OBJECTIVE: To determine the frequency of and the factors related to delayed diagnosis of sarcoidosis in Brazil. METHODS: We evaluated patients with a biopsy-proven diagnosis of sarcoidosis, using a questionnaire that addressed the following: time since symptom onset and since the first medical visit; and the number and specialty of the physicians visited. We divided the patients by the timeliness of the diagnosis-timely (< 6 months) and delayed (≥ 6 months)-comparing the two groups in terms of systemic and pulmonary symptoms; extrathoracic involvement; spirometric data; radiological staging; level of education; income; and tuberculosis (diagnosis and treatment). RESULTS: We evaluated 100 patients. The median number of physicians consulted was 3 (range, 1-14). In 11 cases, sarcoidosis was diagnosed at the first visit. In 54, the first physician seen was a general practitioner. The diagnosis of sarcoidosis was timely in 41 patients and delayed in 59. The groups did not differ in terms of gender; race; type of health insurance; level of education; income; respiratory/systemic symptoms; extrathoracic involvement; and radiological staging. In the delayed diagnosis group, FVC was lower (80.3 ± 20.4% vs. 90.5 ± 17.1%; p = 0.010), as was FEV1 (77.3 ± 19.9% vs. 86.4 ± 19.5%; p = 0.024), misdiagnosis with and treatment for tuberculosis (≥ 3 months) also being more common (24% vs. 7%, p = 0.032, and 20% vs. 0%; p = 0.002, respectively). CONCLUSIONS: The diagnosis of sarcoidosis is often delayed, even when the imaging is suggestive of sarcoidosis. Delayed diagnosis is associated with impaired lung function at the time of diagnosis. Many sarcoidosis patients are misdiagnosed with and treated for tuberculosis.
OBJECTIVE: To determine the frequency of and the factors related to delayed diagnosis of sarcoidosis in Brazil. METHODS: We evaluated patients with a biopsy-proven diagnosis of sarcoidosis, using a questionnaire that addressed the following: time since symptom onset and since the first medical visit; and the number and specialty of the physicians visited. We divided the patients by the timeliness of the diagnosis-timely (< 6 months) and delayed (≥ 6 months)-comparing the two groups in terms of systemic and pulmonary symptoms; extrathoracic involvement; spirometric data; radiological staging; level of education; income; and tuberculosis (diagnosis and treatment). RESULTS: We evaluated 100 patients. The median number of physicians consulted was 3 (range, 1-14). In 11 cases, sarcoidosis was diagnosed at the first visit. In 54, the first physician seen was a general practitioner. The diagnosis of sarcoidosis was timely in 41 patients and delayed in 59. The groups did not differ in terms of gender; race; type of health insurance; level of education; income; respiratory/systemic symptoms; extrathoracic involvement; and radiological staging. In the delayed diagnosis group, FVC was lower (80.3 ± 20.4% vs. 90.5 ± 17.1%; p = 0.010), as was FEV1 (77.3 ± 19.9% vs. 86.4 ± 19.5%; p = 0.024), misdiagnosis with and treatment for tuberculosis (≥ 3 months) also being more common (24% vs. 7%, p = 0.032, and 20% vs. 0%; p = 0.002, respectively). CONCLUSIONS: The diagnosis of sarcoidosis is often delayed, even when the imaging is suggestive of sarcoidosis. Delayed diagnosis is associated with impaired lung function at the time of diagnosis. Many sarcoidosispatients are misdiagnosed with and treated for tuberculosis.
A diagnosis of sarcoidosis is established when clinical and radiological findings are
associated with the histological presence of granulomas and causes other than
sarcoidosis have been excluded.(
,
) The diagnosis of sarcoidosis is often delayed for various reasons. The
disease is uncommon and can affect any organ system. The clinical presentation of
sarcoidosis can be typical(
,
,
) or atypical.In countries in which the number of cases of tuberculosis is high, such as Brazil, a
diagnosis of tuberculosis is always considered in patients in whom a biopsy shows
granulomas, and such patients often receive treatment for tuberculosis (even without
bacteriological confirmation). Similarly to tuberculosis, sarcoidosis predominantly
affects the upper lobes, and parenchymal fibrosis can cause architectural distortion,
hilar retraction, and upper lobe volume loss.(
) Patients with sarcoidosis can present with systemic symptoms, such as
fever, weight loss, and night sweats, which can lead physicians to misdiagnose such
patients as having tuberculosis.(
) Sarcoidosis is asymptomatic in many cases. In symptomatic patients, the
symptoms are generally nonspecific. Finally, economic barriers and difficult access to
specialists can affect the time to diagnosis. In a study conducted in the United States
and involving 189 patients, the diagnosis was delayed (> 6 months after symptom
onset) in 31%.(
) The presence of pulmonary symptoms and radiological findings indicative of
advanced disease were associated with a longer time to diagnosis. In contrast, the
presence of skin lesions resulted in earlier diagnosis.The objective of the present study was to determine the frequency of and the factors
related to delayed diagnosis of sarcoidosis in patients seeking treatment at specialized
centers in Brazil.
Methods
The study protocol was approved by the Research Ethics Committee of the Federal
University of São Paulo, located in the city of São Paulo, Brazil. All patients gave
written informed consent. Clinical and biopsy findings were consistent with sarcoidosis
in all of the patients included in the present study. Other conditions were excluded on
the basis of the results of ancillary tests, including special stains for mycobacteria
and fungi on biopsy slides. Patients were recruited from among those treated at any of
the following facilities in the city of São Paulo, Brazil: the Hospital São
Paulo Outpatient Clinic for Interstitial Diseases (n = 68); the São Paulo
Hospital for State Civil Servants (n = 22); and a private clinic owned by one of the
authors of the present study (n = 10).A standardized questionnaire was administered at diagnosis or retrospectively, up to one
year after diagnosis. Data were collected between June of 2007 and May of 2011.We excluded patients who were unable to remember the time of symptom onset or the number
of physicians consulted, as well as those who were unable to perform acceptable
spirometric maneuvers.In the initial evaluation, patients underwent history taking, detailed physical
examination, and ancillary tests to determine whether there was multiple organ
involvement by sarcoidosis.(
) Chest X-rays and CT scans taken closest to the date of biopsy were
reviewed. The chest X-ray findings were classified in accordance with the criteria
proposed by Scadding(
): stage 0, no thoracic involvement; stage I, hilar adenopathy alone; stage
II, hilar enlargement with interstitial lung disease; stage III, interstitial lung
disease alone; and stage IV, pulmonary fibrosis.All spirometric tests met the acceptability and reproducibility criteria proposed by the
Brazilian Thoracic Association, and the results were expressed as predicted values for
the Brazilian population.(
,
) Abnormal patterns were classified as obstructive (reduced
FEV1/FVC ratio with normal FVC), restrictive (reduced FVC and FEV1
with normal FEV1/FVC ratio), or mixed (reduced FVC and FEV1/FVC
ratio). Dyspnea was assessed by the baseline dyspnea index.(
) Dyspnea was considered severe when the score was ≤ 6.Patients answered questions regarding the date of symptom onset, the date of the first
medical visit, the number of physicians visited (and their specialty), and the specialty
of the physician who made the diagnosis of sarcoidosis. Physicians were classified as
general practitioners or generalists (including emergency room physicians),
pulmonologists, dermatologists, ophthalmologists, or others.The clinical presentation of sarcoidosis was classified as asymptomatic (abnormal
radiological findings without symptoms), pulmonary symptoms, cutaneous symptoms, ocular
symptoms, systemic symptoms (fever, weight loss, and night sweats), or others.Patients were classified as Black or White on the basis of their skin color. The level
of education was classified as illiterate, elementary school, high school, or college.
The monthly individual income was quantified by the number of times the average minimum
wage at the time of the study: 1-3 (600-1,800 Brazilian reals); 4-10; and > 10.
Health insurance was classified as public or private.The diagnosis of sarcoidosis was classified as timely (< 6 months) or delayed (≥ 6
months) on the basis of the time elapsed between the first medical visit and the date of
the visit in which the diagnosis of sarcoidosis was established.Variables with normal distribution were expressed as mean and standard deviation, and
those with non-normal distribution were expressed as median and range. The chi-square
test was used in order to compare the timely diagnosis and delayed diagnosis groups
regarding categorical variables. Regarding variables with normal distribution, the two
groups of patients were compared by the Student's t-test. Regarding those with
non-normal distribution, they were compared by the Kruskal-Wallis test or the
Mann-Whitney test. Correlations of FVC% with the time elapsed between symptom onset and
diagnosis were determined by Spearman's test.A preliminary analysis of sarcoidosispatients showed that approximately half had had a
delayed diagnosis. Considering a difference of 20% in the frequency of findings between
the two groups relevant, with values of α = 0.05 and power (1 − β) of 0.80, we found
that a sample of 97 cases was required.(
) Statistical analysis was performed with the Statistical Package for the
Social Sciences, version 17.0 for Windows (SPSS Inc., Chicago, IL, USA). Statistical
significance was set at p < 0.05 (two-tailed).
Results
One hundred patients were included in the present study. The overall findings are shown
in Table 1. White females constituted the
majority of the sample. The mean age was 48 years, being higher for females than for
males (50.1 ± 12.1 years vs. 43.9 ± 11.4 years; t = 2.57; p = 0.012).
Table 1
General characteristics of 100 patients with sarcoidosis.a
Of the sample as a whole, 79% had public health insurance, 52% had finished college, and
60% had a monthly income of less than 2,000 Brazilian reals.Table 2 shows the clinical, functional, and
radiological findings, as well as the biopsy sites in which granulomatous lesions were
found. The diagnosis was based on incidental chest X-ray findings in 13% of the cases.
Systemic symptoms were common: 46% of the patients reported weight loss; 35% reported
night sweats; and 20% reported fever. Increased dyspnea was reported by 19%.
Table 2
Clinical, functional, and radiological findings, as well as biopsy sites
showing granulomas, in 100 patients with sarcoidosis.a
Mean FVC and FEV1 values (as assessed by spirometry) were close to the lower
limit of the reference range, and the mean FEV1/FVC ratio was normal.
Spirometric values were within the reference range in 44 cases; 29 cases were classified
as restrictive, 19 were classified as obstructive, and 8 were classified as mixed.In 49 cases, the diagnosis of sarcoidosis was based on transbronchial biopsy findings.
Skin lesions and extrathoracic adenopathy were common, and one third of the cases were
diagnosed on the basis of biopsy findings in those sites.The median number of physicians consulted was 3 (range, 1-14). The first physician
consulted was a general practitioner in 54 cases; a pulmonologist, in 13; a
dermatologist, in 7; and other specialists, in 24. Sarcoidosis was diagnosed at the
first visit in only 11 patients; 6 patients were diagnosed by a dermatologist, 3 were
diagnosed a pulmonologist, and 2 were diagnosed by a general practitioner. In 26 cases,
five or more physicians were consulted before the correct diagnosis was made.All patients visited one of our facilities, meaning that, by the end of the study, all
patients had consulted with a pulmonologist. Before that, 68 patients had consulted with
90 pulmonologists, the diagnosis having been established in 28 patients (31%).The median time elapsed from the onset of symptoms or from an abnormal radiological
finding to the first visit was 1 month; the median time elapsed from the first visit to
the diagnosis of sarcoidosis was 7.5 months; and the median time elapsed from the onset
of symptoms or from an abnormal radiological finding to the diagnosis of sarcoidosis was
12 months. Fifty-three patients consulted with a physician within 1 month after the
onset of symptoms, and 82 patients consulted with a physician within 6 months after the
onset of symptoms.Fifty-nine patients (59%) were diagnosed late (≥ 6 months after the first visit),
whereas the remaining 41 (41%) were diagnosed timely. The time elapsed from the onset of
symptoms to the first visit was similar between the two groups. The median duration of
symptoms was 1 month in the group of patients with timely diagnosis group and 2 months
in the delayed diagnosis group (z = 0.24; p = 0.81).The time elapsed from the onset of symptoms to the first visit and from the onset of
symptoms to the diagnosis of sarcoidosis did not differ between the genders, races, or
types of health insurance, or among levels of education or incomes (data not shown).Table 3 shows a comparison between the timely
and delayed diagnosis groups in terms of clinical, functional, and imaging findings.
Those in the delayed diagnosis group had a higher mean age, a lower FVC, and a lower
FEV1, as well as having been more commonly misdiagnosed with and treated
for tuberculosis. Of the 59 patients in the delayed diagnosis group, 24 (40.7%) had
restrictive lung disease, when compared with 6 (14.6%) of the 41 patients in the timely
diagnosis group (p = 0.005).
Table 3
Clinical, functional, and radiological findings in 100 patients with
sarcoidosis, by time to diagnosis-timely (< 6 months) or delayed (≥ 6
months).a
There were no differences between the patients in the timely diagnosis group and those
in the delayed diagnosis group regarding systemic symptoms, respiratory symptoms, or the
frequency of extrathoracic involvement. In addition, there were no differences between
the two groups of patients regarding radiological staging. In 97 patients, CT scans were
available for review. After the application of the Scadding classification system to the
CT scans, there were changes in the radiological staging of 50 patients, 36 of whom
moved to a higher stage. Only 1 of the 8 cases classified as stage 0 on the basis of
routine chest X-ray findings remained at that stage after HRCT staging.In 21 cases, the disease stage changed from I, II, or III to IV. There were no
differences between the patients in the timely diagnosis group and those in the delayed
diagnosis group regarding CT staging (Table 3).
When the disease stages were grouped into 0/I/II vs. III/IV and compared regarding
diagnosis (timely or delayed), no differences were observed.Seventeen patients were diagnosed with tuberculosis, none of whom had bacteriological
confirmation. All 17 initially received treatment for tuberculosis, 12 of whom received
it for 3 months or more. As expected, those who were treated for 3 months or more had a
delayed diagnosis (Table 3). Ten patients were
diagnosed as having tuberculosis by pulmonologists, 4 were diagnosed as having
tuberculosis by general practitioners, and 3 were diagnosed as having tuberculosis by
other professionals. Of the 17 patients who were diagnosed with tuberculosis, 15 (88.2%)
had systemic symptoms, compared with 44 (53.0%) of the 83 patients not having been so
diagnosed (χ2 = 7.24; p = 0.007).
Discussion
The present study confirms that delayed diagnosis of sarcoidosis is common. The
diagnosis of sarcoidosis was delayed by 6 months or more in approximately 60% of the
cases. Only 11% of the patients were diagnosed by the first physician consulted. In 26%
of the cases, five or more physicians were consulted before the correct diagnosis of
sarcoidosis.The present study was inspired by another study,(
) which showed that the time elapsed from the onset of symptoms to diagnosis,
as well as the time elapsed from the first medical visit to diagnosis, was longer in
patients with pulmonary symptoms; the presence of skin involvement was associated with a
shorter time to diagnosis of sarcoidosis. In the present study, the abovementioned
findings had no influence on the time to diagnosis.The time to diagnosis of sarcoidosis (from the onset of symptoms and from the first
medical visit) was not affected by gender, race, individual income, or type of health
insurance. According to the Brazilian Institute of Geography and Statistics,(
) the mean monthly individual income in São Paulo at the time of the study
was approximately 850 Brazilian reals, which is similar to the mean value found in the
present study (approximately 800 Brazilian reals).In most cases, a general practitioner was the first physician consulted. However, the
diagnosis of sarcoidosis was made at the first visit in only 1 case. Pulmonologists were
the first physicians consulted in 13 cases, and the diagnosis of sarcoidosis was made in
only 3 of those cases.Two biases in the present study merit consideration. One is a recruitment bias; it is
possible that the sarcoidosispatients included in the present study are not
representative of sarcoidosispatients in general. Regarding age and gender, our
findings are similar to those of two epidemiological studies, one in Holland and the
other in Japan.(
,
) A study conducted in the United States, designated A Case Control Etiologic
Study of Sarcoidosis,(
) involved a large, representative sample of patients with sarcoidosis. The
results of that study showed similarities and differences in comparison with those of
the present study. As occurred in our study, the mean age was higher for females, and
the proportions of skin involvement, extrathoracic lymph nodes, and ocular involvement
were very similar between the two studies. However, the proportion of patients with
stage III/IV sarcoidosis was higher in our sample (36% vs. 16%), whereas that of those
with stage I sarcoidosis was lower (21% vs. 40%).(
)In Brazil, there have been no population-based studies investigating sarcoidosis. Two
studies conducted at referral centers in two different cities (Porto Alegre and Rio de
Janeiro) showed overall findings in patients with sarcoidosis.(
,
) Both studies showed that the disease is more common in female patients,
most patients having stage II or III sarcoidosis. In one of those studies,(
) 6 of 100 sarcoidosispatients had initially been diagnosed with
tuberculosis. The time to diagnosis was not reported in those studies. Therefore, the
present study is similar to other Brazilian studies conducted at referral centers.
However, the patients with sarcoidosis investigated in those studies are probably not
representative of those in the general population. It is possible that patients with
earlier stage disease were not referred to those centers. However, in the present study,
the number of patients with stage I or II sarcoidosis (n = 28) was the same in both
groups (with and without delayed diagnosis). Therefore, the fact that the number of
patients with stage I or II sarcoidosis was lower in our study than in population-based
studies does not explain the findings.Another potential bias in our study is a recall bias due to the fact that, in many
cases, the questionnaires were administered retrospectively. We tried to minimize that
by excluding patients who were unable to recall the details of their cases. However,
this is a subjective task, and we did not record the number of patients excluded on the
basis of that criterion.In those patients in whom the time elapsed from the onset of symptoms to diagnosis was
longer, FVC was lower. This finding suggests that a delayed diagnosis of pulmonary
sarcoidosis results in worsening of lung dysfunction in the absence of treatment.
Unfortunately, we did not record long-term changes in pulmonary function tests after
treatment.In the present study, we found no relationship between radiological staging and delayed
diagnosis. Although HRCT is superior to chest X-ray for detecting nodules and
mediastinal lymph node enlargement,(
) these findings were not useful for earlier diagnosis. In the absence of
pathologic confirmation, clinical and radiological findings can be useful for the
diagnosis of stage I sarcoidosis (reliability, 98%) and stage II sarcoidosis
(reliability, 89%), being, however, less accurate for the diagnosis of stage III
sarcoidosis (reliability, 52%) and stage 0 sarcoidosis (reliability, 23%).(
,
) Asymptomatic bilateral hilar lymphadenopathy without systemic findings or
with acute symptoms (uveitis, polyarthritis, or erythema nodosum) is highly suggestive
of sarcoidosis,(
,
) histological confirmation being therefore unnecessary. A combination of
pulmonary nodules with hilar/mediastinal adenopathy should immediately raise the
suspicion of sarcoidosis. In this context, nodules with a lymphatic distribution on HRCT
scans constitute further evidence for a diagnosis of sarcoidosis. In a retrospective
review of 91 cases of patients with hilar adenopathy and pulmonary nodules in our
registry of interstitial lung disease cases, sarcoidosis was diagnosed in 76 (84%),
silicosis was diagnosed in 10 (11%), and other diseases were diagnosed in 5 (5%;
unpublished data). In the present study, this combination was not useful for earlier
diagnosis.A total of 17 patients were misdiagnosed as having tuberculosis. Of those, 12 were
treated for more than 3 months and 4 completed the standard 6-month treatment
established in Brazil. This shorter treatment duration probably reflects a
reconsideration of the diagnosis after a lack of response to tuberculosis treatment.During the study period, in the three facilities involved, only 1 patient initially
treated as having sarcoidosis was later confirmed as having tuberculosis. The patient in
question was excluded from the present study.In patients with sarcoidosis, the only finding associated with more frequent diagnosis
of tuberculosis was the presence of systemic symptoms. Weight loss, fever, and night
sweats are common findings in patients with sarcoidosis.(
) Fatigue, which is another common finding in patients with
sarcoidosis,(
) was not evaluated in the present study.The risks of treating sarcoidosis as tuberculosis are not negligible.(
) In addition, potentially serious forms of sarcoidosis, such as cardiac
sarcoidosis, can go undetected because of the lack of systematic investigation. Ten of
the 17 patients treated as having tuberculosis had bilateral hilar lymphadenopathy,
which is a rare finding in patients with tuberculosis and should therefore suggest the
correct diagnosis.One study evaluated the usefulness of PCR in biopsy specimens in differentiating between
patients with sarcoidosis and those with tuberculosis.(
) In all 31 cases of patients with tuberculosis, PCR was positive, as it was
in 20 of 104 cases of patients with sarcoidosis. A quantitative analysis was able to
distinguish between the two groups. However, these findings should be confirmed at other
centers.In conclusion, the diagnosis of sarcoidosis is delayed in many cases, even when there
are imaging findings suggestive of sarcoidosis. Delayed diagnosis is associated with
lower lung function at diagnosis. In some cases, patients are diagnosed with and treated
for tuberculosis. This delays the correct diagnosis. General practitioners and even
pulmonologists should become more familiar with the findings that are suggestive of
sarcoidosis.
Introdução
O diagnóstico de sarcoidose é estabelecido quando achados clínicos e radiológicos
associam-se à presença histológica de granulomas e são excluídas outras causas desses
achados.(
,
) O diagnóstico de sarcoidose é frequentemente retardado por diversas
razões. A doença não é comum e pode afetar qualquer sistema orgânico. A apresentação
clínica pode ser característica da doença(
,
,
) ou atípica.Em países com um grande número de casos de tuberculose, como o Brasil, sempre se
considera o diagnóstico de tuberculose em pacientes cuja biópsia mostre granulomas, e
a doença é frequentemente tratada, mesmo sem comprovação bacteriológica. Assim como a
tuberculose, a sarcoidose é mais comum nos lobos superiores, e fibrose do parênquima
pode causar distorção arquitetural, retração hilar e redução do volume dos lobos
superiores.(
) Na sarcoidose, pode haver achados sistêmicos, tais como febre, perda de
peso e sudorese noturna, os quais podem induzir os médicos a estabelecer um
diagnóstico equivocado de tuberculose.(
) A sarcoidose é assintomática em muitos casos. Quando há sintomas, são em
geral inespecíficos. Finalmente, barreiras econômicas e a dificuldade de acesso a
especialistas podem afetar o tempo até o diagnóstico. Nos Estados Unidos, em um
estudo envolvendo 189 pacientes, houve diagnóstico tardio (> 6 meses após o início
dos sintomas) em 31% dos casos.(
) A presença de sintomas pulmonares e achados radiológicos indicativos de
doença mais avançada associaram-se a um tempo mais prolongado até o diagnóstico. Ao
contrário, a presença de lesões de pele resultou em diagnóstico mais precoce.O objetivo do presente estudo foi determinar a frequência do diagnóstico tardio de
sarcoidose em pacientes atendidos em centros especializados no Brasil e os fatores
relacionados a esse atraso.
Métodos
O protocolo do estudo foi aprovado pelo Comitê de Ética em Pesquisa da Universidade
Federal de São Paulo, em São Paulo (SP). Consentimento informado foi obtido de todos
os pacientes. Todos os pacientes incluídos no estudo apresentavam achados clínicos e
de biópsia compatíveis com sarcoidose. Outras condições foram excluídas por meio de
exames complementares, incluindo colorações especiais para micobactérias e fungos nas
lâminas de biópsia. Os pacientes foram recrutados em três locais, todos em São Paulo
(SP): o Ambulatório de Doenças Intersticiais do Hospital São Paulo (n = 68); o
Hospital do Servidor Público Estadual de São Paulo (n = 22) e a clínica privada de um
dos autores (n = 10).Um questionário padronizado foi aplicado no momento do diagnóstico ou
retrospectivamente, até um ano após o diagnóstico. Os dados foram coletados entre
junho de 2007 e maio de 2011.Foram excluídos os pacientes que não se lembravam do tempo do início dos sintomas e
do número de médicos consultados, bem como aqueles que não conseguiram realizar
manobras adequadas para a espirometria.Na avaliação inicial, os pacientes foram submetidos a anamnese, exame físico
detalhado e exames complementares para verificar se havia acometimento de diversos
órgãos pela sarcoidose.(
) As radiografias de tórax e as tomografias com data mais próxima da
biópsia foram revistas. Os achados nas radiografias de tórax foram classificados de
acordo com os critérios propostos por Scadding(
): estágio 0, sem envolvimento torácico; estágio I, adenopatia hilar
isolada; estágio II, aumento hilar com doença pulmonar intersticial; estágio III,
doença intersticial isolada e estágio IV, fibrose pulmonar.Todos os testes espirométricos preencheram os critérios de aceitabilidade e
reprodutibilidade propostos pela Sociedade Brasileira de Pneumologia e Tisiologia, e
os resultados foram expressos em valores previstos para a população
brasileira.(
,
) Os padrões anormais foram classificados em obstrutivo (relação
VEF1/CVF reduzida com CVF normal), restritivo (CVF e VEF1
reduzidos com relação VEF1/CVF normal) e misto (CVF e relação
VEF1/CVF reduzidas). A dispneia foi avaliada pelo índice de dispneia
basal.(
) A dispneia foi considerada grave quando o escore era ≤ 6.Os pacientes responderam a perguntas para que identificássemos a data de início dos
sintomas, a data da primeira consulta médica, o número de médicos consultados (e sua
especialidade) e a especialidade do médico que fez o diagnóstico de sarcoidose. Os
médicos foram classificados em clínicos gerais ou generalistas (incluindo médicos de
atendimento em emergências), pneumologistas, dermatologistas, oftalmologistas e
outros.A apresentação clínica da sarcoidose foi classificada nas seguintes categorias:
assintomática (achados radiológicos anormais sem sintomas); sintomas pulmonares;
sintomas cutâneos; sintomas oculares; sintomas sistêmicos (febre, perda de peso e
sudorese noturna) e outros.Os pacientes foram classificados em negros e brancos com base na cor da pele. O nível
educacional foi classificado em analfabeto, escolar elementar, escolar secundário e
universitário. A renda individual mensal foi quantificada em múltiplos de salário
mínimo médio na época da duração do estudo: 1-3 (600-1.800 reais); 4-10 e > 10. Os
seguros de saúde foram classificados em públicos e privados.O diagnóstico da sarcoidose foi classificado em precoce (< 6 meses) e tardio (≥ 6
meses) com base no tempo decorrido desde a primeira consulta médica até a data da
consulta na qual se estabeleceu o diagnóstico de sarcoidose.As variáveis com distribuição normal foram expressas em média e desvio-padrão, e as
sem distribuição normal foram expressas em mediana e variação. O teste do
qui-quadrado foi usado para comparar os pacientes com diagnóstico precoce àqueles com
diagnóstico tardio quanto às variáveis categóricas. No tocante às variáveis com
distribuição normal, a comparação entre os grupos foi feita por meio do teste t de
Student, e, no tocante àquelas com distribuição não normal, a comparação foi feita
por meio do teste de Kruskal-Wallis ou Mann-Whitney. Correlações entre o tempo
transcorrido desde o início dos sintomas até o diagnóstico e a CVF% foram realizadas
por meio do teste de Spearman.Uma análise preliminar de casos de sarcoidose mostrou que aproximadamente metade
tinha diagnóstico tardio da doença. Considerando-se relevante uma diferença de 20% na
frequência dos achados entre os dois grupos, com valores de α = 0,05 e potência (1 −
β) de 0,80, seria necessária uma amostra de 97 casos.(
) As análises estatísticas foram realizadas com o programa
Statistical Package for the Social Sciences, versão 17.0 para
Windows (SPSS Inc., Chicago, IL, EUA). Significância estatística foi estabelecida com
um valor de p < 0,05 (bicaudal).
Resultados
Cem pacientes foram incluídos no estudo. Os achados gerais são mostrados na Tabela 1. Mulheres brancas constituíram a
maioria da amostra. A média de idade foi de 48 anos, e a média de idade das mulheres
foi maior que a dos homens (50,1 ± 12,1 anos vs. 43,9 ± 11,4 anos; t = 2,57; p =
0,012).
Tabela 1
Características gerais de 100 pacientes com sarcoidose.a
Do total de pacientes incluídos no estudo, 79% tinham assistência médica pública, 52%
tinham nível superior e 60% tinham renda mensal abaixo de 2.000 reais.A Tabela 2 mostra os achados clínicos,
funcionais e radiológicos, bem como os locais que foram submetidos a biópsia e
revelaram lesões granulomatosas. O diagnóstico foi feito por meio de achados
incidentais na radiografia de tórax em 13% dos casos. Sintomas sistêmicos foram
comuns: 46% dos pacientes referiram perda de peso, 35% referiram sudorese noturna e
20% referiram febre. Dispneia acentuada foi relatada por 19%.
Tabela 2
Achados clínicos, funcionais e radiográficos e locais de biópsia
mostrando granulomas em 100 pacientes com sarcoidose.a
A média dos valores de CVF e VEF1 na espirometria situou-se próximo ao
limite inferior da faixa de referência, e a relação VEF1/CVF foi, em
média, normal. Os valores espirométricos se situaram na faixa de referência em 44
casos; 29 casos foram classificados em restritivos, 19 em obstrutivos e 8 em
mistos.Em 49 casos, o diagnóstico de sarcoidose foi feito por meio de biópsia
transbrônquica. Lesões cutâneas e adenopatias extratorácicas foram comuns, e 1/3 dos
casos foram diagnosticados com base em biópsias desses locais.A mediana do número de médicos consultados foi 3 (variação: 1-14). O primeiro médico
consultado foi um clínico geral em 54 casos, um pneumologista em 13, um
dermatologista em 7 e outros especialistas em 24. A sarcoidose foi diagnosticada na
primeira consulta em apenas 11 pacientes; 6 pacientes foram diagnosticados por um
dermatologista, 3 por um pneumologista e 2 por um clínico geral. Em 26 casos, cinco
ou mais médicos foram consultados antes da realização do diagnóstico.Todos os pacientes foram vistos em nossos centros, de modo que, ao final do estudo,
todos os casos foram vistos por um pneumologista. Antes disso, 68 pacientes
consultaram 90 pneumologistas, sendo o diagnóstico realizado em 28 casos (31%).A mediana do tempo transcorrido desde o início dos sintomas ou desde um achado
radiológico anormal até a primeira consulta foi de 1 mês; desde a primeira consulta
até o diagnóstico foi de 7,5 meses e desde o início dos sintomas ou desde um achado
radiológico anormal até o diagnóstico foi de 12 meses. Cinquenta e três pacientes
consultaram um médico dentro de 1 mês após o surgimento dos sintomas, e 82 pacientes
consultaram um médico dentro de 6 meses após o início dos sintomas.Cinquenta e nove pacientes (59%) receberam diagnóstico tardio (≥ 6 meses após a
primeira consulta), ao passo que os 41 restantes (41%) receberam diagnóstico precoce.
O tempo transcorrido desde o início dos sintomas até a primeira consulta foi
semelhante entre os dois grupos. A mediana do tempo de sintomas foi de 1 mês no grupo
com diagnóstico precoce e de 2 meses no grupo com diagnóstico tardio (z = 0,24; p =
0,81).O tempo transcorrido desde o início dos sintomas até a primeira consulta e desde o
início dos sintomas até o diagnóstico não diferiu entre os gêneros, raças, tipos de
seguro de saúde, níveis educacionais ou rendas (dados não mostrados).A Tabela 3 mostra a comparação dos dois
grupos de pacientes (pacientes com diagnóstico precoce e pacientes com diagnóstico
tardio) quanto aos achados clínicos, funcionais e de imagem. Os pacientes com
diagnóstico tardio tinham maior média de idade, menor CVF e VEF1 e foram
mais frequentemente diagnosticados e tratados como portadores de tuberculose. Dos 59
casos com diagnóstico tardio, 24 (40,7%) tinham distúrbio espirométrico restritivo,
comparados a 6 (14,6%) daqueles com diagnóstico precoce (p = 0,005).
Tabela 3
Achados clínicos, funcionais e radiográficos em 100 pacientes com
sarcoidose, separados de acordo com o tempo até o diagnóstico - precoce
(< 6 meses) ou tardio (≥ 6 meses).a
Sintomas sistêmicos e respiratórios e a frequência de envolvimento extratorácico não
diferiram entre os pacientes com diagnóstico precoce e aqueles com diagnóstico
tardio. Não houve diferença entre os pacientes com diagnóstico precoce e aqueles com
diagnóstico tardio no tocante aos estágios radiográficos. A TC estava disponível para
avaliação em 97 pacientes. Após a aplicação da classificação de Scadding à TC, 50
pacientes apresentaram mudança de estágio, 36 deles para um estágio mais elevado.
Pela TCAR, apenas 1 dos 8 casos classificados em estágio 0 com base na radiografia
simples permaneceu neste estágio.Em 21 casos, os estágios mudaram de I, II ou III para o estágio IV. Não houve
diferença entre os pacientes com diagnóstico precoce e aqueles com diagnóstico tardio
no tocante aos estágios tomográficos (Tabela
3). Quando os estágios foram agrupados em 0/I/II vs. estágios III/IV e
comparados em relação ao diagnóstico (precoce ou tardio), novamente não se observou
diferença.Dezessete pacientes receberam diagnóstico de tuberculose, nenhum deles com
confirmação bacteriológica. Todos os 17 inicialmente receberam tratamento para
tuberculose, 12 deles durante 3 meses ou mais. Como esperado, os pacientes tratados
durante 3 meses ou mais tiveram diagnóstico mais tardio (Tabela 3). Dez pacientes foram diagnosticados como portadores de
tuberculose por pneumologistas, 4 o foram por clínicos gerais e 3 o foram por outros
profissionais. Dos 17 pacientes com diagnóstico de tuberculose, 15 (88,2%)
apresentaram sintomas sistêmicos, contra 44 (53,0%) dos 83 pacientes sem diagnóstico
de tuberculose (χ2 = 7,24; p = 0,007).
Discussão
O presente estudo confirma que o atraso diagnóstico é comum na sarcoidose. O
diagnóstico de sarcoidose foi retardadopor 6 meses ou mais em aproximadamente 60%
dos casos. Apenas 11% dos pacientes foram diagnosticados pelo primeiro médico
consultado. Em 26% dos casos, cinco ou mais médicos foram consultados antes da
confirmação diagnóstica de sarcoidose.No estudo que inspirou o presente trabalho,(
) pacientes com sintomas pulmonares apresentavam maior tempo transcorrido
desde o início dos sintomas até o diagnóstico, bem como maior tempo transcorrido
desde a primeira consulta médica até o diagnóstico; a presença de envolvimento
cutâneo associou-se a um tempo menor até o diagnóstico de sarcoidose. No presente
estudo, esses achados não influenciaram o tempo até o diagnóstico.O tempo até o diagnóstico de sarcoidose (desde o início dos sintomas e desde a
primeira consulta médica) não foi afetado pelo gênero, raça, renda individual e tipo
de seguro de saúde. De acordo com o Instituto Brasileiro de Geografia e
Estatística,(
) a renda individual mensal média em São Paulo na época do estudo era de
aproximadamente 850 reais, semelhante ao valor médio encontrado no presente estudo
(aproximadamente 800 reais).Na maioria dos casos, um médico generalista foi o primeiro médico consultado, mas o
diagnóstico de sarcoidose foi feito na primeira consulta em apenas 1 caso.
Pneumologistas foram consultados na primeira visita em 13 casos, e o diagnóstico de
sarcoidose foi feito em apenas 3 deles.Dois vieses merecem consideração no presente estudo. O primeiro é o viés de
recrutamento; os indivíduos deste estudo podem não ser representativos dos pacientes
com sarcoidose em geral. No tocante à idade e ao gênero, nossos achados são
semelhantes aos de dois estudos epidemiológicos, um na Holanda e o outro no
Japão.(
,
) No estudo denominado A Case Control Etiologic Study of
Sarcoidosis,(
) o qual envolveu uma grande amostra representativa norte-americana de
pacientes com sarcoidose, foram observadas algumas semelhanças e diferenças em
comparação ao presente estudo. Como em nosso estudo, a média de idade foi maior em
mulheres, e as proporções de envolvimento da pele, linfonodos extratorácicos e
envolvimento ocular foram muito semelhantes nos dois estudos. Entretanto, em nossa
amostra, mais pacientes tinham doença em estágios III e IV (36% vs. 16%) e a
proporção de pacientes com sarcoidose em estágio I foi menor (21% vs.
40%).(
)No Brasil, não houve ainda um estudo populacional sobre sarcoidose. Dois estudos
realizados em centros de referência em duas cidades distintas (Porto Alegre e Rio de
Janeiro) mostraram achados gerais em pacientes com sarcoidose.(
,
) Os dois estudos mostraram que a doença é mais comum em pacientes do sexo
feminino, com maior número de casos nos estágios II e III. Em um dos
estudos,(
) houve diagnóstico prévio de tuberculose em 6 de 100 casos. O tempo
transcorrido até o diagnóstico não foi referido nesses estudos. Portanto, o presente
estudo assemelha-se a outros estudos brasileiros conduzidos em centros de referência,
mas as amostras investigadas provavelmente não são representativas da sarcoidose na
população geral. É possível que casos em estágios mais precoces não tenham sido
encaminhados a esses centros. Entretanto, no presente estudo, o número de pacientes
nos estágios I e II (n = 28) foi o mesmo nos grupos com e sem diagnóstico tardio,
indicando que o menor número de pacientes nesses estágios em nosso estudo em
comparação a estudos populacionais não explica os achados encontrados.Outro viés de nosso estudo é que, em muitos casos, os questionários foram aplicados
retrospectivamente, o que pode ter levado a um viés de recordação. Nós tentamos
minimizar esse viés excluindo pacientes que não se lembravam de detalhes de seus
casos. Contudo, essa é uma tarefa subjetiva, e o número de pacientes excluídos com
base nesse critério não foi registrado.A CVF foi menor em pacientes nos quais o tempo transcorrido desde o início dos
sintomas até o diagnóstico foi maior. Esse achado sugere que o retardo no diagnóstico
da sarcoidose pulmonar resulta em agravamento da disfunção pulmonar na ausência de
tratamento. Infelizmente, os resultados da função pulmonar em longo prazo, após
tratamento eventual, não foram registrados.No presente estudo, não encontramos uma relação entre os estágios radiográficos e o
retardo no diagnóstico. A TCAR é superior à radiografia de tórax para a detecção de
nódulos e de aumento dos linfonodos mediastinais,(
) mas esses achados não foram úteis para um diagnóstico mais precoce. Na
ausência de confirmação patológica, achados clínicos e radiológicos podem ser úteis
para o diagnóstico no estágio I (confiabilidade de 98%) ou II (89%), mas são menos
precisos para pacientes em estágio III (52%) ou 0 (23%).(
,
) Linfadenopatia hilar bilateral assintomática sem achados sistêmicos ou
com sintomas agudos (uveíte, poliartrite ou eritema nodoso) é altamente sugestiva de
sarcoidose,(
,
) e a comprovação histológica é portanto desnecessária. Uma combinação de
nódulos pulmonares com adenopatia hilar/mediastinal deveria despertar suspeita
imediata de sarcoidose. Nesse contexto, se os nódulos têm uma distribuição linfática
na TCAR, o diagnóstico de sarcoidose é ainda mais reforçado. Em uma revisão
retrospectiva de 91 casos com adenopatia hilar e nódulos pulmonares do nosso registro
de doenças pulmonares intersticiais, a sarcoidose foi diagnosticada em 76 (84%), a
silicose em 10 (11%) e outras doenças em 5 (5%; dados não publicados). No presente
estudo, essa combinação não se mostrou útil para um diagnóstico mais precoce.Houve diagnóstico equivocado de tuberculose em 17 pacientes, 12 dos quais foram
tratados durante mais de 3 meses e 4 completaram o tratamento-padrão de 6 meses
estabelecido no Brasil. Esse tempo de tratamento mais curto provavelmente reflete uma
reconsideração do diagnóstico após a ausência de resposta ao tratamento para
tuberculose.No período do presente estudo, nos três centros envolvidos, apenas 1 paciente tratado
inicialmente como portador de sarcoidose (excluído do presente estudo) teve
tuberculose comprovada na evolução.Em pacientes com sarcoidose, o único achado associado com diagnóstico mais frequente
de tuberculose foi a presença de sintomas sistêmicos. Perda de peso, febre e sudorese
noturna são achados comuns na sarcoidose.(
) Fadiga, outro achado comum na sarcoidose,(
) não foi avaliada no presente trabalho.Os riscos do tratamento da sarcoidose como tuberculose não são
desprezíveis.(
) Além disso, formas de acometimento potencialmente graves, tais como a
sarcoidose cardíaca, podem não ser detectadas por falta de investigação sistemática.
Dez dos 17 pacientes tratados como se tivessem tuberculose tinham adenomegalia hilar
bilateral, um achado raro em tuberculose e que deveria sugerir o diagnóstico correto.Um estudo avaliou a utilidade da PCR em fragmentos de biópsia para separar pacientes
com sarcoidose daqueles com tuberculose.(
) A PCR foi positiva em todos os 31 casos com tuberculose e em 20 dos 104
com sarcoidose. Uma análise quantitativa foi capaz de separar os dois grupos, porém
esses achados devem ser confirmados em outros centros.Em suma, o diagnostico de sarcoidose é tardio em muitos casos, mesmo quando há
achados de imagem sugestivos. O diagnóstico tardio se associa a menor função pulmonar
na época do diagnóstico. Em alguns casos, os pacientes recebem diagnóstico e
tratamento de tuberculose, o que retarda o diagnóstico correto. Os médicos
generalistas e mesmo os pneumologistas deveriam estar mais familiarizados com os
achados sugestivos de sarcoidose.
Authors: R P Baughman; A S Teirstein; M A Judson; M D Rossman; H Yeager; E A Bresnitz; L DePalo; G Hunninghake; M C Iannuzzi; C J Johns; G McLennan; D R Moller; L S Newman; D L Rabin; C Rose; B Rybicki; S E Weinberger; M L Terrin; G L Knatterud; R Cherniak Journal: Am J Respir Crit Care Med Date: 2001-11-15 Impact factor: 21.405
Authors: Marc A Judson; Bruce W Thompson; David L Rabin; Joanne Steimel; Genell L Knattereud; Daniel T Lackland; Cecile Rose; Cynthia S Rand; Robert P Baughman; Alvin S Teirstein Journal: Chest Date: 2003-02 Impact factor: 9.410
Authors: Y Zhou; H P Li; Q H Li; H Zheng; R X Zhang; G Chen; R P Baughman Journal: Sarcoidosis Vasc Diffuse Lung Dis Date: 2008-12 Impact factor: 0.670
Authors: Marjolein Drent; Jolanda De Vries; Merinke Lenters; Rob J S Lamers; Snjezana Rothkranz-Kos; Emiel F M Wouters; Marja P van Dieijen-Visser; Johny A Verschakelen Journal: Eur Radiol Date: 2003-06-17 Impact factor: 5.315
Authors: Manuel Ramos-Casals; Belchin Kostov; Pilar Brito-Zerón; Antoni Sisó-Almirall; Robert P Baughman Journal: Lung Date: 2019-06-12 Impact factor: 2.584
Authors: Marina Dornfeld Cunha Castro; Carlos Alberto de Castro Pereira; Maria Raquel Soares Journal: J Bras Pneumol Date: 2022-02-02 Impact factor: 2.624