crossmark

Detection of Pneumocystis jirovecii by Quantitative PCR To

Differentiate Colonization and Pneumonia in Immunocompromised
HIV-Positive and HIV-Negative Patients
T. Fauchier,a L. Hasseine,a M. Gari-Toussaint,a V. Casanova,b P. M. Marty,a,c,d C. Pomaresa,c,d
Laboratoire de Parasitologie-Mycologie, CHU l’Archet 2, Nice, Francea; Département d’Information Médicale, Hôpital Cimiez, CHU Nice, Nice, Franceb; INSERM, U1065,
Centre Méditerranéen de Médecine Moléculaire, C3M, Toxines Microbiennes dans la Relation Hôte-Pathogènes, Nice, Francec; Université de Nice Sophia Antipolis, Faculté
de Médecine, Nice, Franced

Pneumocystis jirovecii pneumonia (PCP) is an acute and life-threatening lung disease caused by the fungus Pneumocystis jirove-
cii. The presentation of PCP in HIV-positive patients is well-known and consists of a triad of dyspnea, fever, and cough, whereas
the presentation of PCP in HIV-negative patients is atypical and consists of a sudden outbreak, O2 desaturation, and a rapid le-
thal outcome without therapy. Despite the availability of direct and indirect identification methods, the diagnosis of PCP re-
mains difficult. The cycle threshold (CT) values obtained by quantitative PCR (qPCR) allow estimation of the fungal burden. The
more elevated that the fungal burden is, the higher the probability that the diagnosis is pneumonia. The purposes of the present
study were to evaluate the CT values to differentiate colonization and pneumonia in a population of immunocompromised pa-
tients overall and patients stratified on the basis of their HIV infection status. Testing of bronchoalveolar lavage (BAL) fluid
samples from the whole population of qPCR-positive patients showed a mean CT value for patients with PCP of 28 (95% confi-
dence interval [CI], 26 to 30) and a mean CT value for colonized patients of 35 (95% CI, 34 to 36) (P < 10ⴚ3). For the subgroup of
HIV-positive patients, we demonstrated that a CT value below 27 excluded colonization and a CT value above 30 excluded PCP
with a specificity of 100% and a sensitivity of 80%, respectively. In the subgroup of HIV-negative patients, we demonstrated that
a CT value below 31 excluded colonization and a CT value above 35 excluded PCP with a specificity of 80% and a sensitivity of
80%, respectively. Thus, qPCR of BAL fluid samples is an important tool for the differentiation of colonization and pneumonia
in P. jirovecii-infected immunocompromised patients and patients stratified on the basis of HIV infection status with different
CT values.

P neumocystis jirovecii pneumonia (PCP) is an acute and life-

threatening lung disease caused by the fungus Pneumocystis
jirovecii (1, 2). This disease mainly occurs in immunocompro-
Despite the availability of direct and indirect identification
methods, the diagnosis of PCP remains difficult (30–32). Indeed,
the P. jirovecii fungus is difficult to grow in culture, and the sensi-
mised patients, such as HIV-positive (HIV⫹) patients and patients tivity of direct microscopic examination is low (26, 27, 33–35).
receiving corticosteroid therapy, chemotherapy, or biotherapy. PCP PCR has greatly increased the sensitivity of detection of P. jirovecii
used to be the most common opportunistic infection among AIDS DNA. However, the differentiation between colonization and
patients, occurring at an incidence of 3.9 per 1,000 patients per year in pneumonia can be difficult (36). In quantitative PCR (qPCR),
the 1980s and early 1990s (3). Highly active antiretroviral therapy amplification and, thus, the cycle threshold (CT) value are corre-
(HAART) has prodigiously decreased the rate of PCP; however, it lated with the fungal burden in the sample. The more elevated that
remains a cause of death among AIDS patients (4, 5). Moreover, the fungal burden is, the higher the probability that the diagnosis
immunosuppression due to chemotherapies for cancers and autoim- is pneumonia (37, 38). Thus, the use of qPCR could greatly in-
mune diseases has become a new risk factor for PCP among HIV- crease the specificity of detection with a good sensitivity even if it
negative (HIV⫺) patients (6, 7). The presentation of PCP in HIV- could not detect all cases of PCP (39–44).
positive patients is well-known and consists of a triad of dyspnea, Nowadays, several issues remain with PCP. First, the low fun-
fever, and cough, whereas the presentation of PCP in HIV⫺ patients is gus burden among HIV⫺ immunocompromised patients makes
atypical and consists of a sudden outbreak, O2 desaturation, and a the etiological agent responsible for the respiratory syndrome
rapid lethal outcome without therapy (8–10).
During the 2000s, new diagnostic methods, such as molecular
biology methods, allowed the detection of P. jirovecii in the spu- Received 2 December 2015 Returned for modification 18 December 2015
Accepted 14 March 2016
tum of healthy immunocompetent individuals, highlighting the
Accepted manuscript posted online 23 March 2016
fact that subjects without clinical symptoms of PCP can be colo-
Citation Fauchier T, Hasseine L, Gari-Toussaint M, Casanova V, Marty PM, Pomares
nized by the fungus (11–16). This colonization of nonimmuno- C. 2016. Detection of Pneumocystis jirovecii by quantitative PCR to differentiate
compromised patients has cast doubt on its importance in sudden colonization and pneumonia in immunocompromised HIV-positive and HIV-
infant death syndrome, chronic obstructive pulmonary disease, negative patients. J Clin Microbiol 54:1487–1495. doi:10.1128/JCM.03174-15.
cystic fibrosis, and other pulmonary syndromes (12, 15, 17–28). Editor: D. W. Warnock
Although the clinical conditions and diseases for which P. jirovecii Address correspondence to T. Fauchier, thomas.fauchier@gmail.com.
is responsible are unclear, the rate of colonization among individ- Copyright © 2016, American Society for Microbiology. All Rights Reserved.
uals is underestimated (29).

June 2016 Volume 54 Number 6 Journal of Clinical Microbiology jcm.asm.org 1487

Fauchier et al.

corded: underlying disease (cancer, leukemia, AIDS and HIV infection

status, autoimmune disease), radiological signs (obtained by X-ray anal-
ysis, computed tomography scan), data from a biological workup (lym-
phocyte cell count, lymphocyte CD4/CD8 ratio, HIV DNA burden, results
of direct physical examination), treatments (curative and prophylactic treat-
ments, long-term treatment for the underlying disease [i.e., corticoste-
roids, chemotherapies, HAART]), and clinical outcome.
All samples were isolated from patients as part of routine diagnosis
and treatment, and no unnecessary invasive procedures were performed.
qPCR methods. Extraction of DNA from respiratory samples was per-
formed with a QIAamp DNA minikit (Qiagen, Courtabœuf, France) ac-
cording to the manufacturer’s instructions. The elution volume was 100
␮l. For all BAL fluid samples, a standardized volume of 2 ml was extracted.
For the remaining types of samples, the whole sample was used. Before
extraction, all samples were centrifuged at 3,000 rpm for 5 min. The qPCR
described by Meliani et al. was performed (54). The amplified target was
the mitochondria1 large subunit (mtLSU) rRNA gene. A TaqMan probe
(FAM-AGATAGTCGAAAGGGAAC-TAMRA, where FAM is 6-carboxy-
fluorescein and TAMRA is 6-carboxytetramethylrhodamine) was used
with the following primers: PCW3 (CTTAAAATAAATAATCAGACTAT
GTGCGATAAG) and PCW4 (GGAGCTTTAATTACTGTTCTGGGC).
Negative controls (two samples without any DNA) and positive controls
(P. jirovecii DNA) were included in each run. The CT value was registered
for positive samples and was defined as the cycle number at which the
fluorescence generated within a reaction crossed the fluorescence thresh-
old. In addition to the Pneumocystis qPCR, for each sample, amplification
of the human albumin gene was performed to assess the correct progress

TABLE 1 Characteristics of the 225 qPCR⫹ patientsa

Characteristic Value
FIG 1 Flowchart of the study population. ⫹
Sample size (no. of qPCR patients) 225
Sex ratio (no. of men, no. of women) 1.74 (143, 82)
Mean (95% CI) age (yr) 59.65 (57.39–61.91)
difficult to establish (6). Second, the clinical evolution is dif-
ferent among patients, with severe outcomes being seen among Biological data
No. (%) of HIV-positive patients 45 (19.91)
HIV⫺ patients with nonspecific respiratory syndromes (2). Lastly,
Median (IQR) HIV burden (log10 no. of DNA 5.11 (3.23–6.99)
whereas for pathologies such as AIDS or Wegener granulomatosis copies/ml) for patients with a positive serology
there are recommendations for prevention, for other pathologies Median (IQR) lymphocyte count (no. of cells/mm3)b 1,000 (600–1,700)
the means of prevention remain unclear (45–51). In addition,
because of the possibility of drug resistance and side effects, ap- No. (%) of patients with the following CD4/CD8 ratioc:
propriate antibiotics should be prescribed and an accurate diag- ⬍0.1 40 (43.38)
nosis should be made (6, 7, 12, 52, 53). 0.1–1 30 (32.61)
The purposes of the study were to evaluate the use of CT values ⬎1 22 (23.91)
for the differentiation of colonization and pneumonia among an
No. (%) of patients with the following underlying
overall immunocompromised patient population and among disease or condition:
subgroups of HIV⫹ and HIV⫺ patients. Leukemia 47 (20.89)
Autoimmune diseases 43 (19.11)
MATERIALS AND METHODS HIV infection (main diagnosis) 44 (19.56)
Samples and patients. Respiratory samples from Nice University Hospi- Lung diseases 37 (16.44)
tal and other health care facilities in the southeast of France were analyzed. Cancers 30 (13.33)
qPCR and microscopy were performed in the parasitology-mycology lab- Others 14 (6.22)
oratory of Nice University Hospital. This prospective study was noninter- Transplantation 10 (4.44)
ventional, monocentric, and simple blinded (the patients and physicians
knew the diagnosis and the qualitative results of the qPCR, but neither the No. (%) of patients with the following symptoms:
physicians nor the patients knew the CT values). This inception cohort Fever 77 (34.22)
study was conducted from 1 April 2008 to 1 October 2013 with a minimal Cough 68 (30.22)
follow-up of 3 months. All respiratory samples (sputum, induced sputum, Dyspnea 125 (55.56)
bronchoalveolar lavage [BAL] fluid) from patients with respiratory symp- Two symptoms 76 (33.78)
toms received in our laboratory were analyzed by qPCR and microscopic Three symptoms 27 (12)
assays for the detection of P. jirovecii. All results positive for P. jirovecii by a
CI, confidence interval; IQR, interquartile range.
qPCR were included in the analysis. b
Data were missing for 22 patients, and data for patients with aplasia were excluded.
The following clinical and biological data for each patient were re- c
Data were missing for 133 patients.

1488 jcm.asm.org Journal of Clinical Microbiology June 2016 Volume 54 Number 6

 Pneumocystis jirovecii Detection by qPCR

TABLE 2 Comparison of the characteristics of the PCP and PCP⫺ patients

Value(s) for:
P value by
Characteristic PCP⫹ patients PCP⫺ patients univariate analysis
No. (%) of patients 73 (32.44) 152 (67.56)
No. (%) of male patients 48 (65.75) 96 (63.16) 0.63
Mean (95% CI) age (yr) 55.0 (51.21–59.8) 61.70 (59.04–64.38) 0.01
No. (%) of patients HIV positive 30 (41.10) 15 (9.87) ⬍0.01
Median (IQR) HIV burden (log10 no. of DNA copies/ml) 5.71 (5.31–5.91) 4.73 (3.81–5) 0.01
Biological data
Median (IQR) lymphocyte count (no. of cells/mm3) 800 (500–1,300) 1,100 (700–1,800) 0.03
Median (IQR) CD4 count (no. of cells/mm3) 79 (21–258) 258 (80–794) 0.02

No. (%) of patients with the following CD4/CD8 ratio: ⬍0.01

⬍0.1 17 (35.42) 5 (11.36)
0.1–1 17 (35.42) 13 (29.55)
⬎1 14 (29.16) 26 (59.09)

Mean (95% CI) CT value 28 (26–30) 35 (33–36) ⬍0.01

No. (%) of patients with the following underlying disease:

HIV infection (main comorbidity) 30 (41.10) 14 (9.21) ⬍0.01
Leukemia 19 (26.03) 28 (18.42) 0.18
Autoimmune disease 11 (15.07) 32 (21.05) 0.29
Cancers 6 (8.22) 24 (15.79) 0.14
Transplantations 3 (4.11) 7 (4.61) 1
Lung diseases 3 (4.11) 34 (22.37) ⬍0.01

No. (%) of patients with the following symptoms:

Fever 38 (52.05) 39 (25.66) ⬍0.01
Cough 35 (47.95) 33 (21.71) ⬍0.01
Dyspnea 55 (75.54) 70 (46.05) ⬍0.01
Two symptoms 42 (57.53) 34 (22.37) ⬍0.01
Three symptoms 23 (31.51) 4 (2.63) ⬍0.01

No. (%) of patients receiving the following curative treatment: ⬍10⫺4

Anti-Pneumocystis therapy 73 (100.00) 9 (14.08)
Anti-Pneumocystis therapy with corticosteroids 2 (2.74)
Antibioticsa 0 (0.00) 44 (61.97)
Corticosteroids (main line) 0 (0.00) 5 (7.04)
Antituberculosis drugs 0 (0.00) 9 (12.68)
Other drugsb 0 (0.00) 9 (12.68)
APPc alone 1 (2.17) 0 (0.00)
Combination of immunosuppressive therapies and APP 2 (4.34) 4 (6.49)
Combination of HAART and APP 2 (4.34) 0 (0.00)
Combination of APP and other drugsb 1 (2.17) 1 (1.62)

Total no. (%) of patients receiving APP 6 (13.02) 5 (8.11)

a
Antibiotics other than anti-Pneumocystis therapy.
b
Other drugs consisted of cardiologic drugs.
c
APP, anti-Pneumocystis prophylaxis.

of DNA extraction and to confirm the absence of PCR inhibitors. The The physician’s decision to consider a diagnosis of PCP for the patient
human albumin gene is a housekeeping gene. We used primers GCTGT was first made according to the clinical, biological (microscopy findings,
CATCTCTTGTGGGCTGT (0.3 ␮M) and ACTCATGGGAGCTGCTG lymphocyte CD4/CD8 ratio), and radiological data together with the pa-
GTTC (0.3 ␮M) and a probe specific for the human albumin gene (VIC- tient’s history (presence of a chronic disease, HIV infection status, corti-
GGAGAGATTTGTGTGGGCATGACA-TAMRA [0.25 ␮M]) (55). costeroid use). Before the end of the study, all cases were reviewed by a
Detection by microscopy. Samples were cytocentrifuged, and then collegial multidisciplinary team of pneumologists and infectious disease
the slides were stained with May-Grünwald-Giemsa (MGG) and toluidine specialists for determination of a final diagnosis of PCP or not. This final
blue O stains (56). Experienced personnel examined the slides for cyst and diagnosis was used as the reference, and the CT values obtained by qPCR
vegetative forms. The results were given qualitatively. were compared with this final diagnosis (PCP versus colonization, i.e., no
Comparison of the CT value with the physician’s final diagnosis. In PCP [PCP⫺]). Patients considered to have PCP were treated with curative
order to compare the CT values among standardized samples, only the CT sulfadiazine-trimethoprim or curative atovaquone. Some patients were
values for the BAL fluid samples were included in the analysis of sensitivity also treated with an anti-Pneumocystis drug to treat a pulmonary disease
and specificity and for definition of the CT cutoff values. other than PCP. These patients were considered not to have PCP. Patients

June 2016 Volume 54 Number 6 Journal of Clinical Microbiology jcm.asm.org 1489

Fauchier et al.

FIG 2 ROC curves of CT values obtained by qPCR for different populations. (A) ROC for the whole population of qPCR⫹ patients (HIV⫹ and HIV⫺ patients);
(B) ROC curve for the population of qPCR⫹ and HIV⫹ patients; (C) ROC curve for the population of qPCR⫹ and HIV⫺ patients.

with a positive qPCR result without a final diagnosis of PCP were consid- The long-term treatment for the chronic disease was unknown
ered colonized (PCP⫺). If the qPCR result was negative, the patient was for 102 patients. Among these 102 patients, 26 were HIV positive;
considered to be not infected by P. jirovecii. 12 of these patients were not receiving HAART because HIV in-
To our knowledge, no postmortem study was performed, and a cause fection was newly diagnosed or the patients were noncompliant.
of death was found for all patients who died.
Nine of the 12 patients were PCP positive, and 2 died. Six patients
Statistical analysis. Statistical analyses were performed with SAS soft-
ware (version 9.4). All the results were verified with an alpha risk thresh- had a positive qPCR result, despite anti-Pneumocystis preventive
old of 0.05. The Student and Mann-Whitney tests were used for qualita- therapy, and among them, 4 were diagnosed with PCP.
tive variables, and the chi-square or Fisher test was used for categorical Direct examinations were positive for 30 (13.33%) patients,
variables. Univariate survival analyses were performed by a log rank test. and they were all diagnosed with PCP. Interestingly, at least two
CT value thresholds were estimated with receiver operating characteristic different qPCR results (negative to positive or positive to negative
(ROC) curves, which were used to compare the sensitivity and specificity without a statistically significant difference) were obtained for 13
of the CT values obtained by qPCR with the clinical diagnosis (39). CT (5.78%) patients who were confirmed to have a positive qPCR
values were obtained with 80% sensitivity and 80% specificity. result. They were all PCP⫺.
Analysis of the risk factors for PCP was performed by logistic regres-
Comparison of the PCP and PCPⴚ patients among the 225
sion, with an adjustment for age being used for multivariate analysis. For
the survival analysis, we used the Cox model with an adjustment for sex qPCR-positive patients. The characteristics of the PCP and PCP⫺
and age. The collinearities between the lymphocyte CD4/CD8 ratio and patients among the 225 qPCR-positive patients are compared in
the clinical diagnosis or death were studied by use of the Spearman cor- Table 2. The distribution of the underlying disease and the long-
relation. term treatments were different between the PCP and PCP⫺ pa-
On the basis of the findings of the study of Hughes et al., the prevalence tients (P ⬍ 10⫺4), with the cause of immunosuppression in the
of the risk of development of PCP in the population of HIV⫹ and HIV⫺ majority of PCP patients being a disease or condition other than
patients was 35% (57). Taking this number into account, we estimated HIV infection. There was no difference between the PCP and
that 70 PCP patients and 130 PCP⫺ patients were the minimal numbers of
PCP⫺ patients in the long-term treatments used to treat the
patients required to obtain an expected sensitivity of 90% with a lower
95% confidence limit of greater than 0.75 with a 0.95 probability (58).
chronic disease (PCP prophylaxis, immunosuppressive therapies,
HAART, and others drugs) (P ⫽ 0.06), but there were differences
RESULTS in the curative treatments used. All patients with PCP received
From 1 April 2008 to 1 October 2013, we performed 1,964 qPCRs anti-Pneumocystis therapy. Two HIV⫹ patients with PCP received
for 986 patients, with the results for 226 (11.5%) of the qPCRs adjuvant corticosteroid therapy. Among the nine (14.08%) PCP⫺
being positive (qPCR⫹). Seventy-three patients (33.77%) were di- patients who received anti-Pneumocystis therapy, five were first
agnosed with PCP. No patient with PCP had a negative qPCR suspected of having PCP but the diagnosis was later changed to
result. This whole population was separated into several sub- PCP⫺, and for the other four patients, either the patient had a
groups, as shown in Fig. 1. bacterial infection that was treated with anti-Pneumocystis therapy
Description of the entire qPCRⴙ population. The character- or the decision was made to consolidate P. jirovecii prophylaxis.
istics of the population of qPCR⫹ patients are described in Table 1. The rate of occurrence of PCP was 41.1% among the HIV⫹ pa-
All qPCR⫹ patients suffered from chronic diseases (Table 1). Sev- tients and 58.9% among the HIV⫺ patients (P ⬍ 0.01). The evo-
enty-four percent of the patients were receiving immunosuppres- lution of symptoms was favorable in 82.09% of PCP patients,
sive therapies, and 13.82% were receiving HAART. All patients whereas it was favorable in 69.08% of PCP⫺ patients (P ⫽ 0.03).
were immunocompromised and had low lymphocyte counts. The The vital status at the end of the study was unknown for two PCP⫺
diagnosis for two patients was later corrected and they were con- patients. There was no statistically significant difference between
sidered colonized (i.e., the diagnosis of PCP for two patients be- the times to death between the PCP and PCP⫺ patients (487 days
came PCP⫺). [interquartile range {IQR}, 377 to 597 days] and 492 days [IQR,

1490 jcm.asm.org Journal of Clinical Microbiology June 2016 Volume 54 Number 6

 Pneumocystis jirovecii Detection by qPCR

FIG 3 Graphical representation of CT values (indicated on the gray bar at the top) according to the population studied with the associated exclusion diagnostics.
Blue arrows and dashed blue arrows, mean CT values and 95% CIs, respectively, for the population of PCP and PCP⫺ patients; green, orange, and yellow squares,
CT values for the different populations used to exclude PCP and colonization with 80% sensitivity and 80% specificity, respectively.

423 to 562 days] [P ⫽ 0.15], respectively) and the rates of survival Among the HIV⫹ patients, the ROC curve showed that a CT
between PCP and PCP⫺ patients (23.29% and 28.95%, respec- value of 27 was able to exclude colonization with a sensitivity of
tively; P ⫽ 0.37). Comparison of the results of microscopic anal- 74% (95% CI, 46 to 84%) and a specificity of 100% (95% CI, 72 to
ysis and the clinical decision of whether a patient had a diagnosis 100%) (AUC, 0.90; 95% CI, 0.79 to 1.00) (Fig. 2B).
of PCP or was colonized with PCP confirmed the high specificity For the HIV⫺ patients, the ROC curve showed that a CT value
(100%) but the low sensitivity (41%) of this method. The com- of 35 excluded colonization with a sensitivity of 80% (95% CI, 63
pleteness of the 5-year follow-up in the area studied allowed us to to 88%) and a specificity of 60% (95% CI, 50 to 69%) (AUC, 0.72;
investigate the seasonality issue, and we did not find a statistically 95% CI, 0.62 to 0.82) (Fig. 2C).
significant result. For all the populations described above, we propose that the CT
Cycle threshold values: a clue to differentiate Pneumocystis values are able to exclude colonization and pneumonia with a
jirovecii pneumonia and colonization. The CT values obtained specificity of 80% and a sensitivity of 80%, respectively (Fig. 3;
for the different subgroups were compared by taking into account Table 3). A gray zone between these CT values remained, and for
only the values obtained with the BAL fluid specimens (Fig. 1). patients with these values, physicians will have to decide by the use
The mean CT value for PCP patients was 28 (95% confidence of other parameters whether the patient is colonized or should be
interval [CI], 26 to 30), whereas it was 35 (95% CI, 34 to 36) for considered to have PCP (Fig. 3).
colonized patients (P ⬍ 10⫺3), confirming the low fungus burden Interestingly, 2 PCP patients with CT values of 35 and 36 were
among colonized patients. The ROC curve showed that a CT value retrospectively considered to be PCP⫺ by the physicians. These
of 32 had an ability to discriminate between colonization and PCP data confirmed the good exclusion diagnostic of a CT value above
with a 72% sensitivity (95% CI, 57 to 81%) and a 75% specificity 35 in HIV⫺ patients.
(95% CI, 67 to 83%) (area under the curve [AUC], 0.88; 95% CI, In our study, all the CT values for the subgroups described in
0.80 to 0.96) (Fig. 2A). Table 3 matched the clinical decision to consider a patient to
When we stratified the patients on the basis of their HIV infec- have a diagnosis of PCP or to be colonized with PCP. The
tion status, the mean CT values were 24.5 (95% CI, 21.8 to 27.5) negative predictive values (NPVs) were higher for HIV⫹ pa-
and 30.5 (95% CI, 28.0 to 32.8) (P ⫽ 0.004) for HIV⫹ and HIV⫺ tients than for HIV⫺ patients (90% and 80%, respectively). The
patients, respectively. These data confirmed the high fungus bur- characteristics of the patients with CT values in the gray zone
den in HIV⫹ patients. are described in Table 4.

TABLE 3 CT values for exclusion of a diagnosis and the corresponding values of sensitivity and specificity according to the population studied
Exclusion of PCP Exclusion of colonization
Patient group CT value Sensitivity Specificity CT value Sensitivity Specificity
Whole qPCR⫹ population ⬎35 0.80 0.63 ⬍31 0.66 0.80
HIV⫹ patients ⬎30 0.80 0.79 ⬍27 0.73 1.00
HIV⫺ patients ⬎35 0.80 0.60 ⬍31 0.59 0.80

June 2016 Volume 54 Number 6 Journal of Clinical Microbiology jcm.asm.org 1491

Fauchier et al.

TABLE 4 Clinical data for patients with qPCR⫹ result and CT values in gray zonea
PCP Microscopy Underlying disease
Patient group and no. status Sex result or conditionb Treatment Death
qPCR⫹ HIV⫹ patients with gray zone CT
values between 27 and 30
167 Neg M Neg HIV infection (5.11) HAART No
168 Neg F Neg HIV infection (4.49) Combination of APP and other drugsd No
193 Pos M Neg HIV infection (6.34) Anti-Pneumocystis therapy No

qPCR⫹ HIV⫺ patients with gray zone CT

values between 31 and 35
22 Neg M Neg Lung disease Other drugsd No
26 Neg F Neg Lung disease Antituberculosis drugs No
36 Neg M Neg Cancer Antibiotics No
48 Neg F Neg Leukemia Missing data Yes
52 Neg M Neg Autoimmune disease Antibiotics No
55 Neg F Neg Cancer Other drugsd No
63 Neg M Neg Other Missing data Yes
69 Neg M Neg Leukemia Missing data Yes
74c Neg F Neg Autoimmune disease Anti-Pneumocystis therapy Yes
75 Neg M Neg Transplantation Missing data Yes
77 Neg F Neg Lung disease Missing data No
78 Neg M Neg Leukemia Antibiotics Yes
81 Pos M Neg Leukemia Anti-Pneumocystis therapy No
91 Pos M Neg Leukemia Anti-Pneumocystis therapy No
94 Neg F Neg Autoimmune disease Corticosteroid No
104 Pos F Neg Cancer Anti-Pneumocystis therapy Yes
107 Neg F Neg Leukemia Corticosteroid Yes
121 Neg F Neg Autoimmune disease Antibiotics No
131 Pos M Neg Autoimmune disease Anti-Pneumocystis therapy No
134 Neg F Neg Lung disease Other drugsd No
135 Pos M Neg Leukemia Anti-Pneumocystis therapy No
151 Neg M Neg Autoimmune disease Antibiotics No
152 Neg F Neg Lung disease Antituberculosis drugs No
174 Pos M Pos Leukemia Anti-Pneumocystis therapy Yes
183 Pos F Neg Leukemia Anti-Pneumocystis therapy Yes
185 Neg F Neg Lung disease Corticosteroid Yes
197 Neg M Neg Other Antibiotics No
198 Pos M Neg Leukemia Anti-Pneumocystis therapy No
201 Neg F Neg Leukemia Antibiotics Yes
202 Neg M Neg Transplantation Missing data No
206 Neg M Neg Autoimmune disease Antibiotics No
210 Pos M Neg Leukemia Anti-Pneumocystis therapy Yes
211 Neg M Neg Leukemia Missing data No
213 Neg F Neg Autoimmune disease Antibiotics Yes
226 Neg F Neg Leukemia Missing data No
a
CT values in the gray zone are defined in Fig. 3. Neg, negative; Pos, positive; M, male; F, female; APP, anti-Pneumocystis prophylaxis.
b
The values in parentheses represent the HIV burden (log10 number of DNA copies per milliliter) in HIV⫹ patients.
c
Patient 74 was first diagnosed with PCP, but the final diagnosis was an infection by influenza virus.
d
Other drugs consisted of cardiologic drugs.

Risk factors for Pneumocystis jirovecii pneumonia and sur- TABLE 5 Final analysis of risk factors for PCP
vival analysis. In order to study the risk factors for PCP, we
Variable Adjusted P value Hazard ratio (95% CI)
analyzed the underlying disease and biological parameters by
the use of age-adjusted P values (Table 5). Because of non-log Age 0.82 1.00 (0.97–1.02)
linearity, CT values were classified into the following groups:
CT value
those below 25, those between 25 and 35, and those above 35. ⬍25 ⬍0.01 22.20 (6.53–75.41)
After multivariate block analysis, CT values, hemopathy, and 25–35 0.03 4.69 (1.49–14.78)
HIV-positive status were found to be significant risk factors for ⬎35a
PCP (P ⫽ 0.05).
Collinearity showed an inverse association between the CD4/ HIV ⬍0.01 4.84 (1.82–12.91)
CD8 ratio and patient death. The lower that the ratio was, the Hemopathy ⬍0.01 3.32 (1.42–7.76)
higher that the risk of death was (F ⬍ 0.05). The rate of survival of a
The reference cutoff value.

1492 jcm.asm.org Journal of Clinical Microbiology June 2016 Volume 54 Number 6

 Pneumocystis jirovecii Detection by qPCR

HIV⫹ patients was better than that of HIV⫺ patients (log rank test Whereas some studies estimated the rate of mortality to be 7 to
P ⫽ 0.02). Nevertheless, regression analysis by use of the Cox 10% among HIV⫹ PCP patients, the rate of mortality increased
model did not find any risk factor for mortality. from 10% in renal transplantation patients to 50% in cancer pa-
tients (66–68). These rates of mortality raise the question of the
DISCUSSION condition of prophylaxis and how it should be implemented in
In this study, we demonstrated that the use of qPCR could greatly HIV⫺ patients. Another public health problem is the number of
improve the differential diagnosis of colonization and pneumonia colonized patients in hospital units where other patients may be at
in patients infected with P. jirovecii. Contrary to the findings of high risk for infection or colonization. In our study, 67.6% (n ⫽
some studies, no patient with a negative qPCR result was observed 152) of patients were colonized, which can potentially lead to nos-
among patients with PCP in the area of the Nice University Hos- ocomial infections (23, 69, 70).
pital between April 2008 and October 2013 (39–44). All patients The analysis of the CT value was performed by taking into
with positive microscopy results were diagnosed with PCP and account only BAL fluid samples, for which a standardized proto-
had CT values below the threshold for exclusion of a diagnosis of col of extraction and PCR allowed comparison of the data ob-
PCP. Thus, if microscopy had been chosen as the reference, there tained for the different patients. This represents a selection bias, as
would not have been false-negative qPCR results in our study. not all the patients were able to undergo BAL because of their lung
Whereas other studies used microscopy examination as the refer- function; however, it was necessary to use a standard sample.
ence method, we chose to use clinical presumption to consider The diagnosis of PCP or colonization depends on a complex
whether the patient had PCP or not (37, 38, 59, 60). A rapid diag- algorithm based on patient history, biological and radiological
nosis of PCP is required, as any delay in the initiation of the treat- data, treatment, and patient clinical evolution. If a positive micro-
ment increases the risk of mortality. Once the pulmonary sample scopic examination leads to a high probability of PCP, a negative
is collected, PCR can rapidly be performed; however, a positive qPCR result cannot exclude the diagnosis, especially in HIV⫺ pa-
qPCR result for immunocompromised patients could be mislead- tients. In most cases, qPCR is sensitive enough to allow a diagnosis
ing in the determination of whether the patient had PCP or was of PCP to be made in HIV⫺ patients; however, the presence of a
colonized. This highlights the need for an accurate diagnosis in gray zone of CT values prevents this assay from becoming the
order to avoid unnecessary treatment. It is thus important to de- reference method (44, 71). In the case of a CT value in the gray
termine CT threshold values to give physicians information that is zone, the physician will have to choose between prophylactic or
as accurate as possible. In our study, the CT values obtained by curative treatment according to the findings for the other param-
qPCR showed a high power to differentiate colonization and eters described above.
pneumonia. Moreover, we were able to determine different CT Conclusion. qPCR of BAL fluid samples is an important tool
values according to the population of immunocompromised pa- for the differentiation of colonization and pneumonia in patients
tients: HIV⫹ versus HIV⫺ patients. As already described, we with P. jirovecii infections. The CT values are different after strat-
found a correlation between the fungal burden and CT values: the ification on the basis of the patient’s HIV infection status. When
higher that the fungal burden was, the lower that the CT value was CT values are in the gray zone, the qPCR result, in addition to the
(37, 38, 61). The mean CT value was significantly higher for HIV⫺ findings for other clinical and paraclinical parameters, will help to
immunocompromised patients with PCP than HIV⫹ patients determine whether the patient should be considered to have PCP
with PCP. These data confirm that expression of the disease occurs or to be colonized.
with a lower fungal burden in HIV⫺ immunocompromised pa-
tients than HIV⫹ patients. In our study, the CT values were defined REFERENCES
to obtain a sensitivity of 80% and a specificity of 80%, allowing the 1. Buchacz K, Baker RK, Palella FJ, Chmiel JS, Lichtenstein KA, Novak
RM, Wood KC, Brooks JT, HOPS Investigators. 2010. AIDS-defining
CT values to provide physicians with useful insight. Some studies opportunistic illnesses in US patients, 1994-2007: a cohort study. AIDS
determined various PCR cutoff values that may be used to dis- 24:1549 –1559. http://dx.doi.org/10.1097/QAD.0b013e32833a3967.
criminate PCP from colonization, but very few of them gave dif- 2. Sokulska M, Kicia M, Wesołowska M, Hendrich AB. 2015. Pneumocys-
ferent values according to HIV infection status (62, 63). The CT tis jirovecii—from a commensal to pathogen: clinical and diagnostic re-
view. Parasitol Res 114:3577–3585. http://dx.doi.org/10.1007/s00436-015
values given in our study could be used by other laboratories,
-4678-6.
provided that extraction and qPCR are performed as described in 3. Thomas CF, Limper AH. 2004. Pneumocystis pneumonia. N Engl J Med
our Materials and Methods. 350:2487–2498. http://dx.doi.org/10.1056/NEJMra032588.
Since the use of the HAART and prophylaxis, the number of 4. Pulvirenti J, Herrera P, Venkataraman P, Ahmed N. 2003. Pneumo-
HIV⫹ patients affected by PCP has decreased, but for some pa- cystis carinii pneumonia in HIV-infected patients in the HAART
era. AIDS Patient Care STDs 17:261–265. http://dx.doi.org/10.1089
tients, P. jirovecii pneumonia still precedes a diagnosis of HIV /108729103322108139.
infection and is still a cause of mortality (1, 2, 64). PCP occurs in 5. Carmona EM, Limper AH. 2011. Update on the diagnosis and treatment
patients with a lack of access to medical care or unaware of their of Pneumocystis pneumonia. Ther Adv Respir Dis 5:41–59. http://dx.doi
HIV infection status, as was the case for 12 of 14 HIV⫹ patients in .org/10.1177/1753465810380102.
the present study. In our study, the proportion of HIV⫺ patients 6. Tasaka S, Tokuda H. 2014. Pneumocystis jirovecii pneumonia in non-
HIV-infected patients in the era of novel immunosuppressive therapies.
with PCP was greater than the proportion of HIV⫹ patients with Kansensho៮ gaku Zasshi 88:26 –39.
PCP. We also found a higher rate of mortality among HIV⫺ pa- 7. Morris A, Norris KA. 2012. Colonization by Pneumocystis jirovecii and
tients than HIV⫹ patients (10, 39, 49, 65). its role in disease. Clin Microbiol Rev 25:297–317. http://dx.doi.org/10
P. jirovecii pneumonia represents a serious public health care .1128/CMR.00013-12.
8. Phair J, Muñoz A, Detels R, Kaslow R, Rinaldo C, Saah A. 1990. The risk
problem not only because of the severity of the disease but also of Pneumocystis carinii pneumonia among men infected with human
because P. jirovecii pneumonia is often a concomitant pathology immunodeficiency virus type 1. Multicenter AIDS Cohort Study Group.
aggravating a life-threatening disease in HIV⫺ patients (2). N Engl J Med 322:161–165.

June 2016 Volume 54 Number 6 Journal of Clinical Microbiology jcm.asm.org 1493

Fauchier et al.

9. Helweg-Larsen J, Jensen JS, Dohn B, Benfield TL, Lundgren B. 2002. International HIV-Associated Opportunistic Pneumonias Study. 2012.
Detection of Pneumocystis DNA in samples from patients suspected of Low prevalence of Pneumocystis pneumonia (PCP) but high prevalence
bacterial pneumonia—a case-control study. BMC Infect Dis 2:28. http: of pneumocystis dihydropteroate synthase (dhps) gene mutations in HIV-
//dx.doi.org/10.1186/1471-2334-2-28. infected persons in Uganda. PLoS One 7:e49991. http://dx.doi.org/10
10. Enomoto T, Azuma A, Kohno A, Kaneko K, Saito H, Kametaka M, .1371/journal.pone.0049991.
Usuki J, Gemma A, Kudoh S, Nakamura S. 2010. Differences in the 27. Hartung TK, Chimbayo D, van Oosterhout JJG, Chikaonda T, van
clinical characteristics of Pneumocystis jirovecii pneumonia in immuno- Doornum GJJ, Claas ECJ, Melchers WJG, Molyneux ME, Zijlstra EE.
compromized patients with and without HIV infection. Respirology 15: 2011. Etiology of suspected pneumonia in adults admitted to a high-
126 –131. http://dx.doi.org/10.1111/j.1440-1843.2009.01660.x. dependency unit in Blantyre, Malawi. Am J Trop Med Hyg 85:105–112.
11. Spencer L, Ukwu M, Alexander T, Valadez K, Liu L, Frederick T, http://dx.doi.org/10.4269/ajtmh.2011.10-0640.
Kovacs A, Morris A. 2008. Epidemiology of Pneumocystis colonization in 28. Wong ML, Back P, Candy G, Nelson G, Murray J. 2006. Pneumocystis
families. Clin Infect Dis 46:1237–1240. http://dx.doi.org/10.1086/533449. jirovecii pneumonia in African miners at autopsy. Int J Tuberc Lung Dis
12. Maskell NA, Waine DJ, Lindley A, Pepperell JCT, Wakefield AE, Miller 10:756 –760.
RF, Davies RJO. 2003. Asymptomatic carriage of Pneumocystis jiroveci in 29. Vargas SL, Ponce CA, Gálvez P, Ibarra C, Haas EA, Chadwick AE,
subjects undergoing bronchoscopy: a prospective study. Thorax 58:594 – Krous HF. 2007. Pneumocystis is not a direct cause of sudden infant death
597. http://dx.doi.org/10.1136/thorax.58.7.594. syndrome. Pediatr Infect Dis J 26:81– 83. http://dx.doi.org/10.1097/01.inf
13. Vargas SL, Hughes WT, Santolaya ME, Ulloa AV, Ponce CA, Cabrera .0000247071.40739.fd.
CE, Cumsille F, Gigliotti F. 2001. Search for primary infection by Pneu- 30. Thomas CF, Limper AH. 2007. Current insights into the biology and
mocystis carinii in a cohort of normal, healthy infants. Clin Infect Dis pathogenesis of Pneumocystis pneumonia. Nat Rev Microbiol 5:298 –308.
32:855– 861. http://dx.doi.org/10.1086/319340. http://dx.doi.org/10.1038/nrmicro1621.
14. Chintu C, Mudenda V, Lucas S, Nunn A, Lishimpi K, Maswahu D, 31. Beck JM, Cushion MT. 2009. Pneumocystis workshop: 10th anniver-
Kasolo F, Mwaba P, Bhat G, Terunuma H, Zumla A, UNZA-UCLMS sary summary. Eukaryot Cell 8:446 – 460. http://dx.doi.org/10.1128
Project Paediatric Post-mortem Study Group. 2002. Lung diseases at /EC.00309-08.
necropsy in African children dying from respiratory illnesses: a descriptive 32. Stringer JR, Beard CB, Miller RF, Wakefield AE. 2002. A new name
necropsy study. Lancet 360:985–990. http://dx.doi.org/10.1016/S0140 (Pneumocystis jiroveci) for Pneumocystis from humans. Emerg Infect Dis
-6736(02)11082-8. 8:891– 896. http://dx.doi.org/10.3201/eid0809.020096.
15. Ponce CA, Gallo M, Bustamante R, Vargas SL. 2010. Pneumocystis 33. Chalvardjian AM, Grawe LA. 1963. A new procedure for the identifica-
colonization is highly prevalent in the autopsied lungs of the general pop- tion of Pneumocystis carinii cysts in tissue sections and smears. J Clin
ulation. Clin Infect Dis 50:347–353. http://dx.doi.org/10.1086/649868. Pathol 16:383–384. http://dx.doi.org/10.1136/jcp.16.4.383.
16. Medrano FJ, Montes-Cano M, Conde M, de la Horra C, Respaldiza N, 34. Dournon E, Rajagopalan P, Albert F, Camus F, Marche C, McLauchlin
Gasch A, Perez-Lozano MJ, Varela JM, Calderon EJ. 2005. Pneumocys- J, Samuel D, Taylor AG. 1989. Diagnosis of Pneumocystis carinii pneu-
tis jirovecii in general population. Emerg Infect Dis 11:245–250. http://dx monia by non-experts. Lancet i:107–108.
35. Kovacs JA, Gill V, Swan JC, Ognibene F, Shelhamer J, Parrillo JE,
.doi.org/10.3201/eid1102.040487.
Masur H. 1986. Prospective evaluation of a monoclonal antibody in di-
17. Vargas SL, Ponce CA, Gallo M, Pérez F, Astorga J-F, Bustamante R,
agnosis of Pneumocystis carinii pneumonia. Lancet ii:1–3.
Chabé M, Durand-Joly I, Iturra P, Miller RF, Aliouat EM, Dei-Cas E.
36. Fujisawa T, Suda T, Matsuda H, Inui N, Nakamura Y, Sato J, Toyo-
2013. Near-universal prevalence of Pneumocystis and associated increase
shima M, Nakano Y, Yasuda K, Gemma H, Hayakawa H, Chida K.
in mucus in the lungs of infants with sudden unexpected death. Clin Infect
2009. Real-time PCR is more specific than conventional PCR for induced
Dis 56:171–179. http://dx.doi.org/10.1093/cid/cis870.
sputum diagnosis of Pneumocystis pneumonia in immunocompromised
18. Vargas SL, Ponce CA, Hughes WT, Wakefield AE, Weitz JC, Donoso S,
patients without HIV infection. Respirology 14:203–209. http://dx.doi
Ulloa AV, Madrid P, Gould S, Latorre JJ, Avila R, Benveniste S, Gallo
.org/10.1111/j.1440-1843.2008.01457.x.
M, Belletti J, Lopez R. 1999. Association of primary Pneumocystis carinii
37. Botterel F, Cabaret O, Foulet F, Cordonnier C, Costa J-M, Bretagne S.
infection and sudden infant death syndrome. Clin Infect Dis 29:1489 – 2012. Clinical significance of quantifying Pneumocystis jirovecii DNA by
1493. http://dx.doi.org/10.1086/313521. using real-time PCR in bronchoalveolar lavage fluid from immunocom-
19. Morgan DJ, Vargas SL, Reyes-Mugica M, Walterspiel JN, Carver W, promised patients. J Clin Microbiol 50:227–231. http://dx.doi.org/10
Gigliotti F. 2001. Identification of Pneumocystis carinii in the lungs of .1128/JCM.06036-11.
infants dying of sudden infant death syndrome. Pediatr Infect Dis J 20: 38. Alanio A, Desoubeaux G, Sarfati C, Hamane S, Bergeron A, Azoulay E,
306 –309. http://dx.doi.org/10.1097/00006454-200103000-00017. Molina JM, Derouin F, Menotti J. 2011. Real-time PCR assay-based
20. Nevez G, Magois E, Duwat H, Gouilleux V, Jounieaux V, Totet A. 2006. strategy for differentiation between active Pneumocystis jirovecii pneu-
Apparent absence of Pneumocystis jirovecii in healthy subjects. Clin Infect monia and colonization in immunocompromised patients. Clin Micro-
Dis 42:e99 – e101. http://dx.doi.org/10.1086/503908. biol Infect 17:1531–1537. http://dx.doi.org/10.1111/j.1469-0691.2010
21. Totet A, Gresillon K, Matos O, Lundgren B, Larsen HH, Roux P, .03400.x.
Hauser P, Raccurt C, Nevez G. 2006. Pneumocystis jirovecii genotypes in 39. Robberts FJL, Liebowitz LD, Chalkley LJ. 2007. Polymerase chain reac-
European patients with chronic obstructive pulmonary diseases. J Eu- tion detection of Pneumocystis jiroveci: evaluation of 9 assays. Diagn Mi-
karyot Microbiol 53(Suppl 1):S102–S103. http://dx.doi.org/10.1111/j crobiol Infect Dis 58:385–392. http://dx.doi.org/10.1016/j.diagmicrobio
.1550-7408.2006.00189.x. .2007.02.014.
22. Morris A, Wei K, Afshar K, Huang L. 2008. Epidemiology and clinical 40. Sing A, Trebesius K, Roggenkamp A, Rüssmann H, Tybus K, Pfaff F,
significance of pneumocystis colonization. J Infect Dis 197:10 –17. http: Bogner JR, Emminger C, Heesemann J. 2000. Evaluation of diagnostic
//dx.doi.org/10.1086/523814. value and epidemiological implications of PCR for Pneumocystis carinii
23. Le Gal S, Pougnet L, Damiani C, Fréalle E, Guéguen P, Virmaux M, in different immunosuppressed and immunocompetent patient groups. J
Ansart S, Jaffuel S, Couturaud F, Delluc A, Tonnelier J-M, Castellant P, Clin Microbiol 38:1461–1467.
Le Meur Y, Le Floch G, Totet A, Menotti J, Nevez G. 2015. Pneumo- 41. Alvarez-Martínez MJ, Miró JM, Valls ME, Moreno A, Rivas PV, Solé M,
cystis jirovecii in the air surrounding patients with Pneumocystis pulmo- Benito N, Domingo P, Muñoz C, Rivera E, Zar HJ, Wissmann G, Diehl
nary colonization. Diagn Microbiol Infect Dis 82:137–142. http://dx.doi ARS, Prolla JC, de Anta MTJ, Gatell JM, Wilson PE, Meshnick SR,
.org/10.1016/j.diagmicrobio.2015.01.004. Spanish PCP Working Group. 2006. Sensitivity and specificity of nested
24. Calderón EJ, Friaza V, Dapena FJ, de la Horra C. 2010. Pneumocystis and real-time PCR for the detection of Pneumocystis jiroveci in clinical
jirovecii and cystic fibrosis. Med Mycol 48(Suppl 1):S17–S21. http://dx specimens. Diagn Microbiol Infect Dis 56:153–160. http://dx.doi.org/10
.doi.org/10.3109/13693786.2010.505205. .1016/j.diagmicrobio.2006.03.010.
25. Norris KA, Morris A. 2011. Pneumocystis infection and the pathogenesis 42. Döskaya M, Caner A, Degirmenci A, Wengenack NL, Yolasigmaz A,
of chronic obstructive pulmonary disease. Immunol Res 50:175–180. http: Turgay N, Ozensoy Töz S, Gürüz Y. 2011. Degree and frequency of
//dx.doi.org/10.1007/s12026-011-8218-x. inhibition in a routine real-time PCR detecting Pneumocystis jirovecii for
26. Taylor SM, Meshnick SR, Worodria W, Andama A, Cattamanchi A, the diagnosis of Pneumocystis pneumonia in Turkey. J Med Microbiol
Davis JL, Yoo SD, Byanyima P, Kaswabuli S, Goodman CD, Huang L, 60:937–944. http://dx.doi.org/10.1099/jmm.0.030775-0.

1494 jcm.asm.org Journal of Clinical Microbiology June 2016 Volume 54 Number 6

 Pneumocystis jirovecii Detection by qPCR

43. Arcenas RC, Uhl JR, Buckwalter SP, Limper AH, Crino D, Roberts GD, 58. Flahault A, Cadilhac M, Thomas G. 2005. Sample size calculation should
Wengenack NL. 2006. A real-time polymerase chain reaction assay for be performed for design accuracy in diagnostic test studies. J Clin Epide-
detection of Pneumocystis from bronchoalveolar lavage fluid. Diagn Mi- miol 58:859 – 862. http://dx.doi.org/10.1016/j.jclinepi.2004.12.009.
crobiol Infect Dis 54:169 –175. http://dx.doi.org/10.1016/j.diagmicrobio 59. Fillaux J, Malvy S, Alvarez M, Fabre R, Cassaing S, Marchou B, Linas
.2005.08.006. M-D, Berry A. 2008. Accuracy of a routine real-time PCR assay for the
44. Flori P, Bellete B, Durand F, Raberin H, Cazorla C, Hafid J, Lucht F, diagnosis of Pneumocystis jirovecii pneumonia. J Microbiol Methods 75:
Sung RTM. 2004. Comparison between real-time PCR, conventional 258 –261. http://dx.doi.org/10.1016/j.mimet.2008.06.009.
PCR and different staining techniques for diagnosing Pneumocystis ji- 60. Fily F, Lachkar S, Thiberville L, Favennec L, Caron F. 2011. Pneumo-
roveci pneumonia from bronchoalveolar lavage specimens. J Med Micro- cystis jirovecii colonization and infection among non HIV-infected pa-
biol 53:603– 607. http://dx.doi.org/10.1099/jmm.0.45528-0. tients. Med Mal Infect 41:526 –531. http://dx.doi.org/10.1016/j.medmal
45. Goto N, Oka S. 2011. Pneumocystis jirovecii pneumonia in kidney trans- .2011.07.006.
plantation. Transpl Infect Dis 13:551–558. http://dx.doi.org/10.1111/j 61. Wakefield AE, Pixley FJ, Banerji S, Sinclair K, Miller RF, Moxon ER,
.1399-3062.2011.00691.x. Hopkin JM. 1990. Detection of Pneumocystis carinii with DNA amplifi-
46. Ognibene FP, Shelhamer JH, Hoffman GS, Kerr GS, Reda D, Fauci AS, cation. Lancet 336:451– 453. http://dx.doi.org/10.1016/0140-6736(90)
Leavitt RY. 1995. Pneumocystis carinii pneumonia: a major complication 92008-6.
of immunosuppressive therapy in patients with Wegener’s granulomato- 62. Montesinos I, Brancart F, Schepers K, Jacobs F, Denis O, Delforge M-L.
sis. Am J Respir Crit Care Med 151:795–799. 2015. Comparison of 2 real-time PCR assays for diagnosis of Pneumocys-
47. Jung AC, Paauw DS. 1994. Management of adverse reactions to trim- tis jirovecii pneumonia in human immunodeficiency virus (HIV) and
ethoprim-sulfamethoxazole in human immunodeficiency virus-infected non-HIV immunocompromised patients. Diagn Microbiol Infect Dis 82:
patients. Arch Intern Med 154:2402–2406. http://dx.doi.org/10.1001 143–147. http://dx.doi.org/10.1016/j.diagmicrobio.2015.03.006.
/archinte.1994.00420210028004. 63. Robert-Gangneux F, Belaz S, Revest M, Tattevin P, Jouneau S, Decaux
48. De Castro N, Xu F, Porcher R, Pavie J, Molina J-M, Peraldi M-N. 2010. O, Chevrier S, Le Tulzo Y, Gangneux J-P. 2014. Diagnosis of Pneumo-
Pneumocystis jirovecii pneumonia in renal transplant recipients occur- cystis jirovecii pneumonia in immunocompromised patients by real-time
ring after discontinuation of prophylaxis: a casecontrol study. Clin Micro- PCR: a 4-year prospective study. J Clin Microbiol 52:3370 –3376. http://dx
biol Infect 16:1375–1377. http://dx.doi.org/10.1111/j.1469-0691.2010 .doi.org/10.1128/JCM.01480-14.
.03143.x. 64. López-Sánchez C, Falcó V, Burgos J, Navarro J, Martín MT, Curran A,
49. Sepkowitz KA, Brown AE, Armstrong D. 1995. Pneumocystis carinii
Miguel L, Ocaña I, Ribera E, Crespo M, Almirante B. 2015. Epidemi-
pneumonia without acquired immunodeficiency syndrome. More pa-
ology and long-term survival in HIV-infected patients with Pneumocystis
tients, same risk. Arch Intern Med 155:1125–1128.
jirovecii pneumonia in the HAART era: experience in a university hospital
50. Godeau B, Coutant-Perronne V, Le Thi Huong D, Guillevin L, Maga-
and review of the literature. Medicine (Baltimore) 94:e681. http://dx.doi
dur G, De Bandt M, Dellion S, Rossert J, Rostoker G, Piette JC. 1994.
.org/10.1097/MD.0000000000000681.
Pneumocystis carinii pneumonia in the course of connective tissue dis-
65. Torres HA, Chemaly RF, Storey R, Aguilera EA, Nogueras GM, Safdar
ease: report of 34 cases. J Rheumatol 21:246 –251.
51. Yale SH, Limper AH. 1996. Pneumocystis carinii pneumonia in patients A, Rolston KVI, Raad II, Kontoyiannis DP. 2006. Influence of type of
without acquired immunodeficiency syndrome: associated illness and cancer and hematopoietic stem cell transplantation on clinical presenta-
prior corticosteroid therapy. Mayo Clin Proc 71:5–13. http://dx.doi.org tion of Pneumocystis jiroveci pneumonia in cancer patients. Eur J Clin
/10.4065/71.1.5. Microbiol Infect Dis 25:382–388. http://dx.doi.org/10.1007/s10096-006
52. Helweg-Larsen J, Benfield T, Atzori C, Miller RF. 2009. Clinical efficacy -0149-4.
of first- and second-line treatments for HIV-associated Pneumocystis ji- 66. Walzer PD, Evans HER, Copas AJ, Edwards SG, Grant AD, Miller RF.
rovecii pneumonia: a tri-centre cohort study. J Antimicrob Chemother 2008. Early predictors of mortality from Pneumocystis jirovecii pneumo-
64:1282–1290. http://dx.doi.org/10.1093/jac/dkp372. nia in HIV-infected patients: 1985-2006. Clin Infect Dis 46:625– 633. http:
53. Thomas M, Rupali P, Woodhouse A, Ellis-Pegler R. 2009. Good out- //dx.doi.org/10.1086/526778.
come with trimethoprim 10 mg/kg/day-sulfamethoxazole 50 mg/kg/day 67. Arend SM, Kroon FP, van’t Wout JW. 1995. Pneumocystis carinii
for Pneumocystis jirovecii pneumonia in HIV infected patients. Scand J pneumonia in patients without AIDS, 1980 through 1993. An analysis of
Infect Dis 41:862– 868. http://dx.doi.org/10.3109/00365540903214256. 78 cases. Arch Intern Med 155:2436 –2441.
54. Meliani L, Develoux M, Marteau-Miltgen M, Magne D, Barbu V, Poirot 68. Sepkowitz KA. 2002. Antibiotic prophylaxis in patients receiving hema-
J-L, Roux P. 2003. Real time quantitative PCR assay for Pneumocystis topoietic stem cell transplant. Bone Marrow Transplant 29:367–371. http:
jirovecii detection. J Eukaryot Microbiol 50(Suppl):651. http://dx.doi.org //dx.doi.org/10.1038/sj.bmt.1703366.
/10.1111/j.1550-7408.2003.tb00670.x. 69. Nevez G, Chabé M, Rabodonirina M, Virmaux M, Dei-Cas E, Hauser
55. Mary C, Faraut F, Lascombe L, Dumon H. 2004. Quantification of PM, Totet A. 2008. Nosocomial Pneumocystis jirovecii infections. Para-
Leishmania infantum DNA by a real-time PCR assay with high sensitivity. site 15:359 –365. http://dx.doi.org/10.1051/parasite/2008153359.
J Clin Microbiol 42:5249 –5255. http://dx.doi.org/10.1128/JCM.42.11 70. Hauser PM, Nahimana A, Taffe P, Weber R, Francioli P, Bille J,
.5249-5255.2004. Rabodonirina M. 2010. Interhuman transmission as a potential key pa-
56. Witebsky FG, Andrews JW, Gill VJ, MacLowry JD. 1988. Modified rameter for geographical variation in the prevalence of Pneumocystis ji-
toluidine blue O stain for Pneumocystis carinii: further evaluation of some rovecii dihydropteroate synthase mutations. Clin Infect Dis 51:e28 – e33.
technical factors. J Clin Microbiol 26:774 –775. http://dx.doi.org/10.1086/655145.
57. Hughes WT, Feldman S, Aur RJ, Verzosa MS, Hustu HO, Simone JV. 71. Mühlethaler K, Bögli-Stuber K, Wasmer S, von Garnier C, Dumont P,
1975. Intensity of immunosuppressive therapy and the incidence of Pneu- Rauch A, Mühlemann K, Garzoni C. 2012. Quantitative PCR to diagnose
mocystis carinii pneumonitis. Cancer 36:2004 –2009. http://dx.doi.org/10 Pneumocystis pneumonia in immunocompromised non-HIV patients.
.1002/cncr.2820360912. Eur Respir J 39:971–978. http://dx.doi.org/10.1183/09031936.00095811.

June 2016 Volume 54 Number 6 Journal of Clinical Microbiology jcm.asm.org 1495