The Clinical Characteristics and Risk Factors for Poor Prognosis Among HIV Patients with Talaromyces Marneffei Bloodstream Infection

Background Talaromyces marneffei(TM) bloodstream infection is common in AIDS patients with extreme immunodeciency in Southeast Asia and South China, however, clinical case study on TM bloodstream infection is scarce. We retrospectively analyze the clinical characteristics of TM bloodstream infection in hospitalized AIDS patients and determine the outcomes of hospitalization after diagnosis in our hospital over the past 5 years. Methods From January 2015 to July 2020, cases of TM detected by blood culture in patients admitted to our center were sorted and analyzed. The admission complaints, blood cells, biochemistry, CD4/CD8 count and 1,3-β-D-glucan, procalcitonin, CRP level on the day of blood culture test, and outcomes during hospitalization were analyzed. Logistic regression analysis was performed for the risk factors for poor prognosis. Spearman correlation analysis was used to analyze the correlation between peripheral blood cells, albumin and the time required for TM to become positive in blood culture. The difference was statistically signicant when the P value was <0.05. Results A total of 87 patients were collected, with a median age of 34 years, a median hemoglobin of 94 g/L and CD4 count of 7/µl. The rate of TM bloodstream infection among all in-hospital patients increased from 0.99% in 2015 to 2.09% in 2020(half year). Patients with TM bloodstream infection with a CD8 count <200/µl had a 12.6-fold higher risk of poor prognosis than those with a CD8 count >200/µl (p=0.04), and those with 1,3-β-D-glucan <100 pg/mL had a 34.9-fold higher risk of poor prognosis than those with 1,3-beta-D-glucan >100 pg/mL (p=0.01). TM bloodstream infection in AIDS patients in endemic areas. For those patients with extremely low CD4 and CD8 cell counts below 200/µl, the 1,3-β-D-glucan <100 pg/mL poor Spearman correlation analysis was performed with the time to TM positive conversion in blood culture as the dependent variable and the levels of peripheral leukocytes, neutrophils, hemoglobin, platelets, albumin and PCT, and 1,3-β-D-glucan as independent variables. The results showed that the levels of both platelets (r = 0.37, p = 0.0004) and albumin (r = 0.44, p < 0.0001) were positively correlated with the time required for TM blood culture conversion, and the levels of PCT (r= -0.40, p = 0.0009) and 1,3-β-D-glucan (r= -0.30, p = 0.0076) were negatively correlated with the time required for TM positive in blood culture.

Background AIDS patients are susceptible to multiple opportunistic infections at the late stage, especially when the CD4 count is below 50/µl [1,2]. The spectrum of bloodstream infection varies among people living with AIDS in different geographical settings [3]. Talaromyces marneffei (formerly Penicilliosis marneffei) is endemic in Southeast Asia [4], Northeastern India [5], and Southern China [6]. The rst natural human case of Talaromycosis (Penicillosis) was reported in 1973 and involved an American minister with Hodgkin's disease who lived in Southeast Asia [7]. TM was rst reported in mainland China by Z.L. Deng in 1984 [8] and in Vietnam by T.V. Hien in 2001 [9]. Recently, TM has become an emerging pathogen in immunocompromised patients in mainland China [7]; furthermore, the number of con rmed cases of TM is increasing quickly in mainland China [10]. The incidence rates of TM infection range from 4-14%, with an associated mortality of 10-30% [11]. Reports of TM disease are common in the AIDS population in the late stage. A meta-analysis [12] showed that the prevalence of TM among HIV patients in China ranged from 0.2% (95% CI: 0.1-0.5%) to 26.5% (95% CI: 16.2-43.5%). South China had the highest prevalence, estimated at 15.0% (95% CI: 11.0-20.4%), while Southwest China had the lowest prevalence, estimated at 0.3% [13]. It is estimated that there will be 4,951 TM cases per year in patients with AIDS in southern China by 2050, and the endemic areas are increasing [12]. The disease is mostly con rmed in the lungs and skin and is related to TM contact by inhalation or direct contact [14,15]. However, in clinical settings, because the onset of the disease is relatively insidious and mild, the clinical symptoms are not very typical, and patients often experience delayed diagnosis or misdiagnosis by doctors who are not familiar with the symptoms of TM disease. Although they are nally diagnosed with TM, there are multiple organ lesions, such as super cial lymphadenopathy, abdominal pain, hepatosplenomegaly, mediastinal lymph node enlargement, bone marrow involvement and hematopenia [16]. TM bloodstream infection is a kind of TM disease, especially in AIDS patients with extreme immunode ciency; however, clinical case studies on TM bloodstream infection are scarce. Our hospital is a referral center for AIDS patients in East China and many AIDS patients were admitted in hospital every year. The aim of this study was to retrospectively analyze the clinical characteristics of TM bloodstream infection in hospitalized AIDS patients and determine the outcomes of hospitalization after diagnosis in our hospital over the past 5 years.

Methods
The design and setting of the study Retrospectively collected from January 1st 2015 to July 31th 2020, cases of TM detected by blood culture in patients admitted to the Shanghai Public Health Clinical Center (SPHCC) were sorted and analyzed. Electronic medical records of HIV-infected patients diagnosed with TM bloodstream infection (de ned by a culture positive for TM from patient blood) were searched. Information including place of birth or long-term residence, presenting complaints and physical examinations (fever, umbilical fovea-like rash [17], and super cial lymphadenopathy), and laboratory tests (routine blood test, biochemistry, procalcitonin, CD4/CD8 T-cell count, CRP and outcomes during hospitalization) were analyzed. Blood culture tests were performed on patients with fever, loss of weight, lymphadenopathy or other symptoms need considering infection on the time of getting admission, and routine blood tests and tests for biochemistry, cellular immune function, procalcitonin and CRP were measured at the same time as blood culture. The patients were divided into a survival group and a poor prognosis group (death or having given up treatment due to advanced severe disease and being discharged from the hospital before death) according to their nal clinical outcomes during the hospital admission.
SPHCC is the designated hospital for HIV patients in Shanghai municipality and is also a tertiary referral hospital for refractory and complicated HIV-infected cases in China. Cultures of clinical specimens were established on Sabouraud's dextrose agar at 25 °C and 37 °C [18].
The characteristics of participants Among 8621 cases (from January 1st 2015 to July 31th 2020) admitted to SPHCC, there were 662 positive blood culture results and including 87 cases with TM bloodstream infection.

Statistical methods
The rank sum test was used for the comparison of demographics and clinical characteristic data between the two groups; chi-square analysis was used for the comparison of categorical variables. The predictors included in the multivariate model were selected based on a signi cance level of p < 0.1 in the univariate analyses. The confounding factors retained in the multivariate model were serum procalcitonin (> 0.60 ng/mL), CD8 count level (< 200/µl), 1,3-beta-D-glucan (< 100 pg/mL) and blood urea nitrogen level (> 4 mmol/L). Spearman correlation analysis was used to analyze the correlation between peripheral blood cells, albumin and the time required for TM to become positive in blood culture. The difference was statistically signi cant when the P value was < 0.05. Data analysis was conducted using IBM SPSS version 20.0 (IBM SPSS, Inc., Armonk, NY, USA).

Case distribution and trends in TM bloodstream infection
Among 8621 cases (from January 1st 2015 to July 31th 2020) admitted to SPHCC, the rate of TM bloodstream infection among all in-hospital patients was 13 (0.99%) in 2015, and in 2019, it grew to 18 (1.01%). Since July 2020, the rate was 2.09%. Because SPHCC is a referral hospital in China, all TM bloodstream infection cases were scattered throughout many provinces, Zhejiang (24 cases) and Jiangxi (13 cases) were the two provinces with the largest number of TM bloodstream infection cases in this analysis.
The comparison of clinical and laboratory features between the two groups with different clinical outcomes The median age of patients with TM bloodstream infection was 34 years, the median hemoglobin level was 94 g/L, and the median CD4 count was 7/µl. The median time required for TM culture conversion(from taking blood culture to TM identi ed ) was 8 days. The blood urea nitrogen level, procalcitonin and CRP in the poor prognosis group were signi cantly higher than those in the survival group; the peripheral blood CD8 count, 1,3-β-D-glucan, and serum albumin in the poor prognosis group were signi cantly lower than those in the survival group; the count levels of hemoglobin and CD4 in the poor prognosis group were lower than those in the survival group, but the difference was not signi cantly different. There was no signi cant difference in age or gender rate between the two groups. Among the 87 cases with TM bloodstream infection, there were 10 cases of NTM disease (6 cases of pulmonary NTM disease, 3 cases of NTM blood infection, 1 case of intestinal NTM disease); 10 cases of PCP; 9 cases of oral candidiasis; 8 cases of CMVR; 5 cases of combined tuberculosis infection, including 3 cases of tuberculosis, 1 case of lymphotuberculosis, 1 case of tuberculous pleurisy; and 1 case of cryptococcal meningitis. For patients with poor prognosis, the median length of hospital stay was 5 days, with a minimum of 1 day and a maximum of 22 days; for patients with normal discharge, the median length of hospital stay was 22 days (See Table 1).

Risk factors for TM bloodstream infection patients with poor prognosis
Of the 87 cases in the study, 27 cases with incomplete data (CD4, CD8 count, PCT, albumin) were excluded, and 60 cases remained (11 with poor prognosis and 49 survival cases). The age, peripheral blood cells, serum albumin, CD4 count level, CD8 count level, and the time required for blood culture conversion, and the procalcitonin, 1,3-beta-D-glucan, and blood urea nitrogen levels of the patients were analyzed by logistic test for risk factors for poor prognosis. Univariate analysis showed that serum procalcitonin (> 0.60 ng/mL), CD8 count level (< 200/µl), 1,3-beta-D-glucan (< 100 pg/mL) and blood urea nitrogen level (> 4 mmol/L) were risk factors for poor prognosis. The above four factors were included in the multivariate analysis, and the results showed that the risk of poor prognosis was 12.6 times higher in patients with CD8 count < 200/µl than in those with CD8 count > 200/µl (p = 0.04) and 34.9 times higher in those with 1,3-beta-D-glucan < 100 pg/mL than in those with 1,3-beta-D-glucan > 100 pg/mL (p = 0.01) (See Table 2).

Discussion
The number of TM bloodstream infection cases among in-hospital patients increased from 2015 (13 cases) to 2019 (18 cases) at SPHCC. The rate of TM bloodstream infection among admitted patients increased from 0.99% in 2015 to 2.09% in 2020 (from January to July). According to a systematic review by Qin et al. [13], the estimated pooled prevalence of TM infection in China was 3.3% (95% CI: 1.8-5.8%), and the prevalence in Shanghai was 1.8% (95% CI: 1.3-2.4%), which are in line with our ndings. As a referral center, we treat many kinds of opportunistic infections, including TM, and based on the etiology study of bloodstream infection among in-hospital patients in 2016 [3], TM bloodstream infection accounted for 18.8% (43/299), which was the second most common bloodstream infection pathogen.
With the progress of AIDS treatment and physician awareness of TM, reports of AIDS with TM infection have become increasingly common, and the prognosis depends on a timely diagnosis [19] and proper antifungal treatment [20,21]. Immunode cient patients from endemic areas should visit hospitals, and physicians should remain vigilant about TM symptoms and perform a specimen culture as early as possible and then initiate the antifungal regimen with Amphotericin B, Itraconazole or Voriconazole [21]. In this study, the mortality rate of patients with TM bloodstream infection was 17.2% (15/87), which is similar to 17.5% (191/1093) published by Jiang et al in 2019 [6] and 16.7% [18] in our hospital from 2014 to 2015. In view of the fact that this is a single-center study from a tertiary hospital [22] where patients from all parts of the country obtain further diagnosis and treatment, it cannot represent all the patients with TM disease.
In this study, the admission blood urea nitrogen level, procalcitonin, and CRP in the poor prognosis group were signi cantly higher than those in the survival group; the peripheral blood CD8 count level and 1,3-β-D-glucan level in the poor prognosis group were signi cantly lower than those in the survival group; in the overall presentation of patients with TM bloodstream infection, although they have extreme immunode ciency, poor prognosis patients are often in an extremely severe in ammatory state and have poor nutritional status compared with the survival patients. As a result, the patient's condition still deteriorated with symptomatic support treatment, and eventually, the condition worsened or the patient even died within a few days after admission. Among the 15 cases of worsening illness or death in this study, the median length of hospital stay was only 5 days, which means that at the time of worsening illness or death, TM disease often had not been con rmed, which also indicates that the immunode ciency caused by HIV infection requires early diagnosis and early intervention to avoid the development of severe immunode ciency [23].
The presence of TM bloodstream infection is a predictor of severe immunode ciency [4,24]. When they are in such an immunode cient state, patients often have coinfection, such as cytomegalovirus retinitis, pulmonary TB, NTM disease, and cryptococcal meningitis. [22,25]. There were fewer cases of coinfection in the group with poor prognosis. Considering that the group with poor prognosis had shorter overall hospitalization time and worsened or even died soon after admission, resulting in insu cient time for diagnosis, there may be missed diagnoses and fewer con rmed diseases than those in the group of surviving patients. Patients with poor prognosis are mostly in the late stage of the disease. Although they are given symptomatic supportive treatment after hospitalization, their condition fails to improve, and they lose the chance for antifungal treatment (or other comorbidities fail to control, and their condition is incurable). Therefore, the number of hospitalization days for patients with poor prognosis was signi cantly less than that of normal discharged patients.
In this study, risk factor analysis of patients with poor prognosis after clinical treatment for TM bloodstream infection showed that patients with CD8 count < 200/µl and whole blood 1,3-beta-D-glucan < 100 pg/mL had 12.6 times and 34.9 times the risk of poor prognosis, respectively. The median CD4 count was only 7/µl, which belongs to the extremely immunode cient population; the CD8 cell count was still signi cantly reduced, and the median count was only 202/µl. CD8 counts may predict prognosis independently of CD4 counts [26]. In most cases, HIV infection severely progresses, CD4 and CD8 are depleted [27]. This may explain why a delay in the diagnosis of TM independently predicted the early mortality of the patients [28,29].
For the 1,3-β-D-glucan (BDG) level, this study found that those with a BDG less than 100 pg/mL are at risk for poor prognosis. This unexpected nding may be explained by the fact that late diagnosis is the main reason for the high mortality in TM-infected patients [30]. As mentioned previously, the BDG assay is helpful in the early detection of invasive fugal disease [31], including TM [32,33]. The possible reason for this nding is that in clinical practice, using this marker as important evidence for fungal infection may cause delays in empirical diagnosis and treatment of patients with fever but low BDG levels; after all, blood culture results take time. Antifungal treatment may be initiated earlier in patients with high levels of BDG than in patients with low levels of BDG. Furthermore, in our analysis, when using the Spearman correlation, BDG was negatively correlated with the time required for blood culture to detect TM.
There are some limitations of this study. First, because this is a retrospective analysis, the clinical data, such as treatment regimen administered before admission in our hospital, records of patients' visit experience, as well as the time from onset of symptoms to diagnosis, all affect the accuracy of descriptions and analysis. Second, this was a single-center study, and caution should be paid to making a summary or comparison with the population in areas with a high incidence of TM disease in China.

Conclusions
Bloodstream infections caused by TM are becoming increasingly common in the AIDS population with severe immunode ciency in East China. For those with extremely low CD4 and CD8 cell counts below 200/ul, the 1,3-β-D-glucan < 100 pg/mL are with an increased risk of poor prognosis. Improving the awareness of symptoms such as umbilical fovea rash, fever and lymphadenopathy has a positive effect on early diagnosis of the disease and optimization of treatment e cacy.

Declarations
Ethics approval and consent to participate Ethical approval was granted by the Ethics Committee of Shanghai Public Health Clinical Center (Ethics approval No. 2017-S022-04). This study used retrospective and anonymous data collection methods, which did not involve patient privacy, so informed consent was exempted from ethical review.
Consent for publication Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its supplementary information les(S1 File).
Competing interests The authors declare that they have no competing interests.