- Email: [email protected]
Early diagnosis and treatment are crucial for the survival of Pneumocystis pneumonia patients without human immunodeficiency virus infection
J Infect Chemother (2012) 18:898–905 DOI 10.1007/s10156-012-0441-4
ORIGINAL ARTICLE
Early diagnosis and treatment are crucial for the survival of Pneumocystis pneumonia patients without human immunodeficiency virus infection Nobuhiro Asai • Shinji Motojima • Yoshihiro Ohkuni • Ryo Matsunuma • Kei Nakashima • Takuya Iwasaki • Tamao Nakashita • Yoshihito Otsuka • Norihiro Kaneko
Received: 5 April 2012 / Accepted: 20 May 2012 / Published online: 13 June 2012 Ó Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases 2012
Abstract The mortality of Pneumocystis pneumonia (PCP) patients without human immunodeficiency virus (HIV) infection ranges from 0 to 70 %, whereas that of HIV-infected PCP patients ranges from 10 to 20 %. The reasons for these differences are not known. We retrospectively analyzed factors contributing to the survival of 23 patients with PCP and without HIV infection, in whom PCP developed as community-acquired pneumonia (CAP). The interval from admission to the start of PCP-specific treatment was significantly shorter for survivors (2.71 ± 3.64 days; n = 14) than for non-survivors (8.67 ± 5.5 days; n = 9; p = 0.003). Moreover, although the severity scores/classes assessed by A-DROP, CURB-65, and PSI were no different on admission, scores/classes at the start of PCP-specific treatment were significantly higher for non-survivors. Overall mortality was 39 %, but mortality was approximately 70–100 % for patients classified as severe grade by A-DROP, CURB-65, or PSI scores/classes at the time when PCP-specific treatment was started, which was far higher than expected for these guidelines. In conclusion, early diagnosis and treatment within 3 days are crucial for the survival of PCP patients without HIV infection. We emphasize the limitations of application of guidelines for CAP to patients with PCP.
N. Asai (&) S. Motojima Y. Ohkuni R. Matsunuma K. Nakashima T. Iwasaki T. Nakashita N. Kaneko The Department of Pulmonology, Kameda Medical Center, 929 Higashi-cho, Kamogawa, Chiba, Japan e-mail: [email protected] Y. Otsuka Department of Laboratory Medicine, Kameda Medical Center, Chiba, Japan
123
Keywords Community-acquired pneumonia Human immunodeficiency virus Pneumocystis pneumonia
Introduction Pneumocystis pneumonia (PCP) not related to human immunodeficiency virus (HIV) can develop in patients with malignancies, rheumatic diseases, and other diseases whose immune systems have been compromised by immunosuppressive drugs [1, 2]. The mortality of patients with PCP but without HIV infection is diverse and ranges from 0 to 70 % [1, 3–9], compared with mortality of HIV-infected PCP patients, which ranges from 10 to 20 % [1, 3, 10]. The higher mortality among non-HIV patients has been attributed to severe lung inflammation [1, 3, 5], although the exact etiology of these large mortality differences has not yet been determined. In this retrospective study, we attempted to identify clinical factors that contributed to the survival of non-HIV patients with PCP infection. In addition, because we routinely apply management guidelines for communityacquired pneumonia (CAP) to PCP patients whom we encounter in the outpatient clinic or the emergency room, we also tested whether the application of guidelines for CAP to patients with PCP was suitable; that is, whether the guidelines correctly evaluated the severity of PCP and correctly predicted mortality.
Patients and methods We retrospectively reviewed all PCP cases that developed in the setting of CAP and required admission for treatment in the Kameda Medical Center from 2001 to 2010. A total
J Infect Chemother (2012) 18:898–905
of 23 PCP patients had negative HIV tests. PCP cases that developed as hospital-acquired pneumonia (HAP) were excluded. PCP was diagnosed on the basis of the following: 1.
2. 3.
microbiological analysis of respiratory samples, for example induced sputum or bronchoalveolar lavage (BAL) fluid, by use of polymerase chain reaction and conventional staining with Grocott methenamine silver stain and Diff-QuickTM; radiographic findings on chest X-ray and chest computed tomography (CT) compatible with PCP; and compatible clinical findings, including dyspnea, cough, and fever [1, 3, 7, 11].
All PCP patients were enrolled with both positive PCPPCR and any microscopic findings from the induced sputum or BAL specimen. Radiologic findings and any clinical symptoms were confirmed. Because all PCP cases in this study developed as CAP, CAP severity was assessed on the first visit to the outpatient clinic or the emergency room by use of the A-DROP system developed by the Japanese Respiratory Society (JRS) [12], CURB-65 scores developed by the British Thoracic Society (BTS) [13], and the pneumonia severity index (PSI) developed by the Infectious Disease Society of America (IDSA) [14]. Severity was determined again on the day when PCP-specific treatment was started, because specific treatment for PCP was not started on the first day of admission for most patients, in accordance with recommendations for empirical treatment of CAP [12–14]. It remains a contentious issue whether or not severity determinations by these guidelines at the time of PCPspecific treatment are suitable, because, for most patients, several days will pass in the hospital before PCP-specific treatment is started. However, we used these guidelines at the start of PCP-specific treatment, for two reasons: 1. 2.
we wanted to compare PCP severity by using the same standards; and all of our PCP cases developed in the setting of CAP and no additional diseases developed after admission.
For PCP-specific treatment, all patients received oral or parenteral trimethoprim/sulfamethoxazole (TMP/SMX) and adjunct glucocorticoid therapy, although the day when PCP-specific treatment started after admission differed widely depending on the attending physicians. Statistical analysis Group means were compared by use of unpaired or paired t tests or the Mann–Whitney U test. Contingency tables were evaluated by use of Fisher’s exact probability test. p values \0.05 were regarded as significant.
899
Institutional Review Board approval Each patient has provided permission to publish these features of his/her case, and their identity has been protected. The IRB at Kameda Medical Center approved this retrospective study.
Results The characteristics of the 23 PCP patients are shown in Table 1. Twelve patients (52.2 %) had rheumatic or autoimmune diseases, which were the most common underlying diseases, followed by malignancy (n = 10; 43.5 %). Twenty-one patients (91.3 %) were receiving glucocorticoids or immunosuppressants for their underlying diseases. Prophylaxis consisting of TMP/SMX had been given to 1 patient (4.3 %). Mean CAP severity on admission, as assessed by A-DROP, CURB-65, and PSI scores/classes, is also shown. For all patients, PCP was diagnosed on the basis of any respiratory symptoms and radiology findings. Positive PCP-PCR results were obtained for all induced sputum (n = 18, 78.3 %) or BAL fluid (n = 5, 21.7 %) specimens. Either Diff-QuikÒ (n = 7, 30.4 %) or Grocott staining (n = 18, 78.3 %) was confirmed for all patients. Table 2 shows the laboratory results at the start of PCPspecific treatment of the patients according to their prognostic outcomes: i.e., survivors and non-survivors. Nine patients (39.1 %) died of PCP. There were no significant differences in laboratory results between the survivor and non-survivor groups except for PaO2/FiO2. On admission, the severity of pneumonia assessed by A-DROP, CURB-65, and PSI scores/classes was not significantly different between the survivor and non-survivor groups. However, at the start of PCP-specific treatment, the severity of pneumonia assessed by A-DROP, CURB-65, and PSI scores/ classes was significantly different between the survivor and non-survivor groups. There were also significant differences in the interval from admission to diagnosis of PCP and that from admission to initiation of PCP-specific treatment. In fact, most of the patients in the survivor group were started on PCP-specific treatment before diagnosis of PCP. Table 3 shows the changes in CAP severity between the two phases: on admission and at the start of PCP-specific treatment. In the survivor group, severity had not changed significantly at the start of PCP-specific treatment compared with that on admission. However, in the non-survivor group, severity had changed significantly at the start of PCP-specific treatment compared with that on admission. Table 4 shows the distributions of patients and mortality in each risk class of prediction rules at the start of PCP-specific treatment. Note the very high patient mortality
123
900
J Infect Chemother (2012) 18:898–905
Table 1 Characteristics of PCP patients without HIV infection (n = 23) Characteristic
Patients No.
%
71.3 (9.07),
57–88
Male
13
56.5
Female
10
43.5
10
43.5
6 5
26.1 21.7
Age, mean (SD), range Sex
Underlying diseases Malignancies Solid cancers Lung cancer Brain tumor
1
4.3
4
17.4
Acute myeloblastic leukemia
1
4.3
Adult T cell leukemia
1
4.3
Multiple myeloma
1
4.3
Malignant lymphoma
1
4.3
Hematologic malignancies
Rheumatic and autoimmune diseases
12
52.2
Rheumatoid arthritis
10
43.5
Dermatomyositis
1
4.3
Sarcoidosis
1
4.3
Diabetes mellitus
4
17.4
Chronic pulmonary diseases
6
26.1
Heart diseases
6
26.1
Cerebrovascular diseases Renal diseases
2 3
8.7 13.0
Liver disease
2
8.7
Long-term glucocorticoids alone
6
26.1
Immunosuppressants alone/chemotherapeutic agents alone
5
21.7
Methotrexate
2
8.7
u-RIST
1
4.3
Vinorerbin
1
4.3
TS-1
1
4.3
Long-term glucocorticoids combined with chemotherapeutic/immunosuppressive agents
10
43.5
Glucocorticoid ? methotrexate
7
30.4
Glucocorticoid ? anti-TNF drugs
3
13.0
PCP prophylaxis
1
4.3
Daily glucocorticoids dose at the time of first visit, PSL mg; mean (range) Glucocorticoids alone
20.7 (4–30)
Combination glucocorticoids and any agents
14.1 (2.5–18)
Duration of glucocorticoids immunosuppressive agents at the time of first visit, month; mean (range) Glucocorticoids alone
27.8 (1–128)
Combination glucocorticoids and any agents
23.0 (1–128)
Severity of CAP on admission evaluated by use of the following guidelines expressed as risk groups or net points. mean (SD) A-DROP
1.36 (1.14)
CURB-65
1.59 (1.14)
PSI
94.0 (25.7)
SD standard deviation, PCP Pneumocystis pneumonia, PSI pneumonia severity index, TNF tumor necrosis factor, TS-1 tegafur, gimeracil and oteracil potassium, u-RIST reduced intensity, allogeneic stem cell transplantation from unrelated donors
123
J Infect Chemother (2012) 18:898–905
901
Table 2 Patient characteristics on admission, by prognostic outcome Item
Survivors (n = 14)
Non-survivors (n = 9)
p value
Age
69.2 (8.6)
74.4 (8.8)
0.172
BMI (kg/m2)
22.4 (3.35)
22.2 (1.85)
0.853
b-DG (pg/ml)
87.5 (156)
75.1 (63.5)
0.824
KL-6 (U/ml)
781 (452)
1474 (1469)
0.175
LDH (IU/l)
515 (278)
729 (352)
0.126
CRP (mg/dl)
8.00 (4.25)
11.3 (5.44)
0.135
Alb/BUN
0.209 (0.128)
0.13 (0.072)
0.106
Lymphocytes (/ll)
1134.8 (992)
783.3 (397)
0.356
Neutrophils (/ll)
7149.1 (2900)
7729.3 (2976)
0.67
WBC (/ll)
9200 (3495)
10272 (5404)
0.568
PO2/FiO2 Interval from admission to PCP diagnosis (days)
223.4 (63.4) 4.86 (3.77)
137.5 (77.1) 8.56 (3.7)
0.007 0.012
Interval from admission to start PCP-specific treatment (days)
2.71 (3.64)
8.67 (5.5)
0.003
Severity of CAP on admission evaluated by use of the following guidelines A-DROP
1.07 (1.27)
1.78 (0.67)
0.14
CURB-65
1.36 (1.39)
1.89 (0.33)
0.28
PSI
87.1 (27.1)
105.6 (17.5)
0.09
Severity of CAP on admission evaluated by use of the following guidelines, number (%) A-DROP 0
6 (42.9)
0 (0)
1
4 (28.6)
3 (33.3)
2
2 (14.3)
5 (55.6)
3
1 (7.1)
1 (11.1)
4/5
1 (7.1)
0 (0)
0
5 (35.7)
0 (0)
1 2
3 (21.4) 4 (28.6)
1 (11.1) 8 (88.9)
3
0 (0)
0 (0)
4/5
2 (14.3)
0 (0)
I
0 (0)
0 (0)
II
4 (28.6)
0 (0)
III
3 (21.4)
1 (11.1)
IV
6 (42.9)
7 (77.8)
V
1 (7.1)
1 (11.1)
CURB-65
PSI
a
Severity of CAP at the start of PCP-specific treatment evaluated by use of the following guidelines A-DROP
1.21 (1.19)
CURB-65
1.43 (1.34)
PSI
88.6 (28.3)
Severity of CAP at the start of PCP-specific treatment evaluated by use of the following guidelines, number (%) A-DROP 0
4 (28.6)
0 (0)
1
6 (42.9)
0 (0)
2
2 (14.3)
1 (7.1)
3
1 (7.1)
5 (55.6)
4/5
1 (7.1)
3 (33.3)
123
902
J Infect Chemother (2012) 18:898–905
Table 2 continued Item
Survivors (n = 14)
Non-survivors (n = 9)
0
3 (21.4)
0 (0)
1
4 (28.6)
0 (0)
2
5 (35.7)
0 (0)
3
0 (0)
6 (66.7)
4/5
2 (14.3)
3 (33.3)
I
0 (0)
0 (0)
II
4 (28.6)
0 (0)
III
3 (21.4)
0 (0)
IV
6 (42.9)
3 (33.3)
V
1 (7.1)
6 (66.7)
p value
CURB-65
PSIa
Results are mean (SD) BMI body mass index, b-DG (1 ? 3)-b-D-glucan, KL-6 Krebs von den Lungen 6, LDH lactate dehydrogenase, CRP C-reactive protein, BUN blood urea nitrogen, WBC white blood cells, PCP Pneumocystis pneumonia, CAP community-acquired pneumonia, PSI pulmonary severity index a
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131
Table 3 Changes in CAP severity assessed by established guidelines and expressed as risk groups or net points On admission
At start of PCP-specific treatment
p value
Table 4 Distributions of patients and their mortality in each risk class of prediction rules at the start of PCP-specific treatment Risk groups
Number of patients (%)
Survivors (n = 14) A-DROP
1.07 (1.27)
1.21 (1.19)
0.956
CURB-65 PSI
1.36 (1.39) 87.3 (27.1)
1.43 (1.34) 88.6 (31.8)
0.891 0.893
1.78 (0.67)
3.33 (0.87)
\0.001
CURB-65
1.89 (0.33)
3.44 (0.77)
\0.001
105.6 (17.5)
143.3 (36.7)
0.009
PSI
Mortality (%)
A-DROP
Non-survivors (n = 9) A-DROP
Number of patients with respiratory failure (%)
0
6 (26.1)
3 (50)
0 (0)
1
7 (30.4)
6 (85.7)
3 (42.9)
2
7 (30.4)
7 (100)
5 (71.4)
3
2 (8.7)
2 (100)
1 (50)
4/5
1 (4.3)
1 (100)
0 (0)
CURB-65
Results are mean (SD)
0
5 (21.7)
2 (40)
0 (0)
PCP Pneumocystis pneumonia, CAP community-acquired pneumonia, PSI pulmonary severity index
1
4 (17.4)
3 (75)
1 (25)
2 3
12 (52.2) 0 (0)
12 (100) 0 (0)
2 (8.7)
2 (100)
0 (0)
0 (0)
in risk classes 2 by A-DROP and CURB-65, and class IV by PSI. Prognostic accuracy of each of the guidelines at the start of PCP-specific treatment is compared in Table 5. A-DROP scores of 3–5, CURB-65 scores of 3–5, and PSI risk class V are classified as severe in these guidelines. This also shows the survival predictions when PCP-specific treatment was started within 3, 5, or 7 days after admission. Note that the sensitivity, specificity, positive predictive value, and negative predictive value for survival were very high when PCPspecific treatment was started within 3 days after admission.
123
4/5
8 (66.7) 0 (0)
PSIa I
0 (0)
II
4 (17.4)
2 (50)
III
4 (17.4)
3 (75)
IV
13 (56.5)
12 (92.3)
V
2 (8.7)
2 (100)
1 (25) 7 (53.8) 1 (50)
PCP Pneumocystis pneumonia, PSI pulmonary severity index a
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131
J Infect Chemother (2012) 18:898–905
903
Table 5 Prognostic accuracy of A-DROP, CURB-65, and PSI at the start of PCP-specific treatment for patients classified in severe grades, and survival prediction when PCP-specific treatment was started within 3, 5, or 7 days after admission Sensitivity for mortality (%) A-DROP scores 3–5
11.1
CURB-65 scores 3–5 PSI risk classes V
0 11.1
Specificity for mortality (%) A-DROP scores 3–5
85.7
CURB-65 scores 3–5
85.7
PSI risk classes V
88.9
Positive predictive values for mortality (%) A-DROP scores 3–5
33.3
CURB-65 scores 3–5
0
PSI risk classes V
50.0
Negative predictive values for mortality (%) A-DROP scores 3–5
60.0
CURB-65 scores 3–5
57.1
PSI risk classes V
61.9
Sensitivity for survival (%) Within 3 days Within 5 days
92.9 72.2
Within 7 days
65.0
Specificity for survival (%) Within 3 days
88.9
Within 5 days
80.0
Within 7 days
66.7
Positive predictive values for survival (%) Within 3 days
92.9
Within 5 days
92.9
Within 7 days
92.9
Negative predictive values for survival (%) Within 3 days
88.9
Within 5 days
44.4
Within 7 days
22.2
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131 PCP Pneumocystis pneumonia, PSI pulmonary severity index
Discussion It is well known that PCP can develop not only in patients with HIV infection but also in some immunocompromized patients [1, 2]. Mortality of PCP patients without HIV infection is diverse and ranges from 0 to 70 % [1, 3–9]; for HIV-infected PCP patients mortality ranges from 10 to 20 % [1, 3, 10]. The higher mortality among non-HIV patients has been attributed to severe lung inflammation [1, 3, 5]. However, the reasons for the large differences in mortality have not been determined.
For PCP patients without HIV infection, we found that the interval from admission to the start of PCP-specific treatment was significantly shorter in the survivor group than in the non-survivor group. In addition, the interval from admission to diagnosis was significantly shorter in the survivor group than in the non-survivor group. Our report is the second to point out the relationship between mortality and treatment delay [9]. Overgaard et al. [9] reported that the mean intervals from admission to the start of treatment in non-survivor and survivor groups were 15 and 5 days, respectively. Unfortunately, standard deviations were not shown in their report. In a recent report by Komano et al. [7] survival was 100 % for patients with rheumatoid arthritis in whom PCP developed during treatment with infliximab, a tumor necrosis factor (TNF) blocker, and attributed the good outcome to prompt diagnosis and treatment. However, their paper did not mention how prompt this was. Although a relationship between mortality and treatment delay has not been shown generally, we suspect that treatment delay was responsible for the high mortality in some reports. Moreover, we have shown specifically that the positive predictive value for survival was [90 % when PCP-specific treatment was started within 3 days after admission. This is the first report to show a survival prediction with regard to the starting day of PCP-specific treatment. Although the mean risk classes determined by A-DROP, CURB-65, and PSI scores/classes were not statistically significantly different between the survivor and non-survivor groups on admission, they were significantly different at the start of PCP-specific treatment. This difference seems to be largely related to the treatment delay in the non-survivor group. There was approximately 6 days difference in the mean start of PCP-specific treatment between the survivor and non-survivor groups, and during those days the risk classes increased in the non-survivor group. It is reasonable to assume that mortality will be high among patients in high-risk groups; i.e., those with scores [3 on A-DROP and CURB-65 and class V on PSI. However, at the time when PCP-specific treatment was started, it was striking that mortality was approximately 70 to 100 %. Shindo et al. showed that among CAP patients, for whom most of the pathognomonic microorganisms were presumably bacteria, patient mortality in risk groups [3 when A-DROP and CURB-65 were applied was 23.3 and 21.0 %, respectively [15]. Usui et al. [16] also showed that patient mortality in the high-risk groups when A-DROP and CURB-65 were applied was 11.5 and 11.6 %, respectively. According to Fine et al. [17], the mortality for severe CAP (risk group V; PSI [130) was 29.2 % in a Medis group validation cohort. In the report by Usui et al. [16], the mortality of patients in risk group V
123
904
J Infect Chemother (2012) 18:898–905
was 28.3 %. Chen et al. [18] reported that mortality of CAP patients in risk classes 3–5 on CURB-65 and class IV/V on PSI was 16.6 and 12.5 %, respectively. An additional striking fact is that, although most patients in the nonsurvivor group had risk scores of 3 but not 5, their mortality was extremely high. Accordingly, we emphasize that mortality prediction in PCP is not correct when these management guidelines for CAP are applied, even when PCP develops in the setting of CAP. These guidelines definitely underestimate the severity of PCP as CAP. The important issue is why these guidelines cannot correctly estimate PCP severity. We suppose that the pathophysiology is different between PCP and typical bacterial pneumonia. In experimental models of bacterial pneumonia, mortality is high for animals with high bacterial numbers in the lungs [19]. However, the relationship is the opposite in PCP. According to a report by Limper et al., the number of Pneumocystis cysts in BAL fluid was significantly higher in PCP with AIDS than in PCP without AIDS. The number of cysts in BAL fluid was significantly positively correlated with PaO2 levels, and there was a significant negative correlation between PaO2 and the percentage of neutrophils in BAL fluid. Surviving patients had more cysts in BAL fluid than non-surviving patients, but the difference was marginal. Finally, non-surviving patients have significantly higher percentages of neutrophils in BAL fluid [3]. Tokuda et al. reported that plasma concentrations of b-D-glucan, a constituent of Pneumocystis that is used as a diagnostic marker [20], were 10 times higher in AIDSrelated PCP than in rheumatoid arthritis (RA)-related PCP. Notwithstanding, the mortality for AIDS-related PCP is 0 % and that for RA-related PCP is 14.3 % [21]. Thus, it seems likely that establishment of PCP requires host immune responses to Pneumocystis because Pneumocystis can exist without host invasion [22]. This is in contrast with typical bacterial pneumonia in which bacteria release toxins and enzymes that can damage host tissues. We wish to emphasize that there are limitations to applying the guidelines for CAP to PCP despite the fact that PCP can develop in the setting of CAP. These reasons are: 1.
2.
Pneumocystis is not considered to be a routine pathogen for CAP, as evidenced by the fact that the term ‘‘Pneumocystis’’ is found only twice in the ATS guidelines [14] and only once in the BTS guidelines [13], resulting in TMP/SMX not being included in the first empirical treatment regimens; treatment failure is considered after 3 days’ (72 h) continuation of the empirical treatment in the ATS [14] and JRS guidelines [12], although our results show that PCP-specific treatment should be started within 3 days after admission; and
123
3.
immune responses to Pneumocystis are not always beneficial—in some situations they can be harmful to patients with PCP, as discussed above, and the concept that the correct choice of antibiotics alone is enough to resolve PCP may not be sufficient.
A clue enabling early diagnosis of PCP is chest CT findings; that is, ground glass appearance with a panlobular pattern or so called crazy paving appearance [21, 23]. These findings are also found in viral pneumonias, mycoplasmal pneumonia, alveolar hemorrhage, methotrexate pneumonia, and others. However, in combination with the background characteristics of patients, the presence of PCP should be suspected. There have been reports of prognostic markers of shortterm mortality in AIDS-associated PCP [24–26]. However, by use of univariate analysis we unfortunately did not prove or disprove any risk factors of mortality in non-HIV PCP except for interval both from admission to PCP diagnosis and from admission to the start of PCP-specific treatment, as shown in Table 2. A limitation of our study is that it is a retrospective analysis of a very small population. Retrospective studies may be less reliable in terms of the data collected, particularly physical examination data. A prospective study on more cases is necessary. It has been reported that 20 % [27] of ordinary people and 30 % [28] of patients with chronic respiratory diseases have colonization of P. jirovecii. Although PCP was diagnosed on the basis of microbiological findings, radiology findings, and clinical symptoms, there could be the colonization of P. jirovecii, as mentioned above. In conclusion, we argue that when patients possibly have PCP, as judged from patients’ underlying diseases, concomitant treatments, and radiographic findings, TMP/AMX should be added to routine empirical treatment within 3 days at most, even before the diagnosis of PCP is established, with possible concurrent use of glucocorticoids. Acknowledgments We are grateful for the diligent and thorough critical reading of our manuscript by Mr John Wocher, Executive Vice President and Director, International Affairs/International Patient Services, Kameda Medical Center (Japan). Conflict of interest
The authors have no conflicting interests.
References 1. Thomas CF Jr, Limper AH. Pneumocystis pneumonia. N Engl J Med. 2004;350:2487–8. 2. Kovacs JA, Masur H. Evolving health effects of Pneumocystis: one hundred years of progress in diagnosis and treatment. JAMA. 2009;301:2578–85. 3. Limper AH, Offord KP, Smith TF, Martin WJ 2nd. Pneumocystis pneumonia. Differences in lung parasite number and
J Infect Chemother (2012) 18:898–905
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
inflammation in patients with and without AIDS. Am Rev Respir Dis. 1989;140:1204–9. Festic E, Gajic O, Limper AH, Aksamit TR. Acute respiratory failure due to Pneumocystis pneumonia in patients without human immunodeficiency virus infection: outcome and associated features. Chest. 2005;128:573–9. Yale SH, Limper AH. Pneumocystis carinii pneumonia in patients without acquired immunodeficiency syndrome: associated illness and prior corticosteroid therapy. Mayo Clin Proc. 1996;71:5–13. Ward MM, Donald F. Pneumocystis carinii pneumonia in patients with connective tissue diseases: the role of hospital experience in diagnosis and mortality. Arthritis Rheum. 1999; 42:780–9. Komano Y, Harigai M, Koike R, Sugiyama H, Ogawa J, Saito K, et al. Pneumocystis jiroveci pneumonia in patients with rheumatoid arthritis treated with infliximab: a retrospective review and case-control study of 21 patients. Arthritis Rheum. 2009;61: 305–12. Su YS, Lu JJ, Perng CL, Chang FY. Pneumocystis jirovecii pneumonia in patients with and without human immunodeficiency virus infection. J Microbiol Immunol Infect. 2008;41:478–82. Overgaard UM, Helweg-Larsen J. Pneumocystis jiroveci pneumonia (PCP) in HIV-1-negative patients: a retrospective study 2002–2004. Scand J Infect Dis. 2007;39:589–95. Dworkin MS, Hanson DL, Navin TR. Survival of patients with AIDS, after diagnosis of Pneumocystis carinii pneumonia, in the United States. J Infect Dis. 2001;183:1409–12. Harigai M, Koike R. Pneumocystis Pneumonia under Anti-Tumor Necrosis Factor Therapy (PAT) Study Group. Pneumocystis pneumonia associated with infliximab in Japan. N Engl J Med. 2007;357:1874–6. Miyashita N, Matsushima T. Japanese Respiratory Society. The JRS guidelines for the management of community-acquired pneumonia in adults: an update and new recommendations. Intern Med. 2006;45:419–28. British Thoracic Society Standards of Care Committee. BTS guidelines for the management of community acquired pneumonia in adults. Thorax. 2001;56(Suppl 4):1–64. American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med. 2001;163:1730–54. Shindo Y, Sato S, Maruyama E, Ohashi T, Ogawa M, Imaizumi K, et al. Comparison of severity scoring systems A-DROP and CURB-65 for community-acquired pneumonia. Respirology. 2008;13:731–5. Usui K, Tanaka Y, Noda H, Ishihara T. Comparison of three prediction rules for prognosis in community acquired pneumonia: Pneumonia Severity Index (PSI), CURB-65, and A-DROP. Nihon Kokyuki Gakkai Zasshi 2009; 47:781–5 (Japanese).
905 17. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, Singer DE, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med. 1997;336: 243–50. 18. Chen JH, Chang SS, Liu JJ, Chan RC, Wu JY, Wang WC, et al. Comparison of clinical characteristics and performance of pneumonia severity score and CURB-65 among younger adults, elderly and very old subjects. Thorax. 2010;65:971–7. 19. Wolff M, Joly-Guillou ML, Farinotti R, Carbon C. In vivo efficacies of combinations of beta-lactams, beta-lactamase inhibitors, and rifampin against Acinetobacter baumannii in a mouse pneumonia model. Antimicrob Agents Chemother. 1999;43: 1406–11. 20. Tasaka S, Hasegawa N, Kobayashi S, Yamada W, Nishimura T, Takeuchi T, et al. Serum indicators for the diagnosis of Pneumocystis pneumonia. Chest. 2007;131:1173. 21. Tokuda H, Sakai F, Yamada H, Johkoh T, Imamura A, Dohi M, et al. Clinical and radiological features of Pneumocystis pneumonia in patients with rheumatoid arthritis, in comparison with methotrexate pneumonitis and Pneumocystis pneumonia in acquired immunodeficiency syndrome: a multicenter study. Intern Med. 2008;47:915–23. 22. Maskell NA, Waine DJ, Lindley A, Pepperell JC, Wakefield AE, Miller RF, et al. Asymptomatic carriage of Pneumocystis jiroveci in subjects undergoing bronchoscopy: a prospective study. Thorax. 2003;58:594–7. 23. Kameda H, Tokuda H, Sakai F, Johkoh T, Mori S, Yoshida Y, et al. Clinical and radiological features of acute-onset diffuse interstitial lung diseases in patients with rheumatoid arthritis receiving treatment with biological agents: importance of Pneumocystis pneumonia in Japan revealed by a multicenter study. Intern Med. 2011;50:305–13. 24. Benfield TL, Helweg-Larsen J, Bang D, Junge J, Lundgren JD. Prognostic markers of short-term mortality in AIDS-associated Pneumocystis carinii pneumonia. Chest. 2001;119:844–51. 25. Azoulay E, Parrot A, Flahault A, Cesari D, Lecomte I, Roux P, et al. AIDS-related Pneumocystis carinii pneumonia in the era of adjunctive steroids: implication of BAL neutrophilia. Am J Respir Crit Care Med. 1999;160:493–9. 26. Fei MW, Kim EJ, Sant CA, Jarlsberg LG, Davis JL, Swartzman A, et al. Predicting mortality from HIV-associated Pneumocystis pneumonia at illness presentation: an observational cohort study. Thorax. 2009;64:1070–6. 27. Medrano FJ, Montes-Cano M, Conde MA, de la Horra C, Respaldiza N, Gasch A, et al. Pneumocystis jirovecii in general population. Emerg Infect Dis. 2005;11:245–50. 28. Vidal S, de la Horra C, Martin J, Montes-Cano MA, Rodriguez E, Respaldiza N, et al. Pneumocystis jirovecii colonization in patients with interstitial lung disease. Clin Microbiol Infect. 2006;12:231–5.
123
ORIGINAL ARTICLE
Early diagnosis and treatment are crucial for the survival of Pneumocystis pneumonia patients without human immunodeficiency virus infection Nobuhiro Asai • Shinji Motojima • Yoshihiro Ohkuni • Ryo Matsunuma • Kei Nakashima • Takuya Iwasaki • Tamao Nakashita • Yoshihito Otsuka • Norihiro Kaneko
Received: 5 April 2012 / Accepted: 20 May 2012 / Published online: 13 June 2012 Ó Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases 2012
Abstract The mortality of Pneumocystis pneumonia (PCP) patients without human immunodeficiency virus (HIV) infection ranges from 0 to 70 %, whereas that of HIV-infected PCP patients ranges from 10 to 20 %. The reasons for these differences are not known. We retrospectively analyzed factors contributing to the survival of 23 patients with PCP and without HIV infection, in whom PCP developed as community-acquired pneumonia (CAP). The interval from admission to the start of PCP-specific treatment was significantly shorter for survivors (2.71 ± 3.64 days; n = 14) than for non-survivors (8.67 ± 5.5 days; n = 9; p = 0.003). Moreover, although the severity scores/classes assessed by A-DROP, CURB-65, and PSI were no different on admission, scores/classes at the start of PCP-specific treatment were significantly higher for non-survivors. Overall mortality was 39 %, but mortality was approximately 70–100 % for patients classified as severe grade by A-DROP, CURB-65, or PSI scores/classes at the time when PCP-specific treatment was started, which was far higher than expected for these guidelines. In conclusion, early diagnosis and treatment within 3 days are crucial for the survival of PCP patients without HIV infection. We emphasize the limitations of application of guidelines for CAP to patients with PCP.
N. Asai (&) S. Motojima Y. Ohkuni R. Matsunuma K. Nakashima T. Iwasaki T. Nakashita N. Kaneko The Department of Pulmonology, Kameda Medical Center, 929 Higashi-cho, Kamogawa, Chiba, Japan e-mail: [email protected] Y. Otsuka Department of Laboratory Medicine, Kameda Medical Center, Chiba, Japan
123
Keywords Community-acquired pneumonia Human immunodeficiency virus Pneumocystis pneumonia
Introduction Pneumocystis pneumonia (PCP) not related to human immunodeficiency virus (HIV) can develop in patients with malignancies, rheumatic diseases, and other diseases whose immune systems have been compromised by immunosuppressive drugs [1, 2]. The mortality of patients with PCP but without HIV infection is diverse and ranges from 0 to 70 % [1, 3–9], compared with mortality of HIV-infected PCP patients, which ranges from 10 to 20 % [1, 3, 10]. The higher mortality among non-HIV patients has been attributed to severe lung inflammation [1, 3, 5], although the exact etiology of these large mortality differences has not yet been determined. In this retrospective study, we attempted to identify clinical factors that contributed to the survival of non-HIV patients with PCP infection. In addition, because we routinely apply management guidelines for communityacquired pneumonia (CAP) to PCP patients whom we encounter in the outpatient clinic or the emergency room, we also tested whether the application of guidelines for CAP to patients with PCP was suitable; that is, whether the guidelines correctly evaluated the severity of PCP and correctly predicted mortality.
Patients and methods We retrospectively reviewed all PCP cases that developed in the setting of CAP and required admission for treatment in the Kameda Medical Center from 2001 to 2010. A total
J Infect Chemother (2012) 18:898–905
of 23 PCP patients had negative HIV tests. PCP cases that developed as hospital-acquired pneumonia (HAP) were excluded. PCP was diagnosed on the basis of the following: 1.
2. 3.
microbiological analysis of respiratory samples, for example induced sputum or bronchoalveolar lavage (BAL) fluid, by use of polymerase chain reaction and conventional staining with Grocott methenamine silver stain and Diff-QuickTM; radiographic findings on chest X-ray and chest computed tomography (CT) compatible with PCP; and compatible clinical findings, including dyspnea, cough, and fever [1, 3, 7, 11].
All PCP patients were enrolled with both positive PCPPCR and any microscopic findings from the induced sputum or BAL specimen. Radiologic findings and any clinical symptoms were confirmed. Because all PCP cases in this study developed as CAP, CAP severity was assessed on the first visit to the outpatient clinic or the emergency room by use of the A-DROP system developed by the Japanese Respiratory Society (JRS) [12], CURB-65 scores developed by the British Thoracic Society (BTS) [13], and the pneumonia severity index (PSI) developed by the Infectious Disease Society of America (IDSA) [14]. Severity was determined again on the day when PCP-specific treatment was started, because specific treatment for PCP was not started on the first day of admission for most patients, in accordance with recommendations for empirical treatment of CAP [12–14]. It remains a contentious issue whether or not severity determinations by these guidelines at the time of PCPspecific treatment are suitable, because, for most patients, several days will pass in the hospital before PCP-specific treatment is started. However, we used these guidelines at the start of PCP-specific treatment, for two reasons: 1. 2.
we wanted to compare PCP severity by using the same standards; and all of our PCP cases developed in the setting of CAP and no additional diseases developed after admission.
For PCP-specific treatment, all patients received oral or parenteral trimethoprim/sulfamethoxazole (TMP/SMX) and adjunct glucocorticoid therapy, although the day when PCP-specific treatment started after admission differed widely depending on the attending physicians. Statistical analysis Group means were compared by use of unpaired or paired t tests or the Mann–Whitney U test. Contingency tables were evaluated by use of Fisher’s exact probability test. p values \0.05 were regarded as significant.
899
Institutional Review Board approval Each patient has provided permission to publish these features of his/her case, and their identity has been protected. The IRB at Kameda Medical Center approved this retrospective study.
Results The characteristics of the 23 PCP patients are shown in Table 1. Twelve patients (52.2 %) had rheumatic or autoimmune diseases, which were the most common underlying diseases, followed by malignancy (n = 10; 43.5 %). Twenty-one patients (91.3 %) were receiving glucocorticoids or immunosuppressants for their underlying diseases. Prophylaxis consisting of TMP/SMX had been given to 1 patient (4.3 %). Mean CAP severity on admission, as assessed by A-DROP, CURB-65, and PSI scores/classes, is also shown. For all patients, PCP was diagnosed on the basis of any respiratory symptoms and radiology findings. Positive PCP-PCR results were obtained for all induced sputum (n = 18, 78.3 %) or BAL fluid (n = 5, 21.7 %) specimens. Either Diff-QuikÒ (n = 7, 30.4 %) or Grocott staining (n = 18, 78.3 %) was confirmed for all patients. Table 2 shows the laboratory results at the start of PCPspecific treatment of the patients according to their prognostic outcomes: i.e., survivors and non-survivors. Nine patients (39.1 %) died of PCP. There were no significant differences in laboratory results between the survivor and non-survivor groups except for PaO2/FiO2. On admission, the severity of pneumonia assessed by A-DROP, CURB-65, and PSI scores/classes was not significantly different between the survivor and non-survivor groups. However, at the start of PCP-specific treatment, the severity of pneumonia assessed by A-DROP, CURB-65, and PSI scores/ classes was significantly different between the survivor and non-survivor groups. There were also significant differences in the interval from admission to diagnosis of PCP and that from admission to initiation of PCP-specific treatment. In fact, most of the patients in the survivor group were started on PCP-specific treatment before diagnosis of PCP. Table 3 shows the changes in CAP severity between the two phases: on admission and at the start of PCP-specific treatment. In the survivor group, severity had not changed significantly at the start of PCP-specific treatment compared with that on admission. However, in the non-survivor group, severity had changed significantly at the start of PCP-specific treatment compared with that on admission. Table 4 shows the distributions of patients and mortality in each risk class of prediction rules at the start of PCP-specific treatment. Note the very high patient mortality
123
900
J Infect Chemother (2012) 18:898–905
Table 1 Characteristics of PCP patients without HIV infection (n = 23) Characteristic
Patients No.
%
71.3 (9.07),
57–88
Male
13
56.5
Female
10
43.5
10
43.5
6 5
26.1 21.7
Age, mean (SD), range Sex
Underlying diseases Malignancies Solid cancers Lung cancer Brain tumor
1
4.3
4
17.4
Acute myeloblastic leukemia
1
4.3
Adult T cell leukemia
1
4.3
Multiple myeloma
1
4.3
Malignant lymphoma
1
4.3
Hematologic malignancies
Rheumatic and autoimmune diseases
12
52.2
Rheumatoid arthritis
10
43.5
Dermatomyositis
1
4.3
Sarcoidosis
1
4.3
Diabetes mellitus
4
17.4
Chronic pulmonary diseases
6
26.1
Heart diseases
6
26.1
Cerebrovascular diseases Renal diseases
2 3
8.7 13.0
Liver disease
2
8.7
Long-term glucocorticoids alone
6
26.1
Immunosuppressants alone/chemotherapeutic agents alone
5
21.7
Methotrexate
2
8.7
u-RIST
1
4.3
Vinorerbin
1
4.3
TS-1
1
4.3
Long-term glucocorticoids combined with chemotherapeutic/immunosuppressive agents
10
43.5
Glucocorticoid ? methotrexate
7
30.4
Glucocorticoid ? anti-TNF drugs
3
13.0
PCP prophylaxis
1
4.3
Daily glucocorticoids dose at the time of first visit, PSL mg; mean (range) Glucocorticoids alone
20.7 (4–30)
Combination glucocorticoids and any agents
14.1 (2.5–18)
Duration of glucocorticoids immunosuppressive agents at the time of first visit, month; mean (range) Glucocorticoids alone
27.8 (1–128)
Combination glucocorticoids and any agents
23.0 (1–128)
Severity of CAP on admission evaluated by use of the following guidelines expressed as risk groups or net points. mean (SD) A-DROP
1.36 (1.14)
CURB-65
1.59 (1.14)
PSI
94.0 (25.7)
SD standard deviation, PCP Pneumocystis pneumonia, PSI pneumonia severity index, TNF tumor necrosis factor, TS-1 tegafur, gimeracil and oteracil potassium, u-RIST reduced intensity, allogeneic stem cell transplantation from unrelated donors
123
J Infect Chemother (2012) 18:898–905
901
Table 2 Patient characteristics on admission, by prognostic outcome Item
Survivors (n = 14)
Non-survivors (n = 9)
p value
Age
69.2 (8.6)
74.4 (8.8)
0.172
BMI (kg/m2)
22.4 (3.35)
22.2 (1.85)
0.853
b-DG (pg/ml)
87.5 (156)
75.1 (63.5)
0.824
KL-6 (U/ml)
781 (452)
1474 (1469)
0.175
LDH (IU/l)
515 (278)
729 (352)
0.126
CRP (mg/dl)
8.00 (4.25)
11.3 (5.44)
0.135
Alb/BUN
0.209 (0.128)
0.13 (0.072)
0.106
Lymphocytes (/ll)
1134.8 (992)
783.3 (397)
0.356
Neutrophils (/ll)
7149.1 (2900)
7729.3 (2976)
0.67
WBC (/ll)
9200 (3495)
10272 (5404)
0.568
PO2/FiO2 Interval from admission to PCP diagnosis (days)
223.4 (63.4) 4.86 (3.77)
137.5 (77.1) 8.56 (3.7)
0.007 0.012
Interval from admission to start PCP-specific treatment (days)
2.71 (3.64)
8.67 (5.5)
0.003
Severity of CAP on admission evaluated by use of the following guidelines A-DROP
1.07 (1.27)
1.78 (0.67)
0.14
CURB-65
1.36 (1.39)
1.89 (0.33)
0.28
PSI
87.1 (27.1)
105.6 (17.5)
0.09
Severity of CAP on admission evaluated by use of the following guidelines, number (%) A-DROP 0
6 (42.9)
0 (0)
1
4 (28.6)
3 (33.3)
2
2 (14.3)
5 (55.6)
3
1 (7.1)
1 (11.1)
4/5
1 (7.1)
0 (0)
0
5 (35.7)
0 (0)
1 2
3 (21.4) 4 (28.6)
1 (11.1) 8 (88.9)
3
0 (0)
0 (0)
4/5
2 (14.3)
0 (0)
I
0 (0)
0 (0)
II
4 (28.6)
0 (0)
III
3 (21.4)
1 (11.1)
IV
6 (42.9)
7 (77.8)
V
1 (7.1)
1 (11.1)
CURB-65
PSI
a
Severity of CAP at the start of PCP-specific treatment evaluated by use of the following guidelines A-DROP
1.21 (1.19)
CURB-65
1.43 (1.34)
PSI
88.6 (28.3)
Severity of CAP at the start of PCP-specific treatment evaluated by use of the following guidelines, number (%) A-DROP 0
4 (28.6)
0 (0)
1
6 (42.9)
0 (0)
2
2 (14.3)
1 (7.1)
3
1 (7.1)
5 (55.6)
4/5
1 (7.1)
3 (33.3)
123
902
J Infect Chemother (2012) 18:898–905
Table 2 continued Item
Survivors (n = 14)
Non-survivors (n = 9)
0
3 (21.4)
0 (0)
1
4 (28.6)
0 (0)
2
5 (35.7)
0 (0)
3
0 (0)
6 (66.7)
4/5
2 (14.3)
3 (33.3)
I
0 (0)
0 (0)
II
4 (28.6)
0 (0)
III
3 (21.4)
0 (0)
IV
6 (42.9)
3 (33.3)
V
1 (7.1)
6 (66.7)
p value
CURB-65
PSIa
Results are mean (SD) BMI body mass index, b-DG (1 ? 3)-b-D-glucan, KL-6 Krebs von den Lungen 6, LDH lactate dehydrogenase, CRP C-reactive protein, BUN blood urea nitrogen, WBC white blood cells, PCP Pneumocystis pneumonia, CAP community-acquired pneumonia, PSI pulmonary severity index a
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131
Table 3 Changes in CAP severity assessed by established guidelines and expressed as risk groups or net points On admission
At start of PCP-specific treatment
p value
Table 4 Distributions of patients and their mortality in each risk class of prediction rules at the start of PCP-specific treatment Risk groups
Number of patients (%)
Survivors (n = 14) A-DROP
1.07 (1.27)
1.21 (1.19)
0.956
CURB-65 PSI
1.36 (1.39) 87.3 (27.1)
1.43 (1.34) 88.6 (31.8)
0.891 0.893
1.78 (0.67)
3.33 (0.87)
\0.001
CURB-65
1.89 (0.33)
3.44 (0.77)
\0.001
105.6 (17.5)
143.3 (36.7)
0.009
PSI
Mortality (%)
A-DROP
Non-survivors (n = 9) A-DROP
Number of patients with respiratory failure (%)
0
6 (26.1)
3 (50)
0 (0)
1
7 (30.4)
6 (85.7)
3 (42.9)
2
7 (30.4)
7 (100)
5 (71.4)
3
2 (8.7)
2 (100)
1 (50)
4/5
1 (4.3)
1 (100)
0 (0)
CURB-65
Results are mean (SD)
0
5 (21.7)
2 (40)
0 (0)
PCP Pneumocystis pneumonia, CAP community-acquired pneumonia, PSI pulmonary severity index
1
4 (17.4)
3 (75)
1 (25)
2 3
12 (52.2) 0 (0)
12 (100) 0 (0)
2 (8.7)
2 (100)
0 (0)
0 (0)
in risk classes 2 by A-DROP and CURB-65, and class IV by PSI. Prognostic accuracy of each of the guidelines at the start of PCP-specific treatment is compared in Table 5. A-DROP scores of 3–5, CURB-65 scores of 3–5, and PSI risk class V are classified as severe in these guidelines. This also shows the survival predictions when PCP-specific treatment was started within 3, 5, or 7 days after admission. Note that the sensitivity, specificity, positive predictive value, and negative predictive value for survival were very high when PCPspecific treatment was started within 3 days after admission.
123
4/5
8 (66.7) 0 (0)
PSIa I
0 (0)
II
4 (17.4)
2 (50)
III
4 (17.4)
3 (75)
IV
13 (56.5)
12 (92.3)
V
2 (8.7)
2 (100)
1 (25) 7 (53.8) 1 (50)
PCP Pneumocystis pneumonia, PSI pulmonary severity index a
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131
J Infect Chemother (2012) 18:898–905
903
Table 5 Prognostic accuracy of A-DROP, CURB-65, and PSI at the start of PCP-specific treatment for patients classified in severe grades, and survival prediction when PCP-specific treatment was started within 3, 5, or 7 days after admission Sensitivity for mortality (%) A-DROP scores 3–5
11.1
CURB-65 scores 3–5 PSI risk classes V
0 11.1
Specificity for mortality (%) A-DROP scores 3–5
85.7
CURB-65 scores 3–5
85.7
PSI risk classes V
88.9
Positive predictive values for mortality (%) A-DROP scores 3–5
33.3
CURB-65 scores 3–5
0
PSI risk classes V
50.0
Negative predictive values for mortality (%) A-DROP scores 3–5
60.0
CURB-65 scores 3–5
57.1
PSI risk classes V
61.9
Sensitivity for survival (%) Within 3 days Within 5 days
92.9 72.2
Within 7 days
65.0
Specificity for survival (%) Within 3 days
88.9
Within 5 days
80.0
Within 7 days
66.7
Positive predictive values for survival (%) Within 3 days
92.9
Within 5 days
92.9
Within 7 days
92.9
Negative predictive values for survival (%) Within 3 days
88.9
Within 5 days
44.4
Within 7 days
22.2
In PSI, relationships between risk classes and net points are: I, 0; II, 1–70; III, 71–90; IV, 91–130; V, [131 PCP Pneumocystis pneumonia, PSI pulmonary severity index
Discussion It is well known that PCP can develop not only in patients with HIV infection but also in some immunocompromized patients [1, 2]. Mortality of PCP patients without HIV infection is diverse and ranges from 0 to 70 % [1, 3–9]; for HIV-infected PCP patients mortality ranges from 10 to 20 % [1, 3, 10]. The higher mortality among non-HIV patients has been attributed to severe lung inflammation [1, 3, 5]. However, the reasons for the large differences in mortality have not been determined.
For PCP patients without HIV infection, we found that the interval from admission to the start of PCP-specific treatment was significantly shorter in the survivor group than in the non-survivor group. In addition, the interval from admission to diagnosis was significantly shorter in the survivor group than in the non-survivor group. Our report is the second to point out the relationship between mortality and treatment delay [9]. Overgaard et al. [9] reported that the mean intervals from admission to the start of treatment in non-survivor and survivor groups were 15 and 5 days, respectively. Unfortunately, standard deviations were not shown in their report. In a recent report by Komano et al. [7] survival was 100 % for patients with rheumatoid arthritis in whom PCP developed during treatment with infliximab, a tumor necrosis factor (TNF) blocker, and attributed the good outcome to prompt diagnosis and treatment. However, their paper did not mention how prompt this was. Although a relationship between mortality and treatment delay has not been shown generally, we suspect that treatment delay was responsible for the high mortality in some reports. Moreover, we have shown specifically that the positive predictive value for survival was [90 % when PCP-specific treatment was started within 3 days after admission. This is the first report to show a survival prediction with regard to the starting day of PCP-specific treatment. Although the mean risk classes determined by A-DROP, CURB-65, and PSI scores/classes were not statistically significantly different between the survivor and non-survivor groups on admission, they were significantly different at the start of PCP-specific treatment. This difference seems to be largely related to the treatment delay in the non-survivor group. There was approximately 6 days difference in the mean start of PCP-specific treatment between the survivor and non-survivor groups, and during those days the risk classes increased in the non-survivor group. It is reasonable to assume that mortality will be high among patients in high-risk groups; i.e., those with scores [3 on A-DROP and CURB-65 and class V on PSI. However, at the time when PCP-specific treatment was started, it was striking that mortality was approximately 70 to 100 %. Shindo et al. showed that among CAP patients, for whom most of the pathognomonic microorganisms were presumably bacteria, patient mortality in risk groups [3 when A-DROP and CURB-65 were applied was 23.3 and 21.0 %, respectively [15]. Usui et al. [16] also showed that patient mortality in the high-risk groups when A-DROP and CURB-65 were applied was 11.5 and 11.6 %, respectively. According to Fine et al. [17], the mortality for severe CAP (risk group V; PSI [130) was 29.2 % in a Medis group validation cohort. In the report by Usui et al. [16], the mortality of patients in risk group V
123
904
J Infect Chemother (2012) 18:898–905
was 28.3 %. Chen et al. [18] reported that mortality of CAP patients in risk classes 3–5 on CURB-65 and class IV/V on PSI was 16.6 and 12.5 %, respectively. An additional striking fact is that, although most patients in the nonsurvivor group had risk scores of 3 but not 5, their mortality was extremely high. Accordingly, we emphasize that mortality prediction in PCP is not correct when these management guidelines for CAP are applied, even when PCP develops in the setting of CAP. These guidelines definitely underestimate the severity of PCP as CAP. The important issue is why these guidelines cannot correctly estimate PCP severity. We suppose that the pathophysiology is different between PCP and typical bacterial pneumonia. In experimental models of bacterial pneumonia, mortality is high for animals with high bacterial numbers in the lungs [19]. However, the relationship is the opposite in PCP. According to a report by Limper et al., the number of Pneumocystis cysts in BAL fluid was significantly higher in PCP with AIDS than in PCP without AIDS. The number of cysts in BAL fluid was significantly positively correlated with PaO2 levels, and there was a significant negative correlation between PaO2 and the percentage of neutrophils in BAL fluid. Surviving patients had more cysts in BAL fluid than non-surviving patients, but the difference was marginal. Finally, non-surviving patients have significantly higher percentages of neutrophils in BAL fluid [3]. Tokuda et al. reported that plasma concentrations of b-D-glucan, a constituent of Pneumocystis that is used as a diagnostic marker [20], were 10 times higher in AIDSrelated PCP than in rheumatoid arthritis (RA)-related PCP. Notwithstanding, the mortality for AIDS-related PCP is 0 % and that for RA-related PCP is 14.3 % [21]. Thus, it seems likely that establishment of PCP requires host immune responses to Pneumocystis because Pneumocystis can exist without host invasion [22]. This is in contrast with typical bacterial pneumonia in which bacteria release toxins and enzymes that can damage host tissues. We wish to emphasize that there are limitations to applying the guidelines for CAP to PCP despite the fact that PCP can develop in the setting of CAP. These reasons are: 1.
2.
Pneumocystis is not considered to be a routine pathogen for CAP, as evidenced by the fact that the term ‘‘Pneumocystis’’ is found only twice in the ATS guidelines [14] and only once in the BTS guidelines [13], resulting in TMP/SMX not being included in the first empirical treatment regimens; treatment failure is considered after 3 days’ (72 h) continuation of the empirical treatment in the ATS [14] and JRS guidelines [12], although our results show that PCP-specific treatment should be started within 3 days after admission; and
123
3.
immune responses to Pneumocystis are not always beneficial—in some situations they can be harmful to patients with PCP, as discussed above, and the concept that the correct choice of antibiotics alone is enough to resolve PCP may not be sufficient.
A clue enabling early diagnosis of PCP is chest CT findings; that is, ground glass appearance with a panlobular pattern or so called crazy paving appearance [21, 23]. These findings are also found in viral pneumonias, mycoplasmal pneumonia, alveolar hemorrhage, methotrexate pneumonia, and others. However, in combination with the background characteristics of patients, the presence of PCP should be suspected. There have been reports of prognostic markers of shortterm mortality in AIDS-associated PCP [24–26]. However, by use of univariate analysis we unfortunately did not prove or disprove any risk factors of mortality in non-HIV PCP except for interval both from admission to PCP diagnosis and from admission to the start of PCP-specific treatment, as shown in Table 2. A limitation of our study is that it is a retrospective analysis of a very small population. Retrospective studies may be less reliable in terms of the data collected, particularly physical examination data. A prospective study on more cases is necessary. It has been reported that 20 % [27] of ordinary people and 30 % [28] of patients with chronic respiratory diseases have colonization of P. jirovecii. Although PCP was diagnosed on the basis of microbiological findings, radiology findings, and clinical symptoms, there could be the colonization of P. jirovecii, as mentioned above. In conclusion, we argue that when patients possibly have PCP, as judged from patients’ underlying diseases, concomitant treatments, and radiographic findings, TMP/AMX should be added to routine empirical treatment within 3 days at most, even before the diagnosis of PCP is established, with possible concurrent use of glucocorticoids. Acknowledgments We are grateful for the diligent and thorough critical reading of our manuscript by Mr John Wocher, Executive Vice President and Director, International Affairs/International Patient Services, Kameda Medical Center (Japan). Conflict of interest
The authors have no conflicting interests.
References 1. Thomas CF Jr, Limper AH. Pneumocystis pneumonia. N Engl J Med. 2004;350:2487–8. 2. Kovacs JA, Masur H. Evolving health effects of Pneumocystis: one hundred years of progress in diagnosis and treatment. JAMA. 2009;301:2578–85. 3. Limper AH, Offord KP, Smith TF, Martin WJ 2nd. Pneumocystis pneumonia. Differences in lung parasite number and
J Infect Chemother (2012) 18:898–905
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
inflammation in patients with and without AIDS. Am Rev Respir Dis. 1989;140:1204–9. Festic E, Gajic O, Limper AH, Aksamit TR. Acute respiratory failure due to Pneumocystis pneumonia in patients without human immunodeficiency virus infection: outcome and associated features. Chest. 2005;128:573–9. Yale SH, Limper AH. Pneumocystis carinii pneumonia in patients without acquired immunodeficiency syndrome: associated illness and prior corticosteroid therapy. Mayo Clin Proc. 1996;71:5–13. Ward MM, Donald F. Pneumocystis carinii pneumonia in patients with connective tissue diseases: the role of hospital experience in diagnosis and mortality. Arthritis Rheum. 1999; 42:780–9. Komano Y, Harigai M, Koike R, Sugiyama H, Ogawa J, Saito K, et al. Pneumocystis jiroveci pneumonia in patients with rheumatoid arthritis treated with infliximab: a retrospective review and case-control study of 21 patients. Arthritis Rheum. 2009;61: 305–12. Su YS, Lu JJ, Perng CL, Chang FY. Pneumocystis jirovecii pneumonia in patients with and without human immunodeficiency virus infection. J Microbiol Immunol Infect. 2008;41:478–82. Overgaard UM, Helweg-Larsen J. Pneumocystis jiroveci pneumonia (PCP) in HIV-1-negative patients: a retrospective study 2002–2004. Scand J Infect Dis. 2007;39:589–95. Dworkin MS, Hanson DL, Navin TR. Survival of patients with AIDS, after diagnosis of Pneumocystis carinii pneumonia, in the United States. J Infect Dis. 2001;183:1409–12. Harigai M, Koike R. Pneumocystis Pneumonia under Anti-Tumor Necrosis Factor Therapy (PAT) Study Group. Pneumocystis pneumonia associated with infliximab in Japan. N Engl J Med. 2007;357:1874–6. Miyashita N, Matsushima T. Japanese Respiratory Society. The JRS guidelines for the management of community-acquired pneumonia in adults: an update and new recommendations. Intern Med. 2006;45:419–28. British Thoracic Society Standards of Care Committee. BTS guidelines for the management of community acquired pneumonia in adults. Thorax. 2001;56(Suppl 4):1–64. American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med. 2001;163:1730–54. Shindo Y, Sato S, Maruyama E, Ohashi T, Ogawa M, Imaizumi K, et al. Comparison of severity scoring systems A-DROP and CURB-65 for community-acquired pneumonia. Respirology. 2008;13:731–5. Usui K, Tanaka Y, Noda H, Ishihara T. Comparison of three prediction rules for prognosis in community acquired pneumonia: Pneumonia Severity Index (PSI), CURB-65, and A-DROP. Nihon Kokyuki Gakkai Zasshi 2009; 47:781–5 (Japanese).
905 17. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, Singer DE, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med. 1997;336: 243–50. 18. Chen JH, Chang SS, Liu JJ, Chan RC, Wu JY, Wang WC, et al. Comparison of clinical characteristics and performance of pneumonia severity score and CURB-65 among younger adults, elderly and very old subjects. Thorax. 2010;65:971–7. 19. Wolff M, Joly-Guillou ML, Farinotti R, Carbon C. In vivo efficacies of combinations of beta-lactams, beta-lactamase inhibitors, and rifampin against Acinetobacter baumannii in a mouse pneumonia model. Antimicrob Agents Chemother. 1999;43: 1406–11. 20. Tasaka S, Hasegawa N, Kobayashi S, Yamada W, Nishimura T, Takeuchi T, et al. Serum indicators for the diagnosis of Pneumocystis pneumonia. Chest. 2007;131:1173. 21. Tokuda H, Sakai F, Yamada H, Johkoh T, Imamura A, Dohi M, et al. Clinical and radiological features of Pneumocystis pneumonia in patients with rheumatoid arthritis, in comparison with methotrexate pneumonitis and Pneumocystis pneumonia in acquired immunodeficiency syndrome: a multicenter study. Intern Med. 2008;47:915–23. 22. Maskell NA, Waine DJ, Lindley A, Pepperell JC, Wakefield AE, Miller RF, et al. Asymptomatic carriage of Pneumocystis jiroveci in subjects undergoing bronchoscopy: a prospective study. Thorax. 2003;58:594–7. 23. Kameda H, Tokuda H, Sakai F, Johkoh T, Mori S, Yoshida Y, et al. Clinical and radiological features of acute-onset diffuse interstitial lung diseases in patients with rheumatoid arthritis receiving treatment with biological agents: importance of Pneumocystis pneumonia in Japan revealed by a multicenter study. Intern Med. 2011;50:305–13. 24. Benfield TL, Helweg-Larsen J, Bang D, Junge J, Lundgren JD. Prognostic markers of short-term mortality in AIDS-associated Pneumocystis carinii pneumonia. Chest. 2001;119:844–51. 25. Azoulay E, Parrot A, Flahault A, Cesari D, Lecomte I, Roux P, et al. AIDS-related Pneumocystis carinii pneumonia in the era of adjunctive steroids: implication of BAL neutrophilia. Am J Respir Crit Care Med. 1999;160:493–9. 26. Fei MW, Kim EJ, Sant CA, Jarlsberg LG, Davis JL, Swartzman A, et al. Predicting mortality from HIV-associated Pneumocystis pneumonia at illness presentation: an observational cohort study. Thorax. 2009;64:1070–6. 27. Medrano FJ, Montes-Cano M, Conde MA, de la Horra C, Respaldiza N, Gasch A, et al. Pneumocystis jirovecii in general population. Emerg Infect Dis. 2005;11:245–50. 28. Vidal S, de la Horra C, Martin J, Montes-Cano MA, Rodriguez E, Respaldiza N, et al. Pneumocystis jirovecii colonization in patients with interstitial lung disease. Clin Microbiol Infect. 2006;12:231–5.
123