Ongoing transmission of HCV: Should cesarean section be justified? Data mining discovery

Introduction

Globally, the prevalence of hepatitis C virus (HCV) infection in pregnant women is nearly 2–3% of the general population and around 3–10% of the vertically transmitted infections in children^{[1, 2]}. Although HCV can infect placental cells, most of the children are not regularly tested for perinatal HCV. However, a greater risk of vertical HCV transmission to the offspring may be much higher especially in HCV endemic countries such as Egypt. Egypt recorded the highest prevalence (14.7%) of HCV antibody (Ab) worldwide in 2014^[3]. It is noteworthy to know that the risks involved in cesarean section (CS) delivery have a significant high impact on vertical and horizontal HCV transmission^{[4, 5]}. The risk of transmission (one in every 20 children) delivered by chronic HCV-infected women highlights that vertical transmission likely constitutes the primary transmission route among children. These updated estimates are considered as a basis for decision making in prioritization of the research into risk-reducing measures^[6-8]. Recently, data mining (DM) programming allows clinicians to extract hidden factors that lead to such disease progression or regression in an artificial intelligence manner^[9]. The aim of this study was to estimate the risk of cesarean-section-associated vertical and horizontal HCV transmission in an Egyptian cohort.

Methods

We prospectively followed a cohort of 3,836 pregnant women of age between 18 and 43 years (childbearing period), attending prenatal care, unaware of HCV infection, from urban and rural areas across Egypt from January to June 2016. All the pregnant women were screened for HCV infection by third-generation ELISA and quantitative polymerase chain reaction (qPCR) tests. Patients who showed HCV-positive results were regularly followed up during the study period in randomized groups representing all Egyptian governorates. Furthermore, HCV-related complications such as cirrhosis, hepatocellular carcinoma (HCC), and esophageal varices were also investigated using noninvasive 2D ultrasound (US); additionally, routine laboratory tests were performed during pregnancy. Furthermore, all neonates born to HCV-pregnant women were investigated in the first 24 h postdelivery and before breastfeeding for both HCV Abs and HCV-RNA using ELISA and qPCR tests, respectively. Pregnant women positive to Abs and negative in qPCR test were excluded from the study. Those co-infected with HBV or HIV were also excluded.

All infected pregnant women were HCV genotype 4, the predominant genotype in Egypt. Demographic, laboratory, and clinical information were collected and entered into Microsoft Excel 2010.

Results

Among 3836 pregnant women screened for HCV, 620 were nulliparous (16.1%), 1,891 (49.3%) monoparous, and 1,325 (34.5 %) multiparous. Additionally, 3,682 (96%) were free of any medical diseases, 68 (1.7%) had history of hypertension, 46 (1.0%) had history of diabetes mellitus, 27 (0.7%) had history of cardiac diseases, and only 11 pregnant women (0.3%) had history of thyroid disorders [Table 1]. A case wise identification of HCV vertical transmission among study population was shown in Figure 1.

Modified decision tree modified by Rapid I, Berlin, Germany, showing the most independent factor for HCV ongoing transmission. PCR: polymerase chain reaction; HCV: hepatitis C virus; ELISA: enzyme-linked immunosorbent assay.

Eighty pregnant women (2.08%) showed positive HCV infection by both ELISA and real-time PCR HCV-RNA during third trimester or just before delivery. However, positive HCV antibodies were noticed in 99 pregnant women (2.6%). All HCV-positive women had history of surgical intervention (22.5%) or CS (77.5%). The mean (±SD) HCV-RNA identified by real-time PCR is 878,371 ±153,182.9 in HCV-positive women, ranging from 11,400 to 5,000,000. Ten neonates (21.5%) of all 80 deliveries showed positive HCV by both ELISA and PCR RNA within the first 24 h after delivery.

In the multivariate model, the area under the curve (AUC) value of the receiver operating characteristic curve was 0.709 (95% CI = 0.638–0.781). Calibration of the model by the Hosmer–Lemeshow test showed no significance (P = 1.000), which indicates that the model was correctly calibrated.

DM analysis showed that the only independent risk factor for such HCV transmission was CS; however, high viral load was an additional factor for vertical HCV transmission.

Discussion

HCV can cause acute and chronic hepatitis. Chronic HCV often progresses over many years, resulting in cirrhosis, HCC, liver cell failure, and the need for liver transplantation ^[10]. However, a considerable number of HCV-positive women are giving birth and most of their children are not being tested perinatally for HCV infection. The overall rate of mother-to-child transmission (MTCT) is 3–5%; however, coinfection with HIV increases the MTCT rate of HCV significantly. Mode of delivery does not represent significant risk factor for HCV vertical transmission; hence, infected HCV pregnant women not advised to have CS unless indicated for other causes, whatever if minded theoretically, exposure to HCV-infected blood as the infant passes through the birth canal during normal vaginal delivery (NVD) may lead to HCV transmission. However, exposure to maternal blood can also occur with CS; accordingly, there is no reason to offer elective caesarean section to HCV-infected parturients^[11-14], with respect to special circumstances when CS is preferred to NVD, such as membranous rupture and internal fetal monitoring are confirmed to be associated with HCV transmission.

CS and its complications were the main cause of death in many developing countries ^[15-18]. Egypt has the highest HCV prevalence worldwide; therefore, our results showed that the prevalence of HCV in a sample of Egyptian pregnant women was 2.08% (80 of 3,836 women), based on HCV Ab and PCR testing. All HCV genotype 4 pregnant women had history of surgical operations, likely the route of HCV transmission. Furthermore, Data suggest that CS, which accounted for 77.5% of all surgical interventions, can transmit HCV. In addition, 10 in 80 neonates (12.5%) showed HCV positivity by ELISA and PCR during the first 24 hours after delivery, confirming intrauterine HCV transmission. Our current study showed the significance of CS over other surgical procedures in overall HCV transmission. In addition, CS is more associated with maternal complications than NVD. Furthermore, some morbidity in childhood-associated cesareans seems more significant than what was believed. A recent article from Italy recommended that the most direct way for reducing CS in the country could be the formal prohibition of CS section based on maternal request ^[19-21]. There is little information about the timing of MTCT; currently, no interventions to decrease transmission rates have been identified yet ^{[22, 23]}. However, all vertical transmissions might occurred during intrauterine life because all children are evaluated during the first 24 h after delivery. HCV-RNA should be positive until 1–2 weeks following MTCT, when levels of HCV viremia reach the optimal threshold.

A history of blood transfusion, CS, and other surgical interventions are significant sources of HCV in Egyptian pregnant women using both traditional statistical methods and DM computational analysis. All pregnant women included in this study were from rural or urban areas and were unaware of vertical transmission of HCV or the future complications associated with such transmission to their children; therefore, they did not disclose their status to the obstetrician. Prenatal care and screening of neonates born to HCV-positive women can identify the infected group and determine the factors leading to current HCV transmission. For all reasons, screening babies born to HCV-positive mothers may discover disease early. As a result of all mentioned obstacles and the limits of HCV screening in pregnant women, it is more likely that CS will continue to be a source of such HCV transmission worldwide, especially in developing countries where inadequately tested blood, unsterile medical injections, and un-medical indication of CS without ethical justification still remain. Failure to identify all HCV-infected women may drive the world to adopt universal prenatal HCV screening. Hence, we need a worldwide recommendation to reduce the cesarean section rate by attenuating the birth fear of patients, mass education about benefits of vaginal delivery, encouraging vaginal delivery by social media and teaching national seminars, and providing comprehensive counseling with all health providers about the medical issues of labor.

DM is a breakthrough in computational analysis overcoming mathematical and statistical analyses in space, labor, biology, engineering and different computer sciences, and, recently, medicine. The use of DM in prediction medicine should discover the important factors related to disease progression or regression by extracting hidden factors that have never been identified by the usual statistical programs. The use of DM in clinical medicine is important for morbidity and mortality prediction. For example, whenever we sign up for an Internet shopping website using VISA card, place a purchase using a credit card, or surf the Web, we are creating every purchasing data. Furthermore, these data are stored in large sets on powerful computers owned by the companies we deal with every day that they understand our mind for purchase. The same goes for physicians and researchers who understand each disease behavior ^[24-31].

Simple distribution Naïve Bayes and decision tree created by Rapid I version 4.6 emphasized the level of maternal HCV-related vertical transmission; > 3× 10⁶ IU should be considered as the high level of vertical prediction, reported in our previous work ^[13], and CS should be likely an important route of global HCV transmission especially in developing countries [Figures 1 and 2]. However, the level of maternal viremia is still a matter of scientific debate in such vertical transmission. We have created a figure showing the global HCV transmission [Figure 3].

Simple distribution Naïve Bayes diagram created by Rapid I version 4.6 shows the level of maternal HCV G4 related vertical transmission; > 3× 10⁶ IU should be considered as the high level of vertical prediction. PCR, polymerase chain reaction

Global HCV transmission—based on MTCT.

In conclusion, screening women who had experienced surgical intervention or CS during child bearing period and before pregnancy might prevent HCV MTCT. CS should be ethically justified to decrease global HCV transmission.