Prognostic Nomogram for Death from Pneumocystis Pneumonia in Non-HIV-and HIV-Infected Patients

Background: Pneumocystis pneumonia is a major cause of death in immunocompromised patients. Many risk factors for poor prognosis have been reported, but few studies have created predictive models with these variables to calculate the death rate accurately. This study created nomogram models for the precise prediction of mortality risk in non-human immunode�ciency virus (NHIV)- and human immunode�ciency virus (HIV)-infected patients with Pneumocystis jirovecii pneumonia (PJP). Methods: A retrospective study was performed over a 10-year period to evaluate the clinical characteristics and outcomes of NHIV-PJP at Beijing Chaoyang Hospital and HIV-PJP at Beijing Ditan Hospital in China from 2010 to 2019. Univariate and multivariate logistic regression analyses were used to screen out mortality risk factors to create the nomograms. Nomogram models were evaluated by using a bootstrapped concordance index, calibration plots and receiver operating characteristic (ROC) curves. Results: A total of 167 NHIV-PJP patients and 193 HIV-PJP patients were included in the study. Pneumothorax, febrile days after admission, CD4+ T cells ≤ 100/µl and sulfa combined with caspofungin (CAS) treatment were identi�ed as independent risk factors for death that could be combined to accurately predict mortality risk in NHIV-PJP patients. We created a nomogram for mortality by using these variables. The area under the curve was 0.865 (95% con�dence interval 0.799-0.931). The nomogram had a C-index of 0.865 and was well calibrated. The independent risk factors for death in HIV-PJP patients included in the nomogram were pneumothorax, platelet (PLT) ≤ 80×10 9 /L, haemoglobin (HGB) ≤ 90 g/L, albumin (ALB), cytomegalovirus (CMV) coinfection and sulfa combined with CAS treatment. The nomogram showed good discrimination, with a C-index of 0.904 and excellent calibration. The area under the


Background
Pneumocystis jirovecii pneumonia (PJP), also known as interstitial plasma cell pneumonia, is a fungal infection of the respiratory system caused by Pneumocystis jirovecii (PJ).As one of the most common opportunistic infections in acquired immunode ciency syndrome (AIDS) patients, it is a major cause of morbidity and mortality in them [1].In recent years, with the widespread application of glucocorticoids and cytotoxic drugs, the fast development of tumour chemoradiotherapy, and the frequency of connective tissue diseases and various organ transplantation, the incidence of PJP in non-AIDS immunosuppressed patients has signi cantly increased [2].PJP typically presents with acute and rapid progressive respiratory insu ciency.It has higher mortality in non-HIV patients than in HIV patients (30-60% versus 10-20%) [3,4].Many risk factors for poor prognosis have been reported [5], but it is di cult to predict the individual death rate accurately.More e cient prediction tools for estimating the prognosis of PJP patients are needed.
Nomograms are graphical models that enable users to calculate the overall probability of a speci c clinical outcome for an individual patient [6,7].There are many nomograms used as prediction tools in various diseases, such as cancer [8].The nomogram facilitates the clinical implementation and probability calculation of risk factors or other predictor variables.The objective of this study was to combine clinical manifestations, treatment variables, and laboratory variables that were associated with death into two prediction nomograms: We developed and validated nomograms that predicted death risk in NHIV and HIV patients.

Study Patients
We conducted a retrospective study to collect clinical data from Beijing Chaoyang Hospital and Beijing Ditan Hospital, Capital Medical University, and the study protocol was approved by the research ethics committee of the hospital.Because of the retrospective nature of the observational study, which lacked interventional aspects, informed written consent was waived by the ethics committee.
We retrospectively collected the data of these patients who were con rmed to have PJP and were hospitalized for the rst time between 1 January 2010 and 31 December 2019 at either of the two hospitals of Capital Medical University, Beijing Chaoyang Hospital and Beijing Ditan Hospital, both tertiary care university hospitals in Beijing, China.By screening the eligible adult patients (age ≥ 18 years) from a computerized medical chart search system by ICD-10 (International Classi cation of Diseases, 10th revision) code, their medical records were reviewed, and data were extracted and registered in the research forms.All of these important data were entered into Excel for preservation.We de ned con rmed PJP via three criteria: (1) clinical symptoms, such as fever, dry cough (occasionally expectorant) and progressive dyspnoea; (2) abnormal imaging ndings: CT showed a broad range of features, from ground-glass opacity to nodules, cysts, patchy shadows and diffuse interstitial in ltrates; (3) a positive result for Pneumocystis jirovecii by Gomori-Grocott or toluidine blue stain or positive immuno uorescence test results in an induced sputum [9], low-tracheal-aspiration or bronchoalveolar lavage uid (BALF) specimen.We did not include patients who only had positive PCR results.
De nite PJP cases with a rst episode were included.Exclusion criteria: current pregnancy, allergy to sulfa drugs, and less than one week of hospitalization.The diagnosis of HIV/AIDS according to the Centers for Disease Control and Prevention (CDC) classi cation was based on Western blot conducted by the CDC to detect HIV-1 antibody positivity [10].

Data collection
The electronic medical charts for each enrolled patient were reviewed to obtain the following information: demographical data, underlying diseases, use of immunosuppressive drugs, clinical manifestations, radiology characteristics, laboratory tests, treatment and hospital mortality.Laboratory data used for analysis were those recorded within 48 hours or the worst values after admission.Febrile days meant the unbroken duration of a daily temperature > 37.5°C after admission until discharge or death.
Statistical Analysis SPSS 20.0 (SPSS Institute, Chicago IL, USA) statistical software was used to perform statistical analysis.Nomogram models were created with R software (version 4.0.3;http://www.Rproject.org)using the "rms" package.Measurement data with a normal distribution are expressed as mean ± SD, while measurement data with a nonnormal distribution are expressed as median (interquartile range).Count variables are expressed as a percentage (%).The independent sample T-test was used for the continuous numerical variables obeying a normal distribution.Continuous variables that did not follow a normal distribution or graded variables were tested using the Mann-Whitney U test, and categorical variables were compared by using the Pearson chi-square test.P < 0.05 was statistically signi cant.The variables with statistical signi cance were selected for the binary logistic regression analysis of prognosis, and the variables with P < 0.05 in the univariate regression analysis were incorporated into the multivariate regression analysis model.To identify independent predictive factors of in-hospital mortality in a multivariate logistic regression analysis model, nomograms for hospital mortality risk were created based on the multivariate logistic regression model.The performance of the nomogram was evaluated through the concordance index and calibration plots with bootstrap samples.

Results
A total of 622 adult patients with a rst episode of PJP were screened in the computer system from 2010 to 2019.Among these, 202 patients had probable PJP without microbiological results.Sixty patients were excluded: 26 for a sulfa drug allergy and 34 for hospitalization less than 1 week.Finally, 360 patients were included in the study: 167 patients in the NHIV-PJP group and 193 patients in the HIV-PJP group.

Nomogram for mortality prediction
We investigated the association between clinical factors and all-cause mortality by univariate analysis in both groups.Febrile days after admission, PLT ≤ 80(x10 9 /L), HGB ≤ 90 g/L, CD4 + T cells ≤ 100/µl, PCT, LDH ≥ 500 U/L, ALB, CMV co-infection, EBV co-infection, pneumothorax, sulfa combined with CAS, ICU days, and ECMO were signi cantly associated with mortality in the NHIV-PJP group (Table 3).We then performed multivariate logistic regression analysis with these associated factors.We identi ed febrile days after admission, CD4 + cells ≤ 100/µl, pneumothorax and sulfa combined with CAS as independent risk factors.We found that a combination of these factors most precisely predicted mortality (Table 4).
We then created a nomogram for mortality including all of these factors (Fig. 1).The area under the curve (AUC) was 0.865 (95% con dence interval 0.799-0.931;Fig. 2).The nomogram had a bootstrapped concordance index of 0.865 and was well calibrated (Fig. 3).In the same way, we created a nomogram for mortality in the HIV-PJP group (Fig. 4).The area under the curve (AUC) was 0.910 (95% con dence interval 0.850-0.970;Fig. 5).The nomogram had a bootstrapped concordance index of 0.904 and was well calibrated (Fig. 6).

Discussion
To our knowledge, this is the rst study to create predictive nomogram models to calculate mortality risk precisely in PJP patients.This retrospective study describes the clinical characteristics and outcomes of NHIV-PJP cases and HIV-PJP cases.The results show that the PJP populations suffering from HIV and non-HIV immunosuppression are different depending on their baseline data.These HIV-negative patients were older than those in studies of HIV-positive patients, as found in a previous report [11].Coinfections, most notably with viruses, especially CMV coinfection, were considerably more prevalent among NHIV-PJP patients than HIV-PJP patients in our study, which was consistent with published ndings [12].PJP usually presents with atypical symptoms, such as fever, dry cough, and dyspnoea, occurring in up to 86%, 76%, and 81% of cases, respectively [13].In our study, the fever rates in the HIV-PJP and NHIV-PJP groups were 86% and 89.8%, respectively.In the multivariate regression analysis, febrile days after admission was independently associated with death in NHIV-PJP patients.The results suggested that continuous fever as a predictive factor could enable clinicians to recognize the mortality risk of PJP earlier and avoid further deterioration of the patient's condition.
The main risk factors for immunosuppression in our study were drug-related immunosuppression and transplantation, which are obviously related to de ciencies in cellular immunity.Our study shows that almost 1/3 of PJP patients were renal transplant recipients, and 141 (84.4%) patients of the NHIV group had a low level of CD4 + T cells.This largely correlates to those patients who became organ transplant recipients, who remain at risk for PJP for many years after transplantation [2], but fewer recipients accepted TMP-SMX for PJP prophylaxis.On the other hand, these recipients took hormones and cytotoxic drugs simultaneously, which aggravated their immunode ciency.Glucocorticoid treatment is a wellknown risk factor for PJP in non-HIV patients and accounts for 55-97% of published cases [14,15], including 88.0% in our study.The mechanism could be a decrease in peripheral CD4 + T cells caused by the glucocorticoid therapy [14].Immunosuppressive agents such as thiopurine could reduce the absolute numbers of lymphocytes by inhibiting cell proliferation.Tacrolimus and cyclosporine can inhibit lymphocyte activation, and cytokines can inhibit lymphocyte function.
The main radiologic features of PJP identi ed through CT scanning are extensive ground-glass opacity (GGO) and reticulation [16,17].In our study, the rate of GGOs was higher in the non-HIV-PJP group (59.9%) and in the HIV-PJP group (88.1%) than other radiographic features.Pneumothorax is an unresolved problem in PJP because PJ has enough time to grow in the subpleural spaces and is thus di cult to eradicate by treatment [18].In adults, the incidence of pneumothorax ranges from 4-36% [19].
When barotrauma occurs, it usually indicates a poor prognosis and a high mortality rate of 50%-100% [20][21][22].The pneumothorax rates were 10.2% and 4.7% in our two groups, respectively.Nearly all patients with a poor prognosis developed pneumothorax.In particular, the NHIV-PJP group had a higher rate, and the difference was statistically signi cant.In the multivariate regression model, pneumothorax was an independent risk factor for mortality in both groups.
The auxiliary examination showed that the lymphocyte count, CD4 + T cells, serum ALB, and oxygenation index were below normal and that (1,3)-β-D-glucan and lactate dehydrogenase (LDH) levels were elevated in PJP patients.WBC count was normal in these patients.These ndings are consistent with other studies [23].Previous studies demonstrated that hypoalbuminaemia had a positive correlation with increased lung injury and could be a signi cant indicator of death in critically ill patients [24,25].In our study, the mean ALB level was higher in the HIV-PJP group than the NHIV-PJP group, but both were below normal.
We showed that serum albumin was a signi cant independent factor of poor prognosis in only the HIV-PJP group and not in the NHIV-PJP group.This nding was similar to that of Kim et al., who showed that hypoalbuminaemia was not considered an independent predictor of mortality [26].These results suggest that ALB level might be a predictive factor for the prognosis of PJP patients.Overall, these results re ect that treatment strategies for HIV-PJP patients should be chosen with the awareness that the serum albumin level can be a warning of the risk of death.A low oxygenation index has also been associated with poor outcomes in PJP patients with immunosuppressive disease [27].In our study, the oxygenation index in the NHIV-PJP and HIV-PJP groups was 287.57± 119.28 vs. 310.78± 100.68, respectively.A lower oxygenation index, which was also representative of ventilation-perfusion abnormalities, was related to death in both groups.
A decrease in haemoglobin indicates anaemia, and haemoglobin below 90 g/L is moderate to severe anaemia.Anaemia can also be de ned as a lowered ability of the blood to carry oxygen.Therefore, we thought that anaemia might cause a worse prognosis in PJP patients.We found that HGB ≤ 90 g/L was an important risk factor for mortality in the two groups and an independent predictor of death in HIV-PJP patients.These results were consistent with a previous study [28], which indicated that timely discovery and correction of anaemia were necessary in patients with PJP.
Despite being associated with intolerance and adverse events, TMP-SMZ is still the rst-line therapeutic regimen for PJP [9,29,30].In our study, 55.1% of NHIV-PJP patients and 88.6% of HIV-PJP patients were initially treated with TMP-SMZ within 24 hours, but early anti-PCP treatment did not improve clinical prognosis [31].Caspofungin was recognized as a second-line regimen.Its drug class, the echinocandins, were reported to inhibit the enzyme 1,3-β glucan synthase and improve overall mortality in patients with AIDS-PJP [32].Other studies reported failure of echinocandin salvage therapy to improve survival among non-AIDS patients [33,34].Likewise, we found that sulfa combined with CAS treatment was an independent risk factor for death in multivariate analysis in both groups.This outcome reminds us to be careful about combining these two drugs for PJP treatment.
The overall mortality rate in NHIV-PJP patients is 31%, but it rises to almost 100% when PJP is not properly and quickly treated [35,36].In several studies, PJP was more often fatal in non-HIV-infected patients than in HIV-positive patients [11,20].We also observed that the mortality rate of 29.3% in the non-HIV group was higher than the 18.1% in the HIV group; previous rates on the order of 30-60% non-HIV patients and 10-20% in HIV patients have been reported.However, several studies have also reported mortality rates in the range of 7 to 14% [37,38].To calculate the precise mortality risk of PJP patients, we constructed a nomogram model incorporating clinical parameters and auxiliary examinations.
This study has some limitations.First, it is a retrospective study of two centres with a small population.
Retrospective studies may be biased in terms of the data collected, such as physical examination data and the normal ranges in lab tests.A larger, prospective study is necessary.Second, the underlying diseases/conditions were mostly kidney transplants in the NHIV group, which is not very representative.Third, it only includes PJP patients who were hospitalized for more than 7 days.Some patients were not enrolled in this study because they left within 7 days after admission for any reason.

Conclusions
Our nomogram models provide a useful, convenient and applicable tool to evaluate the individualized prognosis of mortality in NHIV-PJP and HIV-PJP patients.
Receiver operating characteristic curve for the prediction model of the NHIV-PJP group.area under the curve was 0.865 (95% con dence interval 0.799-0.931) Nomogram for mortality in HIV-PJP group.

Figures
Figures

Table 3
Prognostic factors in a univariate regression analysis in patients with PJP