Available via license: CC BY 4.0
Content may be subject to copyright.
Page 1/24
Establishment of a risk classier to predict the in-
hospital death risk of nosocomial infections caused
by fungi in cancer patients
Ruoxuan Wang
The First Aliated Hospital of Xi'an Jiaotong University
Aimin Jiang
The First Aliated Hospital of Xi'an Jiaotong University
Rui Zhang
Baoji Traditional Chinese Medicine Hospital
Chuchu Shi
The First Aliated Hospital of Xi'an Jiaotong University
Qianqian Ding
The First Aliated Hospital of Xi'an Jiaotong University
Shihan Liu
The First Aliated Hospital of Xi'an Jiaotong University
Fumei Zhao
The First Aliated Hospital of Xi'an Jiaotong University
Yuyan Ma
The First Aliated Hospital of Xi'an Jiaotong University
Junhui Liu
The First Aliated Hospital of Xi'an Jiaotong University
Xiao Fu
The First Aliated Hospital of Xi'an Jiaotong University
Xuan Liang
The First Aliated Hospital of Xi'an Jiaotong University
ZhiPing Ruan
The First Aliated Hospital of Xi'an Jiaotong University
Yu Yao
The First Aliated Hospital of Xi'an Jiaotong University
Tao Tian ( taintao0607@163.com )
The First Aliated Hospital of Xi'an Jiaotong University
Research Article
Page 2/24
Keywords: cancer patients, nosocomial infections, fungal infections, risk factors, in-hospital mortality,
nomograms
Posted Date: January 19th, 2023
DOI: https://doi.org/10.21203/rs.3.rs-2486032/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
Page 3/24
Abstract
(1) Background:
Patients with malignancy are more vulnerable to developing nosocomial infections. Limited studies
investigated cancer patients' clinical features and prognostic factors of fungi infections. Herein, this
study was performed to explore the clinical characteristics of nosocomial infections due to fungi and
develop a nomogram to predict the in-hospital death risk of these patients.
(2) Methods: This retrospective observational study analyzed cancer patients with nosocomial infections
caused by fungi from September 2013 to September 2021. The univariate and multivariate logistics
regression analyses were utilized to identify the inuencing factors of in-hospital death risk of
nosocomial infections caused by fungi. A nomogram was developed to predict the in-hospital death risk
of these individuals, with the receiver operating characteristics curve (ROC), calibration curve, and
decision curve being generated to evaluate its performance.
(3) Results: 216 patients with solid tumors developed fungal infections during hospitalization, of which
57 experienced in-hospital death. C.albicans is the most common fungal species(68.0%). The respiratory
system was the most common site of infection(59.0%), followed by intra-abdominal infection (8.8%). The
multivariate regression analysis revealed that ECOG-PS 3–4, pulmonary metastases, thrombocytopenia,
hypoalbuminemia, and mechanical ventilation were independent risk factors of in-hospital death risk. A
nomogram was constructed based on the identied risk factors to predict the in-hospital death risk of
these patients.
(4) Conclusions: Fungi-related nosocomial infections are common in solid tumors and have a bleak
prognosis. The constructed nomogram could help oncologists make a timely and appropriate clinical
decision with signicant net clinical benet to patients.
1. Introduction
Cancer patients are predisposed to developing nosocomial infections due to immunosuppressive caused
by malignancy and long-term antitumor treatment. [1, 2]. Besides, routine diagnostic and therapeutic
procedures, especially invasive operations such as tissue biopsy and catheter placement, signicantly
increase the risk of nosocomial infections in cancer patients[3]. It is reported that surgery is closely related
to the occurrence of nosocomial infections in these individuals[4]. Therefore, nosocomial infections have
become one of the most common complications in patients with tumors. Once a severe infection occurs,
it will undoubtedly affect the initiation of antitumor treatment, prolong the length of hospitalization,
increase healthcare-related costs, and lead to the death of patients in severe cases. As a result, infection
has become the leading non-cancer cause of death in cancer patients[5, 6]
Page 4/24
The clinical features, microbiological distribution, and prognostic factors of nosocomial infections
caused by bacteria are well documented[1, 2, 7–11]. Most importantly, relevant guidelines are also published
to guide the diagnosing and treatment of nosocomial infections caused by bacteria in patients with
malignancy[12–14]. Unlike bacterial pathogens, fungi usually do not produce endotoxins and exotoxins.
The pathogenicity caused by fungi may be related to mechanical damage caused by their reproduction in
the body, as well as the type of enzyme production and acidic metabolites[15]. Patients with malignant
tumors are at high risk of fungal infections due to impaired immune function. In this context, invasive
fungal disease (IFD) will occur in severe cases [16]. Furthermore, the subsequent long-course intervention
of antifungal therapy malnutrition, and secondary infections will further increase the risk of in-hospital
death of these patients [17].
However, in actual clinical work, there are still tumor patients with a fungal infection, but the degree of
infection is not up to the diagnostic criteria for IFD. There is insucient statistical evidence and relevant
guidelines for these patients' infection characteristics, treatment modes, and prognostic factors.
Therefore, it is vital for clinical practice to understand the clinical features and epidemiological
characteristics of solid tumors complicated with fungal infections in hospitals. Meanwhile, there is no
available risk model to predict these patients' prognoses robustly. Therefore, we conducted this
retrospective study to explore the clinical characteristics and prognostic factors of nosocomial infections
caused by fungi in cancer patients and to develop a nomogram to predict the in-hospital mortality of
these patients.
2. Methods
2.1 Study population and design
We conducted this single-center retrospective observational cohort study at the First Aliated Hospital of
Xi'an Jiaotong University in the Northwest of China. Searching for ICD-10 coded diagnoses included
patients with solid tumors with in-hospital fungal infections who received medical care during their
hospitalization from September 2013 to September 2021. This study included patients who met all the
following criteria: 1) age ≥ 18 years; 2) laboratory conrmed diagnosis of infection caused by fungi; 3) a
solid tumor was conrmed by histological pathology or cytological pathology; and 4) patients
hospitalized during the study period with complete electronic medical records (EMR). Patients under 18
years old or without total medical records were excluded. The study was approved by the ethics
committee of the First Aliated Hospital of Xi'an Jiaotong University (No: XJTU1AF2020LSK-049).
Waiving of informed consent was obtained due to the retrospective noninterventional study design.
2.2 Data collection
All data were manually extracted from the EMR and recorded in the Excel of Microsoft. The extracted data
included age, gender, smoking history, Eastern Cooperative Oncology Group (ECOG) Performance Status,
TNM staging, tumor type, and records of distant metastases. Information related to the infection was
Page 5/24
collected simultaneously, including the primary site of infection, the fungal species, whether combined
with bacterial infection and time and types of intravenous antifungal drugs. Other information was also
collected, such as fever, antitumor therapy within 30 days (including but not limited to surgery,
chemotherapy, immune checkpoint inhibitor therapy, and radiotherapy), corticosteroid therapy in the past
30 days, Granulocyte colony-stimulating factor (G-CSF) usage within 30 days, antibiotic therapy within 30
days, invasive procedures in the last 30 days, intensive care unit (ICU) admission during hospitalization,
the experience of septic shock, mechanical ventilation, and outcome after fungi infection (death or
discharge). The worst values of laboratory parameters before infection diagnosis, including blood routine
results (hemoglobin counts, platelet counts, and the white blood cell and differential count), serum
albumin, and electrolyte levels were collected in terms of laboratory indicators.
2.3 Denition
Nosocomial infection caused by fungi was considered if the patient had the following criteria: (a) on the
premise of excluding contamination of clinical specimens, the results of fungi culture indicated that at
least one pathogen was positive (> 48h after hospital admission); (b) there were corresponding clinical
manifestations, laboratory examination results, or radiological results recorded which is in electronic
medical records; or (c) a clear infection type record that was acquired from the electronic medical records.
Otherwise, the case is considered community-onset [18–20]. Clinical samples such as sputum, urine, blood
culture, stool, wounds secreta, ascites, pleural, drainage uid postoperation, and other samples were
collected once patients were suspected of fungal infection. Fever was considered an axillary temperature
of 38.3°C on one occasion or a temperature of > 38.0°C on two or more occasions during 12h [21]. The
shock was referred to as systolic blood pressure90mmHg, and uid therapy and/or vasoactive
medications have no improvement in this condition[20].
2.3 Study outcomes
This study aimed to characterize the clinical features, microbial prole, and prognostic factors of
nosocomial infections caused by fungi in patients with solid tumors and to develop a predictive model to
predict their in-hospital death risk. Thus, in-hospital mortality was the primary outcome of this study. It is
worth noting that only death cases caused by nosocomial infections during hospitalization were selected
for in-hospital mortality calculation.
2.4 Statistical analysis
The extracted clinical data were recorded in a standardized form and compared according to the patient's
survival status after infection during hospitalization. As appropriate, continuous variables were
summarized as means and standard deviation (SD) or median and interquartile range (IQR). Categorical
variables were expressed as frequency and percentage. Continuous variables were analyzed by an
independent-sample t-test or a Mann-Whitney U-test. Categorical variables were analyzed by Chi-Square
Test or Fisher's Exact Test. Univariate and multivariate logistic regression analyses were adopted to
investigate the independent risk factors for in-hospital mortality of nosocomial infections. Variables for
the p-value < 0.05 for the univariate analysis were included in the multivariable logistic regression
Page 6/24
analysis. A two-sided p-value < 0.05 was considered statistically signicant. The nomogram was
constructed based on independent factors identied in the multivariate analysis to predict the probability
of in-hospital death after nosocomial fungal infection. Besides, the receiver operating characteristics
curve (ROC), calibration curve, and decision curve were employed to evaluate its performance. All
statistical analyses were performed in R software (version 4.1.3) for windows 64.0.
3. Results
3.1 The essential characteristics of the participants
A total of 216 patients with solid tumors were diagnosed with nosocomial infections and received
complementary treatment in the First Aliated Hospital of Xi'an Jiaotong University during the eight
years of the study period (Fig.1). Among them, 138 were males (64%), and 78 were females (36%). The
median age was 65-year-old. 90% of patients had an ECOG-PS 0–2, and 74% had a TNM stage of III-IV.
The common diagnoses were respiratory tumors (34%), gastrointestinal tumors (24%), and hepatobiliary
and pancreatic tumors (24%). Regarding the detailed antitumor therapy, 72 patients (33.6%) underwent
surgery, 62 patients (29%) received chemotherapy, and 13 patients (6%) received immune checkpoint
blockade (ICB) treatment within 30 days, respectively. A total of 69 patients (32%) received
glucocorticoids within 30 days. In the past 30 days, 78 patients (36%) received simultaneous antibacterial
therapy (Table1).
Page 7/24
Table 1
The general characteristics of solid patients with fungal infection
Variable Overall, N = 2161Survival, N =
1591Death, N = 571p-
value2
Demographic data
Age (years) 65 (58, 71) 65 (59, 71) 65 (58, 71) >
0.900
Gender 0.200
male 138 (64%) 98 (62%) 40 (70%)
female 78 (36%) 61 (38%) 17 (30%)
Smoking history (Yes) 97 (45%) 66 (42%) 31 (54%) 0.094
Days of
hospitalization(days) 17 (9, 27) 17 (10, 27) 16 (8, 26) 0.400
ECOG-performance status <
0.001
0,1,2 194 (90%) 153 (96%) 41 (72%)
3,4 22 (10%) 6 (3.8%) 16 (28%)
TNM stage 0.014
Stage I-II 57 (26%) 49 (31%) 8 (14%)
Stage III-IV 159 (74%) 110 (69%) 49 (86%)
Underlying cancer type
Head and neck cancer 7 (3.2%) 6 (3.8%) 1 (1.8%)
Lung cancer 73 (34%) 50 (31%) 23 (40%)
Esophago-gastrointestinal
cancer 35 (16%) 26 (16%) 9 (16%)
Colon and rectal cancer 17 (7.9%) 10 (6.3%) 7 (12%)
Hepatobiliary and
pancreatic cancer 52 (24%) 42 (26%) 10 (18%)
Breast cancer 6 (2.8%) 5 (3.1%) 1 (1.8%)
Genitourinary cancer 8 (3.7%) 4 (2.5%) 4 (7.0%)
Gynecological cancer 10 (4.6%) 9 (5.7%) 1 (1.8%)
Lymphoma 4 (1.9%) 4 (2.5%) 0 (0%)
Others 4 (1.9%) 3 (1.9%) 1 (1.8%)
Page 8/24
Variable Overall, N = 2161Survival, N =
1591Death, N = 571p-
value2
Distant metastasis
Liver metastasis 32 (15%) 26 (16%) 6 (11%) 0.300
Lung metastasis 31 (14%) 18 (11%) 13 (23%) 0.034
Brain metastasis 10 (4.6%) 5 (3.1%) 5 (8.8%) 0.130
Bone metastasis 40 (19%) 24 (15%) 16 (28%) 0.030
Other metastasis 41 (19%) 29 (18%) 12 (21%) 0.600
CCI score 0.012
0–3 207 (96%) 156 (98%) 51 (89%)
> 3 9 (4.2%) 3 (1.9%) 6 (11%)
Operation type (within 30
days) 0.017
Unoperated 144 (67%) 99 (62%) 45 (79%)
Curative operation 60 (28%) 52 (33%) 8 (14%)
Palliative operation 12 (5.6%) 8 (5.0%) 4 (7.0%)
Prior treatment (within 30
days)
Chemotherapy 62 (29%) 47 (30%) 15 (26%) 0.600
Radiotherapy 20 (9.3%) 14 (8.8%) 6 (11%) 0.700
Concurrent
chemoradiotherapy 11 (5.1%) 8 (5.0%) 3 (5.3%) >
0.900
Perfusion therapy 11 (5.1%) 10 (6.3%) 1 (1.8%) 0.300
Immunotherapy 13 (6.0%) 7 (4.4%) 6 (11%) 0.110
Targeted therapy 15 (6.9%) 8 (5.0%) 7 (12%) 0.075
Glucocorticoid therapy 69 (32%) 53 (33%) 16 (28%) 0.500
G-CSF usage 47 (22%) 35 (22%) 12 (21%) 0.900
Antibiotic usage 78 (36%) 52 (33%) 26 (46%) 0.082
Laboratory indexes
Hemoglobin(g/L) 103 (90, 120) 106 (93, 120) 98 (85, 115) 0.041
< 110 131 (61%) 94 (59%) 37 (65%) 0.400
Page 9/24
Variable Overall, N = 2161Survival, N =
1591Death, N = 571p-
value2
Platelet count (×109/L) 176 (111, 252) 197 (132, 266) 117 (58, 210) <
0.001
< 100 50 (23%) 26 (16%) 24 (42%) <
0.001
Leucocyte count (×109/L) 8.0 (5.4, 11.2) 8.1 (5.4, 11.1) 7.9 (5.1, 11.9) 0.700
< 4.0 33 (15%) 26 (16%) 7 (12%) 0.500
Neutrophils(×109/L) 6.4 (3.7, 9.3) 6.3 (3.5, 9.2) 6.8 (4.3, 10.8) 0.300
Lymphocyte(×109/L) 0.82 (0.52, 1.11) 0.84 (0.52,
1.14) 0.71 (0.48, 1.06) 0.140
Monocyte(×109/L) 0.42 (0.25, 0.71) 0.42 (0.25,
0.72) 0.40 (0.23, 0.68) 0.800
Albumin(g/L) 30.9 (28.2, 35.0) 31.8 (28.8,
36.0) 29.0 (25.8, 31.2) <
0.001
< 30 90 (42%) 55 (35%) 35 (61%) <
0.001
Serum calcium(mmol/L) 2.06 (1.95, 2.20) 2.09 (1.98,
2.21) 2.01 (1.89, 2.12) 0.006
< 2.0 144 (67%) 114 (72%) 30 (53%) 0.009
Serum corrected
calcium(mmol/L) 2.28 (2.20, 2.38) 2.28 (2.21,
2.38) 2.27 (2.18, 2.40) 0.800
Serum sodium(mmol/L) 138.0(134.0,140.1) 138.4(135.6,
141.0) 135.3(131.0,139.0) 0.020
< 130 61 (28%) 36 (23%) 25 (44%) 0.002
Abbreviations: ECOG Eastern Cooperative Oncology Group, CCI Charlson Co-morbidity Index score, G-
CSF granulocyte colony-stimulating factor
1n (%); Median (IQR),2Pearson's Chi-squared test; Wilcoxon rank sum test; Fisher's exact test
3.2 Data on infection in cancer patients with nosocomial
infections caused by fungi
We reviewed all clinical data on nosocomial infections caused by fungi of the participants. Thirty-four
patients (16%) had a history of previously known infection within 30 days. Respiratory tract infection was
the most predominant primary infection type, accounting for 59% of cases, followed by celiac infections
(8.8%). During hospitalization, 140 patients (65%) received intravenous antifungal therapy. Of these
people, 122 patients (56%) received triazole antifungal drugs, followed by echinocandin antifungal drugs
Page 10/24
(5.1%). At the same time, 3.2% of patients received two or more intravenous antifungal drugs. C. Albicans
were the predominant pathogens (68%), followed by other Candida species (19%). Two patients (0.9%)
were complicated with two or more fungal infections. Of all patients, 135 patients have undergone
invasive procedures in the last 30 days, with indwelling catheterization being the most common (28%). In
addition, 43 patients (20%) were admitted to the ICU, and 29 (13%) were mechanically ventilated.
3.3 Comparison of clinical and infection-related characteristics in the study population based on the
survival status of patients during hospitalization
We used data on nosocomial mortality to assess the primary clinical outcomes of nosocomial infections
caused by fungi in patients with solid tumors. The study participants' overall fatality rate was 26.4%
(57/216). We also analyzed the relationship between these patients' prognoses and clinical features. The
results showed that ECOG-PS, TNM stage, pulmonary metastases, liver metastases, CCI, surgery or
chemotherapy within 30 days, laboratory results (platelet count, serum albumin level, serum calcium, and
serum sodium levels) were varied (p < 0.05; Table1). Meanwhile, the two groups' body temperature,
antifungal therapy, immunoglobulin therapy, admission to ICU, mechanical ventilation, and type of sepsis
varied (p < 0.05; Table2).
Page 11/24
Table 2
The Infection-related characteristics of solid patients with fungal infection
Variable Overall, N =
2161
Survival, N =
1591
Death, N =
571
p-
value2
Primary sites of infection 0.500
Respiratory tract 128 (59%) 93 (58%) 35 (61%)
Digestive tract 15 (6.9%) 11 (6.9%) 4 (7.0%)
Urinary tract 11 (5.1%) 7 (4.4%) 4 (7.0%)
Thoracic cavity 5 (2.3%) 5 (3.1%) 0 (0%)
Enterocoelia 19 (8.8%) 16 (10%) 3 (5.3%)
Temperature(≥ 38℃) 68 (31%) 43 (27%) 25 (44%) 0.019
Infection history (within 30 days) 34 (16%) 23 (14%) 11 (19%) 0.400
FN history (within 30 days) 3 (1.4%) 2 (1.3%) 1 (1.8%) > 0.900
Invasive procedure (within 30
days) 135 (62%) 104 (65%) 31 (54%) 0.140
Biliary stent implantation 7 (3.2%) 7 (4.4%) 0 (0%) 0.200
Ureteral stent implantation 3 (1.4%) 2 (1.3%) 1 (1.8%) > 0.900
Indwelling urinary catheter 60 (28%) 46 (29%) 14 (25%) 0.500
PICC 16 (7.4%) 12 (7.5%) 4 (7.0%) > 0.900
Infusion port implantation 3 (1.4%) 3 (1.9%) 0 (0%) 0.600
Thoracic puncture catheter
drainage 34 (16%) 25 (16%) 9 (16%) > 0.900
Abdominal catheterization 19 (8.8%) 15 (9.4%) 4 (7.0%) 0.600
Arterial catheterization 5 (2.3%) 1 (0.6%) 4 (7.0%) 0.018
Central venous pressure apparatus 11 (5.1%) 8 (5.0%) 3 (5.3%) > 0.900
Postoperative drainage 59 (27%) 50 (31%) 9 (16%) 0.023
Indwelling gastric tube 49 (23%) 37 (23%) 12 (21%) 0.700
Fungi types 0.200
Candida albicans 146 (68%) 114 (72%) 32 (56%)
Mycotoruloides 41 (19%) 26 (16%) 15 (26%)
Aspergillus avus 6 (2.8%) 4 (2.5%) 2 (3.5%)
Page 12/24
Variable Overall, N =
2161
Survival, N =
1591
Death, N =
571
p-
value2
Aspergillus 18 (8.3%) 11 (6.9%) 7 (12%)
Penicillium 1 (0.5%) 1 (0.6%) 0 (0%)
Coinfection 2 (0.9%) 1 (0.6%) 1 (1.8%)
Others 2 (0.9%) 2 (1.3%) 0 (0%)
Types of antifungal drugs 0.005
Unantifungal treatment 76 (35%) 63 (40%) 13 (23%)
Triazole antifungal agent 122 (56%) 88 (55%) 34 (60%)
Echinocandin antifungal agent 11 (5.1%) 6 (3.8%) 5 (8.8%)
Combination therapy 7 (3.2%) 2 (1.3%) 5 (8.8%)
Length of antifungal treatment
(days) 4 (0, 8) 3 (0, 8) 5 (1, 9) 0.110
Combined with bacterial infection 54 (25%) 37 (23%) 17 (30%) 0.300
Immunoglobulin use 40 (19%) 23 (14%) 17 (30%) 0.010
ICU admission 43 (20%) 26 (16%) 17 (30%) 0.029
Mechanical ventilation 29 (13%) 14 (8.8%) 15 (26%) < 0.001
Cardiac arrest 12 (5.6%) 0 (0%) 12 (21%) < 0.001
Sepsis classication < 0.001
None 176 (81%) 140 (88%) 36 (63%)
Sepsis 21 (9.8%) 15 (9.5%) 6 (11%)
Severe sepsis 5 (2.3%) 3 (1.9%) 2 (3.5%)
Septic shock 14 (6.5%) 1 (0.6%) 13 (23%)
Abbreviations: PICC peripherally inserted central catheter, ICU Intensive Care Unit
1n (%); Median (IQR),2Pearson's Chi-squared test; Wilcoxon rank sum test; Fisher's exact test
3.4 Risk factors for nosocomial death
In this study, univariate analysis results showed: ECOG-PS 3–4, TNM stage III-IV, lung metastasis, bone
metastasis, radical surgery with 30 days, CCI, admission to the ICU, mechanical ventilation,
hypoproteinemia, thrombocytopenia, and hyponatremia were signicantly associated with in-hospital
mortality. The multivariate analysis determined that ECOG-PS 3–4 (OR = 6.08, 95%CI: 2.04–18.12, p
Page 13/24
= .001), pulmonary metastases (OR = 2.76, 95%CI: 1.11–6.848, p = .029), thrombocytopenia (OR = 2.58,
95%CI: 1.21–5.47, p = .014), hypoalbuminemia (OR = 2.44, 95%CI: 1.22–4.90, p = .012), and mechanical
ventilation (OR = 2.64, 95%CI: 1.03–6.73, p = 0.42) were independent inuencing factors of nosocomial
death in tumor patients with nosocomial infections caused by fungi (Table3).
Table 3
Logistic regression analysis of in-hospital death in solid patients with fungal infection
Variable OR (univariable) OR (multivariable)
ECOG-PS 0,1,2 REF (1.00) REF (1.00)
3,4 9.95 (3.66–27.04, p
< .001) 6.08 (2.04–18.12, p
= .001)
TNM stage I-II REF (1.00)
III-IV 2.73 (1.20–6.19, p
= .016)
Pulmonary metastasis Yes 2.31 (1.05–5.10, p
= .037) 2.76 (1.11–6.848, p
= .029)
Bone metastasis Yes 2.20 (1.07–4.52, p
= .033)
Operation type Unoperated REF (1.00)
Radical operation 0.34 (0.15–0.77, p
= .010)
Palliative
operation 1.10 (0.31–3.84, p
= .881)
ICU admission Yes 2.17 (1.07–4.40, p
= .031)
CCI ≤ 3 REF (1.00)
> 3 2.69 (1.13–6.40, p
= .026)
Platelet count
(×109/L)
< 100 3.72 (1.90–7.29, p
< .001) 2.58 (1.21–5.47, p
= .014)
Albumin(g/L) < 30 3.01 (1.61–5.62, p
< .001) 2.44 (1.22–4.90, p
= .012)
Serum
sodium(mmol/L) < 130 2.67 (1.41–5.07, p
= .003)
Mechanical
ventilation Yes 3.70 (1.65–8.28, p
= .002) 2.64 (1.03–6.73, p
= .042)
Abbreviations: OR odds ratio, ECOG-PS Eastern Cooperative Oncology Group Performance Status, ICU
intensive care u unit, CCI Charlson Co-morbidity Index score
Page 14/24
3.5 Nomogram establishment and evaluation
Based on the results of multivariate logistic analysis, the nal factors included were: ECOG-PS, lung
metastases, platelet count, serum albumin level, and mechanical ventilation. Thus, we established a
nomogram to predict the risk of nosocomial death from nosocomial infections in patients with oncology
(Fig.2). Multiple methods were performed to assess the discrimination and calibration abilities of the
nomogram, including ROC and calibration curves. The area under the ROC curve (AUC) of the nomogram
was 0.759 (95%CI: 0.682–0.835) (Fig.3), suggesting an excellent discrimination ability in predicting the
in-hospital death risk of these patients. Besides, the calibration curve showed that there was a high
consistency between the predicted and actual in-hospital death risk of the nomogram (Fig.4), indicating
a reliable calibration ability of the nomogram. Due to the ROC curve and calibration curves depending on
the nomogram's sensitivity and specicity, so they could not identify “false negative” and “false positive”
events. Therefore, DCA was conducted to evaluate the net clinical benet of the nomogram when it was
adopted to guide clinical decision-making. The results demonstrated that the nomogram would bring
more net clinical benet to these patients at the whole range of risk threshold compared to other single
factors in the nomogram (Fig.5). Taken together, the constructed nomogram is a reliable risk classier to
predict the in-hospital death risk of nosocomial infections caused by fungi in patients with solid tumors.
4. Discussion
Patients with malignant tumors are more likely to develop infections for various reasons[5, 6]. Therefore,
we conducted this study to fully understand the clinical features of nosocomial fungal infections in
patients with solid tumors and established a nomogram to predict in-hospital mortality in these patients
to accurately estimate the risk of nosocomial infections death in each patient.
In the current study, 1.3% of patients with solid tumor had nosocomial fungal infections over the eight
years study period, which was low compared with the results of studies before[22]. This discrepancy could
be explained by the fact that the prevalence of hospital acquired infections in cancer patients varies
widely from region to region. Compared with other types of tumors, patients with respiratory tumors
accounted for the highest proportion of the total population (34%), followed by gastrointestinal tumors
(24%) and hepatobiliary and pancreatic system tumors (24%). This may be related to the incidence of
these neoplasms[23]. Meanwhile, lung cancer patients inltrate and continuously secrete
immunosuppressive factors by respiratory tumor cells, and the body's natural barrier function is inhibited,
resulting in increased alveoli and bronchial secretions and mass obstruction of bronchi making lung
cancer patients more susceptible to co-infection than non-lung cancer patients[24, 25].
In this study, we observed that C.Albicans is the essential microbe causing fungal infections in patients
with solid tumors, accounting for 68%, followed by other Candida genera (19%). This result is consistent
with the results of previous studies[11, 26]. Unfortunately, the study was retrospective, and whether patients
had a swab to screen for colonizing microbiota is unknown. C.Albicans is the most common fungal
Page 15/24
infection and strain that causes invasive fungal disease. However, in recent years, studies have found
that the proportion of non-C. Albicans detected in Candida and Aspergillus are increasing, and the case
fatality rate is higher[27].
In this retrospective study, 76 patients (36%) received antibacterial therapy within 30 days before the
diagnosis of fungal infection, the most crucial treatment received in the previous 30 days for the general
population. This is consistent with our standard view: antibiotics may lead to dysbacteriosis and fungal
proliferation. So patients who have previously received antibiotics need to be alert to fungal infections.
Nosocomial deaths occurred in 26 of the 76 patients, accounting for 46% of the total deaths. In addition
to fungal pathogenicity and invasiveness, this may also be associated with suppressed immune function
in cancer patients [1, 2]. One hundred thirty-ve patients received invasive procedures within 30 days of the
diagnosis of fungal infection, accounting for 62% of the total population. Invasive operations such as
indwelling catheterization and PICC damage the mucous membrane of the body cavity and the inner wall
of blood vessels, destroying the physiological immune barrier of the human body so that fungal
displacement and colonization result in infection. A prospective study published in 2018 showed a 9.1%
incidence of concurrent infection of central venous catheters[3]. Therefore, patients with solid tumors
should be particularly concerned about potential fungal infections when receiving antibiotic therapy or
invasive procedures to avoid fungal-related deaths.
Above all, we investigated the predictors of nosocomial mortality risk of nosocomial infections in people
with cancer. We found ECOG-PS 3–4, lung metastases, mechanical ventilation, thrombocytopenic, and
hypoalbuminemia to be independent risk factors for in-hospital mortality. This conclusion is similar to a
previous study of nosocomial mortality from bacterial infections[28]. Patients with cancer with poor
ECOG-PS and distant metastases are known to be associated with adverse survival outcomes, as in our
study. It recommends that we pay more attention to patients with higher ECOG-PS and those with
pulmonary metastases for more rened management. Patients who received mechanical ventilation
during hospitalization had a poorer prognosis, consistent with previous studies of bacterial infections[18,
21, 29, 30]. That is due to concomitant circulatory and/or respiratory dysfunction in these patients, resulting
in poor clinical outcomes. We also found that patients with hypoalbuminemia and low platelets were
signicantly associated with higher in-hospital mortality. For one thing, because lower albumin level is
often associated with immunosuppression, decreased muscle mass, malnutrition, and weight loss in
patients with malignancy, these patients have a poor prognosis and an increase in cancer-related
deaths[31–33]. For another, low albumin levels lead to low PNI, which one study conrmed as an
independent risk factor for NSCLC[34]. At the same time, related studies have shown that
thrombocytopenia is associated with a poor prognosis for many diseases[35, 36]. Our ndings further
conrm this view.
In this study, we developed a nomogram to predict the risk of nosocomial death from nosocomial fungal
infections in cancer patients and evaluated its predictive power and clinical utility. This nomogram has
good performance in predicting the risk of in-hospital death in these people. To our knowledge, this is the
Page 16/24
rst study to systematically evaluate the clinical features of nosocomial fungal infections in cancer
patients in northwest China and develop and validate a nomogram that can accurately predict the risk of
nosocomial death from nosocomial infections in these patients. Still, there are some inevitable limits to
our study. First, due to the design of the single-center retrospective analysis, it is challenging to collect
variables such as specic chemotherapy and radiation doses, specic prior antibiotic treatment
information, and more detailed laboratory results. Thus, there may be potential biases in the analysis of
the relationships. Second, although we established a nomogram that effectively predicted the risk of in-
hospital death from nosocomial infections in patients with solid tumors, internal cohort validation was
not possible due to sample size, and there was a lack of independent external validation cohorts.
Therefore, there is an urgent need for multi-center retrospective and well-designed prospective studies to
verify the performance of nomograms.
5. Conclusion
In summary, in our study, Fungi-related nosocomial infections in cancer patients resulted in higher in-
hospital mortality. The most common pathogen is C.Albicans, and the leading infection site is the
respiratory system. ECOG-PS 3–4, pulmonary metastasis, thrombocytopenia, hypoalbuminemia, and
mechanical ventilation were independent prognostic factors for in-hospital death in these patients. In
addition, we constructed a new nomogram that accurately predicts the risk of in-hospital death from
nosocomial fungal infections in cancer patients. Precise management of patients with lung metastases,
high ECOG-PS, mechanical ventilation, and dynamic monitoring of serum albumin levels and platelets
may improve the prognosis of these individuals.
Abbreviations
FN: Febrile neutropenia; OR: Odds ratio; CI: Condence interval; ECOG-PS: Eastern Cooperative Oncology
Group Performance Status; CCI: Charlson comorbidity index; G-CSF: Granulocyte colony-stimulating
factor; ICU: Intensive care unit; PICC: Peripherally inserted central catheter.
Declarations
Author contributions
TT and YY conceived the study. RXW, AMJ, JHL and RZ were involved in data collecting, statistical
analysis, and manuscript drafting. CCS and QQD conducted the data collection and analysis and
provided the critical revision. SHL, FMZ, and YYM were involved in data collecting. XF participated in the
study design and helped with the data collection. XL and ZPR participated in the study design and
manuscript revision. All authors read and approved the nal manuscript.
Funding
Page 17/24
This study was supported by the CSCO-Hengrui Cancer Research Fundation (NO. Y-HR2019-0401),
Medical scientic research project (Medical research project for young and middle-aged oncologist of
lung cancer), and Youth Program of National Natural Science Foundation of China (NO. 82002437).
Acknowledgments
Not applicable.
Conict of Interest
The authors declare that the research was conducted in the absence of any commercial or nancial
relationships that could be construed as a potential conict of interest.
References
1. Alevizakos, M., S. Karanika, M. Detsis, et al., Colonisation with extended-spectrum β-lactamase-
producing Enterobacteriaceae and risk for infection among patients with solid or haematological
malignancy: a systematic review and meta-analysis. Int J Antimicrob Agents, 2016. 48(6): p. 647-
654.
2. Gudiol, C., J.M. Aguado, and J. Carratalà, Bloodstream infections in patients with solid tumors.
Virulence, 2016. 7(3): p. 298-308.
3. Schwanke, A.A., M.T.R. Danski, L. Pontes, et al., Central venous catheter for hemodialysis: incidence
of infection and risk factors. Rev Bras Enferm, 2018. 71(3): p. 1115-1121.
4. Uçkay, I., D. Holy, M. Betz, et al., Osteoarticular infections: a specic program for older patients? Aging
Clin Exp Res, 2021. 33(3): p. 703-710.
5. Anderson, C., J.L. Lund, M.A. Weaver, et al., Noncancer mortality among adolescents and young
adults with cancer. Cancer, 2019. 125(12): p. 2107-2114.
. Zaorsky, N.G., T.M. Churilla, B.L. Egleston, et al., Causes of death among cancer patients. Ann Oncol,
2017. 28(2): p. 400-407.
7. Guo, H., L. Zhao, J. Zhu, et al., Microbes in lung cancer initiation, treatment, and outcome: Boon or
bane? Semin Cancer Biol, 2021.
. Jiang, A.M., N. Liu, R. Ali Said, et al., Nosocomial Infections in Gastrointestinal Cancer Patients:
Bacterial Prole, Antibiotic Resistance Pattern, and Prognostic Factors. Cancer Manag Res, 2020. 12:
p. 4969-4979.
9. Jiang, A.M., N. Liu, R. Zhao, et al., Clinical outcomes and prognostic factors in bloodstream
infections due to extended-spectrum β-lactamase-producing Enterobacteriaceae among patients with
malignancy: a meta-analysis. Ann Clin Microbiol Antimicrob, 2020. 19(1): p. 53.
10. Jiang, A.M., X. Shi, N. Liu, et al., Nosocomial infections due to multidrug-resistant bacteria in cancer
patients: a six-year retrospective study of an oncology Center in Western China. BMC Infect Dis,
Page 18/24
2020. 20(1): p. 452.
11. Laroumagne, S., B. Lepage, C. Hermant, et al., Bronchial colonisation in patients with lung cancer: a
prospective study. Eur Respir J, 2013. 42(1): p. 220-9.
12. Alamri, M., H. Alghamdi, S. Althawadi, et al., Invasive fungal infection of the brain caused by
Neoscytalidium dimidiatum in a post-renal transplant patient: A case report. Med Mycol Case Rep,
2021. 34: p. 27-31.
13. Chen, X., J. Wang, S. Wang, et al., Real-world assessment of the effectiveness of posaconazole for
the prophylaxis and treatment of invasive fungal infections in hematological patients: A
retrospective observational study. Medicine (Baltimore), 2021. 100(30): p. e26772.
14. De Pauw, B., T.J. Walsh, J.P. Donnelly, et al., Revised denitions of invasive fungal disease from the
European Organization for Research and Treatment of Cancer/Invasive Fungal Infections
Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study
Group (EORTC/MSG) Consensus Group. Clin Infect Dis, 2008. 46(12): p. 1813-21.
15. Mack, D., Molecular mechanisms of Staphylococcus epidermidis biolm formation. J Hosp Infect,
1999. 43 Suppl: p. S113-25.
1. Rinaldi, M., M. Bartoletti, A. Ferrarese, et al., Breakthrough invasive fungal infection after liver
transplantation in patients on targeted antifungal prophylaxis: A prospective multicentre study.
Transpl Infect Dis, 2021. 23(4): p. e13608.
17. José, P., F. Alvarez-Lerma, E. Maseda, et al., Invasive fungal infection in crtically ill patients: hurdles
and next challenges. J Chemother, 2019. 31(2): p. 64-73.
1. Palacios-Baena, Z.R., B. Gutiérrez-Gutiérrez, M. De Cueto, et al., Development and validation of the
INCREMENT-ESBL predictive score for mortality in patients with bloodstream infections due to
extended-spectrum-β-lactamase-producing Enterobacteriaceae. J Antimicrob Chemother, 2017. 72(3):
p. 906-913.
19. Jenks, J.D., J.P. Gangneux, I.S. Schwartz, et al., Diagnosis of Breakthrough Fungal Infections in the
Clinical Mycology Laboratory: An ECMM Consensus Statement. J Fungi (Basel), 2020. 6(4).
20. Singer, M., C.S. Deutschman, C.W. Seymour, et al., The Third International Consensus Denitions for
Sepsis and Septic Shock (Sepsis-3). Jama, 2016. 315(8): p. 801-10.
21. Chen, C.Y., F.M. Tien, W.H. Sheng, et al., Clinical and microbiological characteristics of bloodstream
infections among patients with haematological malignancies with and without neutropenia at a
medical centre in northern Taiwan, 2008-2013. Int J Antimicrob Agents, 2017. 49(3): p. 272-281.
22. Angarone, M., Fungal infections in cancer patients. Cancer Treat Res, 2014. 161: p. 129-55.
23. Sung, H., J. Ferlay, R.L. Siegel, et al., Global Cancer Statistics 2020: GLOBOCAN Estimates of
Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin, 2021. 71(3): p.
209-249.
Page 19/24
24. Nagy, A., V. Müller, A.M. Kolonics-Farkas, et al., Worse lung cancer outcome in patients with lower
respiratory tract infection conrmed at time of diagnosis. Thorac Cancer, 2019. 10(9): p. 1819-1826.
25. Wang, L., Y. Li, X. Zhang, et al., Characteristics of nosocomial infection and its effects on the survival
of chemotherapy patients with advanced non-small cell lung cancer. Oncol Lett, 2017. 14(6): p. 7379-
7383.
2. Vieira de Melo, A.P., D.L. Zuza-Alves, W.P. da Silva-Rocha, et al., Virulence factors of Candida spp.
obtained from blood cultures of patients with candidemia attended at tertiary hospitals in Northeast
Brazil. J Mycol Med, 2019. 29(2): p. 132-139.
27. Xiao, M., S.C. Chen, F. Kong, et al., Distribution and Antifungal Susceptibility of Candida Species
Causing Candidemia in China: An Update From the CHIF-NET Study. J Infect Dis, 2020. 221(Suppl 2):
p. S139-s147.
2. Jiang, A., X. Shi, H. Zheng, et al., Establishment and validation of a nomogram to predict the in-
hospital death risk of nosocomial infections in cancer patients. Antimicrob Resist Infect Control,
2022. 11(1): p. 29.
29. Antonio, M., C. Gudiol, C. Royo-Cebrecos, et al., Current etiology, clinical features and outcomes of
bacteremia in older patients with solid tumors. J Geriatr Oncol, 2019. 10(2): p. 246-251.
30. Tu, R.H., J.X. Lin, J. Desiderio, et al., Does Intra-Abdominal Infection after Curative Gastrectomy
Affect Patients' Long-Term Prognosis? A Multi-Center Study Based on a Large Sample Size. Surg
Infect (Larchmt), 2019. 20(4): p. 271-277.
31. Soeters, P.B., R.R. Wolfe, and A. Shenkin, Hypoalbuminemia: Pathogenesis and Clinical Signicance.
JPEN J Parenter Enteral Nutr, 2019. 43(2): p. 181-193.
32. Paccagnella, A., I. Morassutti, and G. Rosti, Nutritional intervention for improving treatment tolerance
in cancer patients. Curr Opin Oncol, 2011. 23(4): p. 322-30.
33. Fruchtenicht, A.V., A.K. Poziomyck, G.B. Kabke, et al., Nutritional risk assessment in critically ill cancer
patients: systematic review. Rev Bras Ter Intensiva, 2015. 27(3): p. 274-83.
34. Liu, N., A. Jiang, X. Zheng, et al., Prognostic Nutritional Index identies risk of early progression and
survival outcomes in Advanced Non-small Cell Lung Cancer patients treated with PD-1 inhibitors. J
Cancer, 2021. 12(10): p. 2960-2967.
35. Thiery-Antier, N., C. Binquet, S. Vinault, et al., Is Thrombocytopenia an Early Prognostic Marker in
Septic Shock? Crit Care Med, 2016. 44(4): p. 764-72.
3. Venkata, C., R. Kashyap, J.C. Farmer, et al., Thrombocytopenia in adult patients with sepsis:
incidence, risk factors, and its association with clinical outcome. J Intensive Care, 2013. 1(1): p. 9.
Figures
Page 20/24
Figure 1
Flow chart of the study.
Page 21/24
Figure 2
A nomogram to predict the risk of in-hospital death from fungal infections in cancer patients. This
patient’s albumin level was 35g/L, platelet count was 88×109/L, without mechanical ventilation, no
pulmonary metastasis and ECOG-PS 1. According to the nomogram, we can calculate that the total point
for this patient is 139 and its corresponding in-hospital death risk is 21.2%.
Page 22/24
Figure 3
The ROC curve to evaluate the discrimination ability of the nomogram. AUC = 0.759 (95%CI: 0.682-0.835).
ROC, receiver operating characteristic curve.
Page 23/24
Figure 4
The calibration curve of the nomogram for predicting in-hospital death risk of nosocomial infections
caused by fungi in cancer patients.
Page 24/24
Figure 5
Decision curve analysis of the nomogram for predicting in-hospital death risk of nosocomial infections
caused by fungi in cancer patients.
