AJTCCM VOL. 29 NO. 2 2023 67
ORIGINAL RESEARCH: BRIEF REPORT
Diuse alveolar haemorrhage (DAH) is considered a rare condition in children. ere is no consensus on the management of DAH
syndromes in Africa or other low- and middle-income countries. In this brief report, the clinical characteristics, management and outcomes
of children treated for DAH in the Chris Hani Baragwanath Academic Hospital paediatric pulmonology unit in Johannesburg, South
Africa are described. Fieen children were included in this case series, of whom 11 (73.3%) presented with severe microcytic anaemia. Of
the 11 children who had bronchoalveolar lavage, 9 (81.8%; 60.0% of the total) had haemosiderin-laden macrophages on microscopy. Only
5 children had a lung biopsy, of whom 3 (60.0%) had capillaritis. All the children were started on oral prednisone at presentation, and 11
(73.3%) received additional complementary treatment. Nine children (60.0%) had normal haemoglobin levels 1 year aer initiation of
treatment. Our series supports previous reports that DAH is uncommon in children. A large proportion of our patients responded well to
treatment despite some resource limitations.
Keywords. Diuse alveolar haemorrhage, pulmonary haemosiderosis, pulmonary capillaritis.
Afr J Thoracic Crit Care Med 2023;29(2):e282. https://doi.org/10.7196/AJTCCM.2023.v29i2.282
Diuse alveolar haemorrhage (DAH) is a potentially life-threatening
condition that can present acutely with respiratory failure or more
insidiously with microcytic anaemia and diffuse radiological
changes.[1,2] DAH in children is considered to be a rare condition,
and most of the literature is limited to case reports or case series from
developed countries.[3]
DAH is suspected if >20% of alveolar macrophages (minimum
count of 200) stain positive for haemosiderin on a bronchoalveolar
lavage (BAL) or sputum sample.[1] e Golde score may also be used
to diagnose and assess the severity of DAH.[4] ere are many causes
of DAH, which can broadly be catergorised into those with or without
immune-mediated pulmonary capillaritis.
Pulmonary capillaritis describes a histological pattern of lung injury
in which autoantibodies are generated against neutrophilic components,
causing enzymatic damage to the endothelium of the alveolar capillary.
Pulmonary capillaritis may present as part of a systemic condition (such
as systemic lupus erythematosus, granulomatosis with polyangiitis
(GPA), microscopic polyangiitis (MPA) or Goodpasture syndrome),
or occur in isolation.[1,5,6] Antineutrophilic cytoplasmic antibodies
(ANCAs), antiglomerular basement membrane antibodies (AGBMAs),
anti-smooth-muscle antibodies (ASMAs), antinuclear antibodies
(ANAs) and rheumatoid factor may be measured, but generally lack
specicity.[7] DAH not associated with pulmonary capillaritis can either
be idiopathic pulmonary haemosiderosis (IPH) or secondary to causes
such as mitral stenosis, pulmonary venous hypertension or veno-
occlusive disease.
Corticosteroid therapy is generally accepted as the initial treatment
for DAH. Additional therapies such as azathioprine, cyclophosphamide
and hydroxychloroquine are often required.[1] Another option is
rituximab, a monoclonal antibody that acts on the CD20 receptor on
B lymphocytes, preventing immunoglobulin binding to antigens in
pulmonary capillaries.[1]
There are currently no published data on children hospitalised
with DAH in South Africa (SA). We therefore describe the clinical
characteristics, management and outcomes of children treated for
DAH in the Chris Hani Baragwanath Academic Hospital (CHBAH)
paediatric pulmonology unit in Johannesburg, SA.
A retrospective audit of children (aged 28 days - 16 years) with DAH
who were managed in the CHBAH paediatric pulmonology unit from 1
January 2011 to 30 April 2022 was undertaken. Children referred from
Diuse alveolar haemorrhage in children hospitalised in a
tertiary‑level hospital: A retrospective descriptive study
K Mopeli,1 MD; T Mabaso,1 MD; N Alli,2 PhD; Z Dangor,1 PhD; C Ver wey,1 PhD
1 Department of Paediatrics and Child Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
2 Department of Molecular Medicine and Haematology, School of Pathology, Faculty of Health Sciences, University of the Witwatersrand,
Johannesburg, South Africa; National Health Laboratory Service, Johannesburg, South Africa
Corresponding author: K Mopeli (reloe.mopeli@wits.ac.za)
Study synopsis
What the study adds
e study provides additional data on children presenting with diuse alveolar haemorrhage in a South African tertiary hospital.
What are the implications of the ndings
ere is a need for South African pulmonologists to come together and conduct a national audit of these patients in dierent hospitals to
determine the incidence in our country, as well as to inform a management plan in the presence or absence of specialised tests.
68 AJTCCM VOL. 29 NO. 2 2023
ORIGINAL RESEARCH: BRIEF REPORT
the general paediatric wards at CHBAH or from other hospitals in the
CHBAH’s catchment area are seen in the paediatric pulmonology unit
and their details are captured on a database.
The paediatric pulmonology database was searched for patients
fullling the description of DAH. e criteria used were any two of the
following: bilateral diuse inltrates on the chest radiograph, anaemia,
and haemosiderin-laden macrophages (HLMs) in sputum or BAL uid.
Clinical, laboratory and radiological information was extracted from
patient records and entered onto data collection forms before being
entered onto an Excel sheet (Windows 10; Microso Inc., USA).
ere are no guidelines or protocols for investigation or management
of children with DAH at CHBAH. Generally, the following laboratory
investigations are carried out: a full blood count, serum urea and
electrolytes, erythrocyte sedimentation rate, tests for immune-mediated
diseases such as cytoplasmic ANCAs (c-ANCAs) and perinuclear
ANCAs (p-ANCAs), AGBMAs, ASMAs and ANAs, and a screen for
coeliac disease with anti-gliadin, anti-tissue transglutaminase and anti-
endomysial antibodies. A bronchoscopy with BAL is undertaken to
identify HLMs. Chest radiographs and computed tomography (CT)
scans are performed as indicated. Paediatric surgery and postoperative
intensive care unit beds required for lung biopsies are limited, so
most children are diagnosed on the basis of clinical, haematological
and radiological ndings, and treated without complete histological
investigations.
Data are reported as proportions for categorical variables, and
medians and interquartile ranges (IQRs) for continuous variables.
Ethical approval was obtained from the Human Research Ethics
Committee (Medical) of the University of the Witwatersrand (ref. no.
M200828).
Of 2 614 children in the database, 18 (0.7%) who met the description
of DAH were identied. ree cases were excluded: 2 had severe iron
deciency anaemia only, and 1 was lost to follow-up with no clinical
data. A total of 15 children (0.6%) were therefore included in the
analysis.
Twelve of the children (80.0%) were female, with a median (IQR)
age of 35 (15 - 84) months (Table1). All the children were of black
African descent and HIV uninfected. irteen children (86.7%) lived
in Gauteng Province, 1 child (6.7%) was from North West Province
and 1 (6.7%) was from Limpopo Province. ree children (20.0%)
were underweight and 4 (26.7%) were stunted. Most children (n=12;
80.0%) had had multiple previous admissions (median of 4) to
hospital. Eleven (73.3%) presented with severe anaemia. ree children
(20.0%) were referred from other institutions where they had received
blood transfusions and were therefore not anaemic on presentation
(Supplementary Table1, https://www.samedical.org/le/2036). Four
children (26.7%) were hypoxic at presentation, and 1 child required
invasive ventilation.
All the children had diffuse bilateral opacification on the chest
radiograph noted by the attending clinician. Two children had CT
scans on presentation which were reported by radiologists; the rst
showed multiple tiny cysts throughout the lung elds, and the second
bilateral alveolar opacication. No organisms were cultured on sputa.
Four children (26.7%) were c-ANCA positive and 1 (6.7%) was
p-ANCA positive (Table1). Of the 13 children who had an ANA,
C3 and C4 test, all were negative or had results within the normal
range. Rheumatoid factor was measured in 4 cases, and all the results
were within the normal range. Five children screened negative for
coeliac disease. Tests for autoimmunity were repeated annually for
symptomatic children.
Five (33.3%) of the 15 children had HLMs observed on sputum
microscopy. Eleven children (73.3%) had a BAL, of whom 9 (81.8%;
60.0% of the total) were positive for HLMs. Five children (33.3%)
had a lung biopsy; 3 (60.0%) had evidence of capillaritis, while 2
(40.0%) had intra-alveolar HLMs without any evidence of capillaritis
(Supplementary Table1,
https://www.samedical.org/le/2036
). ose
who had no antibodies detected and had no capillaritis, or had not had
a biopsy done yet, were treated for IPH. One child was c-ANCA positive
with capillaritis on histology, and was diagnosed with GPA.
All the children were started on oral prednisone 2 mg/kg/d at
presentation; 5 (33.3%) also received three pulses of intravenous
methylprednisolone 500 mg/m2 per day on alternate days for a duration
of 5 days. Four (26.7%) remained on prednisone alone throughout
their management, while 11 (73.3%) received further complementary
treatment, including azathioprine (n=7; 46.7%), cyclophosphamide
(n=5; 33.3%), rituximab (n=2; 13.3%), mycophenolate mofetil (MMF)
(n=2; 13.3%) and hydroxychloroquine (n=7; 46.7%) (Table1). e
decision to initiate complementary therapy was made at the discretion
of the treating clinician. It was generally based on a continuous
dropping of haemoglobin coupled with elevated reticulocytes, and
clinical symptoms such as hypoxia (Supplementary Table2, https://
www.samedical.org/le/2037). One patient complained of blurring
vision, and was reported to have developed retinopathy of the le eye
that was thought to be caused by hydroxychloroquine. e drug was
subsequently stopped and her vision returned to normal.
Nine children (60.0%) had normal haemoglobin concentrations
1year aer initiation of treatment. ree (20.0%) were discharged from
the service, 5 (33.3%) are still being followed up and on treatment, and
7 (46.7%) were lost to follow-up.
In this retrospective study, we investigated the clinical presentation
and management of 15 (0.5%) children with DAH over an 11-year
period at a tertiary-level paediatric pulmonology service. Similar to
reports in high-income countries,[3,8] DAH is an uncommon clinical
condition in SA children, and there is a paucity of studies describing
the clinical characteristics and outcomes of these patients. In our study,
most children had a delayed diagnosis and presented with multiple
hospital admissions for anaemia or lower respiratory tract infections
before DAH was suspected. e delay in referral and presentation may
be due to lack of awareness about this condition among clinicians.
Five children (33.3%) had serological findings suggestive of an
immune-mediated condition. Negative serology does not exclude an
immune-mediated cause of DAH, so repeat testing aer initial screening
may be necessary.[5,9] In cases where an immune-mediated capillaritis is
suspected and serological investigations are not helpful, a lung biopsy
is indicated.[10]
Only 5 children had a lung biopsy; 3 had histological features of
pulmonary capillaritis, of whom only 1 was c-ANCA positive. A further
3 children without a biopsy were c-ANCA positive and another was
p-ANCA positive. We were therefore only condent of an underlying
immune-mediated cause in 7 (46.7%) of the 15 children with DAH;
the main differential diagnoses were GPA and MPA. A systematic
review found that >90% of children with GPA or MPA had ANCAs
detected and that they were predominantly female.[11] Despite its being
AJTCCM VOL. 29 NO. 2 2023 69
ORIGINAL RESEARCH: BRIEF REPORT
Table1. Characteristics, investigations, treatment and outcomes of children with diuse alveolar haemorrhage (N=15)
n (%)*
Characteristics
Female gender 12 (80.0)
Black African 15 (100)
HIV negative 15 (100)
Stunting (HAZ <–2) 4 (26.7)
Age at presentation (months), median (IQR) 35 (15 - 84)
Number of admissions before diagnosis was made (n=12), median (IQR) 4 (2 - 5)
Investigations
Haemoglobin (g/dL), median (IQR) 5.8 (2.7 - 9.5)
Platelets (× 109/L), median (IQR) 438 (361 - 542)
Creatinine (µmol/L), median (IQR) 35 (22 - 44)
INR (n=11), median (IQR) 1.03 (0.92 - 1.10)
ESR (mm/h) (n=9), median (IQR) 10 (8 - 11)
C3 (g/L) (n=13), median (IQR) 1.26 (1.17 - 1.41)
C4 (g/L) (n=13), median (IQR) 0.30 (0.21 - 0.40)
ANA
Positive 0
Negative 13 (86.7)
Unknown 2 (13.3)
c-ANCA
Positive 4 (26.7)
Negative 8 (53.3)
Unknown 3 (20.0)
p-ANCA
Positive 1 (6.7)
Negative 11 (73.3)
Unknown 3 (20.0)
AGBMA
Positive 1 (6.8)
Negative 7 (46.7)
Unknown 7 (46.7)
HLMs
Sputum 5 (33.3)
Bronchoalveolar lavage (n=11) 9 (81.8)
Negative 1 (6.7)
Histological ndings (n=5)
Capillaritis 3 (60.0)
No capillaritis 2 (40.0)
Treatment†
Steroids 15 (100)
Cyclophosphamide 5 (33.3)
Hydrochloroquine 7 (46.7)
Azathioprine 7 (46.7)
MMF 2 (13.3)
Rituximab 2 (13.3)
Outcome
On treatment 4 (26.7)
In remission 3 (20.0)
Transfer to another hospital 1 (6.7)
Lost to follow-up 7 (46.7)
HAZ = height-for-age z-score; IQR = interquartile range; INR = international normalised ratio; ESR = erythrocyte sedimentation rate; C3 = complement component 3; C4 = complement component 4;
ANA = antinuclear antibody; c-ANCA = cytoplasmic antineutrophilic cytoplasmic antibody; p-ANCA = perinuclear antineutrophilic cytoplasmic antibody; AGBMA = antiglomerular basement membrane
antibody; HLMs = haemosiderin-laden macrophages; MMF = mycophenolate mofetil.
*Except where otherwise indicated.
†Some children were on more than one treatment.
70 AJTCCM VOL. 29 NO. 2 2023
ORIGINAL RESEARCH: BRIEF REPORT
a tertiary-level hospital, there are resource constraints at CHBAH.
Consequently, a lung biopsy in all children with DAH was not possible.
In some instances, parents refused this procedure for cultural reasons,
or concerns around pain and the risk of complications. Some experts
believe that a lung biopsy should be mandatory and the gold standard
for identifying the cause of DAH.[5,12]
Immunosuppressive therapy was initiated even when the cause of
DAH was not known. All the patients were started on oral or intravenous
steroids, and those who remained symptomatic, with dropping
haemoglobin, respiratory distress and/or bilateral diuse inltrates
on the chest radiograph, received additional immunosuppressive
treatment in a stepwise manner. is was done in consultation with the
rheumatologist when an underlying immune-mediated condition was
present or suspected. Two challenging patients were ultimately treated
with MMF and rituximab. e rst was a girl who had histological
features of capillaritis and was serologically positive for c-ANCAs. e
working diagnosis was GPA, as she subsequently developed symmetrical
polyarthritis of the small joints as well as a diuse pattern of nodular
inltrates on the chest radiograph.[1] She has not had haematuria and her
renal function has remained normal. She has had multiple relapses with
a low haemoglobin level and respiratory distress. e second patient,
also a girl, was serologically positive for p-ANCAs. Despite receiving
intravenous methylprednisolone and cyclophosphamide pulsing, plus
azathioprine, her haemoglobin level would drop to as low as 3g/dL
during bleeding episodes. Both patients were clinically stable at the time
of submission of this report.
Most children in our series had a normal haemoglobin concentration
1 year aer treatment, with no other signs to suggest continued alveolar
haemorrhaging. However, long-term follow up is required, as those
who do not go into remission may develop brosis with reduced lung
capacity.[13] We recommend investigating all cases with a meticulous,
systematic and stepwise approach to avoid essential tests being omitted.
Antibody tests should ideally be repeated annually, as some children
with immune-mediated disease may not initially be seropositive. Eorts
should be made to conrm a histological diagnosis, but where there are
resource constraints, treatment can be initiated based on the ndings of
clinical and laboratory investigations.
ere were many limitations to this study. As it was a retrospective
review, there was missing information, and no uniform investigational
approach was undertaken. We had a small number of cases, and almost
half (n=7) were lost to follow-up. Possible reasons for loss to follow-up
may include the eect of poor socioeconomic circumstances, which make
it dicult for patients to come for ongoing check-ups, as well as inadequate
counselling or understanding of the seriousness of the condition.
DAH is an uncommon clinical condition in children that oen
responds to immunosuppressive therapy. However, there are no
evidence-based guidelines in the literature on how to investigate and
treat children with DAH, particularly in a setting where there are
resource constraints. Pragmatically, children need to be investigated
based on the availability of specialised testing, bronchoscopy and
surgical lung biopsy. Treatment should ideally be in collaboration
with rheumatologists when an immune-mediated condition is the
likely cause. ere is a need for SA pulmonologists to come together
and conduct a national audit of these patients in dierent hospitals
to determine the incidence in our country, as well as to inform a
management plan in the presence or absence of specialised tests.
Declaration. CV is a member of the editorial board.
Acknowledgements. None.
Author contributions. KM conceptualised the study and participated in
its design and performance, in the statistical analysis, and in draing and
revising the manuscript. CV and ZD participated in the statistical analysis,
and in draing and revising of the manuscript. TM and NA participated in
draing and revising of the manuscript. All the authors read and approved
the nal manuscript.
Funding. None.
Conicts of interest. None.
1. Parris D, van Niekerk A, Jeevarathnum AC, Green RJ. An approach to pulmonary
haemorrhage in children. S Afr Respir J 2017;23(3):63-70. https://doi.org/10.7196/
SARJ.2017.v23i3.152
2. Collard HR, Schwarz MI. Diuse alveolar hemorrhage. Clin Chest Med 2004;25(3):583-
592. https://doi.org/10.1016/j.ccm.2004.04.007
3. Nuesslein TG, Teig N, Rieger CH. Pulmonary haemosiderosis in infants and children.
Paediatr Respir Rev 2006;7(1):45-48. https://doi.org/10.1016/j.prrv.2005.11.003
4. Golde D, Drew W, Klein H, Finley T, Cline M. Occult pulmonary haemorrhage in
leukaemia. BMJ 1975;2(5964):166-168. https://doi.org/10.1136/bmj.2.5964.166
5. Susarla S, Fan L. Diuse alveolar hemorrhage syndromes in children. Curr Opin Pediatr
2007;19(3):314-320. https://doi.org/10.1097/MOP.0b013e3280dd8c4a
6. Wang H, Sun L, Tan W. Clinical features of children with pulmonary microscopic
polyangiitis: Report of 9 cases. PLoS ONE 2015;10(4):e0124352. https://doi.org/10.1371/
journal.pone.0124352
7. Taytard J, Nathan N, de Blic J, etal. New insights into pediatric idiopathic pulmonary
hemosiderosis: e French RespiRare® cohort. Orphanet J Rare Dis 2013;8:161. https://
doi.org/10.1186/1750-1172-8-161
8. Fullmer JJ, Langston C, Dishop MK, Fan LL. Pulmonary capillaritis in children:
Areview of eight cases with comparison to other alveolar hemorrhage syndromes.
JPediatr2005;146(3):376-381.https://doi.org/10.1016/j.jpeds.2004.10.025
9. Chin CI, Kohn SL, Keens TG, Margetis MF, Kato RM. A physician survey reveals
dierences in management of idiopathic pulmonary hemosiderosis. Orphanet J Rare
Dis 2015;10:98. https://doi.org/10.1186/s13023-015-0319-5
10. Specks U. Diuse alveolar hemorrhage syndromes. Curr Opin Rheumatol 2001;13(1):12-
17. https://doi.org/10.1097/00002281-200101000-00003
11. Ludici M, Quartier P, Terrier B, Mouthon L, Guillevin L, Guillevin L. Childhood-onset
granulomatosis with polyangiitis and microscopic polyangiitis: Systematic review and
meta-analysis. Orphanet J Rare Dis 2016;11:141. https://doi.org/10.1186/s13023-016-
0523-y
12. Zhang Y, Luo F, Wang N, Song Y, Tao Y. Clinical characteristics and prognosis of
idiopathic pulmonary hemosiderosis in pediatric patients. J Int Med Res 2019;47(1):293-
302. https://doi.org/10.1177/0300060518800652
13. Saha BK. Idiopathic pulmonary hemosiderosis: A state of the art review. Respir Med
2021;176:106234. https://doi.org/10.1016/j.rmed.2020.106234
Submitted 20 October 2022. Accepted 3 May 2023. Published xx July 2023.
