Élie AZOULAY
Hôpital Saint-Louis, Service de Réanimation Médicale
Université Paris-Diderot, Sorbonne Paris-Cité
Groupe de Recherche Respiratoire en Réanimation Onco-Hématologique (GRRR-OH)
Respiratory events in hematology patients
Author
Patients Incidence Time Finding
Mortality
Azoulay 2004
Medicine
All
7-12%
(203/3782)
~
Undetermined
diagnosis
47.7%
Puig. 2007
Leuk Lymphoma
AutoBMT
15%
(49/326)
J11
Myeloma
Neutropenia
51% vs. 8%
Specchia 2003
Leuk Lymphoma
AL
27.7%
(80/288)
/
AML>ALL
No remission
Age
Sortie
d’aplasie
Chaoui 2004
Leukemia
AL
ICU admission
35% vs. 6%
46%
(30/65)
/
Three major advances:
- Leukemic infiltrates
- Non invasive diagnostic strategy
- ARDS
Pneumonia during chemotherapy
in Patients with Acute Leukemia
Garcia et al. Annals of the ATS, Vol. 10, No. 5 (2013), pp. 432-440.
85 Early Stage = 11%
148 throughout
induction = 21%
MDS > AML > ALL
801
AML/
ALL/
MDS
28-day
Risk
factor
mortality
for pneumonia are
Age>60
6%
without
years,
pneumonia
AML,with pneumonia
26%
↑Low
ICUplatelet
length count,
of stay
↑Low
Hospital
albumin
length
level,
of stay
↑Neutropenia
Costs
Leukemic infiltrates: Vincent F, Rev Mal Respir 2012
Pulmonary leukostasis
Leukemic infiltration
Lysis pneumopathy
Pulmonary leukostasis: ALI
 Quantitative
 “Pulmonary hyperviscosity”
 Rapidly growing hyperleukocytic AML.
 Constant if >200 000/mm3.
 Endothelial injury and activation from
microvascular invasion related to
hyperviscosity, leukocytic microthrombi,
oxygen steal and hypoxia.
Clinical manifestations of leukostasis
Central nervous system circulation
Headache, confusion, somnolence, dizziness, slurred speech,
impaired hearing, tinnitus, diplopia, delirium, stupor, coma
Retinal vein distension, retinal hemorrhage, papilledema
Intracranial hemorrhage
Penile circulation
Priapism
Cardiovascular system
Acute myocardial infarction
Right ventricular overload
Acral lividosis
 Rapidly rising white-cell count
 Cell type (monoblasts)
 A postmortem examination revealed profound
leukostasis in the heart, lungs, adrenal glands, liver,
and spleen.
Sludging of high numbers of leukocytes in
small vessels (brain, lungs, and kidneys)
Leukemic pulmonary infiltrates
 Blasts are aggregated in vascular
lumens
 The infiltrates typically follow the
lymphatic routes along the
bronchovascular bundles, interlobular
septa, and pleural interstitial tissue.
Lysis pneumopathy: DAD
 Occurs immediately or early after
chemotherapy
 Diffuse alveolar damage
LEUKEMIC
INFILTRATION
leukemic infiltration
of the alveolar capillaries
LEUKOSTASIS
Filling of a pulmonary artery with
leukemic cells, as well as leukemic
infiltration of alveolar capillaries.
Catching up errors
Valentino et al. CHEST. 2005,128(5):3629-3633
monoblaste
Timing
Poumon leucémique: prise en charge
 Infiltration pure:
 VNI, chimiothérapie classique,
 inhibition de l’adhésion du monoblaste sur l’endothélium
pulmonaire (DXM?)
 Leucostase ou leucostase + infiltration:
 Essai de VNI, cytoréduction (Hydréa)
 DXM?
 Poumon de lyse (DAD)
 Intubation et optimisation des échanges gazeux
 Attendre et éviter l’infection
Moreau AS et al. Submitted 2012
Mamez & Lengliné. In preparation
Three major advances:
- Leukemic infiltrates
- Non invasive diagnostic strategy
- ARDS
Invasive or noninvasive ?
 Invasive:
 Biopsy
 Semi-invasive:
 FO-BAL
 Non invasive:










Blood cultures
Sputa
Induced sputa
NPA/swabs
Viral PCR
Pan-PCR
Mass spectrometry
Antigenes
Echography
HRCT
Clinical Assessment for Identifying Causes of
Acute Respiratory Failure in Cancer Patients
Schnell D et al. European Respiratory Journal 2012
Culture based and non-culture based
tools
 Culture-based




 Non culture-based
 Sang
 Scanner
 Crachats
 Aspergillus
 LBA
galactomannan (GM),
 1,3-beta-D-glucan
 Méthodes moléculaires
Pathogènes colonisant
Faux positifs
Faux négatifs
Pas de gold standard
 Trop sensibles?
 Colonisation ou infection
 Pertinence?
The MiniMax study, Azoulay et al. 2010
Etiologies
MAXI
N=113
MINI
N=106
Bacteria
47 (41%)
39 (37%)
Viruses
7 (6%)
19 (18%)
Fongi
14 (12%)
9 (8.5%)
Pneumocystis
9 (8%)
10 (9.4%)
Malignant infiltration
10 (9%)
6 (6%)
Cardiac Pulm. Edem
7 (6%)
3 (3%)
More than one diagnosis
9 (8%)
9 (8.4%)
Diagnoses
 CPO: 10
Diagnosis made by NIT
 10 (100%)
 Fungi: 23
 22 (95.6%)
 Aspergilus: 18
 17 (94%)
 Virus: 26
 24 (92.3%)
 Malignancy: 16
 13 (81%)
 Pneumocystis: 19
 15 (79%)
 Bacteria: 86
 57 (66.3%)
Need for mechanical ventilation
P=0.62
0.
0.2
Cumulative in
and mec
0.0
Day-28 survival
0
5
10
15
20
25
28
0.6
0.4
0.2
0.0
Cumulative Survival
0.8
1.0
Time (days) after ICU admission
0
5
10
15
20
Time (days) after ICU admission
25
28
MiniMax dans le détail
Rentabilité
du LBA
Rentabilité
du LBA
TOUS
18%
Nodules
centrolobulaires
5%
Neutropéniques
10%
Nodules sous
pleuraux
4%
Non
Neutropéniques
20%
Consolidations
alvéolaires
20%
Lymphoïde
40%
Verre dépoli diffus
isolé
40%
MiniMax 2
 Corrélation clinico – radiographique – morphologique
 Stratégie diagnostique adaptative
 Utilisation de PCR pan-pathogène: bactéries,
fungiques, virus …
 Critères de jugement clinique pour éviter d’inclure
2000 patients: morbi-mortalité
Three major advances:
- Leukemic infiltrates
- Non invasive diagnostic strategy
- ARDS
Five evidence in the literature
 Heterogeneous (at least) disease
 Few positive trials, good negative trials
 Clinical – pathological discrepancies
 Berlin Definition
 Scarce data
ARDS and neutropenia
 Funny paradox (Frederic Ognibene)
 Pathophysiology
 Infection / infiltration / toxicity
 Used to be associated with poor outcomes
 Neutropenia recovery
Two recent papers
Turkoglu et al. (2013) Acute respiratory distress syndrome in patients
with hematological malignancies. Hematology 18:123-130
• 68 patients with hematological malignancies and ARDS. PO2/FiO2
was 104 (74-165). 10 (15%) NIV, 21 (31%) VM, and 36 (53%) both
• ICU mortality was 77% in the cohort.
• >two organ failures was the only independent risk factor for
mortality (P = 0.045), whereas NIV was associated with low mortality
(P = 0.001).
Soubani et al. (2014) The outcome of cancer patients with acute
respiratory distress syndrome. J Crit Care 29:183 e187-183 e112
• ARDS Network randomized controlled trials: 116/ 2515 patients
• Patients with cancer had significantly higher mortality (55.2%)
compared with those without cancer (24.3%) (P < .0001): OR 2.54
(95% confidence interval [CI], 1.570-4.120).
• APACHE III and age
6084 cancer patients
admitted to 14 ICUs
3042(50%) patients with
acute respiratory failure
1117 (36.7%) ARDS patients
according to the Berlin definition
Grrr-OH, 1990-2011
113 patients not
included ╬
1004 ARDS
patients included
Mild ARDS:
252 (25.1%)
Moderate
ARDS: 426
(42.4%)
Severe ARDS:
326 (32.5%)
Mortality: 59%
Mortality: 63%
Mortality:
68.5%
Study
population
(N=1004)
Survivors
(N=364)
Nonsurvivors
(N=640)
P value
Male gender
642 (63.9%)
240 (65.9%)
402 (62.8%)
0.32
Age (y)
58 [48-67]
57 [47-67]
58 [48-67]
0.33
Median (IQR) or n (%)
Underlying malignancy
Acute leukemia
298 (29.7%)
96 (26.4%)
202 (31.6%)
0.08
Non-Hodgkin’s lymphoma
318 (31.7%)
115 (31.6%)
203 (31.7%)
0.97
Myeloma
113 (11.3%)
34 (9.3%)
79 (12.3%)
<0.0001
Solid tumor
147 (14.6%)
60 (16.5%)
87 (13.6%)
0.21
Miscellaneous
95 (9.5%)
46 (12.6%)
48 (7.7%)
0.01
Allogeneic BMT/HSTC*
115 (11.5%)
36 (9.9%)
79 (12.3%)
0.23
Neutropenia
444 (44.2%)
148 (40.7%)
296 (46.3%)
0.08
Stage
Progressive
Partial/complete
remission
Newly diagnosed
Unknown
0.0003
458 (45.6%)
171 (47.0%)
287 (44.8%)
237 (23.6%)
100 (27.4%)
137 (21.4%)
72 (7.2%)
33 (9.1%)
39 (6.1%)
237 (23.6%)
60 (16.5%)
177 (27.7%)
Etiologies
Median (IQR) or n (%).
Cause of ARDS
Pulmonary infection #
Secondary ARDS#
Fungal infection&
Pneumocystis
No definite diagnosis§
Organ Support
NIV
NIV failure
Endotracheal MV
Vasopressors
Renal replacement therapy
Study
Survivors (N=364)
populati
on
(N=1004)
Non-
P value*
662
(65.9%)
281 (77.2%)
381 (59.5%)
<0.0001
225
(22.4%)
55 (15.1%)
170 (26.6%)
<0.0001
293
(30.7%)
83 (23.2%)
210 (35.1%)
0.0001
64
(6.4%)
41
(5.7%)
30 (8.2%)
34 (5.3%)
0.07
12 (4.5%)
29 (6.4%)
0.29
survivors (N=640)
387 (38.6%) 174 (47.8%)
213 (33.3%)
<0.0001
276 (27.5%)
173 (27.0%)
0.67
600 (93.8%)
490 (76.6%)
207 (32.3%)
<0.0001
0.0004
0.09
103 (28.3%)
893 (88.9%) 293 (80.5%)
731 (72.8%) 241 (66.2%)
306 (30.5%) 99 (27.2%)
Berlin definition
Figure 3.
1
P<0.0001
0.8
0.6
Mild ARDS
0.4
0.2
Severe ARDS
0
0
20
40
60
Days
80
100
1990-2001: trends in outcomes
Hospital mortality
1
0.8
0.6
0.4
0.2
0.0
Determinants of hospital mortality
Solid tumor
Need for emergency surgery
Allogeneic BMT/HSCT
mSOFA (per point)
Cause of respiratory involvement
No definite diagnosis
Primary ARDS
Secondary ARDS
Invasive fungal infection
Ventilation
NIV
NIV failure
Endotracheal MV
ARDS severity
Mild
Moderate
Severe
OR
95%CI
0.51
(0.34-0.77)
0.002
0.61
(0.35-1.05)
0.07
1.71
(1.07-2.71)
0.04
1.11
(1.06-1.16)
<0.001
0.41
(Reference)
(0.20-0.88)
0.02
0.90
(0.41-2.01)
0.80
1.72
(1.25-2.37)
0.001
2.93
(Reference)
(1.80-4.79)
<0.001
3.24
(2.02-5.24)
<0.001
1.25
(Reference)
(0.88-1.78)
0.22
1.61
(1.10-2.36)
0.01
1
1
1
P value
ARDS and Aspergillus
Monocyte de-activation
Copland, AJRCCM, 03
T cells switch to TH2/TH3
PLötz et al, Crit Care Med, 04
Increase anti-inflammatory CK (IL-10,
TNF-r1, IL-1ra)
Park et al, AJRCCM, 01
In a case-control study, ARDS was
independently associated with positive
Aspergillus in the LRT OR= 2,36 [1,144,89] (p=0,02)
Vandewoude, Crit Care, 06
8/64 patients with IPA
Invasive fungal infection and ARDS
 Need for additional descriptive data
 Need for better understanding of the risk factors
 Differences across IFI
 IPA
 Zygo
 Other mould / fungi
 Pneumocystis
 Towards primary preventive strategies?
 Noninvasive fungal diagnostic kit
Second confirmatory data set
 First study to be launched in the European group
 Observational data
 EORTC criteria (customized)
 Relationship between ARDS etiology and outcomes
 Prophylaxis
Three major advances:
- Leukemic infiltrates
- Non invasive diagnostic strategy
- ARDS