The Prevalence of Human Cytomegalovirus Infection in Children Leukemia by Chip Digital PCR

Wen-Jun Wang National Institute for Viral Disease Control and Prevention Miao Feng Capital Institute of Pediatrics Feng He Capital Institute of Pediatrics Juan Song National Institute for Viral Disease Control and Prevention Qin-Qin Song National Institute for Viral Disease Control and Prevention Xia Dong National Institute for Viral Disease Control and Prevention Bing-Tian Shi National Institute for Viral Disease Control and Prevention Ping Chen National Institute for Viral Disease Control and Prevention Hai-Lan Yao Capital Institute of Pediatrics Jun Han (  hanjun_sci@163.com ) National Institute for Viral Disease Control and Prevention https://orcid.org/0000-0002-0803-4403


Abstract Background
To establish a method for detecting HCMV viral load to guide clinical treatment by chip digital PCR (cdPCR). Methods 5.67×10 6 TCID50/ml of HCMV AD169 was serially diluted to evaluate sensitive of cdPCR, HSV-1, HSV-2, VZV, EBV, HHV-6 and HHV-7 were used to evaluate the speci city of HCMV cdPCR. HCMV infection were analyzed in 110 children leukemia whole blood by RT-qPCR and cdPCR.

Results
The sensitive of HCMV cdPCR was up to 71 ± 32 copies/ml, which is higher than that of RT-qPCR. HCMV cdPCR did not cross react with other herpesviruses. The cdPCR effectively detected 7 HCMV positive samples, making the laboratory diagnosis rate of HCMV increased by 6.36% (7/110) for children leukemia patients. And the prevalence of HCMV infection increased from 28.18-34.54% in 110 children leukemia patients by cdPCR.
Conclusion cdPCR is more sensitive to detect viral load than RT-qPCR. The cdPCR may be used to evaluate relationship between viral load and progression of HCMV infection in patients.

Background
Human cytomegalovirus (HCMV), a ubiquitous β-herpesvirus, infects as high as 90% of the human population worldwide in developing countries [1]. HCMV infection remains largely asymptomatic for healthy, immunocompetent individuals, persistent infection or recurrent infection often occur immunocompromised individuals [2][3][4]. Like all herpesvirus, HCMV stablishes latency in haematopoietic cells and persists for lifelong of the host after primary infection [5,6]. HCMV infection may result in signi cant morbidity and mortality in immunode ciency patients, especially transplant patients including allogenic stem cell transplantation(allo-SCT) [7][8][9]. HCMV infection is also one of the most leading cause of infant birth defects [5,10].
Diagnostic testing to con rm HCMV and to monitor viral loads and immune responses among solid-organ transplant (SOT) and hematopoietic stem cell transplant (HSCT) recipients is crucial to effective patient care. Laboratory tests that directly detect HCMV are recommended for surveillance, diagnosis, and monitoring, while assays of immune status are relied upon for HCMV risk assessment and strati cation of risk factors. Virus isolation is highly speci c for the diagnosis of HCMV infection. Culture can also be used to assess phenotypic antiviral drug testing. The most common approach for direct virus detection is commercial quantitative nucleic acid ampli cation tests (Real-time uorescent quantitative PCR, RT-qPCR), which are higher sensitive and offer rapid turn-around times. Trends in viral loads over time (viral load kinetics) directly correlate with the likelihood of sever HCMV disease [10][11][12]. The higher viral loads or more rapid viral load increase correlate with both development of HCMV disease and a higher risk for severe HCMV disease [10][11][12][13]. The quantitative nature allows for assessment of the degree of HCMV replication, which is expressed as the absolute viral load value.
However, because the quantify of RT-qPCR relies on the standard curve, the diversity of RT-qPCR result in signi cant variability of the reported quantitative and qualitative data among different laboratories [14][15][16]. The divergences in qualitative data may result in misjudge development and severity of disease, and initiation or termination of antiviral therapies.
In this study, we established a new method for detection HCMV by Chip Digital PCR (cdPCR), which is a new potential generation digital PCR technology. The reliability of cdPCR in quantifying HCMV was evaluated and it was used in HCMV diagnosis in children leukemia.

Patients selection
110 children leukemia were enrolled in this study. The subjects were divided into 2 groups: leukemia group of 51 cases (M/F: 29/22, the median age is 7.5). Hematopoietic Stem Cell transplantation (HCT)group of 59 cases (M/F: 39/20, the median age is 7.5). Children patients were collected from May 2016 to May 2019 in Beijing Capital Institute of Pediatrics' Children's Hospital. All patients were con rmed according to diagnostic criteria. The subject was approved by Ethics Committee of National Institute for Viral Disease Control and Prevention. Whole blood samples from 110 patients were tested by RT-qPCR and cdPCR.

Primers and Probe
In the experiment, the primers and probe of cdPCR and RT-qPCR were designed according to the gene sequence (MK425187.1) as follows.

Statistical analyses
SPSS20.0 software was used to analyze data. Measurements were expressed by mean ± standard deviation (SD). Counting data were compared by the X 2 -test; measurement data were compared by the T-test, with statistically signi cant difference at P < 0.05.

Determination of optimal annealing temperature of chip digital PCR
To look for optimal annealing temperature of cdPCR, the temperature was set to increase progressively from 59℃ to 61℃ with an interval of 0.5℃. After viral DNA was extracted from 140 µL HCMV AD169 strain (5.67×10 6 TCID50/ml) according to Kit's instruction, 2µL DNA was used in each PCR reaction. The results showed that virus copies/25µL were 287.20 ± 36.85, 280.73 ± 11.71, 296.43 ± 3.11, 262.00 ± 28.16, 280.00 ± 13.72 at each temperature. The largest average copies was ampli ed at 60℃ of the annealing temperature. Thus, 60℃ was determined as the optimal annealing temperature of cdPCR.
To determine the speci city of cdPCR, 8 herpesviruses, including HSV-1, HSV-2, VZV, EBV, HCMV, HHV-6A, HHV-6B, HHV-7 were detected by HCMV cdPCR, respectively. Our results showed that DNA of 7 herpesvirus were not detected but DNA of HCMV. The results showed that HCMV cdPCR method were speci city with no cross-reaction with other herpesviruses.

Clinical samples validation of chip digital PCR
To understand whether cdPCR is suitable for sample detection, 110 whole blood samples from children leukemia were conducted by RT-qPCR and cdPCR, respectively. 31 samples were positive by RT-qPCR and cdPCR. The copy number was from 98 copies/ml to 3208 copies/ml. 79 samples were negative by RT-qPCR. However, seven out of 79 RT-qPCR negative samples were positive by cdPCR. The copy number of 7 samples was from 15 copies/ml to 73 copies/ml. The 7 positive samples were con rmed further by nest PCR and sequencing. The results showed that the method of cdPCR was more sensitive than RT-qPCR for HCMV detection (Table 1).   (Table 3), the range of copy number is from 179 copies/ml -3208 copies/ml. The prevalence of HCMV infection were 40.00% in the HCT patients aged 0-6 years old. And the prevalence of HCMV infection were 36.84% in the patients aged 7-12 years old ( Table 2). For leukemia patients, the prevalence of HCMV infection were 34.78% in the leukemia patients aged 0-6 years old and 26.09% in the patients aged 7-12 years old (Table 3).

Discussion
It has been con rmed that the HCMV viral load and growth in clinical samples can predict the risk of disease in patients [11]. But the lack of well-established viral load thresholds limits HCMV RT-qPCR in clinical application. Due to the results of RT-qPCR cannot be directly compared, the results of HCMV viral load cannot be used after change hospital [14]. No viral load value can be used to initiate preemptive therapy for patients infected HCMV without consensus standardization of HCMV RT-qPCR [10,14]. There is widely variability of viral load in 33 laboratories using different HCMV RT-qPCR in a study [18]. Though a World Health Organization (WHO) provide an international standard for calibration of HCMV RT-qPCR, viral load variability remains because of assay performance (limits of detection and quanti cation), sample type, method for nucleic acid extraction, gene target, and amplicon size, even the type of patients [19][20][21].
Digital PCR solves this problem well. Droplet digital PCR (ddPCR) and cdPCR are two types of commercial digital PCR platforms. Other studies have shown that the sensitive of digital PCR is signi cantly higher than that of RT-qPCR [22,23]. Our results also gave a similar conclusion. ddPCR mainly forms water-in-oil droplets, and each droplet is an independent PCR reaction system. Furthermore, the sensitivity of ddPCR for HCMV is 100 copies/ml [24]. And cdPCR limits detection of HCMV viral load is 15 copies/ml in clinical sample. cdPCR complete PCR reaction through 2D array of microchamber, and cdPCR can realize three-color multiplexing ampli cation [25,26]. Due to simpli ed steps, cdPCR effectively reduce risk of contamination.
Thus, we established a HCMV cdPCR method to evaluate HCMV infection in this study.
It is reported that RT-qPCR alone is inadequate for the accurate diagnosis of virus infection [27]. In this study, the cdPCR effectively detected 7 HCMV positive samples with low copies nucleic acid, making the laboratory diagnosis rate of HCMV increased by 6.36% (7/110). Our results showed that the lowest copies of detection by cdPCR was 14.58 copies/ml. The results suggested that cdPCR was suitable for low loads nucleic acid detection.
In addition, we also found that HCMV was more easily detected in whole blood than in serum. In this study, both serum and whole blood samples were collected in 44 of 110 patients (31 of 59 HCT patients and 13 of 51 leukemia patients). We found there were 3(3/31 = 9.67%) cases of HCMV positive in whole blood not in the serum from 31 HCT patients. There was 1 (1/13 = 7.69%) case of HCMV positive in whole blood not in serum from 13 leukemia patients. The results made the laboratory diagnosis rate of HCMV increased by 3.63% (4/ 110).
We still found that cdPCR was more sensitive in the whole blood samples than in serum. And in our study, the lowest copies number is 15 copies/ml in whole blood samples and 27 copies/ml in serum samples, respectively. In a word, we found that the detection rate of HCMV in whole blood was slightly higher than in serum both in leukemia and HCT patients. The reason maybe that HCMV can replicate in many cells, including epithelial cells, endothelial cells and leukocytes of peripheral blood [28][29][30]. Only in the presence of viremia, high levels of HCMV viral load can be detected in the serum.
As we all known, HCMV is the most common virus infection after transplantation, it is considered to be the major risk factor for transplantation. Studies have con rmed that almost all HCMV viremia after bone marrow transplantation occurs in HCMV-positive recipients, and only in few patients can be transmitted from donor [31]. In other words, it is necessary to pay attention to leukemia patients with HCMV and HCMVinfected bone marrow transplant patients. In our study, 34.54% of samples were positive for HCMV by cdPCR. More than 90% of HCMV infections occurred before the age of 12 in these patients with leukemia or HCT. Continuous monitoring and timely medication should be conducted to prevent postoperative viremia.
Among HCT patients, the infection rate of HCMV in male 13.56% (8/59) is higher than that in female 10.17% (6/59). But in leukemia patients, the infection rate in female 17.65% (9/51) is nearly 3 times than that of males 5.88% (3/51). This phenomenon needs to be further con rmed by large samples.

Conclusion
In conclusion, the cdPCR was established for detecting HCMV viral load in this study and veri ed through clinical samples, thus cdPCR may be used to evaluate clinical progress of HCMV infection.