS Results Deaths. Eighteen of the 59 patients who fulfilled
the study criteria of major and massive hemoptysis died from hemoptysis. These patients were compared November 1987 Volume 294 Number 5
Corey an? Hla
to the 41 patients who lived (Table 2). There was little difference in the ages or the predominance of males in both groups. There was a lower whitelblack ratio in the patients who died when compared to the survivors (1.6 vs. 4.1); however, it was not statistically significant. Previous episodes of hemoptysis were found less frequently in the patients who died (26% vs. 46%). Unfortunately, data were unavailable in 12 of the 59 patients, and the difference between the remaining patients again did not achieve a significant level. Hemoptysis as a presenting symptom was significantly less frequent in nonsurvivors (56% vs. 85%; p< 0.05). Hypotension was more frequent in the patients who died (33% vs. 12%; P < 0.05), but tachypnea was seen more commonly in the survivors (54% vs. 17%; p < 0.01). Again, the database was incomplete on 15 patients. Of the 18 patients who died, 13 (72%) had neoplastic disease, one of which was complicated by aspergillus infection (Table 3). Lung cancer was responsible for seven of the deaths, and leukemia was responsible for five others. Other causes of fatal hemoptysis included severe cystic fibrosis (CF) in two patients, complicated by sepsis in one and pneumonia in the other; hyper-IgE syndrome and Aspergillus pneumonia in one patient; sarcoid and sudden
TABLE 2 Clinical Characteristics of 59 Patients with Major/Massive Hemoptysis Number Age-mean (range In years) Sex (M/F) Race (W/B) Previous episodes of hemoptysis Hemoptysis as presenting symptom Hypotension (BP, 100/70) Respiratory rate of 20 Evidence of aspiration onCXR
Yes No Yes No Yes No Yes No Possible No
Lived
Died
41
18
51 (15-75) 46 (18-73) 4.1: 1 2.6 :1 1.6: 1 4.1: 1 19 (46%) 5(28%) 14 9 35 (85%) 10 (56%) 6 8 5 (12%) 34 22 (54%) 10 9(22%) 32
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
6 (33%) 11 3 (17%) 12 10 (56%) 4
TABLE 3 Comparison of Underlying Etiologic Factors in Survivors and Nonsurvivors with Major/Massive Hemoptysis
Bronchitis/Bronchiectasis Tuberculosis Neoplasia CA of lung Leukemia Metastatic CA Cystic fibrosis Anticoagulation Unknown Miscellaneous
Living (41)
Died (18)
Total (59)
42 7 20 12 5 2 5 10 7 10
0 0 72 39 28 6 11 0 6 11
28 5 36 20 12 3 7 7 7 10
All values given in percentages.
exsanguinating hemoptysis in one patient; and an unknown cause of hemoptysis in one patient. Interestingly, because of the endstage nature of the illnesses, only three of the nonsurvivors underwent bronchoscopy (Table 4). 'l\vo had positive cytologies and one had extrinsic compression of a bron-
TABLE 4 Diagnostic and Therapeutic Procedures Lived (41) Operable Surgery Transfusions Antibiotics Bronchoscopy Diagnosis on bronchoscopy
Localization of bleed on bronchoscopy
Yes No ND* Yes No ND
Died (18)
63 5 27 63 78
28 11 72 67 17
51 27 22 39 39 22
0 17 83 0 17 83
All values given as percentages. • ND = not done.
303
Major and Massive Hemoptysis: Reassessment ot Conservative Management
TABLE 5 Analysis of Mortality According to Rate of Bleeding and Underlying Disease Survivors (41 Patients)
Nonsurvivors (18 Patients)
Bleeding Anti- Lung Met Bran- BronchiCA Met Grand Rate (m1/24 hr) chitis ectasis TB coag CA CA Leuk CF Unk Misc Total Lung CA Leuk CF Misc Unk Total Total 200-499 500-999 1.000 Total
9 5
1 1
15
2
1 3
3 4
1 3 1 5
3 2' 2
1 2
1 3
4
17 13 11 41
1 6 7
5 5
1 22 15 18
2 2
2
18 15 26 59
'Both had pulmonary Aspergillosis: chronic lymphocytic leukemia (CLLj-aspergi/loma; acute myelogenous leukemia (AMLj-aspergi/lus pneumonia,
chus, but location of bleeding was not identified in any of the three. Conservative management with bed rest, sedatives, cough suppressants, and airway control was considered standard therapy and was not quantitated. Among the nonsurvivors, 67% received antibiotics and 72% received transfusions. More non survivors received transfusions than survivors (72% vs. 27%; p < 0.01). Of the five not receiving transfusions, three died suddenly and two had terminal illnesses. As can be seen from Table 5, the amount of bleeding in the patients who died was always greater than 1,000 m1l24 hr, except in the two patients with CF who had "complicated" hemoptysis where the cause of death was multifactorial, and in one patient with nonresectable bronchogenic carcinoma. Only five (28%) patients in this group of nonsurvivors were considered surgical candidates, and two of these underwent a thoractomy (Table 6). Three patients, however, died without the possible benefit of surgery. '!\vo of these had "late but not terminal" chronic leukemia and one died suddenly and unexpectedly with Stage IV sarcoid lung disease and hemoptysis. The two patients who were operated on died during or shortly after surgery with hemoptysis secondary to hyper-IgE syndrome complicated by Aspergillus pneumonia and unresectable CA of the lung, respectively. Survivors. Forty-one patients survived their episode of major or massive. hemoptysis. Seventeen (41 %) of these had bronchitis or bronchiectasis; eight (20%) had neoplasia; four (10%) were receiving anticoagulants (one of whom had had a documented pulmonary embolus); three (7%) had tuberculosis; two had cystic fibrosis (CF); two had aspiration pneumonia (one with foreign body removed at bronchoscopy); one had congestive heart failure (CHF); and one had an arteriovenous malformation (AVM). No diagnosis was made in three (Table 3). In this group, 32 (78%) of the 41 patients under304
went bronchoscopy and a bleeding site was seen in 50% of these (Table 4). Antibiotics were given to 26 (63%), but only 11 (29%) received transfusions. The severity of the bleeding is seen in Table 5 with 17 patients having 200 to 499 m1l24 hr, 13 having 500 to 999 ml/24 hr, and 11 having more than 1,000 ml/24 hr. Although 26 (63%) of these patients were considered operable, only two patients underwent a surgical procedure. One patient with AML lived, only to
TABLE 6 Mortality in the Operable and Nonoperable Groups of Patients with Major/Massive Hemoptysis Operable Surgery
No Surgery
Nonoperable
Patients Died Patients Died Patients Died Bronchitis Bronchiectasis CAof lung Metastatic CA Leukemia Tuberculosis Anticoagulation Cystic fibrosis Unknown Miscellaneous Total 4
11 1
3 3 2
4
2
11 2
6
3
3
2
2
4
2
1
1
5
2 (50%)
27
3 (11%)
28
13 (46%)
November 1987 Volume 294 Number 5
Corey and Hie
TABLE 7 Mortality According to Underlying Cause and Rate of Bleeding Mortality Bronchitls/BronchiectasisfTBI Anticoagulation CA of lung Leukemia Neoplasia <1.000 ml 2::1.000 ml 2::1.000 ml and neoplasia Overall
o 59 71
59 9
58 80 31
All values given in percentages.
die of other causes two months later during the same hospitalization. A second patient left the hospital with the diagnosis still unknown. Factors Influencing Mortality. The overall mortality rate was 31% (18/59). The presence of carcinoma of the lung or other neoplasia increased the mortality rate to 59% (Table 7). Also, if the bleeding was ;:::1,000 ml/24 hr, the mortality rate was 58%, while it was only 9% if bleeding was <1,000 ml/24 hr. If the patient had both neoplasia and bleeding ;:::1,000 ml/24 hr, the mortality rate rose to 80%. On the other hand, if the patient had bronchitis, bronchiectasis, tuberculosis, or was anticoagulated, the mortality rate was 0% (0/24). As expected, operable patients had a higher survival rate than nonoperable patients (84% vs. 54%; P < 0.05). Operative intervention was undertaken in only four patients despite the fact that 31/59 (53%) were surgical candidates (Table 6). Of these patients, one lived to leave the hospital (diagnosis unknown), one died at surgery (hyper-IgE with Aspergillus pneumonia), one died shortly after surgery (CA of lung), and one died from AML two months later while still in the hospital. Five (16%) of 31 of the operable patients died, three without surgery (see above). Discussion
The importance of hemoptysis in significant intrathoracic disease has been known since Hippocrates. 19 The frequency of hemoptysis as a symptom of intrathoracic disease has ranged from 11% to 38%,20.21 and in patients evaluated for hemoptysis as a presenting symptom, an underlying lesion will be found in 55% to 96% of the cases. 20 -28 In reviewing the causes of hemoptysis in several series from 1946 THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
to the present,20-28 it becomes apparent that there is significant variation in the underlying pathologic features (Table 1). In recent years, bronchitis has been diagnosed much more frequently, perhaps as a result offiberoptic bronchoscopy. Bronchiectasis and tuberculosis, on the other hand, have decreased markedly in frequency, while carcinoma of the lung has remained fairly stable. Death from hemoptysis is not common and is felt to occur primarily in those patients who are bleeding briskly. This form of hemoptysis has been referred to in the literature as major,2 massive,9 significant,t or life-threatening8 hemoptysis. The treatment of major and massive hemoptysis has been a much debated issue in the medical literature for the past 20 years. Despite several reports showing a high mortality rate with conservative treatment,t-9 the Surgery Department at our institution steadfastly refused to operate on these patients, basing the decisionon personal experience. In an attempt to critically evaluate this policy, we undertook this retrospective assessment of patients with major and massive hemoptysis. In reviewing the literature, it is apparent that there is no generally accepted definition of major or massive hemoptysis. Amounts have ranged from 100 ml/24 hr2 to 600 ml/16 hr.4 By defining major hemoptysis as 200 ml or greater in a 24-hr period, we hoped to eliminate a large group of patients with significant but clearly not life-threatening hemoptysis. This was, of necessity, an arbitrary decision. Fifty-nine patients were identified with bleeding rates of ;:::200 ml/24 hr. Twenty-six were hemoptysizing at a rate of 1,000 ml/24 hr or more, a rate we deemed to be massive hemoptysis. In this case series, because of the extremely low incidence of surgery (four cases), we observed the natural history of conservative treatment of hemoptysis in 55 medical patients. Unfortunately, the uniformity of this group of patients is limited because of the 10-year span of this case series, during which time significant improvements were made in bronchoscopy as well as supporting care on intensive case units for these patients. However, since new therapeutic modalities such as bronchial or pulmonary artery embolization,13.14 which are now coming into more widespread use, were not used during this 10-year period, our study is not further complicated by multiple therapeutic options. In trying to predict outcome, we were unable to determine any differences between the two groups in age or sex, although being of black race tended to be an adverse predictive factor . As expected, previous episodes of hemoptysis were more common in the surviving group of patients, suggesting chronic, more benign underlying causes in this group. Hypotension, on the other hand, was more common in the patients who died, indicating serious compromise of cardio305
Major and Massive Hemoptysis: Reassessment of Conservative Management
v~scular
stability, whether from blood loss or hypqxia. Surprisingly, an unexplained increase in the incidence of tachypnea was seen in the survivors. Because of the difficulties in delineating clear-cut e\iidence of aspiration in our patients, we do not feel tll.at any strong conclusions could be drawn from the fact that 56% of the 19 patients with possible aspiration died. This does, however, support the findings of other seriesl.3.5.10 that asphyxiation may be more important than exsanguination as a cause of death in patients with major and massive hemoptysis. In contrast to the rather weak associations above, etiologic factors and operability were excellent indicators of outcome. No patients with bronchitis or bronchiectasis died. However, the mortality rate of hemoptysis in patients with lung cancer was almost 60%, and that of patients with leukemia was over 70%. Similarly, in operable patients, the overall mortality was 16% (11 % in patients who did not have surgery, while patients who were nonoperable had a 46% mortality rate. Only five (28%) of the 18 patients who died were operable, and two of these succumbed despite aggressive surgical management. Furthermore, quantity of hemoptysis did appear to playa role in outcome as it had in previous series. In patients with ;:;:1,000 m1l24 hr of hemoptysis, the mortality rate was more than six times the rate in patients who bled
306
surgery, however, is a major advantage of the present series because it allowed us to look at the conservative management of major/massive hemoptysis. As can be seen from Table 8, the surgical mortality rates in these series ranges from 0% to 50%, but realistically is probably between 9% and 23% (our 50% mortality rate reflects only four operations). The medical mortality rate on the other hand varies from 0% to 100%, and dividing patients into operable/nonoperable/undefined groups fails to bring uniformity to this disparity. However, when one looks at only the two most-recent series, more order is apparent. Mortality rates in the medicallymanaged-but-operable groups are 2% (>100 m1l24 hr), 11% (>200 m1l24 hr), and 30% (> 1,000 m1l24 hr), increasing with rate of bleeding; similarly, the mortality rates in the medically-managed-butnonoperable groups are 29%, 46%, and 65%, again increasing with bleeding rate. Thus, it is apparent that in the operable patients the surgical and medical mortality rates are quite comparable. Furthermore, it is suggested that at lower bleeding rates (100-1,000 mll24 hr), medical therapy may well be slightly superior, while at higher bleeding rates (;:;:1,000 m1l24 hr) surgical intervention may improve survival. Nonoperable patients, on the other hand, have a consistently higher mortality rate, which may well have skewed the medical mortality rate considerably in the previous studies where mortality in the medically treated patients was often not stratified according to operability. The remarkably high mortality rate in medically treated yet operable patients in the two series by Crocco and Sehhat remains unexplained, although it is interesting to note that the operable patients managed medically by Crocco did worse than the nonoperable patients. This may have resulted from the higher rate of bleeding in the operable patients. When one compares the etiologic factors of he moptysis in previous studies with those of the present study, the largest discrepancies are noted (Table 9). In most previous series, tuberculosis was the overwhelmingly predominant cause of hemoptysis, accounting for 59% of all cases, with bronchiectasis providing an additional 24%. In contrast, in the present series, the incidence of tuberculosis was 5%; bronchiectasis, 3%; bronchitis, 25%; and neoplasia, 36%. This marked difference in cause may well help explain the differences in mortality with nonsurgical therapy. The source of bleeding in pulmonary tuberculosis is most commonly a ruptured bronchial artery aneurysm or bronchiopulmonary anastomosis in the wall of a tuberculous cavity.29-32 Less commonly, cavernoliths, broncholiths, and aspergillomas cause rupture of the same bronchial arteries. All of these lead to a relatively high pressure November 1987 Volume 294 Number 5
Corey and Hla
TABLE 8 Operability and Mortality Rates in 11 Series of Major/Massive Hemoptysis
Name Yeoh (1967) Crocco (1968) (Subsets) Amlrana (1968) Gaurin/ Garzon (1974) Garzonl Gourln (1977) McCollum (1975) Sehhat (1978) Yang (1978) Conlan (1983) Bobrowltz (1983) Corey (major) (massive)
Definition of Major/Massive Hemoptysis
Number Number of Operable ("10) Patients
Mortality
Number Undergoing Operation ("10)
Surgical
Medical Operable
Nonoperable
Undefined
200 ml/24 hr
56
NA
23"10
15"10
600 ml/48 hr 600 mll16 hr 600 m1l16-48 hr
67 46 21
61"10 65"10 52"10
48"10 46"10 52"10
19"10 23"10 9"10
100 ml/24 hr
17 45
NA NA
0 33"10
0"10
600 ml/24 hI'
62
92"10
92"10
18"10
80"10
600 ml/24 hr
75
91"10
91"10
18"10
71"10
150 ml/24 hr
15
100"10
87"10
23"10
600 ml/48 hr
146
91"10
73"10
0.9"10
200 ml/24 hr
20
85"10
15"10
0"10
18"10
600 ml/24 hr
100*
NA
34"10
18"10
32"10
100 ml/24 hr
113
65"10
27"10
13"10
2"10
29"10
200 ml/24 hr 1000 ml/24 hr
59 26
53"10 41"10
7"10 8"10
50"10 50"10
11 "10 30"10
46"10 65"10
23"10 78"10 78"10
46"10 75"10 0"10 29"10 0"10
50"10 85"10
69"10
100"10
• Twenty·three patients were treated with endobronchial iced saline lavage, with a 4% mortality rate.
hemorrhage which may not stop without surgical intervention. In bronchitis and carcinoma of the lung, on the other hand, hemorrhage is usually from venous sources that are more likely to clot with withdrawal of irritating stimuli such as cough, smoke, or infection. Importantly, Bobrowitz et apo recently reported a series of 113 patients, most of whom had tuberculosis or bronchiectasis, and concluded that conservative management was the treatment of choice. Unfortunately the rate of bleeding used in this series was :::=100 m1l24 hr, which may explain the low (2%) mortality rate seen in the operable patients. This low bleeding rate also makes it impossible to apply THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
the results of that study to the patient with lifethreatening hemoptysis. In summary, we feel that our surgical colleagues may well have been correct in refusing to operate on many of our patients with major and massive hemoptysis. Our data suggest that a conservative approach is reasonable and that, indeed, patients with non-tuberculosis causes of hemoptysis often do quite well without surgery. There are two subsets of patients, however, where the mortality of conservative management is quite high: (1) in patients with bleeding rates> 1,000 mll 24 hI'; and (2) in patients with active or inactive tuberculosis. Further studies will be needed to deter307
Major and Massive Hemoptysis: Reassessment of Conservative Management
TABLE 9 Incidence of Underlying Pulmonary Disease in Different Series of Major/Massive Hemoptysis Yeoh (1967) Crocco (1968) Amirana (1968) Geurin/Garzon (1974) Garzon/Gourln (1977) McCollum (1975) Sehhat (1978) Yang (1978) Conlan (1983) Bobrowltz (1983) Corey (1984) (Major) (Massive)
Tuberculosis
Bronchiectasis
100 73 100 73 71 33 20 50 38 71
0 10 0 11 12 20 63 15 30 18
5 4
3 4
mine the role of surgery in these subgroups and to look at the efficacy of local measures such as bronchial or pulmonary artery occlusion, endobronchial tamponade, and endobronchial iced saline lavage. References 1. Amirana M, Frater R, Tirschwell P, Janis M, Bloomberg A, State 0: An aggressive surgical approach to significant hemoptysis in patients with pulmonary tuberculosis. Am Rev Respir Dis 97:187-192, 1968. 2. Conlan AA, Hurwitz SS, Krige L, Nicolaou N, Pool R: Massive hemoptysis: Review of 123 cases. J Thorae Cardiovasc Surg 85:120-124, 1983. 3. Crocco JA, Rooney JJ, Fankushen OS, DiBenedetto RJ, Lyons HA: Massive hemoptysis. Arch Intern Med 121: 495-498, 1968. 4. Garzon AA, Gourin A: Surgical management of massive hemoptysis. Ann Surg 137:267-271, 1977. 5. Gourin A, Garzon AA: Operative treatment of massive hemoptysis. Ann Thorac Surg 18:52-60,1974. 6. McCollum WB, Mattox KL, Guinn GA, Beall AC: Immediate operative treatment for massive hemoptysis. Chest 67: 152-155, 1975. 7. Sehhat S, Oreizie M, Moinedine K: Massive pulmonary hemorrhage: Surgical approach as choice of treatment. Ann Thorac Surg 25:12-15, 1978. 8. Yang CT, Berger HW: Conservative management of lifethreatening hemoptysis. Mount Sinai J Med 45:329-333, 1978. 9. Yeoh CB, Hubaytar RT, Ford JM, Wylie RH: Treatment of massive hemorrhage in pulmonary tuberculosis. J Thorae Cardiovase Surg 54:503-510,1967. 10. Bobrowitz ID, Ramakrishna S, Shim Y-S: Comparison of medical vs. surgical treatment of major hemoptysis. Arch Intern Med 143:1343-1346,1983. 11. Porter OK, Van Every MJ, Anthracite RF, MackJW: Massive
308
Bronchitis
Neoplasia
Other
0 0 0 0 0 0 0 0 0 2
0 8 0 3 4 7 12 15 5 3
0 10 0 13 13 40 5 20 27 6
25 4
36 58
31 30
hemoptysis in cystic fibrosis. Arch Intern Med 143:287-290, 1983 12. Stern RC, Wood RE, Boat TF, Matthews LW, Tucker AS, Doershuk CF: Treatment and prognosis of massive hemoptysis in cystic fibrosis. Am Rev Respir Dis 117:825, 1978. 13. Remy J, Arnaud A, Fardou H, Giraud R, Voisin C: Treatment of hemoptysis by embolization of bronchial arteries. Radiology 122:33-37, 1977. 14. Ferris EJ: Pulmonary hemorrhage. Vascular evaluation and interventional therapy. Chest 80:710-714,1981. 15. Feloney JP: Repeated massive hemoptysis. Successful control using multiple balloon-tipped catheters for endobronchial tamponade. Chest 683-685, 1978. 16. Centennial: Koch's discovery of the tubercle bacillus. MMWR. US Public Health Service 31(10):121-123, 1982. 17. Epidemiology of Respiratory Diseases-Task Force Report. US Department of Health and Human Services, NIH #81-2019, October, 1980. 18. Armitage P: Statistical Methods in Medical Research. London, Blackwell Scientific Publications, 1971. 19. The Aphorisms of Hippocrates, Special Edition. Divis.ion of Gryphon Editions, Ltd., Birmingham, AL, 1982, P 230. 20. Heller R: The significance of hemoptysis. Tubercle 27:70, 1946. 21. Abbot OA: The clinical significance of pulmonary hemorrhage: A study of 1316 patients with chest disease. Dis Chest 14:824, 1948. 22. Levitt N: Clinical significance of hemoptysis. JMSMS June:606-610, 1951. 23. Souders CR, Smith AT: The clinical significance of hemoptysis. N Engl J Med 247:790-793,1952. 24. Moersch HJ: Clinical significance of hemoptysis. JAMA 148:1461, 1952. 25. Johnston RN, Lockhart W, Ritchie RT: Haemoptysis. Br Med J 1:592, 1960. 26. Boucot KR, Cooper DA, Weiss W: Hemoptysis in older men. Geriatrics 14:67, 1959. 27. Pursel SE, Lindskog GE: Hemoptysis: A clinical evaluation of
November 1987 Volume 294 Number 5
Corey end Hie
105 patients examined consecutively on a thoracic surgical service. Am Rev Respir Dis 84:329, 1961. 28. Soll B, Selecky PA, Chang R, Cleary MG, Casaburi R: The use of the flberoptic bronchoscope in the evaluation of hemoptysis. AARD 15(2):165,1978. 29. Rasmussen V: On hemoptysis, especially when fatal in its anatomical and clinical aspects. Edinburgh Med J 14:384, 1868.
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
30. Thompson JR: Mechanisms of fatal pulmonary hemorrhage in tuberculosis. Am J Surg 80:637-644, 1955. 31. Liebow AA, Hales MR, Lindskog GE: Enlargement of the bronchial arteries and then anastamosis with the pulmonary arteries in bronchiectasis. Am J Pa/hol 25:211-220, 1949. 32. Wood DA, Miller M: The role of the dual pulmonary circulation in various pathologic conditions of the lungs. J Thoracic Surg 7:649-670, 1937-38.
309