Laurent PAPAZIAN Réanimation médicale Hôpital Sainte-Marguerite Doit-on adopter une stratégie diagnostique des pneumonies en réanimation spécifique à " l’immunodéprimé " ? Herpesviruses have an envelope surrounding an icosahedral capsid, approximately 100nm in diameter, which contains the dsDNA genome Laurent PAPAZIAN Réanimation médicale Hôpital Sainte-Marguerite U R laurent.papazian@ap-hm.fr

Incidence des pneumonies de l’ID Tx médullaires: 30-60% Tx hépatique: 11-15% Tx cardiaque: 14-38% Tx pulmonaire: 50% Neutropénie: 0,5-10% VIH, cancer…

Immunodépression Chimiothérapie Corticoïdes Imurel et Endoxan Immunité cellulaire et immunité humorale Corticoïdes Altération fonction PNN et macrophages Imurel et Endoxan Inhibition production PNN et lympho Cyclosporine, tacrolimus Modulation fonctions lympho T

Autres facteurs Mécaniques Nutrition Colonisation Barrières Hypoalbuminémie  adhésion BGN Altération fonctions PNN Colonisation Oro-pharynx Sinus Estomac

Etiologies des pneumonies en hémato-cancéro Bactéries Communes (SA, Pneumocoque, P. aeruginosa, entérobact, HI) Intra-cellulaires Legionella Chlamydia Mycoplasma Nocardia sp., Actinomyces israeli Mycobactéries

P. jiroveci Fungiques Virales Aspergilloses Candida (candidémie) Fusarioses, zygomycetes Histoplasmose, blastomycose, coccidioidomycose Virales Resp: Rhino, Influenza, Parainfluenza, VRS Herpesvirus: HSV, CMV, HVZ, HHV6 Adenovirus

Germes, LBA et type de patients Joos et al. Respir Med 2007

Germes, LBA et type de patients Joos et al. Respir Med 2007

Germes, LBA et type de patients Joos et al. Respir Med 2007

Germes, LBA et type de patients Joos et al. Respir Med 2007

Etiologies chez HIV

Etiologies en Tx médullaire et chimio lourdes

Etiologies en cas de tumeurs solides

Cinétique des infections au décours de la Tx médullaire Soubani A.O. CCM 2006

Cinétique des infections au décours de la Tx médullaire Causes non-infectieuses HR intra-alvéolaire DAH = diffuse alveolar hemorrhage / IPS = idiopathic pneumonia syndrome: symptômes pneumonie mais sans infection documentée / engraftment syndrome = sortie de neutropénie avec aug CK et ALI / BO = bronchiolite oblitérante Soubani A.O. CCM 2006

Connaître le type de déficit immunitaire - GdM

Pneumonies et Tx médullaire

Neutropénie Transplantation Greffe de Moelle 1 Mois 3 Mois 6 Mois Bactérien Fungique Neutropénie Rejet Chronique Virus Transplantation Greffe de Moelle Virus - Champignons > Autres CMV > P jirovecii Bactéries Virus Respiratoires Temps 1 Mois 3 Mois 6 Mois

Pneumocystis BAL Gargarismes: PCR Coloration de référence: Papanicolaou Gomori (moins sensible) quand Papanicolaou est négatif ? PCR: plus sensible, moins spécifique Gargarismes: PCR

Le contexte Patient pris en charge en réanimation pour détresse respiratoire et échec de la prise en charge probabiliste

Difficultés du diagnostic Manque de spécificité des signes cliniques et radiologiques "Invasivité" des techniques de diagnostic Traitement anti-infectieux antérieur

Diagnostic chez immuno-déprimés,VIH- Rano et al. Thorax 2001 200 patients

Rentabilité LBA dans la vraie vie… Hémopathie maligne, n = 95 Antibiothérapie à large spectre + anti-fungique LBA contributif: 29/95 BGN, 40% CGP, 35% Mycobactérie, 11% Infection fungique, 11% CMV, 3% Hohenadel et al. Thorax 2001

Pneumonie - pas pneumonie ? 99 patients hémato 122 épisodes, 135 LBA Examens « classiques » + PCR: Pneumocystis, CMV, Legionella sp., mycobacteries, Mycoplasma pneumoniae, and Chlamydia pneumoniae Aspergillus antigénémie Hohenthal et al. Eur J Haematol 2005

Pneumonie chez le malade hémato Identification microbienne: 35,6% (48/135) Bactéries cultures quantitatives: 3 (2,2%) cultures spéciales: 4 (3%) Virus respiratoire: 10 (8,2%) PCR Pneumocystis: 21 PCR CMV positive: 18 Aspergillus antigénémie: 7 Modification traitement anti-infectieux: 27 épisodes (22,1%)

Pneumonie chez le malade hémato Identification microbienne: 35,6% (48/135) Bactéries cultures quantitatives: 3 (2,2%) cultures spéciales: 4 (3%) Virus respiratoire: 10 (8,2%) PCR Pneumocystis: 21 PCR CMV positive: 18 Aspergillus antigénémie: 7 Modification traitement anti-infectieux: 27 épisodes (22,1%)

Pneumonie chez le malade hémato Identification microbienne: 35,6% (48/135) Bactéries cultures quantitatives: 3 (2,2%) cultures spéciales: 4 (3%) Virus respiratoire: 10 (8,2%) PCR Pneumocystis: 21 PCR CMV positive: 18 Aspergillus antigénémie: 7 Modification traitement anti-infectieux: 27 épisodes (22,1%)

Causes non-infectieuses

Rentabilité fibroscopie 104 immunodéprimés non-VIH BW = bronchial washing Jain et al. Chest 2004

Contribution OLB: 100% ! Jain et al. Chest 2004

Rosen et Narasimhan CCM 2006 DR chez un patient VIH+ Rosen et Narasimhan CCM 2006

LBA - BTB Bulpa et al. ERJ 2003

Complications… Bulpa et al. ERJ 2003

TDM… Pneumonie à Pneumocystis chez VIH. Drain tho en place Pneumocystis jiroveci (formerly classified Pneumocystis carinii) has always been a major cause of illness and death in patients with HIV infection. Once thought to be a parasite, genomic analysis revealed that P. jiroveci is in fact a fungus that infects only humans, whereas P. carinii is pathogenic only in immunodeficient rats. Despite the change in taxonomy of this pathogen, the term PCP is still acceptable shorthand for Pneumocystis

Présentation RX et orientation étiologique Rosen et Narasimhan CCM 2006

Park DR Respir Care 2005

Localisation Rouby et al. ARRD 92

HSV et patients non-immunodéprimés Porteous et al. Crit Care Med 84 % oral shedding

HSV et patients non-immunodéprimés Porteous et al. Crit Care Med 84 % oral shedding

HSV et patients non-immunodéprimés Porteous et al. Crit Care Med 84 % oral shedding

Critères cliniques Clinical Pulmonary Infection Score * temperature °C > 36.5 and < 38.4 : 0 point > 38.5 and < 38.9 : 1 point > 39 or < 36 : 2 points * WBC, mm-3 < 4,000 and < 11,000 : 0 point < 4,000 or > 11,000 : 1 point + band forms > 500 = + 1 point * tracheal secretions < 14 + of tracheal secretions = 0 point > 14 + of tracheal secretions = 1 point + purulent secretions = + 1 point * PaO2/FiO2, mmHg > 240 or ARDS = 0 point < 240 and no evidence of ARDS = 1 point * pulmonary radiography no infiltrate = 0 point diffused (or patchy) infiltrate = 1 point localized infiltrate = 2 points * culture of TA (semiquantitative : 0,1,2 or 3 +) pathogenic bacteria cultured < 1 + or no growth = 0 point pathogenic bacteria cultured > 1 = 1 point + same bacteria on Gram stain > 1 + = + 1 point Pugin et al. ARRD 91

Luna et al. CCM 2003

BAL under fiberscopy technical considerations oro- or naso-tracheal tube > 7 mm FiO2 = 1, sedation ± curarisation aliquots (10 - 60 ml) 1st aliquot discarded

BAL under fiberscopy tolerance Papazian et al. Chest 93 PaO2/FIO2 * p < 0.05

Non-directed BAL : clinical studies n VAP threshold sens. spec. DPC 69 40 SQ 70% 69% PAC 13 9 no 100% 75% duodenography 28 13 BI > 5 73% 96% Ballard 36 14 103 100% 96%

Reproductibilité des examens invasifs BTP 24% des micro-organismes de part et d’autre du seuil de 103 cfu/ml LBA 33% des micro-organismes de part et d’autre du seuil de 104 cfu/ml Timsit et al. Chest 93 Gerbeaux et al. AJRCCM 98

Défaut de sensibilité de la brosse Johanson et al. ARRD 88

Tracheal aspirates diagnostic accuracy El-Ebiary et al. ARRD 93 * reference : blood or pleural cultures, histology * 54 patients (26 with pneumonia) TA 105 TA 105 PSB 103 PSB 103

Etudes comparatives avec histologie pour le diagnostic de VAP bactérienne ! sensibilité/spécificité BTP LBA AT Comb CPIS seuil =103 104 104 105 106 103 6 Torrès AJRCCM 94 36/50 50/45 - - - - - Marquette AJRCCM 95 58/89 47/10067/75 67/75 53/87 - - Chastre AJRCCM 95 82/89 91/78 - - - - - Papazian AJRCCM 95 33/95 50/95 72/80 56/95 44/10067/8072/85

Pathogens associated with inadequate antimicrobial treatment Kollef MH CID 2000

Présentation RX J0 J15

Groundglass attenuation J15

25 pneumonies à CMV prouvées histologiquement CMV et VAP Etude rétrospective (n = 2795 patients) SDRA et VM > 7 jours + histologie Exclusion : hémopathie maligne, VIH, corticoïdes au long cours, chimiothérapie Autopsies (n = 60), OLB (n = 26) 25 pneumonies à CMV prouvées histologiquement - CMV seul dans 88% des cas Papazian et al. Anesthesiology 1996

Diagnostic à l’admission début: 18 j (10 - 40) IgG+ à l’admission 13/18 Diagnostic à l’admission Défaillance respiratoire post-op cancer (oesophage, estomac, poumon) : n = 6 chirurgie cardiaque : n = 4 Péritonite: n = 3 BPCO: n = 5 Coma: n = 4 Méningite: n = 1 Myocardite: n = 1 EP: n = 1

Ventilation mécanique et HSV Plus âgés et plus souvent ventilés que les immunodéprimés Pas d’herpesvirus dans le LBA de sujets sains 26% des LBA de 132 patients Schuller et al. Am J Med 93 Tarp et al. Eur Respir J 2001 Prellner: 47% burns Prellner et al. Scand J Infect Dis 92

Portage HSV-1 chez les patients ventilés Oro-pharynx et trachée (PCR) 393 patients HSV-1: 27% Corrélation age ou APACHE II – portage HSV-1 Trachée et sang (culture) 95 patients HSV-1: 23% Oro-pharynx (culture) 617 patients HSV-1: 21% Ong et al. J Med Virol 2004 Increased hospital mortality in shedders Cook et al. Crit Care Med 2003 Bruynseels et al. The Lancet 2003

HSV et pathogénicité Patients SDRA: double-aveugle acyclovir-placebo % Tuxen et al. ARRD 87

HSV et pathogénicité Patients SDRA: double-aveugle acyclovir-placebo % Tuxen et al. ARRD 87

Bruynseels et al. The Lancet 2003 HSV et PAV ? Bruynseels et al. The Lancet 2003 mortalité p = 0.003 Weekly mini BAL MLR showed that the difference in mortality was related to disease severity TRT ?

Bruynseels et al. The Lancet 2003 HSV et PAV ? Bruynseels et al. The Lancet 2003 mortalité p = 0.45 p = 0.003 Weekly mini BAL MLR showed that the difference in mortality was related to disease severity TRT ?

Cytomégalovirus

Gerna and Lilleri Herpes 2006 Antigénémie Identifie les GB qui expriment la phosphoprotéine pp65 dans leur noyau Précocité Sensibilité In high-risk SOT recipients (i.e. the HCMV-seronegative), primary HCMV infection was treated on first antigenaemia positivity in blood Gerna and Lilleri Herpes 2006

CMV = HSV ? Cook et al. Crit Care Med 2003 *

CMV = HSV ? Cook et al. Crit Care Med 2003 *

CMV = HSV ? Cook et al. Crit Care Med 2003 *

CMV: signification ? 237 patients: au moins 1 antigénémie Incidence: 40/237 (17%) Type d’admission: surg,med,trauma Jaber et al. Chest 2005

Quels malades ? Jaber et al. Chest 2005

Morbidité – Mortalité et réanimation * * * * Mais: + de steroids, + d’IRA et CMV n’est pas FdR indépendant Jaber et al. Chest 2005

Tableau “typique”… - Vers la 2ème – 3ème semaine - Fièvre Détérioration de l’état respiratoire Bactériologie négative - Petite note cholestatique

Performance diagnostique Papazian et al. Anesthesiology 1998

Photomicrograph (original magnification, 400; hematoxylin-eosin stain) shows three large nuclei containing eosinophilic inclusion bodies (arrows) within hyperplastic pneumocytes. x 400 HES

OLB: résultats 100 SDRA 100 OLB

OLB et SDRA n PEEP PaO2/FiO2 morbidité Hill JTCVS 76 42 bed 6.5 (0 - 15) 84 (30 - 350) 1 air leak 1HR 1inf Ashbaugh AS 85 10 ? 10 - 20 42 - 74 0 Costa Auler EJRD 86 5 OR 5 - 12 123 (50 - 255) ? Warner ARRD 88 20 OR ? ? ? Meduri Chest 91 7 OR ? ? ? Canver JCVS 94 27 OR 9 ± 1 ? 6 air leaks 2 PNO Meduri Chest 94 12 OR ? ? 1 air leak Papazian Anesth 98 37 bed/OR 10 ± 3 118 (60 - 190) 5 air leak 1HR PNO Patel Chest 2004 57 OR 10 ± 4 145 ± 61 1 death, 1 HR, 12 air leaks

Modifications thérapeutiques n modifications Hill JTCVS 76 42 33% Warner ARRD 88 20 70% Canver JCVS 94 27 67% Papazian Anest 98 37 81% Patel Chest 2004 57 60%

OLB biopsie contributive biopsie non-contributive Papazian et al. CCM 2007

Mechanical complications

New treatment after OLB results 78 patients

Photomicrograph (original magnification, x 40; hematoxylin-eosin stain) shows diffuse interstitial and intraalveolar fibroblastic proliferation (arrows) with some mononuclear cell infiltration (diffuse alveolar damage, organizing stage) x 40 HES

CMV et fibrose Souris Péritonite Evaluation à 3 semaines CMV – Réactivation CMV Réactivation CMV + Gancyclovir Boxplot comparison of pulmonary fibrosis in latently infected (CMV), noninfected (CMV), and ganciclovir-treated (CMVGCV) mice 3 wks after cecal ligation and puncture. White belts indicate the median value. Boxes give the range between quartiles (25th and 75th percentiles). Notched black regions are 95% confidence intervals for the medians, and the dots indicate the full range of the distributions. Because distributions of pulmonary fibrosis scores are nonnormal, nonparametric tests were used to discern differences among study groups at day 21. Post hoc comparisons of pairs of groups were made using the Mann-Whitney-Wilcoxon test (nonparametric t-test). CMV reactivation was related to significant increases in pulmonary fibrosis compared with healthy controls after CLP (p .0007). Blocking CMV reactivation with ganciclovir after CLP appeared to prevent this increased level of fibrosis, as fibrosis was significantly lower in GCV-treated mice than in nontreated mice (p .0002) Cook et al. Crit Care Med 2006

CMV et fibrose Souris Péritonite Evaluation à 3 semaines CMV – Réactivation CMV Réactivation CMV + Gancyclovir Boxplot comparison of pulmonary fibrosis in latently infected (CMV), noninfected (CMV), and ganciclovir-treated (CMVGCV) mice 3 wks after cecal ligation and puncture. White belts indicate the median value. Boxes give the range between quartiles (25th and 75th percentiles). Notched black regions are 95% confidence intervals for the medians, and the dots indicate the full range of the distributions. Because distributions of pulmonary fibrosis scores are nonnormal, nonparametric tests were used to discern differences among study groups at day 21. Post hoc comparisons of pairs of groups were made using the Mann-Whitney-Wilcoxon test (nonparametric t-test). CMV reactivation was related to significant increases in pulmonary fibrosis compared with healthy controls after CLP (p .0007). Blocking CMV reactivation with ganciclovir after CLP appeared to prevent this increased level of fibrosis, as fibrosis was significantly lower in GCV-treated mice than in nontreated mice (p .0002) Cook et al. Crit Care Med 2006

CMV et fibrose Souris Péritonite Evaluation à 3 semaines CMV – Réactivation CMV Réactivation CMV + Gancyclovir Boxplot comparison of pulmonary fibrosis in latently infected (CMV), noninfected (CMV), and ganciclovir-treated (CMVGCV) mice 3 wks after cecal ligation and puncture. White belts indicate the median value. Boxes give the range between quartiles (25th and 75th percentiles). Notched black regions are 95% confidence intervals for the medians, and the dots indicate the full range of the distributions. Because distributions of pulmonary fibrosis scores are nonnormal, nonparametric tests were used to discern differences among study groups at day 21. Post hoc comparisons of pairs of groups were made using the Mann-Whitney-Wilcoxon test (nonparametric t-test). CMV reactivation was related to significant increases in pulmonary fibrosis compared with healthy controls after CLP (p .0007). Blocking CMV reactivation with ganciclovir after CLP appeared to prevent this increased level of fibrosis, as fibrosis was significantly lower in GCV-treated mice than in nontreated mice (p .0002) Cook et al. Crit Care Med 2006

Other micro-organisms BAL⊕ 4-fold increase in antibody titer stable increased antibody titer Acanthamoeba, Bosea Water-associated microorganisms Amebal pathogens VAP episodes, n=120 Berger et al. Emerg Infect Dis 2006

La Scola et al. Emerg Infect Dis 2005 Mimivirus La Scola et al. Emerg Infect Dis 2005 Transmission electron microscopy: non-enveloped icosahedral virions surrounded by fibrils Mimivirus pour Mimicking Microbes A l’intérieur d’amibes qui vivent dans l’eau Intra-cellulaire obligatoire

Co-infections N Série de 338 CAP chez des non-immunodéprimés De Roux et al. Chest 2004

Apport de la biologie moléculaire Legoff et al. J Clin Microbiol 2005 N Avec Fagon

VAP, adequacy of antibiotic therapy and mortality %

Adequate early antibiotic therapy Disease severity LOD  4 LOD > 4 Clec’h et al. Intensive Care Med 2004 ICU mortality % Inadequate empirical treatment seemed to be associated with a poor prognosis only in patients with Logistic Organ Dysfunction score < or = to 4.

Delay in the initiation of appropriate antibiotic treatment VAP, n=107 P < .01 % Appropriate but delayed antibiotic therapy Iregui et al. Chest 2002

Resistant strains even in EOP Giantsou et al. Intensive Care Med 2005 Micro-organisms responsible for 408 episodes of VAP, % EOP: More surgical patients, more patients who have received previous antibiotics, more corticosteroids Modification of the initial antibiotic therapy: 58% vs. 36%, p < .001

Diagnosis = BAL + blood + urine but initial antibiotherapy = TA !

One or two TA per week

Conclusions Stratégie diagnostique Stratégie thérapeutique Any difference ? Stratégie thérapeutique Quelques différences VAP précoce  Pn Immuno VAP tardive ? Biologie moléculaire laurent.papazian@ap-hm.fr