Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Rev Port Pneumol. 2011;17(1):7-14
ISSN 0873-2159
volume 17 / número 1 / janeiro/fevereiro 2011
EDITORIAIS
Moldando o futuro da Revista
Portuguesa de Pneumologia
Exposição involuntária
ao fumo do tabaco em crianças.
Tabagismo nos jovens
Todos os anos o ano do pulmão!
DESTAQUES
Hemoptises-etiologia, avaliação e tratamento num hospital universitário
HLA class II alleles as markers of tuberculosis susceptibility and resistance
Exposição ao fumo do tabaco (EFT) e morbilidade respiratória
em crianças em idade escolar
Representações sociais do comportamento de fumar em adolescentes
de 13 anos
www.revport pneumol.org
www.revportpneumol.org
ORIGINAL ARTICLE
Hemoptysis —etiology, evaluation and treatment in a university
hospital
F. Soares Pires*, N. Teixeira, F. Coelho, and C. Damas
Pulmonology Service, Hospit al de São João, Port o, Port ugal
Received January 19, 2010; accept ed August 2, 2010
KEYWORDS
Hemopt ysis;
Et iology;
Diagnosis;
Treat ment
Abstract
Obj ect i ve: Evaluat e pat ient s admit t ed f or hemopt ysis, it s et iology, use of diagnost ic t est s,
t reat ment and out come.
Mat eri al and met hods: A ret rospect ive analysis was done, checking clinical fi les of pat ient s
admit t ed for hemopt ysis, bet ween 1st January 2004 and 31st December 2008.
Result s: Two hundred and t hirt y seven pat ient s were included in t his st udy, wit h a mean age of
57.9 years. In pat ient s under 18 years, t he most f requent diagnoses were bronchiect asis and
congenit al cardiopat hy. In adult s, pulmonary t uberculosis sequelae and bronchiect asis were t he
dominant diagnoses (22.2 %and 15.8 %, respect ively), followed by lung cancer. Act ive infect ion
was responsible for bleeding in 51 pat ient s, especially pulmonary t uberculosis, pneumonia and
t racheobronchit is. The et iology of hemopt ysis was not est ablished in 6.3 %. All pat ient s performed
chest X-Ray. Chest CT was performed in 81.4 %of pat ient s and fi beropt ic bronchoscopy in 52.7 %,
t he lat t er locat ing t he source of bleeding in 38.4 %and est ablishing a fi nal diagnosis in 17.6 %. In
most pat ient s, effect ive cont rol of bleeding was achieved by medical t reat ment (90.7 %). During
t hese 5 years, art erial embolizat ion was performed in 11.8 %of pat ient s. The main reason for
embolizat ion was recurrence prevent ion. There were 14 deat hs (5.9 %), being ident ifi ed as poor
prognost ic fact ors, hemodynamic inst abilit y and malignancy.
Concl usi ons: Hemopt ysis is st ill a f requent sympt om, being chronic inf ect ion sequelae and
lung cancer t heir main causes. In t his st udy, f act ors associat ed wit h a worse prognosis were
hemodynamic inst abilit y and malignancy.
© 2010 Published by Elsevier España, S.L. on behalf of Sociedade Port uguesa de Pneumologia.
All right s reserved.
*Corresponding aut hor.
E-mail: fi [email protected] (F. Soares Pires).
0873-2159/ $ - see front mat t er © 2010 Published by Elsevier España, S.L. on behalf of Sociedade Port uguesa de Pneumologia. All right s reserved.
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
8
F. Soares Pires et al
PALAVRAS-CHAVE
Hemopt ises;
Et iologia;
Diagnóst ico;
Trat ament o
Hemoptises–etiologia, avaliação e tratamento num hospital universitário
Resumo
Obj ect ivo: Avaliar doent es admit idos por hemopt ises quant o à et iologia, meios diagnóst icos,
t rat ament o e evolução.
Mat erial e mét odos: Realizou-se um est udo ret rospect ivo dos doent es admit idos no Hospit al de
São João, por hemopt ises, ent re 1 de j aneiro de 2004 e 31 de dezembro de 2008.
Resul t ados: Foram est udados 237 doent es, com idade média de 57, 9 anos. Nos doent es com
idade inferior a 18 anos, os diagnóst icos mais frequent es foram bronquiect asias e cardiopat ia
congénit a. Nos adult os, as sequelas de t uberculose pulmonar e as Bronquiect asias f oram os
diagnóst icos predominant es (22, 2 % e 15, 8 %, respect ivament e), seguidos pel a neopl asia
pulmonar. A inf ecção f oi responsável pelas perdas hemát icas em 51 doent es, essencialment e
t uberculose pulmonar, pneumonia e t raqueobronquit e. A et iologia fi cou por esclarecer em 6,3 %.
Todos os doent es realizaram radiografi a de t órax. A TAC foi realizada em 81,4 %dos doent es e a
broncof ibroscopia em 52, 7 %, est a úl t ima l ocal izando a f ont e da hemorragia em 38, 4 % e
est abelecendo diagnóst ico defi nit ivo em 17,6 %. Na maioria dos casos (90,7 %), a hemorragia foi
cont rolada com t rat ament o médico. Ao longo dest es 5 anos, 11,8 %dos doent es foram submet idos
a embolização art erial, sendo a prevenção de recidiva o principal mot ivo da sua realização.
Foram regist ados 14 óbit os (5,9 %), sendo ident ifi cados, como fact ores de mau prognóst ico, a
inst abilidade hemodinâmica e a et iologia neoplásica.
Concl usões: As hemopt ises cont inuam a ser um sint oma f requent e, sendo act ualment e as
sequelas associadas a infecções crónicas e as neoplasias as principais et iologias. Nest e est udo,
os fact ores de pior prognóst ico foram a inst abilidade hemodinâmica e a et iologia neoplásica.
© 2010 Publicado por Elsevier España, S.L. em nome da Sociedade Port uguesa de Pneumologia.
Todos os direit os reservados.
Introduction
Hemopt ysis is defi ned as t he spit t ing or coughing up of blood,
caused by bleeding of t he lungs or from t he t racheobronchial
t ree. It may be a sympt om of several diseases, more or less
severe, but it s appearance induces concern t o t he pat ient
and requires a full diagnost ic invest igat ion.
Bronchiect asis, pulmonary t uberculosis and lung cancer
ar e t he mai n causes of hemopt ysi s. The f r equency of
each disease as a cause of hemopt ysis varies in dif f erent
seri es, accordi ng t o t he geographi cal area. Pul monary
t uber cul osi s i s an i mpor t ant cause of hemopt ysi s i n
developing count ries, 1-4 whereas, in developed count ries,
bronchiect asis, l ung cancer and bronchit is are t he main
causes. 5-7 In Port ugal , pul monary t ubercul osi s i s st i l l a
f requent di sease, but i t has si gni f i cant l y decreased i n
recent years. 8,9
In children, bronchiect asis (part icularly in cyst ic fi brosis)
and congenit al cardiopat hies are responsible for most cases
of hemopt ysi s. 10, 11 Some st udi es have f ound pul monary
t uber cul osi s as an i mpor t ant cause i n pat i ent s under
18 years old. 12,13
Besides a correct anamnesis and a det ailed physical exam,
diagnost ic t est s are an import ant part of t he diagnost ic
eval uat i on, such as chest X- Ray, chest CT scan and
bronchoscopy. 6, 14-17 Ot her t est s may be import ant , namely
bronchial art eriography and echocardiogram. 16 St udies also
refer t hat t he demand of t he diagnost ic invest igat ion should
be based on how hi gh or l ow t he probabi l i t y f or a l ung
cancer diagnost ic is. 6,15,18-20
Even af t er an adequat e invest igat ion, no cause can be
est ablished in 20-30 %of pat ient s. 16,21,22
In most pat ient s, cont rol of bl eeding is achieved wit h
conser vat i ve measur es. 4, 7 However, ot her t her apeut i c
measur es may be necessar y, such as br onchoscopi c
int ervent ion, 23 surgery or bronchial art ery embolizat ion.
Embol i zat i on’ s ef f i cacy i n t he cont r ol of bl eedi ng
has al r eady been val i dat ed, w i t h success r at es of
85-99 %, 24-29 even higher t han surgery. 30 Bronchial art ery
embol i zat i on has al so an i mpor t ant r ol e i n r ecur r ent
chronic hemopt ysis, f or recurrence prevent ion. However,
i t has not yet been consi dered a defi ni t i ve t herapeut i c
measure, because of it s recurrence rat e which is higher
t han 25 %. 28, 31-33
Objective
Evaluat e pat ient s admit t ed for hemopt ysis, it s et iology, use
of diagnost ic t est s, t reat ment and evolut ion.
Methods
A ret rospect ive st udy was performed, analyzing clinical fi les
of pat ient s admit t ed for hemopt ysis, in t he Hospit al de São
João, Port o, bet ween 1st January of 2004 and 31 December
of 2008.
Two hundred and eight y pat ient s were included. Of t hese,
43 were excl uded: 7 f or hemorrhage of non-respirat ory
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Hemopt ysis —et iology, evaluat ion and t reat ment in a universit y hospit al
origin (digest ive t ract , mout h and ot orhinolaryngological);
3 f or lack of adequat e inf ormat ion; and 33 f or repeat ed
admi ssi ons (onl y t he 1st admi ssi on of each pat i ent was
analyzed).
Two hundred and t hirt y seven clinical fi les were analyzed.
Clinical informat ion was reviewed: age, sex, smoke st at us,
medi cal hi st ory, use of ant i -agregant or ant i coagul ant
medi cat i on, sever i t y cr i t er i a, use of di agnost i c t est s,
defi nit ive diagnosis, t reat ment and evolut ion. Due t o t he
lack of informat ion about t he volume of blood loss, severit y
assessment was based on cl inical crit eria: reduct ion of
hemoglobin concent rat ion > 1 g, hemodynamic inst abilit y
(def ined as syst ol ic art erial bl ood pressure < 90 mmHg)
and respirat ory f ailure (defi ned as art erial oxygen part ial
pressure < 60 mmHg).
Tr acheobr onchi t i s di agnosi s was def i ned as l ower
r espi r at or y i nf ect i on, wi t hout associ at ed r adi ol ogi c
image.
Dat a were analyzed using st at ist ical sof t ware Epi Inf o,
ver si on 3. 5. 1. Exact Fi sher Test was used t o eval uat e
st at i st i cal si gni f i cance (a p val ue l ess t han 0. 01 was
considered st at ist ically signifi cant ).
9
Bronchiectasis
25%
Tuberous sclerosis
6%
Trauma
6%
Invasive
aspergillosis
6%
Pulmonary
tuberculosis
6%
Congenital
vascular
abnormality
6%
Congenital
cardiopathy
19%
Tracheobronchitis
13%
Pneumonia
13%
Figure 1
(n = 16).
Hemopt ysis’ et iol ogy in pat ient s under 18 years
Results
Patients’ characterization
Two hundr ed and t hi r t y seven pat i ent s wer e st udi ed,
165 male (69.6 %) and 72 f emale (30.4 %), wit h an average
age of 57.9 years old. Pediat ric age (less t han 18 years) was
f ound in 6. 8 % of pat ient s. Predominant age groups were
50-59 years and 60-69 years, bot h account ing for 21.1 %of
pat ient s.
Most adul t s (57. 9 %) were current or f ormer smokers.
Thirt y-t hree pat ient s (13. 9 %) were under ant icoagulant s
( w ar f ar i n) and 31 ( 13. 1 %) und er ant i - agr egant s.
Coagul at i on bl ood t est s wer e nor mal i n most pat i ent s
(69.2 %) and showed alt erat ions in 53 (22.4 %). Of t hese, 14
(26 %) were under ant i-agregant s and 33 (62.3 %) under oral
ant icoagulant s.
Etiology
I n p ed i at r i c age, m ost f r eq uent d i agnoses w er e
bronchiect asis and congenit al cardiopat hy (25 %and 18.8 %,
respect ively), as shown in Figure 1.
In adult s (Table 1), pulmonary t uberculosis (PT) sequelae
and bronchiect asis were t he main diagnoses f ound (22. 2 %
and 15. 8 %, r espect i vel y). Lung cancer was di agnosed
i n 33 pat i ent s, bei ng t he 3r d most f r equent di agnosi s.
Most (87. 9 %) were f ound t o have pri mary l ung cancer,
and adenocarci noma was t he predomi nant hi st ol ogi cal
t ype. Only 4 pat ient s (12. 1 %) had secondary lung cancer
(Figure 2) – pul monary met ast asis of neopl asm in ot her
l ocat ion (breast , esophagus and l arynx). We woul d l ike
t o acknowl edge t he l arge number of pat ient s wit hout a
f ul l hist ol ogical charact erizat ion (7 pat ient s – 21. 2 % of
all pat ient s wit h lung cancer). In 5 pat ient s, hist ological
char act er i zat i on was absent of t he cl i ni cal f i l e and
2 pat ient s were diagnosed wit h non-small cell lung cancer
wit hout furt her charact erizat ion. Infect ion was responsible
Table 1
Hemopt ysis’ et iology in adult s (n = 221)
Diagnosis
Pat ient s, n (%)
PT sequelae
Bronchiect asis
Lung cancer
Pulmonary t uberculosis
Unknown
Tracheobronchit is
Pneumonia
Silicosis
Mycet oma
Coagulopat hy
Cardiac failure
At ypical mycobact eriosis
Pulmonary t hromboembolism
Vasculit is
Ot hers*
49 (22.2 %)
35 (15.8 %)
33 (13.9 %)
22 (10 %)
14 (6.3 %)
10 (4.5 %)
10 (4.5 %)
6 (2.7 %)
5 (2.3 %)
4 (1.8 %)
4 (1.8 %)
4 (1.8 %)
3 (1.4 %)
3 (1.4 %)
19 (8.6 %)
*Art eriovenous fi st ulae (2), post -lobect omy complicat ion (2),
mit ral st enosis (2), COPD (2), pulmonary abscess (1),
congenit al cardiopat hy (1), t racheal foreign body (1),
post -radiat ion pulmonary fi brosis (1), foreign body granuloma
(1), t racheal invasion for medast inal mass (1), benign
endobronchial lesion (1), vascular malformat ion (1),
int erst it ial pneumonit is (1), radiat ion pneumonit is (1),
aort ic aneurysm rupt ure (1).
f or bl eedi ng i n 51 pat i ent s (23 %), 22 wi t h pul monar y
t uberculosis, followed by pneumonia, t racheobronchit is and,
less commonly, mycet oma and at ypical mycobact eriosis.
In 4 pat ient s, hemopt ysis’ et iology was coagulopat hy; of
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
10
F. Soares Pires et al
Adenocarcinoma
42.4%
Squamous cell carcinoma
15.2%
No histological characterization
15.2%
Secondary lung cancer
12.1%
Non-small cell lung cancer
6.0%
Small cell lung cancer
6.0%
Large cell lung cancer
3.0%
0
Figure 2
Table 2
2
4
6
8
10
12
14
16
Lung cancer as a cause of hemopt ysis (n = 33).
Main et iologies of hemopt ysis, according t o t he smoking hist ory of adult s (n = 188)
Smokers (n = 72)
Diagnosis
PT sequelae
PT
Lung cancer
n (%)
18 (25 %)
15 (20.8 %)
9 (12.5 %)
Former smokers (n = 56)
Diagnosis
Lung cancer
PT sequelae
Bronchiect asis
n (%)
13 (23.2 %)
13 (23.2 %)
5 (8.9 %)
Non smokers (n = 60)
Diagnosis
Bronchiect asis
PT sequelae
Lung cancer
n (%)
20 (33.3 %)
13 (21.7 %)
5 (8.3 %)
Not e: A smoking hist ory was not found in 33 pat ient s (14.9 %).
t hese, coagulopat hy was iat rogenic in 3. They were pat ient s
medicat ed wit h ant icoagulant s. No diagnosis was reached
in 6.3 %(14 pat ient s).
By age group, it appears t hat pul monary t ubercul osis
sequelae (PT sequelae) was one of t he main diagnoses in all
of t hem. Malignancy had an increasing predominance wit h
t he increasing of age, being t he main diagnosis in t he group
of pat ient s of 70-79 years.
The 3 main diagnoses according t o smoking hist ory of adult
pat ient s are present ed in Table 2. It ’s obvious t he great er
import ance of malignancy in smokers and former smokers.
In t he pat ient s wit h a posit ive smoking hist ory (smokers and
former smokers), lung cancer was responsible for 17.2 %of
hemopt ysis, and for only 8.3 %in non-smokers. No signifi cant
dif f erences were f ound in hemopt ysis’ et iology in f ormer
smokers whose dat e of smoking cessat ion was superior or
inferior t o 10 years.
Sixt y-eight percent of pat ient s wit h pulmonary t uberculosis
were current smokers.
In pat i ent s medi cat ed wi t h ant i coagul ant s (33), t he
main diagnosis was l ung cancer (18. 2 %), predominant l y
adenocarcinoma.
In 55. 7 % (132 pat ient s), t his was t he f irst episode of
hemopt ysi s, whi l e 35 % al ready had previ ous epi sodes.
In ot hers (9. 3 %), no inf ormat ion was obt ained. The main
diagnoses were not signifi cant ly different in pat ient s wit h or
wit hout previous episodes of hemopt ysis, wit h t he except ion
of pul monar y t uber cul osi s. In f act , t hi s was t he f i r st
episode in approximat ely 82 % of pat ient s wit h pulmonary
t uberculosis.
Hemopt ysi s’ cl assi f i cat i on based on t he amount of
bleeding, besides being a non-specifi c dat a, was missing
i n most cl i ni cal f i l es. So, sever i t y was based on t he
magni t ude of cl i ni cal ef f ect , eval uat i ng some cl i ni cal
crit eria, such as reduct ion of hemogl obin concent rat ion
(superi or t o 1 g), respi rat ory f ai l ure and hemodynami c
inst abilit y. Table 3 present s t he main diagnoses f or each
severit y crit eria. When al l severit y crit eria were f ound
(16 pat ient s), malignancy was predominant (80 %).
Diagnostic evaluation
All pat ient s performed a chest X-Ray, which was described
as normal in only 30 (12.7 %). Of t hese, t he main diagnoses
were t racheobronchit is (6), bronchiect asis (4), unknown
et i ol ogy (4) and PT sequel ae (3). Among pat i ent s wi t h
unknown et iol ogy of hemopt ysis whose chest X-Ray was
normal, chest CT scan was also described as normal in 2
(t he ot her 2 pat ient s had ground gl ass pat t ern in chest
CT scan) and br onchoscopy di dn’ t show abnor mal i t i es
in t he 3 pat ient s who were submit t ed t o it . One pat ient
wit h pul monary t ubercul osis had a normal chest X-Ray.
No mal ignancy was diagnosed in pat ient s wit h a normal
chest -X-Ray.
Chest CT scan was perf ormed in 193 pat ient s (81. 4 %),
and abnormalit ies were f ound in most of t hem (187). The
predominant pat t ern was linear ret ract ile opacit ies (24.1 %),
i n accordance wi t h t he f requency of di agnoses. Among
pat ient s wit h lung cancer, CT scan had a good diagnost ic
yield, showing abnormalit ies in all of t hem. Only 1 pat ient
had CT scan abnormal i t i es t hat were not suggest i ve of
malignancy. Among pat ient s wit h a normal chest CT scan (6),
t he main diagnosis was t racheobronchit is (3 pat ient s) and
no diagnosis was f ound in t he rest of t hem. Four pat ient s
wit h a normal CT scan perf ormed bronchoscopy, and no
endobronchial abnormalit ies were found.
Bronchoscopy was perf ormed in 125 pat ient s (52. 7 %).
Among pat ient s wit h a higher probabil it y of mal ignancy
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Hemopt ysis —et iology, evaluat ion and t reat ment in a universit y hospit al
Table 3
crit eria
Hemopt ysis’ et iology, according t o severit y
Diagnosis
No severit y crit eria (n = 85)
PT sequelae
Bronchiect asis
Tracheobronchit is
Unknown
Reduct ion Hb ≥ 1 g (n = 140)
Lung cancer
PT sequelae
Bronchiect asis
Pulmonary t uberculosis
Respirat ory failure (n = 66)
Lung cancer
PT sequelae
Bronchiect asis
Pulmonary t uberculosis
Hemodinamic inst abilit y (n = 25)
Lung cancer
Congenit al cardiopat hy
Vascular malformat ion
Art eriovenous fi st ulae
n (%)
19 (22.4 %)
16 (18.8 %)
8 (9.4 %)
8 (9.4 %)
26 (18.6 %)
24 (17.1 %)
19 (13.6 %)
15 (10.7 %)
13 (19.7 %)
9 (13.6 %)
6 (9.1 %)
5 (7.6 %)
6 (24 %)
2 (8 %)
2 (8 %)
2 (8 %)
(age > 50 years and a posit ive smoking hist ory — current
or f or mer ) , excl udi ng t hose whose mal i gnancy was
already known, we f ound t hat 67. 4 %were submit t ed t o a
bronchoscopy. The higher probabilit y of lung cancer was
associat ed wit h t he perf ormance of a bronchoscopy; t his
associat ion was st at ist ically signifi cant (p < 0. 01).
Bronchoscopy was able t o localize t he source of bleeding
in 38. 4 % of pat ient s and allowed t he est ablishment of a
defi nit ive diagnosis in 17.6 %, mainly pulmonary t uberculosis
and lung cancer. In 55 pat ient s (44 %), no endobronchial
abnormal i t y was f ound. Bronchoscopi c eval uat i on was
perf ormed by a fl exible fi beropt ic bronchoscope in almost
every pat ient . Rigid bronchoscopy was perf ormed in only
5 pat i ent s, f or di f f er ent r easons: di agnosi s (when t he
amount of bleeding was not compat ible wit h a fi beropt ic
bronchoscopy), rest orat ion of an adequat e airway (in a case
of obst ruct ion of t he airway by a t umor) and replacement of
a t racheal st ent .
The diagnost ic yield of each diagnost ic t est is present ed
in Table 4. We did not include angiography as a diagnost ic
t est , because in t his group of pat ient s, it was used almost
exclusively as a t herapeut ic met hod. Three pat ient s were
diagnosed wit h bronchial circulat ion abnormalit ies (2 wit h
art eri ovenous f i st ul a and 1 wi t h an abnormal ori gi n of
bronchial art ery).
In 6. 3 % (14), no diagnosis was reached. Among t hese
pat ient s, 5 (35. 7 %) were t ransf erred, af t er st abilizat ion,
t o anot her hospit al (according t o t he area of residence),
whi ch i s why t hey di d not compl et e t he di agnost i c
eval uat ion in t his hospit al ; 1 pat ient (7. 1 %) died bef ore
complet ing t he st udy. Among t he 8 pat ient s t hat complet ed
t he di agnost i c eval uat i on i n t hi s hospi t al , 5 (62. 5 %)
perf ormed a chest CT scan and a bronchoscopy; chest CT
scan was normal in 2 and showed ground-glass pat t ern in
11
Table 4 Diagnost ic yield of each diagnost ic t est
performed in t he group of pat ient s of t his st udy (n = 237)
Diagnost ic t est
Chest X-Ray
Chest CT scan
Bronchoscopy
Chest CT + bronchoscopy
Performed
(%)
Posit ive
result * (%)
237 (100 %)
193 (81.4 %)
125 (52.7 %)
118 (49.8 %)
121/ 237 (51.1 %)
148/ 193 (76.7 %)
56/ 125 (44.8 %)
103/ 118 (87.3 %)
*Posit ive result : any abnormal fi nding t hat cont ribut ed t o t he
fi nal diagnosis.
3 pat ient s; endobronchial abnormalit ies were not f ound in
any pat ient . In t he ot her 3 pat ient s, diagnost ic evaluat ion
was per f or med onl y wi t h a chest CT scan, showi ng
abnormalit ies in 2 (1 wit h ground-glass pat t ern and 1 wit h
a discret e linear opacit y).
Treatment
Most pat ient s went on medical t reat ment (90.7 %), as seen
on Figure 3. Medical t reat ment was based, rout inely, on t he
administ rat ion of ant it ussives (codeine) and ant ifi brinolyt ics
(aminocaproic acid), and also morphine, in cases in which
cont rol of bl eedi ng was not possi bl e wi t h medi cat i ons
previously described.
The main diagnoses, according t o each t herapeut ic opt ion,
are list ed in Table 5.
Only 1 pat ient needed a bronchoscopic int ervent ion – a
7-year-old child wit h a congenit al cardiopat hy. Cont rol of
bl eeding was init ial l y possibl e wit h medical t reat ment .
However, a rigid bronchoscopy performed for replacement
of a t racheal st ent (t he child also had a t racheal st enosis)
was complicat ed by severe bleeding, which was cont rolled
wit h t amponade.
Thr ee pat i ent s (1. 3 %) wer e submi t t ed t o a sur gi cal
int ervent ion: lobect omy in a case of a pulmonary abscess;
marsupializat ion of an aspergiloma cavit y and replacement
of mit ral valve in a case of mit ral st enosis.
In al l 237 admi ssi ons, ar t er i ogr aphy wi t h br onchi al
art ery embolizat ion was perf ormed in 7. 2 %. During t hese
5 years, considering every admission of each pat ient , we
f ound t hat 11. 8 % of pat i ent s went on bronchi al art ery
embol i zat i on. The mai n di agnoses of t hese pat i ent s
w er e PT sequel ae and br onchi ect asi s, f ol l ow ed by
vascul ar abnormal i t i es and mycet oma. The reasons f or
embol izat ion were recurrence prevent ion in 60. 7 % and
cont rol of bleeding in t he ot hers. When embolizat ion was
perf ormed f or recurrence prevent ion, t he admission was
programmed in 64. 7 %.
Outcome
Pat ient s’ out come was f avorable in 94. 1 %, wit h hospit al
di scharge af t er cont rol of bl eedi ng. Fourt een pat i ent s
di ed (5. 9 %). We f ound a hi gher mor t al i t y r at e f or al l
severit y crit eria (wit h st at ist ical signifi cance – p < 0. 01),
mainly hemodynamic inst abilit y wit h a mort alit y rat e of
24 %. The main diagnoses in pat ient s who died were lung
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
12
F. Soares Pires et al
1
3 1
Conservative
treatment
Bronchial artery
embolization
Surgical
treatment
Bronchoscopic
treatment
Radiation therapy
17
Table 6
opt ion
Pat ient s’ out come according t o each t herapeut ic
Treat ment
Medical (n = 215)
Embolizat ion (n = 17)
Surgery (n = 3)
Bronchoscopic (n = 1)
Radiot herapy (n = 1)
215
Figure 3 Treat ment s perf ormed f or cont rol of bl eedi ng
(n = 237).
Table 5 Main diagnoses found for each t herapeut ic opt ion
(n = 237)
Treat ment
Medical
Embolizat ion
Pat ient s,
n (%)
Main diagnoses
215 (90.7 %) PT sequelae
Bronchiect asis
Lung cancer
45 (20.9 %)
35 (16.3 %)
28 (13 %)
17 (7.2 %) PT sequelae
Bronchiect asis
Mycet oma
4 (23.5 %)
4 (23.5 %)
2 (11.8 %)
Surgical
3 (1.3 %) Aspergiloma
Pulmonary abscess
Mit ral st enosis
1 (33.3 %)
1 (33.3 %)
1 (33.3 %)
Bronchoscopic
1 (0.4 %) Congenit al cardiopat hy
(post rigid bronchoscopy
complicat ion)
Radiat ion
1 (0.4 %) Lung cancer
cancer (5 pat ient s – 35.7 %), bronchiect asis and pulmonary
t uberculosis (2 pat ient s each – 14.3 %).
In Table 6 t he out comes f or each t herapeut ic approach
are list ed.
Discussion
Pulmonary t uberculosis sequelae, bronchiect asis, lung cancer
and pulmonary t uberculosis were t he main diagnoses in t he
adult populat ion st udied. The import ance of t uberculosis,
act i ve or pr evi ous, as a cause of hemopt ysi s, was t he
Out come
Cont rol
Deat h
Cont rol
Deat h
Cont rol
Deat h
Cont rol
Deat h
Cont rol
Deat h
of bleeding
of bleeding
of bleeding
of bleeding
of bleeding
204 (94.9 %)
11 (5.1 %)
15 (88.2 %)
2 (11.8 %)
2 (66.7 %)
1 (33.3 %)
1 (100 %)
0
1 (100 %)
0
main dif f erence compared t o ot her st udies, in developed
count ries. 5-7 In fact , t uberculosis is st ill endemic in Port ugal,
al t hough a si gni f i cant decrease has occurred i n recent
years. 8,9 Among malignancies, in most series, squamous cell
carcinoma is t he main responsible f or hemopt ysis. 5, 34 This
hist ological t ype, due t o it s cent ral localizat ion, is more
f requent ly associat ed wit h t he occurrence of hemopt ysis,
compared t o adenocarcinoma, which has usually a peripheral
localizat ion. However, in t his st udy, adenocarcinoma was
t he main hist ological t ype. Two main reasons can explain
t his fi nding: a high number of malignancies wit hout a f ull
hist ological charact erizat ion and t he higher prevalence of
adenocarcinoma compared t o squamous cell carcinoma,
nowadays. 35,36 In fact , in 5 pat ient s (15.2 %of all pat ient s wit h
lung cancer), hist ological t ype was not found in t he clinical
fi le and 2 pat ient s (6 %) were diagnosed wit h non-small cell
lung cancer wit hout furt her charact erizat ion.
By age group, it is not orious t he increasing f requency
of malignancies keeping pace wit h t he increasing of age,
in accordance wit h reviewed lit erat ure. 3 Smoking hist ory,
being t he main est ablished risk fact or for lung cancer, is a
fundament al clue in pat ient s’ hist ory. In t his st udy, diagnoses
of lung cancer in pat ient s wit h a posit ive smoking hist ory
(current or former smokers) were more t han t wice as much
as t hose wit hout a smoking hist ory.
Most pat i ent s w i t h pul monar y t uber cul osi s w er e
current smokers. In f act , recent st udies have described
an associ at i on bet ween t obacco smoke and t he ri sk of
developing t uberculosis. 37,38
In pedi at r i c popul at i on, t he mai n et i ol ogi es of
hemopt ysis (bronchiect asis and congenit al cardiopat hies)
are in accordance wit h ot her st udies. 10,11 In Europe, cyst ic
fi brosis is one of t he main diseases t hat is responsible f or
t he occurrence of hemopt ysis. 12 In t his st udy, 1 child wit h
bronchiect asis was diagnosed wit h t his genet ic disease.
Hemopt ysis’ severit y is classically classifi ed based on t he
amount of bleeding. In t his st udy, such a classifi cat ion was
not possible, because t his t ype of informat ion was absent in
most clinical fi les. Nevert heless, severit y crit eria considered
are inevit ably associat ed wit h a higher amount of bleeding.
Several aut hors sust ain t hat hemopt ysis’ severit y should be
based, not in t he amount of bleeding, but in t he magnit ude of
t he clinical effect –clinical and analyt ic crit eria t hat refl ect a
severe bleeding, necessit y of t ransfusion, among ot hers. 39,40
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Hemopt ysis —et iology, evaluat ion and t reat ment in a universit y hospit al
In most series, massive hemopt ysis const it ut e 5-23 %of all
hemopt yses4,19,41 and t heir mort alit y rat e varies, from 23 t o
75 %. 4, 41 In t his st udy, hemodynamic inst abil it y occurred
in 10. 5 % of pat ient s and was associat ed wit h a mort alit y
rat e of 24 %. Clinical crit eria used in t his st udy (reduct ion
of hemoglobin concent rat ion > 1 g; respirat ory failure and
hemodynamic inst abilit y) were good indicat ors of severit y,
showing a st at ist ically signifi cant associat ion wit h mort alit y.
The main diseases responsible f or massive hemopt ysis
vary among series: bronchiect asis, pulmonary t uberculosis,
lung cancer, bronchit is, pulmonary abscess, aspergiloma,
among ot hers. 42 Some st udies f ound t hat lung cancer was
associat ed wit h a small bleeding. 19,43 However, in t his st udy,
t his diagnosis was t he most f requent f or al l considered
severit y crit eria.
Chest X-Ray, chest CT scan and bronchoscopy are t he main
diagnost ic t est s in et iologic invest igat ion. 6,14-16 Their use vary
among series, wit h t he except ion of t he almost universal use
of chest X-Ray. It seems consensual t hat a higher probabilit y
of mal ignancy shoul d be ref l ect ed in a more exhaust ive
diagnost ic invest igat ion. 6, 15, 18, 19 In t his st udy, t he f inding
of charact erist ics relat ed t o a higher risk of lung cancer
was associat ed wit h t he perf ormance of bronchoscopy, in
a st at ist ically signifi cant manner. This associat ion was not
f ound in chest CT scan. The perf ormance of chest CT scan
in t his group of pat ient s (81. 4 %) was below t o what was
desirable. Several recent st udies on t he diagnost ic impact
of diagnost ic t est s suggest , as a prot ocol, t he performance
of chest X-Ray and chest CT scan i n al l pat i ent s wi t h
hemopt ysis and bronchoscopic evaluat ion if t here are risk
fact ors for lung cancer, such as a smoking hist ory. 6,15
Diagnost ic evaluat ion perf ormed in t his st udy is st ill f ar
from what ’s suggest ed in such prot ocol. In fact , t here was a
low performance of chest CT scan (only 81.4 %of all). Among
pat ient s wit h a higher risk of lung cancer, chest CT scan was
perf ormed in 84.7 %, and chest CT scan plus bronchoscopy
in onl y 62. 2 %, f ar bel ow t he desirabl e 100 %. However,
some of t he pat ient s included in t his st udy had already had
previous admissions for hemopt ysis, and t heir diagnosis was
already known – so, in t hese cases, t here was a lower use of
diagnost ic t est s.
Chest CT scan was t he most sensit ive diagnost ic t est when
used alone. However, t he combined use of chest CT scan
t oget her wit h a bronchoscopy had t he best diagnost ic yield.
Bot h t hese result s have already been described. 19
In accordance wit h t he experience of several aut hors, 4,7
hemopt ysis’ cont rol was possible mainly t hrough conservat ive
measures. Lit erat ure validat es t he import ance of bronchial
art ery embolizat ion, eit her in t he cont rol of bleeding, or in
recurrence prevent ion. The lat t er was t he main reason for
it s performance in t his st udy. Success rat e in t he cont rol of
bleeding was 72.7 %, below what ’s referred on ot her st udies
(85-99 %). 24-28 Some f act ors may have cont ribut ed t o t his,
such as t he small amount of pat ient s (11), and t heir higher
severit y (54.5 % of pat ient s had already had hemodynamic
inst abilit y). Anot her fact or t hat seems t o limit t he success
rat e in t he cont rol of bleeding is some post ponement of
embolizat ion’s performance. In fact , conservat ive measures
are t he init ial t herapeut ic opt ions and t he advance t o ot her
t herapies occurs when t he fi rst fails. This management may
be j ust ifi ed for limit ed resources, such as t rained t eams for
t he procedure.
13
This st udy was not able t o evaluat e success rat e of bronchial
art ery embolizat ion about recurrence prevent ion.
Conclusions
Nowadays, t he main diseases responsible for hemopt ysis are
chronic infect ions sequelae and malignancies.
Hemodynamic inst abilit y and malignancy were associat ed
wit h a bad prognosis.
Bronchi al art ery embol i zat i on has been used wi t h an
increasing f requency, and it s ef f icacy has al ready been
est ablished in t he cont rol of bleeding.
This st udy and t he review of lit erat ure on t his subj ect
highl ight ed t he necessit y of a great er unif ormizat ion in
hemopt ysis’ management . So, t he elaborat ion of prot ocols
i n order t o uni f ormi ze, ei t her di agnost i c approach, or
t herapeut ic approach is import ant .
Conflict of interest
Aut hors st at e t hat t hey don’ t have any confl ict of int erest .
References
1. Reechaipichit kul W, Lat ong S. Et iology and t reat ment out comes
of massive hemopt ysis. Sout heast Asian. J Trop Med Public
Healt h. 2005;36:474-80.
2. Unsal E, Köksal D, Cimen F, et al. Analysis of pat ient s wit h
hemopt ysis in a reference hospit al for chest diseases. Tuberk
Toraks. 2006;54:34-42.
3. Wong CM, Lim KH, Liam CK. The causes of hemopt ysis in
Malaysian pat ient s aged over 60 and t he diagnost ic yield of
different invest igat ions. Respirology. 2003;8:65-8.
4. Prasad R, Garg R, Singhal S, et al. Lessons from pat ient s wit h
hemopt ysis at t ending a chest clinic in India. Ann Thorac Med.
2009;4:10-2.
5. Porzezinska M, Gorzewska A, Drozdowski J, et al. Assessment
of hemopt ysi s et i ol ogy among pat i ent s hospi t al i zed i n
Pneumonology Depart ment of Medical Universit y of Gdansk in
t he years 1998-2002. Pol Arch Med Wewn. 2005;114:658-63.
6. Tsoumakidou M, Chrysofakis G, Tsiligianni I, et al. A prospect ive
analysis of 184 hemopt ysis cases: diagnost ic impact of chest
X-ray, comput ed t omography, bronchoscopy. Respirat ion. 2006;
73:808-14.
7. Takahashi N, Akusawa H, Kisohara A, et al. A clinical review
of hemopt ysis. Journal of t he Japan Broncho-Esophagological
Societ y. 1999;50:391-5.
8. Direcção Geral de Saúde, Programa Nacional de Lut a Cont ra a
Tuberculose (PNT), 2006. Available from: www.dgs.pt
9. Observat ório Port uguês dos Sist emas de Saúde. Cont rolo da
Tuberculose em Port ugal. Apreciação crít ica epidemiológica
rel at iva a 2006/ 2007. Avail abl e f rom: www. observaport .
org/ OPSS/ Menus/ Sist ema/ Tracers/
10. Coss-Bu JA, Sachdeva RC, Bricker JT, et al. Hemopt ysis: a
10-year ret rospect ive st udy. Pediat rics. 1997;100:e7.
11. Godf rey S. Pulmonary hemorrhage / hemopt ysis in children.
Pediat r Pulmonol. 2004;37:476-84.
12. Roebuck DJ, Barnacle AM. Hemopt ysis and bronchial art ery
embolizat ion in children. Paediat r Respir Rev. 2008;9:95-104.
13. Wong KS, Lien R, Hsia SH. Maj or hemopt ysis in adolescent s.
Indian J Pediat r. 2005;72:537-8.
Document downloaded from http://www.elsevier.es, day 18/06/2017. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
14
14. Thirumaran M, Sundar R, Sut clif f e IM, et al. Is invest igat ion
of pat ient s wit h haemopt ysis and normal chest radiograph
j ust ifi ed? Thorax. 2009;64:854-956.
15. Laurent F, Mart i ns M, Sant y A. Hemopt ysi s: met hods of
localizat ion. Rev Med Suisse. 2005;1:2659-63.
16. Allewelt M, Lode H. Diagnosis oh haemopt oe / haemopt ysis.
Dt sch Med Wochenschr. 2005;130:450-2.
17. Moura e Sá J. Hemopt ises graves / maciças – a int ervenção do
pneumologist a. Rev Port Pneumol. 2008;6 Suppl 4:S227.
18. Law GT, Wong CY, Chen YC, et al. Cost -ef f ect iveness of CT
t horax and bronchoscopy in haemopt ysis wit h normal CXR for
exclusion of lung cancer. Aust ralas Radiol. 2002;46:381-3.
19. Hirshberg B, Biran I, Glazer M, et al. Hemopt ysis: et iology,
evaluat ion, and out come in a t erciary referral hospit al. Chest .
1997;112:440-4.
20. McGuinness G, Beacher JR, Harkin TJ, et al . Hemopt ysis:
prospect ive high-resolut ion CT / bronchoscopic correlat ion.
Chest . 1994;105:1155-62.
21. Andersen PE. Imaging and int ervent ional radiological t reat ment
of hemopt ysis. Act a Radiol. 2006;47:780-92.
22. Weinberger SE. Et iology and evaluat ion of hemopt ysis in adult s.
In: Basow DS, edit or. Upt odat e. Walt ham, MA: Upt odat e; 2009.
23. Valipour A, Kreuzer A, Koller H, et al. Bronchoscopy-guided
t opical hemost at ic t amponade t herapy for t he management of
life-t hreat ening hemopt ysis. Chest . 2005;127:2113-8.
24. Ramakant an R, Bandekar VG, Gandhi MS, et al . Massive
hemopt ysis due t o pul monary t ubercul osis: cont rol wit h
bronchial art ery embolizat ion. Radiology. 1996;200:691-4.
25. Tanaka N, Yamakado K, Murashima S, et al. Superselect ive
bronchial art ery embolizat ion f or hemopt ysis wit h a coaxial
microcat het er syst em. J Vasc Int erv Radiol. 1997;8:65-70.
26. Chan VL, So LK, Lam JY, et al. Maj or hemopt ysis in Hong Kong:
aet iologies, angiographic fi ndings and out comes of bronchial
art ery embolisat ion. Int J Tuberc Lung Dis. 2009;13:1167-73.
27. Chun JY, Bel l i AM. Immediat e and l ong-t erm out comes of
bronchial and non-bronchial syst emic art ery embol isat ion
f or t he management oh haemot pysis. Eur Radiol. 2009 Sep 2
[e-pub].
28. Lee S, Chan JW, Chan SC, et al. Bronchial art ery embolisat ion
can be equally safe and effect ive in t he management of chronic
recurrent haemot pysis. Hong Kong Med J. 2008;14:14-20.
F. Soares Pires et al
29. Menchini L, Remy-Jardin M, Faivre JB, et al . Crypt ogenic
haemopt ysi s i n smoker s: angi ogr aphy and r esul t s of
embolisat ion in 35 pat ient s. Eur Respir J. 2009;34:1031-9.
30. Parrot A, Khalil A, Roques S, et al. Management of severe
hemopt ysis: experience in a specialized cent er. Rev Pneumol
Clin. 2007;63:202-10.
31. Kim YG, Yoon HK, Ko GY, et al. Long-t erm effect of bronchial
art ery embol izat ion in Korean pat ient s wit h haemopt ysis.
Respirology. 2006;11:776-81.
32. Osaki S, Nakanishi Y, Wat aya H, et al. Prognosis of bronchial
art ery embol i sat i on i n t he management of hemopt ysi s.
Respirat ion. 2000;67:412-6.
33. Lordan JL, Gascoigne A, Corris PA. The pulmonary physician in
crit ical care * Illust rat ive case 7: assessment and management
of massive haemopt ysis. Thorax. 2003;58:814-9.
34. Ros Lucas JA, Malia Alvarado D, Fernández Suárez B, et al.
Hemopt isis: revisión de 70 casos. Emergencias. 2006;18:87-93.
35. Molinié F, Velt en M, Remont et L, et al. The progression of
lung cancer incidence in France (1978-2000). Rev Mal Respir.
2006;23:127-34.
36. Wahbah M, Boroumand N, Cast ro C, et al. Changing t rends in
t he dist ribut ion of t he hist ologic t ypes of lung cancer: a review
of 4,439 cases. Ann Diagn Pat hol. 2007;11:89-96.
37. Kol appan C, Gopi PG. Tobacco smoke and pul monar y
t uberculosis. Thorax. 2002;57:964-6.
38. Bat es MN, Khalakdina A, Pai M, et al. Risk of t uberculosis
f rom exposure t o t obacco smoke. A syst emat ic review and
met a-analysis. Arch Int ern Med. 2007;167:335-42.
39. Garzon A, Cerrut i MM, Golding ME. Exsanguinat ing hemopt ysis.
J Thor Cardiovasc Surg. 1982;84:829-33.
40. Endo S, Ot ani S, Hasegawa T. Management of massi ve
hemopt ysis in a t horacic surgical unit . Eur J Cardiot horac Surg.
2003;23:467-72.
41. Jean-Bapt ist e E. Cl inical assessment and management of
massive hemopt ysis. Crit Care Med. 2000;28:1642-7.
42. Ingbar DH. Causes and management of massive hemopt ysis in
adult s. In: Basow DS, edit or. Upt odat e. Walt ham, MA: Upt odat e;
2009.
43. Tunçözgür B, Isik AF, Nacak I, et al. Dilemma on t he t reat ment
of haemopt ysis: an analysis of 249 pat ient s. Act a Chir Belg.
2007;107:302-6.