Severe Pneumocystis jirovecii pneumonia in an idiopathic CD4+ lymphocytopenia patient: case report and review of the literature

Introduction: When diagnosing Pneumocystis jirovecii pneumonia (PJP), the clinical suspicion must be confirmed by laboratory tests. PJP is rarely described in patients with idiopathic CD4+ lymphocytopenia (ICL), a rare T-cell deficiency of unknown origin with persistently low levels of CD4+ T-cells (<300 µl−1 or <20 % of total lymphocytes) but repeated negative human immunodeficiency virus (HIV) tests. We retrospectively analysed a case of an ICL patient with severe PJP associated with multiple opportunistic infections (OIs). We also reviewed the literature since 1986. Case presentation: A laboratory-confirmed case of PJP associated with invasive candidiasis and cytomegalovirus infection was reported in an ICL patient. Despite early treatment, the patient died of respiratory failure under polymicrobial pneumonia. According to the literature, the mortality rate of ICL patients is 10.4 % (33/316). In ICL patients, the risk of OI is 83.2 % (263/316), with viral infections being the most prevalent (58.2 %, 184/316), followed by fungal infections (52.2 %, 165/316) and mycobacterial infections (15.5 %, 49/316). Dysimmunity is reported in 15.5 % (49/316) of ICL patients. Among the fungal infections, cryptococcal infections are the most prevalent (24.1 %, 76/316), followed by candidiasis (15.5 %, 49/316) and PJP (7.9 %, 25/316). Conclusions: The high risk of OIs underlines the importance of more vigorous preventative actions in hospitals. The response to therapy and the detection of early relapse of PJP may be monitored by several laboratory tests including quantitative PCR. It is essential to treat the ICL and to follow the guidelines concerning therapy and prophylaxis of OIs as given to HIV patients.


Introduction
Pneumocystis jirovecii pneumonia (PJP) is infrequently reported in idiopathic CD4 + lymphocytopenia (ICL) patients (Ahmad et al., 2013). ICL is a rare and heterogeneous T-cell immunodeficiency syndrome of unknown origin with repeatedly low levels of CD4 + T-cells (,300 ml 21 or ,20 % of total lymphocytes), and no evidence of human immunodeficiency virus (HIV) infection (Régent et al., 2014;Smith et al., 1993).The differential diagnosis of ICL remains a challenge as it involves a wide range of analyses in different specialty areas such as immunology, haematology, rheumatology and infectious diseases (Zonios et al., 2012). Moreover, the differential diagnosis is also complicated by the fact that ICL patients may have underlying AIDS (Smith et al., 1993). At the time when multiple opportunistic pathogens are identified, the CD4 + counts are generally #150 cells ml 21 (Ahmad et al., 2013;Denis et al., 2014;Marukutira et al., 2014). We report a case of PJP associated with cytomegalovirus (CMV), Candida parapsilosis, Streptococcus pneumoniae and Pseudomonas aeruginosa infection. The aim of the present review was to place our rare PJP case in perspective and to evaluate the prevalence of opportunistic infections (OIs) in ICL patients, in particular the importance of fungal infections.

Methods
This was a retrospective case study. Bronchoalveolar lavage fluid (BALF) samples from the ICL patient were used for in vitro culture for Candida, microscopy for Pneumocystis identification and quantitative PCR (q-PCR) for the detection of P. jirovecii DNA; candidaemia was also evaluated in this patient (Chumpitazi et al., 2011(Chumpitazi et al., , 2014Fricker-Hidalgo et al., 2004). The q-PCR target used was the gene encoding the Pneumocystis major surface glycoprotein (MSG) (Chumpitazi et al., 2011;Cushion & Stringer, 2010;Khot & Fredricks, 2009). The amplicon DNA was sequenced, and pair-wise sequence alignments of the amplicons were performed using CEQ2000 DNA analysis software (Beckman). Further identification of MSG genes of P. jirovecii was performed by BLAST from both nucleotide and amino acid sequences of the three successive isolates from BALF samples with at least 99 % nucleotide identity for each.
For the review of the literature, a Medline search was performed using the keyword combinations: 'idiopathic CD4 + lymphocytopenia' and 'opportunistic infections', or 'idiopathic CD4 + lymphocytopenia' and 'Pneumocystis pneumonia'. We found a systematic review of ICL from 1986 to April 2012 (Ahmad et al., 2013). We added to this review the studies from May 2012 to April 2014.

Case report
A 52-year-old man, with a medical history of surgery to remove an eyelid basocellular carcinoma, presented with dyspnoea on exertion and dry cough without fever at day 23. The patient had hypogammaglobulinaemia, repeated negative HIV tests and autoimmune serologies and was referred to the intensive care unit with asthenia, dyspnoea, dry cough and low-grade fever (day 0). Care of the patient required mechanical ventilation for 38 days. The patient had lymphopenia and low CD4 + counts (#70 cells ml 21 ) during the follow-up (Table 1). The diagnosis of PJP, Table 1. Treatment and biological monitoring of an ICL patient with PJP complicated by viral, bacterial and fungal infections. The ICL patient had lymphopenia, high serum concentrations of C-reactive protein (CRP) and lactate dehydrogenase (LDH), plus low CD4 + counts. CRP levels diminished after treatment but remained relatively high. which revealed the patient's ICL (Smith et al., 1993), was confirmed by microscopic analysis of a BALF sample at day 4 and the patient was treated with trimethoprimsulfamethoxazole (TMP-SMZ). At day 15, the Pneumocystis infection resolved, but the patient presented with severe pancytopenia. TMP-SMZ treatment was stopped and pentamidine isethionate was prescribed at day 16. However, as relapse of PJP was confirmed by q-PCR (day 19) and by microscopy (day 24), TMP-SMZ was reintroduced from day 24 to day 51 (Table 1). To assess the response to therapy, microscopic analysis and q-PCR were performed on BALF samples (Table 1) (Chumpitazi et al., 2011). In a context of multiple polymicrobial pneumonia from day 4 to day 46 (e.g. invasive candidiasis, CMV viraemia and P. aeruginosa infection from day 28 to day 34), the specificity of the Pneumocystis q-PCR used was confirmed by the amplicon nucleotide sequence and BLAST analysis. The q-PCR from BALF was positive for three analysed samples (Table 1). The MSG genes/amino acid sequences identified were AF033209.1/AAC34972.1, AF033210.1/AAC34973.1, AF372980.1/AAL23912.1, DQ000981.1/AAY18808.1 and DQ000983.1/AAY18810.1 (GenBank accession numbers). The first three MSG genes were common to all isolates. At day 51, the patient died of respiratory failure in the context of severe immunodeficiency.

Discussion
The differential diagnosis of ICL is difficult given its scarcity and the extensive number of analyses that need to be performed in different specialties (Zonios et al., 2012). By the time an OI is identified, the CD4 + counts are generally #150 cells ml 21 and the patient has become susceptible to multiple opportunistic pathogens (Ahmad et al., 2013;Denis et al., 2014;Marukutira et al., 2014). As seen in the present review, the risk of OIs in ICL patients is very high, and prophylaxis against these pathogens is essential. This high risk of OIs emphasizes the requirement for more vigorous preventative measures to be taken in hospitals (Walzer, 2013). Cryptococcosis has been described as the most common opportunistic disease in ICL patients; however, mycobacterial disease and progressive multifocal leukoencephalopathy are also mentioned (Ahmad et al., 2013;Zonios et al., 2007Zonios et al., , 2012. Other OIs may also occur such as tuberculosis, histoplasmosis and dermatomal VZV (Duncan et al., 1993;Luo & Li, 2008;Zonios et al., 2007Zonios et al., , 2012. The clinical presentation of PJP can vary from one patient to another, although common characteristics are present such as progressive dyspnoea, dry cough, hypoxia and fever. Chest radiography and highresolution assisted tomography can give valuable data in the diagnosis of Pneumocystis pneumonia. However, the clinical suspicion of OI needs to be confirmed by further laboratory analyses, such as in vitro culture, microscopic observation and q-PCR from BALF samples, as other invasive fungal infections may occur, as in the present case. Amplicon nucleotide sequences after PCR and BLAST analysis may provide a new method to validate the suspicion of clinical PJP. The particularity of our case report resided in severe PJP complicated by viral, bacterial and fungal infection, which is a lethal and very rare event in patients with ICL. A marker of PJP severity and poor prognosis was the high serum concentrations of C-reactive protein (CRP), which were in the range of those found elsewhere in PJP patients (Sage et al., 2010). Venzor et al. (1997) described the case of an ICL patient with PJP, CMV and Candida infection. During the follow-up, the common factor of polymicrobial pneumonia between this case and ours was the CD4 + counts of #70 cells ml 21 . This suggests that multiple OIs may occur below this CD4 + count cut-off. The misdiagnosis of PJP and invasive candidiasis may have fatal consequences in ICL patients given its low prevalence and the diagnostic dilemmas it poses.
In our clinical case, the Pneumocystis MSG load (at day 42) was in favour of renewed infection and failure of the treatment with TMP-SMZ. The limit of the q-PCR used at day 42 was mainly the differentiation of active P. jirovecii infection and residual Pneumocystis colonization, which we confirmed here by amplicon nucleotide sequence and BLAST analysis in the absence of a positive microscopic observation (Alanio et al., 2011;Chumpitazi et al., 2011;Roux et al., 2014). We suggest that both q-PCR and microscopy are used to assess the response to therapy and to detect an eventual relapse early on. PJP relapse is seldom described in the literature. Only one out of 91 PJP patients had a severe relapse (1.1 %; 95 % CI 0.3-5.9 %) (Duncan et al., 1993;Kaczmarski et al., 1994;Matsuyama et al., 1998;Sinicco et al., 1996;Venzor et al., 1997;Zicklerova et al., 2012;Zonios et al., 2007). Adverse effects of TMP-SMZ were reported in 198 out of 1188 cases (16.7 %) (Helweg-Larsen et al., 2009). Pentamidine prophylaxis was prescribed after the severe pancytopenia due to TMP-SMZ, as microscopic observation of the BALF sample became negative for P. jirovecii at that time. Alternative therapies include a low dose of TMP-SMZ associated with caspofungin, dapsone plus trimethoprim, atovaquone or clindamycin administered with primaquine and pentamidine (Castro & Morrison-Bryant, 2010;Esteves et al., 2014;Helweg-Larsen et al., 2009;Tu et al., 2013). However, the development of new molecules against OIs having high efficacy and minimum side effects for CD4 cell recovery is required.
To decrease the number of potential OIs, it is essential to treat idiopathic CD4 + lymphocytopenia as has been done for HIV patients, using an appropriated antiviral therapy. With this aim, several treatments have been proposed for ICL patients such as IL-2, IL-7 and IFN-c (Régent et al., 2012, 2014Zonios et al., 2012). Bone-marrow transplantation may also be an optional treatment for normal recovery of CD4 + counts (Zonios et al., 2012). ICL patients may also have AIDS-related diseases, and in this case the guidelines and/or expert opinions concerning therapy and prophylaxis of opportunistic infections given for HIV patients should also be applied to ICL patients (Masur et al., 2014;Thoden et al., 2013).