Seroprevalence and risk factors of Toxoplasma gondii infection in children with leukemia in Shandong Province, Eastern China: a case—control prospective study

Department of Pediatrics, The Affiliated Hospital of Qingdao University, Qingdao, Shandong, China
Department of Pathology, Affiliated Hospital of Qingdao University, Qingdao, China
Department of Clinical Laboratory, The Affiliated Hospital of Qingdao University, Qingdao, China
Department of Urinary Surgery, The Affiliated Hospital of Qingdao University, Qingdao, China
Academic Editor
Subject Areas
Hematology, Infectious Diseases, Pediatrics
Toxoplasma gondii, Risk factors, Leukemia children, Seroprevalence
© 2019 Zhou et al.
This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ) and either DOI or URL of the article must be cited.
Cite this article
Zhou N, Fu H, Wang Z, Shi H, Yu Y, Qu T, Wang L, Zhang X, Wang L. 2019. Seroprevalence and risk factors of Toxoplasma gondii infection in children with leukemia in Shandong Province, Eastern China: a case—control prospective study. PeerJ 7:e6604


Limited information is available concerning the epidemiology of Toxoplasma gondii infection in children with leukemia in Eastern China. Therefore, a case-control study was conducted to estimate the seroprevalence of toxoplasmosis in this patient group and to identify risk factors and possible routes of infection. Serum samples were collected from 339 children with leukemia and 339 age matched health control subjects in Qingdao from September 2014 to March 2018. Enzyme linked immunoassays were used to screen anti- T. gondii IgG and anti- T. gondii IgM antibodies. Forty-eight (14.2%) children with leukemia and 31 (9.1%) control subjects were positive for anti-T. gondii IgG antibodies (P < 0.05), while 13 (3.8%) patients and 14 (4.1%) controls were positive for anti-T. gondii IgM antibodies (P = 0.84). Multivariate analysis showed exposure to soil and a history of blood transfusion were risk factors for T. gondii infection. Compared with IgG, patients with a history of blood transfusion were more likely to present anti- T. gondii IgM (P = 0.003). Moreover, patients with chronic lymphocytic leukemia and acute lymphocytic leukemia had higher T. gondii seroprevalence in comparison to control subjects (P = 0.002 and P = 0.016, respectively). The results indicated that the seroprevalence of T. gondii infection in children with leukemia is higher than that of healthy children in Eastern China. This information may be used to guide future research and clinical management, and further studies are necessary to elucidate the role of T. gondii in children with leukemia.


Leukemia is one of the most common malignant tumors in children, with more than 8,000 new cases and 4,000 disease-related deaths in China in 2013 (Chen et al., 2017). The morbidity has increased steadily due to changes in lifestyle in recent years. Complications of infection are one of the most severe risk factors that lead to death for leukemia patients. There have been comprehensive surveys for infections conducted in patients with malignancy including bacteria, virus and fungus (Mori et al., 2014; Vehreschild et al., 2014; Wang et al., 2017), but parasites, especially opportunistic infection parasites, in patients who are immunocompromised by virtue of neoplastic disease have received relatively little attention (Cong et al., 2015; Zueter et al., 2014).

One of the most frequent protozoan parasites causing opportunistic pathogenic infection in immunocompromised hosts is Toxoplasma gondii, which is an intracellular protozoan that chronically infects approximately a third of the world’s population (Montoya & Liesenfeld, 2004). Human infection results from ingestion of oocysts from insanitary food or water containing T. gondii tissue cysts (The Lancet Infectious, 2012). In individuals with normal immunity, acute acquired T. gondii infection is generally self-limiting and asymptomatic due to efficient immunity that limits the propagation of the multiplying tachyzoite stage. During this period, the tissue cysts are contained by the humoral and cellular immune system, including macrophages and T-lymphocytes (Montoya & Liesenfeld, 2004; Shaw et al., 2009). However, T. gondii still holds viable in the form of tissue cysts throughout the whole life of its host (Montoya & Liesenfeld, 2004). In immunocompromised individuals, such as patients with malignant tumors or collagen tissue disease, or transplant recipients under immunosuppressive therapy, cysts have greater propensity to relapse and disseminate, which can cause serious clinical consequences and even death (Donahoe et al., 2017; McLeod et al., 2009; Nimir et al., 2010; Syrogiannopoulos et al., 2002).

The treatment methods for leukemia are chemotherapy, radiotherapy, immunotherapy and combination therapy. These managements may cause immune system dysfunction, which predisposes the patient to the development of toxoplasmosis. Moreover, T. gondii infections are often ignored in the process of clinical diagnosis and treatments. Therefore, knowing the factors that increase susceptibility to T. gondii and recognising the early symptoms of T. gondii infection in patients with leukemia will promote the prevention of T. gondii infection and ideally increase the ability to assess patients’ needs. Additionally, several questions remain unclear: first, the prevalence of T. gondii infection in children with leukemia in Eastern China is still unknown; second, it is unclear whether demographic or lifestyle variables increase the risk of T. gondii infection in children with leukemia. Thus, we conducted this study to investigate the risk of toxoplasmosis in children with leukemia and the association between T. gondii infection and different risk factors.



During September 2014 to March 2018, serum samples were collected from 339 primary children with leukemia who presented to the Affiliated Hospital of Qingdao University for diagnosis and treatment. No patients had received IVIG therapy and/or immunotherapy before blood collection. The ages of the children with malignancy were 0–14 years old. For control subjects, 339 children who participated in health screenings at the hospitals were recruited, matched with children with malignancy by age, gender, and residence. The study was approved by the Ethics Committee of the Affiliated Hospital of Qingdao University (No. 20141349) and all patients/guardians had signed informed consent.

Sample collection

Approximately 2 mL of venous blood were drawn from participants who gave their consent to participate in this study. Blood samples were left for 2 h at room temperature to allow clotting, and were centrifuged at 3,000 rpm for 10 min. The sera were collected in EP tubes and stored at −80 °C until tested.

Socio-demographic and clinical data

Socio-demographic data including age, sex, residence and parent’s occupation were obtained from all participants. Behavioral data included any history of contact with stray cats, contact with pigs, any cats or dogs kept at home, consumption of raw/undercooked meat, consumption of raw vegetables and fruits, exposure to soil, and source of drinking water (Cong et al., 2015). Clinical data collected in patients included the histological type of leukemia, and history of blood transfusion and chemotherapy. Data were obtained from the patients/guardians and medical records, and patients/guardians were blinded to infectious status before data were obtained.

Serological assay

Sera were tested for the presence of IgG and IgM antibodies against T. gondii using commercially available enzyme immunoassay kits (ELISA) (Demeditec Diagnostics GmbH, Germany). This procedure was performed according to the manufacturer’s instructions. Positive and negative serum controls were included in every plate. Samples from children with leukemia and control subjects were randomly mixed.


Statistical analysis was performed using the statistical software SPSS 19.0. Chi-square test or Fisher exact test were used to analyze the anti-T. gondii antibody seroprevalence in socio-demographic data. Multivariate logistic regression models were used to adjust for potential confounders. Variables associated with T. gondii infection in univariate analysis (P ≤ 0.25) were included in a multivariate logistic regression analysis. Adjusted odds ratio (OR) and 95% confidence interval (CI) were calculated by multivariate analysis to identify independent risk factors for T. gondii infection. Results with P-value <0.05 were considered statistically significant.


Epidemiology and risk factors for children with leukemia with T. gondii infection

A total of 678 individuals (339 children with leukemia and 339 health controls) were tested between September 2014 and March 2018. Fifty-seven and 33 (16.8% vs 9.7%, P < 0.05) samples were positive for anti-T. gondii antibodies in children with leukemia and control subjects, respectively. A significant difference (P < 0.05) was detected in the seroprevalence of anti-T. gondii IgG antibodies in 48 children with leukemia (14.2%) versus 31 control subjects (9.1%). Thirteen patients tested positive for T. gondii IgM antibodies (3.8%), compared with 14 controls (4.1%) (P = 0.84). The details of the children with leukemia and control subjects, including age distribution, gender, resident area, parent’s occupation and data of lifestyle and clinical treatment, are shown in Table 1. Our analysis revealed significant differences between the 6∼9-year-old age group (8/81, 9.9%) and the ≤ 2-year-old age group (22/101, 21.8%) (P = 0.031). T. gondii seroprevalence was higher in male leukemia patients (38/188, 20.2%) than in female patients (19/151, 12.6%), but there was no statistical difference (P = 0.06).

Univariate analysis showed some lifestyle and clinical variables where P ≤ 0.25, including contact with swine, exposure to soil, and history of transfusion and chemotherapy. In multivariate analysis, only exposure to soil (OR, 1.82; 95% CI [0.92–2.89]; P = 0.046) and history of transfusion (OR, 2.27; 95% CI [1.27–4.08]; P = 0.006) were significantly associated with T. gondii infection in leukemia patients (Table 2). Additionally, we analyzed the different results between IgG and IgM against T. gondii in children with leukemia who received blood transfusions. Compared with IgG, patients with blood transfusion were more likely to present anti- T. gondii IgM (Table 3). No significant associations between T. gondii status and other variables were found in the present study.

Seroprevalence among children with leukemia

The seroprevalence of different histological types of leukemia is presented in Table 4. The highest seroprevalence of T. gondii was detected in patients with chronic lymphocytic leukemia (CLL) (28.9%), followed by acute lymphocytic leukemia (ALL) (18.1%), acute myeloblastic leukemia (AML) (15.9%), and chronic myeloblastic leukemia (CML) (7.5%). Compared with control subjects, patients with CLL and ALL had significantly higher seroprevalence (P = 0.002 and P = 0.016, respectively).


Toxoplasma gondii infection is increasingly being reported among patients with leukemia, and the higher seroprevalence of anti-T. gondii antibodies among leukemia patients reflects latent association between T. gondii infection and leukemia (Gharavi, Roozbehani & Mandeh, 2017; Huang et al., 2016; Yazar et al., 2004); however, both the epidemiology and risk factors for this are unclear. Moreover, information about T. gondii infection in children with leukemia is also limited. Therefore, we assessed the seroprevalence of anti-T.gondii antibodies in 339 children with leukemia from Shandong province, Eastern China, and explored the risk factors of T. gondii infection in children with leukemia.

Table 1:
Seroprevalence of T. gondii infection in children with leukemia and control subjects in Eastern China.
Leukemia Children (N = 339) Controls (N = 339)
Prevalence of T. gondii infection Prevalence of T. gondii infection
Characteristic No. tested No. positive % P No. tested No. positive % P
Age (years)
≤2 101 22 21.8 Reference 98 6 6.1 Reference
3∼6 101 19 18.8 0.6 119 12 10.1 0.29
7∼10 81 8 9.9 0.031 66 11 16.7 0.03
11∼14 56 8 14.3 0.15 56 4 7.1 0.81
Male 188 38 20.2 0.06 180 17 9.4 0.85
Female 151 19 12.6 159 16 10.1
Residence area
Urban 182 30 16.5 0.29 254 25 9.8 0.91
Rural 157 27 17.2 85 8 9.4
Contact with cats
Yes 75 14 18.7 0.63 69 9 13.1 0.29
No 264 43 16.3 270 24 8.9
Contact with dogs
Yes 69 10 14.5 0.56 47 2 4.3 0.28
No 270 47 17.4 292 31 10.6
Contact with swine
Yes 112 15 13.4 0.24 59 3 5.1 0.19
No 227 42 18.5 280 30 10.7
Consumption of raw/ undercooked meat
Yes 36 6 16.7 0.98 83 8 9.6 0.97
No 303 51 16.8 256 25 9.8
Consumption of raw vegetables
Yes 84 16 19.1 0.53 212 19 9 0.54
No 255 41 16.1 127 14 11
Exposure to soil
Yes 95 23 24.2 0.04 58 4 6.9 0.42
No 244 34 13.9 281 29 10.3
Source of drinking water
Well + river 30 4 13.3 0.59 29 3 10.3 0.75
Tap 309 53 17.2 310 30 9.7
Parent’s occupation
Farmer 105 21 20 0.29 119 14 11.8 0.35
Worker 234 36 15.4 220 19 8.6
Surgery history
Yes 75 14 18.7 0.63
No 264 43 16.3
Blood transfusion history
Yes 102 26 25.5 0.005
No 237 31 13.1
Chemotherapy history
Yes 236 36 15.3 0.25
No 103 21 20.4
DOI: 10.7717/peerj.6604/table-1

Our results demonstrated that children with leukemia were more likely to be infected with T. gondii when compared with healthy controls (16.8% vs 9.7%, P = 0.002), though there was no significant difference between children with leukemia and controls with regard to the level of IgM antibodies, which was consistent with previous studies (Gharavi, Roozbehani & Mandeh, 2017). In healthy children, T. gondii infection rate may vary by age, and older children may have higher propensity for T. gondii infection (Fan et al., 2012; Fu et al., 2014; Marchioro et al., 2015), but in this present study, children with leukemia ≤ 2-years-old had the highest seroprevalence than any other subgroup. T. gondii is an opportunistic pathogen—in healthy individuals, T. gondii infection rate may increase with age because there is more possibility of T. gondii infection with time (Meng et al., 2015). However, the mechanisms by which younger leukemia patients are inclined to develop toxoplasmosis have not been comprehensively explained. One speculation for this is that patients with leukemia are immunocompromised and young patients may have immunities too frail to control opportunistic infections (Cong et al., 2015). This suggests that younger leukemia patients might be more susceptible to T. gondii infection, and further studies based on larger and well-distributed sample sizes are needed to confirm this conclusion.

Table 2:
Multivariate analysis of selected characteristics of leukemia patients and their association with T. gondii infection.
Characteristica Adjusted odds ratiob 95% CI P
Gender 1.43 1.08–2.22 0.09
Contact with swine 1.21 0.64–1.71 0.21
Exposure to soil 1.82 0.92–2.89 0.046
Blood transfusion history 2.27 1.27–4.08 0.006
DOI: 10.7717/peerj.6604/table-2


The variables included were those with a P < 0.25 obtained in the univariate analysis.
Adjusted by age.
Table 3:
Seroprevalence of T. gondii antibodies in children with leukemia with history blood transfusion.
Blood transfusion history (No.) P
Yes No
Anti- T. gondii IgG antibodies
No. Positive (%) 19(18.6) 29(12.2) 0.122a
No. Negative (%) 83(81.4) 208(87.8)
Anti- T. gondii IgM antibodies
No. Positive (%) 9(8.8) 4(1.7) 0.003b
No. Negative (%) 93(91.2) 233(98.3)
DOI: 10.7717/peerj.6604/table-3


Chi-square tests.
Fisher exact test were used.
Table 4:
Clinical diagnosis and seroprevalence of T. gondii in children with leukemia in Eastern China.
Clinical diagnosis No. tested No. positive % P
Acute myeloblastic leukemia 132 21 15.9 0.06
Acute lymphoblastic leukemia 116 21 18.1 0.016
Chronic myeloblastic leukemia 53 4 7.5 0.61
Chronic lymphocytic leukemia 38 11 28.9 0.002
DOI: 10.7717/peerj.6604/table-4


As compared with 9.7% seroprevalence of anti- T. gondii antibodies in controls (33/339).

Previous reports indicated that contact with pigs and cats, consumption of raw/undercooked meat and exposure to soil were risk factors for T. gondii infection in cancer patients (Alvaradoesquivel et al., 2010; Cong et al., 2015; Zhou et al., 2018). In the present study, multivariate analysis showed that in leukemia patients, T. gondii seroprevalence was not associated with any lifestyle and feeding habits, except exposure to soil. As the definitive hosts of T. gondii, felids play a vital role in the transmission of this parasite (Webster, 2010). Cats infected with T. gondii shed millions of oocysts that can survive in soil for many years, and a single oocyst has the potential to cause full-blown infection and toxoplasmosis if ingested by humans (Tian et al., 2017a; Webster, 2010). Moreover, contact with cats and drinking water contaminated with cat feces also increases the risk of T. gondii infection (Tian et al., 2017a; Tian et al., 2017b). Although humans can be infected with T. gondii through exposure to soil contaminated with T. gondii oocysts, little attention has been paid to their role. Therefore, there is an urgent need to educate the general public and medical professionals about this risk factor of T. gondii infection in immunocompromised individuals, particularly leukemia patients, in order to reduce the risk of toxoplasmosis.

In our study, patients with a history of blood transfusion had a significantly higher seroprevalence than control subjects, which is a similar result to a study conducted by Alvarado-Esquivel et al. (2018). Moreover, compared with IgG, patients with blood transfusions were more likely to present anti- T. gondii IgM. Previous studies have demonstrated that intracellular tachyzoites can disseminate into all organs through blood circulation (Harker, Ueno & Lodoen, 2015), and hosts might be infected with T. gondii by infusing blood infected with tachyzoites (Alvarado-Esquivel et al., 2018). Positive T. gondii IgM is often defined as an acute infection, but isolated IgM without an IgG positive is interpreted as a chronic infection or false positive (Reshika et al., 2015). However, anti-T. gondii IgG seropositivity appears early after infection and high levels of specific anti-T. gondii IgG antibodies can be present in patients with recent infections (Liesenfeld et al., 2001). In this study, 11 samples (nine children with leukemia and two healthy controls) were detected with isolated positive IgM, and all of them were diagnosed with toxoplasmosis after a few days by clinicians according to clinical features. Chemotherapy is frequently used for the treatment of leukemia, and most children with leukemia who received transfusion were immunocompromised and struggled to produce enough immunoglobulin (e.g., A and G) to resist parasite infections. Moreover, microbiotics (such as amoxicillin, azithromycin and sulfamethoxazole) were also used for patients to prevent against pathogen infection. Sulfamethoxazole might contribute in preventing toxoplasmosis and turning positive antibodies into negative ones, so, it is within reason that there are significant differences in IgM, rather than in IgG antibodies, for patients with a history of transfusion.

Among the various histological types of leukemia, seroprevalence of CLL and ALL were significantly higher than controls. The progression of apoptosis and anti-apoptosis was regulated by miRNAs (He et al., 2015). Toxoplasma gondii can phagocytose these miRNAs, which might result in the hosts’ gene expression irregularity, and can cause carcinogenesis (Gharavi, Roozbehani & Mandeh, 2017; Saçar, Bağcı& Allmer, 2014). Nevertheless, malignant tumors were associated with defects in cell-mediated immunity, which was defined as a lymphocytic exhaustion phenomenon. When strengthened by treatment with defects in cell-mediated immunity, the patients struggled to resist intracellular pathogen infections and inclined toward development of toxoplasmosis (Cong et al., 2015; Yazar et al., 2004). This can partially explain the higher seroprevalence in lymphocytic leukemia than in granulocytic leukemia.

Toxoplasma gondii infections are often ignored in the process of clinical diagnosis and treatment, which may result in severe toxoplasmosis. Although further studies should be conducted to explore the causation between T. gondii infection and blood transfusion in leukemia patients, it is a pressing matter of the moment to encourage medical professionals to pay attention to the important role that exposure to soil and blood transfusion play in transmitting T. gondii infection to leukemia patients, and to test T. gondii routinely before donor blood is infused to patients. Moreover, a recent study conducted by Quinn et al. (2018) suggested that pentamidine effectively prevents Pneumocystis jirovecii pneumonia in pediatric oncology patients receiving immunosuppressive chemotherapy. Children with leukemia are immunosuppressive and struggle to resist parasite infections. Therefore, patients with leukemia should be monitored for early signs of toxoplasmosis and secondary prevention against T. gondii using trimethoprim or sufamethoxazole is necessary for patients with leukemia.

Additionally, the limitations of our results should be considered: first, the sample size was relatively small, rendering some findings inconclusive; second, the donors’ sera antibody statuses were unclear, so the positives for T. gondii antibody caused by donor derived antibody could not be excluded; third, we did not collect the clinical information on the treatment history of study participants, therefore, the effect of immunosuppressive and anti- pathogen therapies for antibody seroprevalence in children with leukemia was uncertain.


In this study, we reported serological evidence of an association between T. gondii infection and children with leukemia. Exposure to soil and blood transfusion history were identified as risk factors for T. gondii infection. This information may be used to guide future research and clinical management. However, further studies are necessary to elucidate the role of T. gondii in children with leukemia.

Supplemental Information

Socio-democlgraphic data of 339 children with leukemia and 339 health subjects were included

DOI: 10.7717/peerj.6604/supp-1