Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Pneumocystis jirovecci pneumonia in connective tissue diseases: Comparison with other immunocompromised patients Andrew J. Teichtahl, (MBBS, BPhysio, FRACP, PhD)a,b,*, Kathleen Morrisroe, (MBBS)a,1, Sabina Ciciriello, (MBBS, FRACP, PhD)a, Ian Jennens, (MBBS, FRACP)c, Susan Tadros, (MBBS, BSc, MRCP)a, Ian Wicks, (MBBS, FRACP, PhD)a,d,e,n a
Department of Rheumatology, The Royal Melbourne Hospital, Parkville 3050, Victoria, Australia Baker IDI Heart and Diabetes Institute, Melbourne 3004, Victoria, Australia Victorian Infectious Diseases Services, The Royal Melbourne Hospital, Parkville 3050, Victoria, Australia d Inflammation Division, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville 3050, Victoria, Australia e University of Melbourne, Parkville, Victoria, Australia b c
a r t i c l e in fo
Keywords: Immunocompromised Lymphocyte Connective tissue Pneumocystis jirovecci
a b s t r a c t Introduction: Pneumocystis jirovecci pneumonia (PJP) is an opportunistic fungal infection occurring in immunocompromised patients, such as those with human immunodeficiency virus (HIV), organ transplantation, malignancies and connective tissue diseases (CTDs). Risk factors for PJP are not well characterised, leading to uncertainty regarding the indications for antimicrobial prophylaxis and monitoring. This study compared differences between patients with and without CTDs who developed PJP. Methods: Retrospective data was collected for all subjects with a positive toludine blue O stain or a positive P. jirovecci PCR and a concurrent respiratory illness that was clinically consistent with PJP between 2002 and 2013 at the Royal Melbourne Hospital, Australia. Sub-groups were assigned according to the underlying disease. Peripheral blood results were retrieved from an in-house pathology database. Results: Eleven of 90 subjects (12.2%) diagnosed with PJP had underlying CTDs. The CTDs group was more likely to have been exposed to corticosteroids (100% versus 35.2%, p o 0.001) and other iatrogenic immunosuppression (90.9% versus 24.6%, p o 0.001). After adjusting for age and gender, the CTDs group had greater lymphopaenia (0.17 versus 0.58 109/L; p ¼ 0.034) and were older (69.6 versus 50.6 years; p o 0.001) than the non-CTD group. Excluding renal transplant recipients, people with CTDs also had lower eGFR than the non-CTD group (65 versus 80; p ¼ 0.015). Conclusions: CTDs contributed to a significant proportion of total PJP diagnoses. Clinicians treating CTDs must be vigilant for PJP, particularly in older patients with exposure to corticosteroids or other iatrogenic immunosuppression, lymphopaenia and renal impairment; factors which may lower the clinical threshold for initiating prophylaxis. & 2015 Elsevier Inc. All rights reserved.
Introduction Pneumocystis jirovecci (formerly Pneumocystis carinii) pneumonia (PJP) is an opportunistic fungal infection most commonly occurring in acquired immunodeficiency syndrome (AIDS)/human
Acknowledgements: This work was supported by the Reid Charitable Trusts, the NHMRC (Clinical Practitioner Fellowship 1023407 to I.P.W.; Program grants 1016647 to I.P.W.). A.J.T. is the recipient of the NHMRC Early Career Fellowship (ID 1073284). n Corresponding author at: Inflammation Division, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville 3050, Victoria, Australia. E-mail address: 1 Denotes joint first author http://dx.doi.org/10.1016/j.semarthrit.2015.01.007 0049-0172/& 2015 Elsevier Inc. All rights reserved.
immunodeficiency virus (HIV). However, iatrogenic immunosuppression is now commonplace for organ transplantation, malignancy and systemic inflammatory illnesses, giving rise to another at-risk patient population. A potentially underappreciated group within the non-HIV-infected cohort is that of connective tissue diseases (CTDs), including patients with conditions such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), inflammatory muscle diseases and systemic vasculitides. The incidence of PJP infection in the CTD population has been estimated at 1–2% [1,2], but with new detection methods such as polymerase chain reaction (PCR), the prevalence of PJP in CTDs populations may be even higher than previously considered [3]. Although PJP is relatively uncommon in CTDs patients, it is often fatal and mortality rates may be higher than in the HIV population [1].
2
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
Unlike HIV, there are no evidence-based guidelines regarding the use of PJP prophylaxis in the CTDs population and recommendations have been extrapolated from studies in the non-CTDs setting [2,4]. Nevertheless, relative to other conditions, a higher rate of adverse events have been noted among people receiving PJP prophylaxis for CTDs [5], raising safety concerns for clinicians. Lymphopaenia is prevalent in cohorts developing PJP, including those with haematological malignancies, renal transplants and HIV [6–8]. Lymphopaenia has also been associated with PJP infection in CTDs and may be associated with increased mortality [9–11]. However, it is unclear whether the degree of lymphopaenia or other laboratory parameters differ significantly between CTDs and other immunocompromised states. Determining such factors may help to inform recommendations for PJP prophylaxis. The aims of this study were therefore (i) to examine the contribution of CTDs to the total number of PJP diagnoses and (ii) to determine any between group differences in subject characteristics and basic laboratory parameters between people with non-CTDs and CTDs who had been diagnosed with PJP.
Methods Data collection This single-center Australian study was conducted at the Royal Melbourne Hospital (RMH), which is a public, university-affiliated tertiary referral adult teaching hospital. Retrospective data collection covered the period January 2002 to September 2013. Inclusion criteria comprised all patients who had a positive toludine blue O stain or a strongly positive (as judged by low polymerase chain reaction (PCR) cycle number, i.e., o32 cycles) PCR on sputum or bronchoalveolar lavage (BAL) specimens. Although toludine blue O stains were available from study inception in 2002, P. jirovecci PCR was only available at RMH from 2006 onwards. All cases were reviewed to ensure the presence of a respiratory illness consistent with PJP, together with either a positive toludine blue O stain or a positive P. jirovecci PCR. Where a subject was found to have recurrent admissions for PJP (n ¼ 3), only the initial episode was included in the study. Where a subject was found to have both a positive toludine blue O and P. jirovecci PCR, the first positive result was considered to be the time of PJP diagnosis. For each episode, patient characteristics (age at diagnosis and gender) and blood results were retrieved from the hospital pathology database. The lowest peripheral blood cell counts, eGFR and albumin and the highest liver function test abnormalities and inflammatory markers, documented within a week either side of the microbiological diagnosis of PJP, were recorded. Mortality was attributed if the patient died within the admission where PJP was diagnosed. Medical histories were also examined to determine exposure to PJP prophylaxis, as well as exposure to corticosteroids or other immunosuppressive drugs (e.g., chemotherapy and antirejection medications) within the 3 months preceding PJP diagnosis. The study was approved by the Royal Melbourne Hospital Human Research Ethics Committee. Statistical analyses Independent t-tests were used to compare unadjusted differences in the characteristics between subject groups. The CTD group was compared with the remainder of the cohort in its entirety and with the major non-CTD sub-groups of HIV, renal transplant, haematological malignancy and solid malignancy. To quantify the between group differences for blood parameters such as the lymphocyte count, general linear models were used and
estimated marginal means calculated, adjusting for age and gender. Chi-squared analyses were used to compare between group differences with dichotomous outcomes (e.g., exposure to corticosteroids). A p-value of less than 0.05 (two-tailed) was considered statistically significant. All analyses were performed using SPSS statistical package (standard version 21.1; SPSS, Chicago, IL, USA).
Results A total of 90 subjects who fulfilled the diagnostic criteria of either a positive P. jirovecii PCR (64%) or toludine blue O stain (36%) with a concurrent respiratory illness were included in the study. Four other cases were excluded because no information regarding an underlying disease process or the presence of a concurrent respiratory illness could be determined. Of the 90 subjects, 39 (43.3%) had HIV/AIDS, 16 (17.8%) had haematological malignancies, 14 (15.6%) had solid malignancies, 11 (12.2%) had CTDs, 7 (7.8%) had renal transplants, 2 (2.2%) had inflammatory bowel disease and 1 (1.1%) had chronic airways disease. The characteristics of the 11 patients with CTDs are shown in Table 1. All CTD patients were exposed to corticosteroids in the 3 months prior to PJP diagnosis, and only subject 8 had not been taking any other biological or nonbiological disease modifying anti-rheumatic drug (DMARD). The maximum dose of corticosteroids in the 3 months preceding PJP diagnosis for each participant is shown in Table 1. The mean dose of oral corticosteroids in the CTD group was 15.5 mg ( 712.3 mg) (range: 5–37.5 mg). One CTD subject (subject 11) had been pulsed with intravenous methylprednisolone in the month prior to developing PJP. Three of the 11 patients had been exposed to cyclophosphamide, but in 2 of these patients, it was 3 years prior to the diagnosis of PJP. Patient 8 had been exposed to cyclophosphamide 3 months prior to the diagnosis of PJP. Although one CTD patient had PJP prophylaxis (Pentamidine), compliance with this therapy is unknown. Clinical information was examined in 57 of the 79 non-CTDs patients where the medical record was available. Fewer patients in the non-CTDs group had been exposed to any corticosteroids in the 3 months preceding PJP diagnosis (35.1% versus 100%, p o 0.001). The mean corticosteroid dose tended to be greater in the CTD sub-group (15.5 mg versus 7.6 mg, p ¼ 0.11). Moreover, the CTD group were more likely than the non-CTD group to have been exposed to other forms of iatrogenic immunosuppression in the 3 months preceding PJP diagnosis (90.1% versus 24.6%, p o 0.001). Exposure to corticosteroid and other immunosuppression was limited to the non-HIV patients. When the HIV sub-group was excluded, the CTD group still had a higher percentage of people exposed to corticosteroids (100% versus 40.5%, p ¼ 0.01) and iatrogenic immunosuppression (100% versus 24.6%, p o 0.001). Two patients in the non-CTDs group had received PJP propyhlaxis (Trimethoprim-sulfamethoxazole). Five people died during the admission for PJP, 3 in the CTDs group and from the non-CTDs group (27.3% versus 2.5%, p o 0.001), both of whom had underlying haematological malignancies. The laboratory and subject characteristics of the total PJP cohort, as well as the CTD and non-CTD sub-groups are shown in Table 2. On average, the total cohort was lymphopaenic (0.53 7 0.56 109/L) but was not leukopaenic or neutropaenic and was mildly anaemic (108.5 7 20.3 g/L) and hypoalbuminaemic (26.1 7 5.9 g/L). Moreover, despite having raised inflammatory markers (CRP, 109.5 7 95.7 mg/L; ESR, 83.7 7 45.0 mm in an hour), the total cohort did not have a reactive thrombocytosis (platelet count: 240 7 133 109/L). The total cohort demonstrated renal impairment, with a mild reduction in the eGFR (74 7 23).
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
3
Table 1 Participant characteristics among people with CTDs diagnosed with PJP Subject Gender CTD
Age Prophylaxis
Prednisolone dose
DMARDs or bDMARDs
Exposure to cyclophosphamide (ever)
Mortality
1 2 3 4
Male Female Female Male
Behcet’s disease Myositis Psoriatic arthritis Rheumatoid arthritis
78 74 71 78
No No No No
5 mg 5 mg 5 mg 5 mg
No No No No
No No No No
5
Male
Rheumatoid arthritis
65
No
25 mg
No
Yes
6
Male
Systemic vasculitis
65
Pentamidine 15 mg
Yes
No
7
Male
57
No
15 mg
Yes
No
8 9
Eosinophilic granulomatosis with polyangitis Male Granulomatosis with polyangitis Female Relapsing polychondritis
AZA, 100 mg AZA, 75 mg Etanercept MTX, 20 mg LEF, 20 mg HCQ 200 mg MTX, 20 mg LEF, 20 mg Myocristin, 50 mg AZA, 50 mg MTX, 10 mg MTX, 25 mg
78 67
No No
35 mg 17.5 mg
Yes No
No No
10
Female Rheumatoid arthritis
76
No
No
Yes
11
Male
57
No
No
Yes
Dermatomyositis
No MTX, 17.5 mg Adalimumab 5 mg MTX, 10 mg LEF, 20 mg HCQ, 200 mg 37.5 mg þ IVMP AZA, 50 mg
AZA, azathioprine; bDMARD, biological disease-modifying anti-rheumatic drug; DMARD, disease modifying anti-rheumatic drug; HCQ, hydroxychloroquine; IVMP, intravenous methylprednisolone (1 g daily for 3 days); LEF, leflunomide; MTX, methotrexate. Prednisolone dose refers to the maximal dose of corticosteoid used within the 3 months prior to PJP diagnosis. Prednisolone, DMARD or bDMARD derived from dose documented in medical records over the 3 months preceding PJP.
Table 3 shows unadjusted analyses using independent t-tests and the estimated marginal means for both CTDs and pooled nonCTDs. In unadjusted analyses, people with CTDs were older (69.6 versus 50.6 years; p o 0.001), more lymphopaenic (0.18 versus 0.57 109/L; p ¼ 0.028) and had a lower eGFR (61 versus 76; p ¼ 0.036) than the pooled non-CTD group at the time of PJP diagnosis. After adjusting for age and gender, the CTDs group still had significantly greater lymphopaenia (0.17 versus 0.58 109/L; p ¼ 0.034) than the pooled non-CTD group. When the renal
transplant sub-group was excluded from the pooled population, people with CTDs had a significantly lower eGFR than the non-CTD group (65 versus 80; p ¼ 0.015) at the time of PJP diagnosis. Table 4 presents sub-group analyses according to the underlying diagnoses. The CTDs sub-group had significantly worse renal function (eGFR) at the time of diagnosis of PJP than the HIV (p ¼ 0.002) sub-group. The renal transplantation group had a lower eGFR than the CTDs (27 versus 62; p ¼ 0.022). The degree of lymphopaenia was greater in the CTDs sub-group when compared
Table 2 Between group differences between people with and without connective tissue diseases
Age (years) Gender (% females)a Toludine blue O stain/PCR (%)a Lymphocytes ( 109/L) Haemoglobin (g/L) Leucocytes ( 109/L) Platelets ( 109/L) Neutrophils ( 109/L) eGFR ESR (n ¼ 24) (mm/h) CRP (n ¼ 71) (mg/L) ALP (n ¼ 80) (IU/L) AST (n ¼ 76) (IU/L) ALT (n ¼ 80) (IU/L) Albumin (n ¼ 80) (g/L) PJP prophylaxis (%)a Corticosteroid exposurea Corticosteroid dose (mg)b Iatrogenic exposure to other immunosuppressantsa Mortality, n (%)a
Total cohort (N ¼ 90)
CTD (n ¼ 11)
Non-CTD (n ¼ 79)
p
52.9 (17.1) 35.6% 36%/64% 0.53 (0.56) 108.5 (20.3) 7.9 (4.6) 240 (133) 6.5 (4.2) 74 (23) 83.7 (45.0) 109.5 (95.7) 141.9 (159.1) 44.6 (27.6) 38.9 (44.7) 26.1 (5.9) 2 (3.0)d 31 (45.6)c c 8.9 (20.1)d 24 (26.7)d 5 (5.6)
69.6 (8.0) 40.0% 30%/70% 0.18 (0.16) 109.7 (26.0) 7.1 (3.6) 236 (130) 5.8 (3.5) 61 (20) 95.7 (50.1) 174 (124) 98 (71) 47 (33) 34 (23) 27 (7) 1 (9.1) 11 (100) 15.5 (12.3) 10 (90.9) 3 (27.3)
50.6 (16.8) 35.4% 36.7%/63.3% 0.57 (0.57) 108.4 (19.6) 8.0 (4.8) 241 (135) 6.6 (4.2) 76 (22) 78.1 (42.7) 98 (86) 148 (167) 44 (27) 37 (24) 26 (6) 1 (1.8)c 20 (35.1)c 7.6 (21.2) c 14 (24.6)c 2 (2.5)
o 0.001 0.952 0.540 o 0.001 0.871 0.455 0.917 0.488 0.039 0.406 0.076 0.107 0.796 0.562 0.858 0.20 o 0.001 0.11 o 0.001 0.001
All results displayed as mean (7SD) unless stated otherwise. p Values for between group differences for the CTD and non-CTD subjects. Normal ranges: lymphocytes (1.2–4), leucocytes (4–11), platelets (140–400), neutrophils (2–8), CRP (o5), ALP (30–120), AST (o31), ALT (o34), albumin (30–50), creatinine (45–90), and eGFR (490). a
Chi-square test. Corticosteroid dose refers to the maximal dose of corticosteroid use within the 3 months prior to PJP diagnosis. Data available for 57 participants. d Data available for 68 participants. b c
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
4
Table 3 Estimated marginal means and between group differences between PJP cases with and the pooled population without CTDs Multivariate 1a
Unadjusted
Lymphocytes eGFR eGFRb
CTD (n ¼ 11)
Pooled (n ¼ 79)
p
CTD (n ¼ 11)
Pooled (n ¼ 79)
p
0.18 (0.16) 61 (7) 61 (5)
0.57 (0.06) 76 (2) 81 (2)
0.028 0.036 o0.001
0.17 (0.18) 67 (7) 65 (5)
0.58 (0.06) 75 (2) 80 (2)
0.034 0.274 0.015
Normal ranges: lymphocytes (1.2–4) and eGFR (490). Renal refers to transplantation, while haematology and oncology refer to patients treated for malignancy. a b
Estimated marginal means (standard error of the mean) adjusted for gender and age. eGFR analyses excluding the renal transplant sub-group.
with the renal transplants (0.19 versus 0.60 109/L; p ¼ 0.002), with a trend toward significance in the solid malignancies (0.17 versus 0.40 109/L; p ¼ 0.057), haematological malignancies (0.20 versus 0.67 109/L; p ¼ 0.094) and HIV (0.15 versus 0.59 109/L; p ¼ 0.096) sub-groups.
with CTDs survive longer than their haematological malignancy and renal transplant counterparts. When the renal transplant group were excluded, people with underlying CTDs also had a greater degree of renal impairment than the non-CTDs cohort. This result was independent of age and gender, and suggests that renal impairment may be a risk factor for PJP. Prospective longitudinal data are required to confirm this conclusion in CTDs. In this study, patients were lymphopaenic at the time of a diagnosis of PJP (0.53 7 0.56 109/L), with significantly greater lymphopaenia demonstrable in the CTDs group (mean ¼ 0.17 109/L; range: 0.00–0.50 109/L). While we cannot determine causality between lymphopaenia and the risk of PJP from this study, our results emphasise the likely importance of lymphopaenia in its pathogenesis. It has been documented that lymphopaenia is common among people diagnosed with PJP with co-existing CTDs, haematological malignancies, renal transplantation and HIV [6–11], however no previous study has compared the degree of lymphopaenia between CTDs and non-CTDs cohorts. This study demonstrates that subjects with CTDs are significantly more lymphopaenic than non-CTD sub-groups at the time of PJP diagnosis. Recently, it had been suggested that prophylaxis be considered in CTDs when certain risk factors for PJP exist, including an absolute lymphocyte count of o0.3 109/L [12]. Results from our study would support this recommendation, but larger prospective studies are required to better delineate a potential “lymphopaenic threshold” where PJP prophylaxis is recommended. Indeed, rationalisation of PJP prophylaxis may also be dependent on other factors, such as immunomodulatory
Discussion This study has demonstrated that CTDs contributed to a relatively large proportion of PJP diagnoses (12.2%) in a tertiary care hospital. Around the time of PJP diagnosis, people with CTDs were significantly older, had greater lymphopaenia and worse renal function than non-CTD subjects and were more likely to have been exposed to corticosteroids or other forms of iatrogenic immunosuppression. These results suggest that clinicians must be vigilant for PJP in patients with CTDs, particularly in patients with corticosteroid or other iatrogenic immunosuppresive exposure, lymphopaenia, older age and renal impairment. No previous study has examined the incidence of PJP in CTDs, relative to other immunocompromised groups. When patients with HIV were excluded, CTDs contributed 21.6% of all documented PJP diagnoses, which dispels the notion that CTDs contribute only a small proportion of PJP diagnoses in tertiary care settings. In all sub-groups, excluding solid malignancies, people with CTDs were significantly older than their immunocompromised counterparts, independent of gender. Age may therefore be a risk factor for PJP in CTDs. Alternatively, it may be that people
Table 4 Estimated marginal means and between group differences between PJP cases with and without CTDs based on sub-groups Multivariate 1a
Unadjusted
Lymphocytes eGFR
Lymphocytes eGFR
Lymphocytes eGFR
Lymphocytes eGFR
CTD
HIV (n ¼ 39)
p
CTD
HIV (n ¼ 39)
p
0.18 (0.16) 61 (7)
0.58 (0.09) 84 (2)
0.031 o 0.001
0.15 (0.22) 61 (6)
0.59 (0.10) 84 (3)
0.096 0.002
CTD
Renal (n ¼ 7)
p
CTD
Renal (n ¼ 7)
p
0.18 (0.16) 61 (7)
0.60 (0.10) 30 (9)
0.16 (0.07) 62 (8)
0.64 (0.09) 27 (10)
0.002 0.022
CTD
Haematology (n ¼ 16)
CTD
Haematology (n ¼ 16)
p
0.18 (0.16) 61 (7)
0.69 (0.15) 76 (4)
0.20 (0.20) 63 (6)
0.67 (0.16) 74 (5)
0.094 0.167
CTD
Oncology (n ¼ 14)
CTD
Oncology (n ¼ 14)
p
0.18 (0.16) 61 (7)
0.39 (0.08) 74 (5)
0.17 (0.09) 62 (5)
0.40 (0.08) 72 (5)
0.057 0.170
0.004 0.014 p 0.045 0.039 p 0.091 0.109
Normal ranges: lymphocytes (1.2–4) and eGFR (490). Renal refers to transplantation, while haematology and oncology refer to patients treated for malignancy. a
Estimated marginal means (standard error of the mean) adjusted for gender and age.
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
medications. We have shown that people with CTDs were more likely to have been exposed to corticosteroids and other iatrogenic immunosuppression than non-CTD patients. A recent retrospective cohort study suggested that patients with CTDs who had PJP had a greater risk for lymphopaenia and mortality when treatment also involved mycophenolate mofetil [13]. Another study suggested that people with asymptomatic carriage of P. jirovecci were most likely to benefit from short-term prophylaxis [14]. Nevertheless, institution of PJP prophylaxis is not without risk of adverse reactions. Emerging data suggests that it may however be possible to stratify people at risk of adverse events from PJP prophylaxis, with the anti-RNP antibody being a potential risk factor for adverse events in a sub-group of patients with CTDs [5]. Impaired cellular immunity is regarded as an important factor in the development of P. jirovecii infection [15]. Why people with CTDs may have more severe lymphopaenia at the time of diagnosis of PJP is speculative. Lymphopaenia may reflect the activity of an underlying CTD such as SLE [16], but surprisingly, our PJP cohort did not comprise any subjects with SLE. Alternatively, lymphopaenia may result from treatment of CTDs with agents such as corticosteroids or cyclophosphamide [17]. Irrespective of the mechanisms accounting for the marked lymphopaenia seen in CTDs, this profile is of concern, because non-survivors of PJP in CTDs have been shown to have significantly greater lymphopaenia than survivors [18]. In this study, 3 of the 11 CTDs patients (27.3%) died during their admission, as opposed to 2 of the 79 non-CTDs patients with PJP (2.5%) (p o 0.01 for difference). These data are consistent with observations that the PJP mortality rate is greater in CTDs [9–11]. Although this study has examined exposure to corticosteroid and other iatrogenic immunosuppressive therapies in the 3 months preceding PJP diagnosis, future studies would benefit from assessment of cumulative dose, duration and compliance with pharmacotherapy. Further analyses of lymphocyte T helper subsets may also help to better understand the contribution and kinetics of lymphopaenia in the pathogenesis of PJP. Moreover, future studies would benefit from temporal analyses of total and subset lymphocyte counts to determine whether the lymphopaenia at the time of PJP infection was in fact chronic, or was an acute phenomenon at the time of illness. Limited numbers may have led to sub-group analyses not achieving statistical significance for differences in some variables and our analysis was limited by the unavailability of medical records for 22 participants. Nevertheless, all sub-group analyses demonstrated a trend for CTDs to have a greater magnitude of lymphopaenia and renal impairment (excluding the renal transplant sub-group). Finally, the aim of this study was to compare differences between patients with and without CTDs who had developed PJP. We did not aim to identify risk factors for the acquisition of PJP, which would have required a control arm, and ideally, a prospective longitudinal study design. This study has demonstrated that people with CTDs contribute a significant proportion of PJP diagnoses in a large, tertiary care hospital. Clinicians must maintain vigilance for PJP when managing CTDs, particularly in the setting of older patients exposed to corticosteroids or other iatrogenic immunosuppression, marked lymphopaenia and renal impairment—factors that appear to distinguish cases of PJP in CTDs from other immunocompromised states. These data may help guide rational introduction and
5
duration of PJP prophylaxis in the CTDs population, although further examination of this is required.
References [1] Sepkowitz KA. Opportunistic infections in patients with and patients without acquired immunodeficiency syndrome. Clin Infect Dis 2002;34(8):1098–107 (Epub 2002 Mar 21). [2] Green H, Paul M, Vidal L, Leibovici L. Prophylaxis of Pneumocystis pneumonia in immunocompromised non-HIV-infected patients: systematic review and meta-analysis of randomized controlled trials. Mayo Clin Proc 2007;82(9): 1052–9. [3] Saito K, Nakayamada S, Nakano K, Tokunaga M, Tsujimura S, Nakatsuka K, et al. Detection of Pneumocystis carinii by DNA amplification in patients with connective tissue diseases: re-evaluation of clinical features of P. carinii pneumonia in rheumatic diseases. Rheumatology (Oxford) 2004;43(4): 479–85 (Epub 2004 Feb 3). [4] Sepkowitz KA, Brown AE, Armstrong D. Pneumocystis carinii pneumonia without acquired immunodeficiency syndrome. More patients, same risk. Arch Intern Med 1995;155(11):1125–8. [5] Maezawa R, Kurasawa K, Arai S, Okada H, Owada T, Fukuda T. Positivity for anti-RNP antibody is a risk factor for adverse effects caused by trimethoprimsulfamethoxazole, a prophylactic agent for P. jiroveci pneumonia, in patients with connective tissue diseases. Mod Rheumatol 2013;23(1):62–70, http://dx. doi.org/10.1007/s10165-012-0625-x (Epub 2012 Mar 31). [6] Haeusler GM, Slavin MA, Seymour JF, Lingaratnam S, Teh BW, Tam CS, et al. Late-onset Pneumocystis jirovecii pneumonia post-fludarabine, cyclophosphamide and rituximab: implications for prophylaxis. Eur J Haematol 2013;91 (2):157–63, http://dx.doi.org/10.1111/ejh.12135 (Epub 2013 Jun 15). [7] De Castro N, Xu F, Porcher R, Pavie J, Molina JM, Peraldi MN. Pneumocystis jirovecii pneumonia in renal transplant recipients occurring after discontinuation of prophylaxis: a case–control study. Clin Microbiol Infect 2010;16 (9):1375–7, http://dx.doi.org/10.111/j.469-0691.2009.03143.x. [8] Wang HW, Lin CC, Kuo CF, Liu CP, Lee CM. Mortality predictors of Pneumocystis jirovecii pneumonia in human immunodeficiency virus-infected patients at presentation: experience in a tertiary care hospital of northern Taiwan. J Microbiol Immunol Infect 2011;44(4):274–81, http://dx.doi.org/10.1016/j. jmii.2010.08.006 (Epub 1 Jan 20). [9] Li J, Huang XM, Fang WG, Zeng XJ. Pneumocystis carinii pneumonia in patients with connective tissue disease. J Clin Rheumatol 2006;12(3):114–7. [10] Bachelez H, Schremmer B, Cadranel J, Mouly F, Sarfati C, Agbalika F, et al. Fulminant Pneumocystis carinii pneumonia in 4 patients with dermatomyositis. Arch Intern Med 1997;157(13):1501–3. [11] Godeau B, Coutant-Perronne V, Le Thi Huong D, Guillevin L, Magadur G, De Bandt M, et al. Pneumocystis carinii pneumonia in the course of connective tissue disease: report of 34 cases. J Rheumatol 1994;21(2):246–51. [12] Demoruelle MK, Kahr A, Verilhac K, Deane K, Fischer A, West S. Recent-onset systemic lupus erythematosus complicated by acute respiratory failure. Arthritis Care Res (Hoboken) 2013;65(2):314–23, http://dx.doi.org/10.1002/ acr.21857. [13] Zhang Y, Zheng Y. Pneumocystis jirovecii pneumonia in mycophenolate mofetil-treated patients with connective tissue disease: analysis of 17 cases. Rheumatol Int 2014;20:20. [14] Mori S, Cho I, Sugimoto M. A followup study of asymptomatic carriers of Pneumocystis jiroveci during immunosuppressive therapy for rheumatoid arthritis. J Rheumatol 2009;36(8):1600–5, http://dx.doi.org/10.3899/ jrheum.081270 (Epub 2009 Jun 16). [15] Reid AB, Chen SC, Worth LJ. Pneumocystis jirovecii pneumonia in non-HIVinfected patients: new risks and diagnostic tools. Curr Opin Infect Dis 2011;24 (6):534–44, http://dx.doi.org/10.1097/QCO.0b013e32834cac17. [16] Petri M, Orbai AM, Alarcon GS, Gordon C, Merrill JT, Fortin PR, et al. Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus. Arthritis Rheum 2012;64(8):2677–86, http://dx.doi.org/10.1002/art.34473. [17] Martin-Suarez I, D’Cruz D, Mansoor M, Fernandes AP, Khamashta MA, Hughes GR. Immunosuppressive treatment in severe connective tissue diseases: effects of low dose intravenous cyclophosphamide. Ann Rheum Dis 1997;56 (8):481–7. [18] Iikuni N, Kitahama M, Ohta S, Okamoto H, Kamatani N, Nishinarita M. Evaluation of Pneumocystis pneumonia infection risk factors in patients with connective tissue disease. Mod Rheumatol 2006;16(5):282–8.
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Pneumocystis jirovecci pneumonia in connective tissue diseases: Comparison with other immunocompromised patients Andrew J. Teichtahl, (MBBS, BPhysio, FRACP, PhD)a,b,*, Kathleen Morrisroe, (MBBS)a,1, Sabina Ciciriello, (MBBS, FRACP, PhD)a, Ian Jennens, (MBBS, FRACP)c, Susan Tadros, (MBBS, BSc, MRCP)a, Ian Wicks, (MBBS, FRACP, PhD)a,d,e,n a
Department of Rheumatology, The Royal Melbourne Hospital, Parkville 3050, Victoria, Australia Baker IDI Heart and Diabetes Institute, Melbourne 3004, Victoria, Australia Victorian Infectious Diseases Services, The Royal Melbourne Hospital, Parkville 3050, Victoria, Australia d Inflammation Division, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville 3050, Victoria, Australia e University of Melbourne, Parkville, Victoria, Australia b c
a r t i c l e in fo
Keywords: Immunocompromised Lymphocyte Connective tissue Pneumocystis jirovecci
a b s t r a c t Introduction: Pneumocystis jirovecci pneumonia (PJP) is an opportunistic fungal infection occurring in immunocompromised patients, such as those with human immunodeficiency virus (HIV), organ transplantation, malignancies and connective tissue diseases (CTDs). Risk factors for PJP are not well characterised, leading to uncertainty regarding the indications for antimicrobial prophylaxis and monitoring. This study compared differences between patients with and without CTDs who developed PJP. Methods: Retrospective data was collected for all subjects with a positive toludine blue O stain or a positive P. jirovecci PCR and a concurrent respiratory illness that was clinically consistent with PJP between 2002 and 2013 at the Royal Melbourne Hospital, Australia. Sub-groups were assigned according to the underlying disease. Peripheral blood results were retrieved from an in-house pathology database. Results: Eleven of 90 subjects (12.2%) diagnosed with PJP had underlying CTDs. The CTDs group was more likely to have been exposed to corticosteroids (100% versus 35.2%, p o 0.001) and other iatrogenic immunosuppression (90.9% versus 24.6%, p o 0.001). After adjusting for age and gender, the CTDs group had greater lymphopaenia (0.17 versus 0.58 109/L; p ¼ 0.034) and were older (69.6 versus 50.6 years; p o 0.001) than the non-CTD group. Excluding renal transplant recipients, people with CTDs also had lower eGFR than the non-CTD group (65 versus 80; p ¼ 0.015). Conclusions: CTDs contributed to a significant proportion of total PJP diagnoses. Clinicians treating CTDs must be vigilant for PJP, particularly in older patients with exposure to corticosteroids or other iatrogenic immunosuppression, lymphopaenia and renal impairment; factors which may lower the clinical threshold for initiating prophylaxis. & 2015 Elsevier Inc. All rights reserved.
Introduction Pneumocystis jirovecci (formerly Pneumocystis carinii) pneumonia (PJP) is an opportunistic fungal infection most commonly occurring in acquired immunodeficiency syndrome (AIDS)/human
Acknowledgements: This work was supported by the Reid Charitable Trusts, the NHMRC (Clinical Practitioner Fellowship 1023407 to I.P.W.; Program grants 1016647 to I.P.W.). A.J.T. is the recipient of the NHMRC Early Career Fellowship (ID 1073284). n Corresponding author at: Inflammation Division, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville 3050, Victoria, Australia. E-mail address: 1 Denotes joint first author http://dx.doi.org/10.1016/j.semarthrit.2015.01.007 0049-0172/& 2015 Elsevier Inc. All rights reserved.
immunodeficiency virus (HIV). However, iatrogenic immunosuppression is now commonplace for organ transplantation, malignancy and systemic inflammatory illnesses, giving rise to another at-risk patient population. A potentially underappreciated group within the non-HIV-infected cohort is that of connective tissue diseases (CTDs), including patients with conditions such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), inflammatory muscle diseases and systemic vasculitides. The incidence of PJP infection in the CTD population has been estimated at 1–2% [1,2], but with new detection methods such as polymerase chain reaction (PCR), the prevalence of PJP in CTDs populations may be even higher than previously considered [3]. Although PJP is relatively uncommon in CTDs patients, it is often fatal and mortality rates may be higher than in the HIV population [1].
2
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
Unlike HIV, there are no evidence-based guidelines regarding the use of PJP prophylaxis in the CTDs population and recommendations have been extrapolated from studies in the non-CTDs setting [2,4]. Nevertheless, relative to other conditions, a higher rate of adverse events have been noted among people receiving PJP prophylaxis for CTDs [5], raising safety concerns for clinicians. Lymphopaenia is prevalent in cohorts developing PJP, including those with haematological malignancies, renal transplants and HIV [6–8]. Lymphopaenia has also been associated with PJP infection in CTDs and may be associated with increased mortality [9–11]. However, it is unclear whether the degree of lymphopaenia or other laboratory parameters differ significantly between CTDs and other immunocompromised states. Determining such factors may help to inform recommendations for PJP prophylaxis. The aims of this study were therefore (i) to examine the contribution of CTDs to the total number of PJP diagnoses and (ii) to determine any between group differences in subject characteristics and basic laboratory parameters between people with non-CTDs and CTDs who had been diagnosed with PJP.
Methods Data collection This single-center Australian study was conducted at the Royal Melbourne Hospital (RMH), which is a public, university-affiliated tertiary referral adult teaching hospital. Retrospective data collection covered the period January 2002 to September 2013. Inclusion criteria comprised all patients who had a positive toludine blue O stain or a strongly positive (as judged by low polymerase chain reaction (PCR) cycle number, i.e., o32 cycles) PCR on sputum or bronchoalveolar lavage (BAL) specimens. Although toludine blue O stains were available from study inception in 2002, P. jirovecci PCR was only available at RMH from 2006 onwards. All cases were reviewed to ensure the presence of a respiratory illness consistent with PJP, together with either a positive toludine blue O stain or a positive P. jirovecci PCR. Where a subject was found to have recurrent admissions for PJP (n ¼ 3), only the initial episode was included in the study. Where a subject was found to have both a positive toludine blue O and P. jirovecci PCR, the first positive result was considered to be the time of PJP diagnosis. For each episode, patient characteristics (age at diagnosis and gender) and blood results were retrieved from the hospital pathology database. The lowest peripheral blood cell counts, eGFR and albumin and the highest liver function test abnormalities and inflammatory markers, documented within a week either side of the microbiological diagnosis of PJP, were recorded. Mortality was attributed if the patient died within the admission where PJP was diagnosed. Medical histories were also examined to determine exposure to PJP prophylaxis, as well as exposure to corticosteroids or other immunosuppressive drugs (e.g., chemotherapy and antirejection medications) within the 3 months preceding PJP diagnosis. The study was approved by the Royal Melbourne Hospital Human Research Ethics Committee. Statistical analyses Independent t-tests were used to compare unadjusted differences in the characteristics between subject groups. The CTD group was compared with the remainder of the cohort in its entirety and with the major non-CTD sub-groups of HIV, renal transplant, haematological malignancy and solid malignancy. To quantify the between group differences for blood parameters such as the lymphocyte count, general linear models were used and
estimated marginal means calculated, adjusting for age and gender. Chi-squared analyses were used to compare between group differences with dichotomous outcomes (e.g., exposure to corticosteroids). A p-value of less than 0.05 (two-tailed) was considered statistically significant. All analyses were performed using SPSS statistical package (standard version 21.1; SPSS, Chicago, IL, USA).
Results A total of 90 subjects who fulfilled the diagnostic criteria of either a positive P. jirovecii PCR (64%) or toludine blue O stain (36%) with a concurrent respiratory illness were included in the study. Four other cases were excluded because no information regarding an underlying disease process or the presence of a concurrent respiratory illness could be determined. Of the 90 subjects, 39 (43.3%) had HIV/AIDS, 16 (17.8%) had haematological malignancies, 14 (15.6%) had solid malignancies, 11 (12.2%) had CTDs, 7 (7.8%) had renal transplants, 2 (2.2%) had inflammatory bowel disease and 1 (1.1%) had chronic airways disease. The characteristics of the 11 patients with CTDs are shown in Table 1. All CTD patients were exposed to corticosteroids in the 3 months prior to PJP diagnosis, and only subject 8 had not been taking any other biological or nonbiological disease modifying anti-rheumatic drug (DMARD). The maximum dose of corticosteroids in the 3 months preceding PJP diagnosis for each participant is shown in Table 1. The mean dose of oral corticosteroids in the CTD group was 15.5 mg ( 712.3 mg) (range: 5–37.5 mg). One CTD subject (subject 11) had been pulsed with intravenous methylprednisolone in the month prior to developing PJP. Three of the 11 patients had been exposed to cyclophosphamide, but in 2 of these patients, it was 3 years prior to the diagnosis of PJP. Patient 8 had been exposed to cyclophosphamide 3 months prior to the diagnosis of PJP. Although one CTD patient had PJP prophylaxis (Pentamidine), compliance with this therapy is unknown. Clinical information was examined in 57 of the 79 non-CTDs patients where the medical record was available. Fewer patients in the non-CTDs group had been exposed to any corticosteroids in the 3 months preceding PJP diagnosis (35.1% versus 100%, p o 0.001). The mean corticosteroid dose tended to be greater in the CTD sub-group (15.5 mg versus 7.6 mg, p ¼ 0.11). Moreover, the CTD group were more likely than the non-CTD group to have been exposed to other forms of iatrogenic immunosuppression in the 3 months preceding PJP diagnosis (90.1% versus 24.6%, p o 0.001). Exposure to corticosteroid and other immunosuppression was limited to the non-HIV patients. When the HIV sub-group was excluded, the CTD group still had a higher percentage of people exposed to corticosteroids (100% versus 40.5%, p ¼ 0.01) and iatrogenic immunosuppression (100% versus 24.6%, p o 0.001). Two patients in the non-CTDs group had received PJP propyhlaxis (Trimethoprim-sulfamethoxazole). Five people died during the admission for PJP, 3 in the CTDs group and from the non-CTDs group (27.3% versus 2.5%, p o 0.001), both of whom had underlying haematological malignancies. The laboratory and subject characteristics of the total PJP cohort, as well as the CTD and non-CTD sub-groups are shown in Table 2. On average, the total cohort was lymphopaenic (0.53 7 0.56 109/L) but was not leukopaenic or neutropaenic and was mildly anaemic (108.5 7 20.3 g/L) and hypoalbuminaemic (26.1 7 5.9 g/L). Moreover, despite having raised inflammatory markers (CRP, 109.5 7 95.7 mg/L; ESR, 83.7 7 45.0 mm in an hour), the total cohort did not have a reactive thrombocytosis (platelet count: 240 7 133 109/L). The total cohort demonstrated renal impairment, with a mild reduction in the eGFR (74 7 23).
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
3
Table 1 Participant characteristics among people with CTDs diagnosed with PJP Subject Gender CTD
Age Prophylaxis
Prednisolone dose
DMARDs or bDMARDs
Exposure to cyclophosphamide (ever)
Mortality
1 2 3 4
Male Female Female Male
Behcet’s disease Myositis Psoriatic arthritis Rheumatoid arthritis
78 74 71 78
No No No No
5 mg 5 mg 5 mg 5 mg
No No No No
No No No No
5
Male
Rheumatoid arthritis
65
No
25 mg
No
Yes
6
Male
Systemic vasculitis
65
Pentamidine 15 mg
Yes
No
7
Male
57
No
15 mg
Yes
No
8 9
Eosinophilic granulomatosis with polyangitis Male Granulomatosis with polyangitis Female Relapsing polychondritis
AZA, 100 mg AZA, 75 mg Etanercept MTX, 20 mg LEF, 20 mg HCQ 200 mg MTX, 20 mg LEF, 20 mg Myocristin, 50 mg AZA, 50 mg MTX, 10 mg MTX, 25 mg
78 67
No No
35 mg 17.5 mg
Yes No
No No
10
Female Rheumatoid arthritis
76
No
No
Yes
11
Male
57
No
No
Yes
Dermatomyositis
No MTX, 17.5 mg Adalimumab 5 mg MTX, 10 mg LEF, 20 mg HCQ, 200 mg 37.5 mg þ IVMP AZA, 50 mg
AZA, azathioprine; bDMARD, biological disease-modifying anti-rheumatic drug; DMARD, disease modifying anti-rheumatic drug; HCQ, hydroxychloroquine; IVMP, intravenous methylprednisolone (1 g daily for 3 days); LEF, leflunomide; MTX, methotrexate. Prednisolone dose refers to the maximal dose of corticosteoid used within the 3 months prior to PJP diagnosis. Prednisolone, DMARD or bDMARD derived from dose documented in medical records over the 3 months preceding PJP.
Table 3 shows unadjusted analyses using independent t-tests and the estimated marginal means for both CTDs and pooled nonCTDs. In unadjusted analyses, people with CTDs were older (69.6 versus 50.6 years; p o 0.001), more lymphopaenic (0.18 versus 0.57 109/L; p ¼ 0.028) and had a lower eGFR (61 versus 76; p ¼ 0.036) than the pooled non-CTD group at the time of PJP diagnosis. After adjusting for age and gender, the CTDs group still had significantly greater lymphopaenia (0.17 versus 0.58 109/L; p ¼ 0.034) than the pooled non-CTD group. When the renal
transplant sub-group was excluded from the pooled population, people with CTDs had a significantly lower eGFR than the non-CTD group (65 versus 80; p ¼ 0.015) at the time of PJP diagnosis. Table 4 presents sub-group analyses according to the underlying diagnoses. The CTDs sub-group had significantly worse renal function (eGFR) at the time of diagnosis of PJP than the HIV (p ¼ 0.002) sub-group. The renal transplantation group had a lower eGFR than the CTDs (27 versus 62; p ¼ 0.022). The degree of lymphopaenia was greater in the CTDs sub-group when compared
Table 2 Between group differences between people with and without connective tissue diseases
Age (years) Gender (% females)a Toludine blue O stain/PCR (%)a Lymphocytes ( 109/L) Haemoglobin (g/L) Leucocytes ( 109/L) Platelets ( 109/L) Neutrophils ( 109/L) eGFR ESR (n ¼ 24) (mm/h) CRP (n ¼ 71) (mg/L) ALP (n ¼ 80) (IU/L) AST (n ¼ 76) (IU/L) ALT (n ¼ 80) (IU/L) Albumin (n ¼ 80) (g/L) PJP prophylaxis (%)a Corticosteroid exposurea Corticosteroid dose (mg)b Iatrogenic exposure to other immunosuppressantsa Mortality, n (%)a
Total cohort (N ¼ 90)
CTD (n ¼ 11)
Non-CTD (n ¼ 79)
p
52.9 (17.1) 35.6% 36%/64% 0.53 (0.56) 108.5 (20.3) 7.9 (4.6) 240 (133) 6.5 (4.2) 74 (23) 83.7 (45.0) 109.5 (95.7) 141.9 (159.1) 44.6 (27.6) 38.9 (44.7) 26.1 (5.9) 2 (3.0)d 31 (45.6)c c 8.9 (20.1)d 24 (26.7)d 5 (5.6)
69.6 (8.0) 40.0% 30%/70% 0.18 (0.16) 109.7 (26.0) 7.1 (3.6) 236 (130) 5.8 (3.5) 61 (20) 95.7 (50.1) 174 (124) 98 (71) 47 (33) 34 (23) 27 (7) 1 (9.1) 11 (100) 15.5 (12.3) 10 (90.9) 3 (27.3)
50.6 (16.8) 35.4% 36.7%/63.3% 0.57 (0.57) 108.4 (19.6) 8.0 (4.8) 241 (135) 6.6 (4.2) 76 (22) 78.1 (42.7) 98 (86) 148 (167) 44 (27) 37 (24) 26 (6) 1 (1.8)c 20 (35.1)c 7.6 (21.2) c 14 (24.6)c 2 (2.5)
o 0.001 0.952 0.540 o 0.001 0.871 0.455 0.917 0.488 0.039 0.406 0.076 0.107 0.796 0.562 0.858 0.20 o 0.001 0.11 o 0.001 0.001
All results displayed as mean (7SD) unless stated otherwise. p Values for between group differences for the CTD and non-CTD subjects. Normal ranges: lymphocytes (1.2–4), leucocytes (4–11), platelets (140–400), neutrophils (2–8), CRP (o5), ALP (30–120), AST (o31), ALT (o34), albumin (30–50), creatinine (45–90), and eGFR (490). a
Chi-square test. Corticosteroid dose refers to the maximal dose of corticosteroid use within the 3 months prior to PJP diagnosis. Data available for 57 participants. d Data available for 68 participants. b c
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
4
Table 3 Estimated marginal means and between group differences between PJP cases with and the pooled population without CTDs Multivariate 1a
Unadjusted
Lymphocytes eGFR eGFRb
CTD (n ¼ 11)
Pooled (n ¼ 79)
p
CTD (n ¼ 11)
Pooled (n ¼ 79)
p
0.18 (0.16) 61 (7) 61 (5)
0.57 (0.06) 76 (2) 81 (2)
0.028 0.036 o0.001
0.17 (0.18) 67 (7) 65 (5)
0.58 (0.06) 75 (2) 80 (2)
0.034 0.274 0.015
Normal ranges: lymphocytes (1.2–4) and eGFR (490). Renal refers to transplantation, while haematology and oncology refer to patients treated for malignancy. a b
Estimated marginal means (standard error of the mean) adjusted for gender and age. eGFR analyses excluding the renal transplant sub-group.
with the renal transplants (0.19 versus 0.60 109/L; p ¼ 0.002), with a trend toward significance in the solid malignancies (0.17 versus 0.40 109/L; p ¼ 0.057), haematological malignancies (0.20 versus 0.67 109/L; p ¼ 0.094) and HIV (0.15 versus 0.59 109/L; p ¼ 0.096) sub-groups.
with CTDs survive longer than their haematological malignancy and renal transplant counterparts. When the renal transplant group were excluded, people with underlying CTDs also had a greater degree of renal impairment than the non-CTDs cohort. This result was independent of age and gender, and suggests that renal impairment may be a risk factor for PJP. Prospective longitudinal data are required to confirm this conclusion in CTDs. In this study, patients were lymphopaenic at the time of a diagnosis of PJP (0.53 7 0.56 109/L), with significantly greater lymphopaenia demonstrable in the CTDs group (mean ¼ 0.17 109/L; range: 0.00–0.50 109/L). While we cannot determine causality between lymphopaenia and the risk of PJP from this study, our results emphasise the likely importance of lymphopaenia in its pathogenesis. It has been documented that lymphopaenia is common among people diagnosed with PJP with co-existing CTDs, haematological malignancies, renal transplantation and HIV [6–11], however no previous study has compared the degree of lymphopaenia between CTDs and non-CTDs cohorts. This study demonstrates that subjects with CTDs are significantly more lymphopaenic than non-CTD sub-groups at the time of PJP diagnosis. Recently, it had been suggested that prophylaxis be considered in CTDs when certain risk factors for PJP exist, including an absolute lymphocyte count of o0.3 109/L [12]. Results from our study would support this recommendation, but larger prospective studies are required to better delineate a potential “lymphopaenic threshold” where PJP prophylaxis is recommended. Indeed, rationalisation of PJP prophylaxis may also be dependent on other factors, such as immunomodulatory
Discussion This study has demonstrated that CTDs contributed to a relatively large proportion of PJP diagnoses (12.2%) in a tertiary care hospital. Around the time of PJP diagnosis, people with CTDs were significantly older, had greater lymphopaenia and worse renal function than non-CTD subjects and were more likely to have been exposed to corticosteroids or other forms of iatrogenic immunosuppression. These results suggest that clinicians must be vigilant for PJP in patients with CTDs, particularly in patients with corticosteroid or other iatrogenic immunosuppresive exposure, lymphopaenia, older age and renal impairment. No previous study has examined the incidence of PJP in CTDs, relative to other immunocompromised groups. When patients with HIV were excluded, CTDs contributed 21.6% of all documented PJP diagnoses, which dispels the notion that CTDs contribute only a small proportion of PJP diagnoses in tertiary care settings. In all sub-groups, excluding solid malignancies, people with CTDs were significantly older than their immunocompromised counterparts, independent of gender. Age may therefore be a risk factor for PJP in CTDs. Alternatively, it may be that people
Table 4 Estimated marginal means and between group differences between PJP cases with and without CTDs based on sub-groups Multivariate 1a
Unadjusted
Lymphocytes eGFR
Lymphocytes eGFR
Lymphocytes eGFR
Lymphocytes eGFR
CTD
HIV (n ¼ 39)
p
CTD
HIV (n ¼ 39)
p
0.18 (0.16) 61 (7)
0.58 (0.09) 84 (2)
0.031 o 0.001
0.15 (0.22) 61 (6)
0.59 (0.10) 84 (3)
0.096 0.002
CTD
Renal (n ¼ 7)
p
CTD
Renal (n ¼ 7)
p
0.18 (0.16) 61 (7)
0.60 (0.10) 30 (9)
0.16 (0.07) 62 (8)
0.64 (0.09) 27 (10)
0.002 0.022
CTD
Haematology (n ¼ 16)
CTD
Haematology (n ¼ 16)
p
0.18 (0.16) 61 (7)
0.69 (0.15) 76 (4)
0.20 (0.20) 63 (6)
0.67 (0.16) 74 (5)
0.094 0.167
CTD
Oncology (n ¼ 14)
CTD
Oncology (n ¼ 14)
p
0.18 (0.16) 61 (7)
0.39 (0.08) 74 (5)
0.17 (0.09) 62 (5)
0.40 (0.08) 72 (5)
0.057 0.170
0.004 0.014 p 0.045 0.039 p 0.091 0.109
Normal ranges: lymphocytes (1.2–4) and eGFR (490). Renal refers to transplantation, while haematology and oncology refer to patients treated for malignancy. a
Estimated marginal means (standard error of the mean) adjusted for gender and age.
A.J. Teichtahl et al. / Seminars in Arthritis and Rheumatism ] (2015) ]]]–]]]
medications. We have shown that people with CTDs were more likely to have been exposed to corticosteroids and other iatrogenic immunosuppression than non-CTD patients. A recent retrospective cohort study suggested that patients with CTDs who had PJP had a greater risk for lymphopaenia and mortality when treatment also involved mycophenolate mofetil [13]. Another study suggested that people with asymptomatic carriage of P. jirovecci were most likely to benefit from short-term prophylaxis [14]. Nevertheless, institution of PJP prophylaxis is not without risk of adverse reactions. Emerging data suggests that it may however be possible to stratify people at risk of adverse events from PJP prophylaxis, with the anti-RNP antibody being a potential risk factor for adverse events in a sub-group of patients with CTDs [5]. Impaired cellular immunity is regarded as an important factor in the development of P. jirovecii infection [15]. Why people with CTDs may have more severe lymphopaenia at the time of diagnosis of PJP is speculative. Lymphopaenia may reflect the activity of an underlying CTD such as SLE [16], but surprisingly, our PJP cohort did not comprise any subjects with SLE. Alternatively, lymphopaenia may result from treatment of CTDs with agents such as corticosteroids or cyclophosphamide [17]. Irrespective of the mechanisms accounting for the marked lymphopaenia seen in CTDs, this profile is of concern, because non-survivors of PJP in CTDs have been shown to have significantly greater lymphopaenia than survivors [18]. In this study, 3 of the 11 CTDs patients (27.3%) died during their admission, as opposed to 2 of the 79 non-CTDs patients with PJP (2.5%) (p o 0.01 for difference). These data are consistent with observations that the PJP mortality rate is greater in CTDs [9–11]. Although this study has examined exposure to corticosteroid and other iatrogenic immunosuppressive therapies in the 3 months preceding PJP diagnosis, future studies would benefit from assessment of cumulative dose, duration and compliance with pharmacotherapy. Further analyses of lymphocyte T helper subsets may also help to better understand the contribution and kinetics of lymphopaenia in the pathogenesis of PJP. Moreover, future studies would benefit from temporal analyses of total and subset lymphocyte counts to determine whether the lymphopaenia at the time of PJP infection was in fact chronic, or was an acute phenomenon at the time of illness. Limited numbers may have led to sub-group analyses not achieving statistical significance for differences in some variables and our analysis was limited by the unavailability of medical records for 22 participants. Nevertheless, all sub-group analyses demonstrated a trend for CTDs to have a greater magnitude of lymphopaenia and renal impairment (excluding the renal transplant sub-group). Finally, the aim of this study was to compare differences between patients with and without CTDs who had developed PJP. We did not aim to identify risk factors for the acquisition of PJP, which would have required a control arm, and ideally, a prospective longitudinal study design. This study has demonstrated that people with CTDs contribute a significant proportion of PJP diagnoses in a large, tertiary care hospital. Clinicians must maintain vigilance for PJP when managing CTDs, particularly in the setting of older patients exposed to corticosteroids or other iatrogenic immunosuppression, marked lymphopaenia and renal impairment—factors that appear to distinguish cases of PJP in CTDs from other immunocompromised states. These data may help guide rational introduction and
5
duration of PJP prophylaxis in the CTDs population, although further examination of this is required.
References [1] Sepkowitz KA. Opportunistic infections in patients with and patients without acquired immunodeficiency syndrome. Clin Infect Dis 2002;34(8):1098–107 (Epub 2002 Mar 21). [2] Green H, Paul M, Vidal L, Leibovici L. Prophylaxis of Pneumocystis pneumonia in immunocompromised non-HIV-infected patients: systematic review and meta-analysis of randomized controlled trials. Mayo Clin Proc 2007;82(9): 1052–9. [3] Saito K, Nakayamada S, Nakano K, Tokunaga M, Tsujimura S, Nakatsuka K, et al. Detection of Pneumocystis carinii by DNA amplification in patients with connective tissue diseases: re-evaluation of clinical features of P. carinii pneumonia in rheumatic diseases. Rheumatology (Oxford) 2004;43(4): 479–85 (Epub 2004 Feb 3). [4] Sepkowitz KA, Brown AE, Armstrong D. Pneumocystis carinii pneumonia without acquired immunodeficiency syndrome. More patients, same risk. Arch Intern Med 1995;155(11):1125–8. [5] Maezawa R, Kurasawa K, Arai S, Okada H, Owada T, Fukuda T. Positivity for anti-RNP antibody is a risk factor for adverse effects caused by trimethoprimsulfamethoxazole, a prophylactic agent for P. jiroveci pneumonia, in patients with connective tissue diseases. Mod Rheumatol 2013;23(1):62–70, http://dx. doi.org/10.1007/s10165-012-0625-x (Epub 2012 Mar 31). [6] Haeusler GM, Slavin MA, Seymour JF, Lingaratnam S, Teh BW, Tam CS, et al. Late-onset Pneumocystis jirovecii pneumonia post-fludarabine, cyclophosphamide and rituximab: implications for prophylaxis. Eur J Haematol 2013;91 (2):157–63, http://dx.doi.org/10.1111/ejh.12135 (Epub 2013 Jun 15). [7] De Castro N, Xu F, Porcher R, Pavie J, Molina JM, Peraldi MN. Pneumocystis jirovecii pneumonia in renal transplant recipients occurring after discontinuation of prophylaxis: a case–control study. Clin Microbiol Infect 2010;16 (9):1375–7, http://dx.doi.org/10.111/j.469-0691.2009.03143.x. [8] Wang HW, Lin CC, Kuo CF, Liu CP, Lee CM. Mortality predictors of Pneumocystis jirovecii pneumonia in human immunodeficiency virus-infected patients at presentation: experience in a tertiary care hospital of northern Taiwan. J Microbiol Immunol Infect 2011;44(4):274–81, http://dx.doi.org/10.1016/j. jmii.2010.08.006 (Epub 1 Jan 20). [9] Li J, Huang XM, Fang WG, Zeng XJ. Pneumocystis carinii pneumonia in patients with connective tissue disease. J Clin Rheumatol 2006;12(3):114–7. [10] Bachelez H, Schremmer B, Cadranel J, Mouly F, Sarfati C, Agbalika F, et al. Fulminant Pneumocystis carinii pneumonia in 4 patients with dermatomyositis. Arch Intern Med 1997;157(13):1501–3. [11] Godeau B, Coutant-Perronne V, Le Thi Huong D, Guillevin L, Magadur G, De Bandt M, et al. Pneumocystis carinii pneumonia in the course of connective tissue disease: report of 34 cases. J Rheumatol 1994;21(2):246–51. [12] Demoruelle MK, Kahr A, Verilhac K, Deane K, Fischer A, West S. Recent-onset systemic lupus erythematosus complicated by acute respiratory failure. Arthritis Care Res (Hoboken) 2013;65(2):314–23, http://dx.doi.org/10.1002/ acr.21857. [13] Zhang Y, Zheng Y. Pneumocystis jirovecii pneumonia in mycophenolate mofetil-treated patients with connective tissue disease: analysis of 17 cases. Rheumatol Int 2014;20:20. [14] Mori S, Cho I, Sugimoto M. A followup study of asymptomatic carriers of Pneumocystis jiroveci during immunosuppressive therapy for rheumatoid arthritis. J Rheumatol 2009;36(8):1600–5, http://dx.doi.org/10.3899/ jrheum.081270 (Epub 2009 Jun 16). [15] Reid AB, Chen SC, Worth LJ. Pneumocystis jirovecii pneumonia in non-HIVinfected patients: new risks and diagnostic tools. Curr Opin Infect Dis 2011;24 (6):534–44, http://dx.doi.org/10.1097/QCO.0b013e32834cac17. [16] Petri M, Orbai AM, Alarcon GS, Gordon C, Merrill JT, Fortin PR, et al. Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus. Arthritis Rheum 2012;64(8):2677–86, http://dx.doi.org/10.1002/art.34473. [17] Martin-Suarez I, D’Cruz D, Mansoor M, Fernandes AP, Khamashta MA, Hughes GR. Immunosuppressive treatment in severe connective tissue diseases: effects of low dose intravenous cyclophosphamide. Ann Rheum Dis 1997;56 (8):481–7. [18] Iikuni N, Kitahama M, Ohta S, Okamoto H, Kamatani N, Nishinarita M. Evaluation of Pneumocystis pneumonia infection risk factors in patients with connective tissue disease. Mod Rheumatol 2006;16(5):282–8.
