Significant Improvement in Diagnosis of Hepatitis C Virus Infection by a One-Step Strategy in a Central Laboratory: an Optimal Tool for Hepatitis C Elimination?

DISCUSSION

The results from this study show how the application of a viral load reflex testing (or one-step) protocol for the diagnosis of HCV infections in a central laboratory allows an outstanding increase in diagnostic rates, particularly in the primary care setting, which in our discipline entails the main entry point for new cases of HCV infection, and drug treatment centers, where hepatitis C has the highest prevalence in our area (21%).

To implement this screening protocol, it was essential to first ensure the suitability of the sample in which the viral load was to be determined. The fact that the flow cytometry analyzer keeps samples capped while processing them—thus preventing contact with aerosols and environmental RNAses—strongly reduced the likelihood of contamination and deterioration of viral RNA. Furthermore, the probability of a sample free of HCV RNA matching a sample with a detectable viral load in the same analytical sequence was in fact much lower than the probability of both samples coexisting in the same anti-HCV antibody determination sequence in the serology analyzer (even taking into account the low probability in the Cobas 8000 analyzer due to the use of disposable tips). Despite the a priori favorable conditions to implement the reflex testing protocol, an occasional error in the washing system of the blood count analyzer could cause minimal contamination, like that observed in two of the nine analyzed samples in which HCV RNA was detected, although in amounts below the quantification limit of 15 IU/ml. In this regard, it is important to mention that, in naive patients, viral loads have been previously reported to be >3 logs in a very high proportion (EASL guidelines) or even >2 logs based on our own experience (14). In fact, in the cases analyzed in this study, after application of the HCV RNA reflex testing, only nine cases (one in the primary care setting and eight in the CAS setting) showed HCV RNA levels of <2 logs. Among them, only five cases with levels of <15 IU/ml (four in CAS and one in primary care) were observed, which, after further investigation, were all related to antiviral treatment. Hence, a value of <15 IU/ml should be considered irrelevant in the context of a new diagnosis (10). Still, faced with this potential risk, the need to evaluate the viral load in a new sample in cases with an HCV RNA level of ≤2 logs was included in the protocol (Fig. 2). Given the sampling path in our area, the sample collected for blood count analysis seems to be optimal for viral load reflex testing; however, other type of samples processed before could be used, provided that their adequacy (i.e., stability and lack of contamination) has been assessed. This possibility is also indicated in our recommended algorithm, which in fact is currently being implemented in all Catalonia.

In our experience, the traditional protocol, in which patients needed six visits for a full diagnosis (Fig. 1A), had a low ratio of HCV RNA confirmatory tests, especially in the ambulatory setting (non-hospital care), where the HCV RNA analysis was omitted in 9 to 82% of cases, depending on the source of the sample. This omission was particularly worrisome in the primary care setting, where HCV RNA analysis was omitted in the 28% of the around 2,000 anti-HCV-positive subjects from a total population of 58,000 individuals. These data confirm the low percentage of HCV RNA confirmatory test performance (46% to 73%) observed in protocols with multiple visits (8). Improvements observed after implementing the new one-step protocol were particularly notorious in the ambulatory setting (outpatients), with increases from 18% to 95% of HCV infection cases in CAS and from 72% to 92% in primary care centers: the contributions of reflex testing to the identification of new cases in these care settings were 54% and 28%, respectively. The spectacular increase in diagnostic rates in drug treatment centers is particularly relevant in terms of public health, due to the high prevalence of HCV infection among drug users. An effective diagnosis in this setting enables starting antiviral treatment, likely to result in HCV curation (15), a key cornerstone for HCV infection elimination (2, 7). However, a low rate of anti-HCV-positive cases remain undiagnosed due to lack of viral load testing; in our case, most are due to lack of hematimetry requests, meaning no samples for HCV RNA testing. In these cases, the validation of other types of samples, such as one’s own sample previously processed for serology, could be used for viral load reflex testing. However, this would require conducting a prior study ensuring its use, such as that performed here regarding hematimetry tubes.

Aside from the benefits of this technique at the level of public health, reflex-testing-based diagnosis means fewer visits to the doctor and unnecessary blood draws. This, in turn, translates into not only more comfortable patients, but also fewer subjects lost to follow-up and a shorter delay time for diagnosis and referral to a specialized center to receive treatment. It is noteworthy that the reflex testing protocol provided a diagnosis within a mean of 2 days, in contrast with the six-step protocol, which lasted weeks—or even months—due to administrative barriers to scheduling extra appointments. In addition to a shorter time for diagnosis, previous studies have found that the application of similar protocols are cost-effective (16). In this regard, it is important to highlight that the cost of a viral RNA determination (less than 30 euros in our laboratory) is much lower than the cost of a physician appointment and a blood extraction, so overall cost reduction may be of note. Furthermore, our algorithm includes request containment tools, such as avoiding duplicated determinations for patients with previous positive anti-HCV test results.

Alternative methods for testing for HCV infection outside of a central laboratory, such as the rapid diagnostic test (RDT) with capillary finger blood (e.g., OraQuick HCV) (17), have higher costs (i.e., 15 euros per test or higher) and poorer outcomes (<98% of correspondence). However, these methods can be beneficial in areas or population groups where the implementation of central laboratory strategies is unreliable (e.g., active drug users). Notwithstanding the benefits of other methods, central laboratory methodologies are strongly recommended by international guidelines for populations outside of the normal health system as well, thus allowing the implementation of one-step strategies, irrespective of the type of sample used. An example of an alternative sampling has been described by Lazarus et al., who implemented a one-step diagnosis strategy using whole blood sampled on dried blood spots (18). In this regard, the use of the GenXpert platform (Cepheid, Sunnyvale, CA, USA), which has shown an extremely optimal correlation with high-throughput systems (19, 20), represents an adequate alternative for HCV viral load determination when it is not possible to perform it in a central laboratory. This is the case of special populations like, for instance, active intravenous drug users, in which it seems advisable to conduct any viral load testing at the same center and start antiviral treatment as soon as possible. In these cases, the extreme high prevalence of HCV infections (>70% seroprevalence according to a recent meta-analysis) (21) would even make direct screening of viral loads cost-effective by means of GenXpert technology (Cepheid) in the same drug treatment center. However, in addition to the possibility of a concomitant determination of blood count and HCV RNA, HCV diagnosis with venipuncture for standard blood sampling in central laboratories had the advantage of the possibility of assessing patients’ liver fibrosis, which is necessary for their management in case of liver cirrhosis or advanced fibrosis (10, 22). In this regard, an HCV diagnosis strategy encompassing the FIB-4 index assessment promotes the patient’s start of treatment, while raising awareness on patients with advanced fibrosis. Clinical Practice Guidelines recommend combining noninvasive biomarker-based measurements, such as the FIB-4 index, with physical examinations, such as transient elastography (TE); however, not all sites have this technology (23). Aside from the high positive and negative predictive values of FIB-4 for the diagnosis of advanced fibrosis and cirrhosis (24, 25), combining the FIB-4 index and the TE examination in sequence increases the accuracy of the diagnosis of the fibrosis stage (26). The results found in this study show the high probability of advanced fibrosis in patients with a detectable viral load (17%) and those over 65 years of age (28%). These data provide a global vision of the significance of the degree of liver fibrosis in these patients and the importance of diagnosing the state of the infection early.

The extremely positive development of hepatitis C therapies is being darkened by the sad reality that the majority (80%) of people with chronic HCV infections remain undiagnosed (2). Achieving the WHO objective of eliminating viral hepatitis as a major public health threat by 2030 will require urgent strategies facilitating a massive scale-up of timely diagnoses as an essential step to drive infected individuals to treatment. It must be kept in mind that late diagnosis and treatment of chronic HCV infections may result in severe consequences for the health of both the individual and the population. In this regard, 30% of HCV-infected patients with late diagnosis will benefit less from achieving viral elimination, as well as those with advanced liver disease who are at high risk of developing decompensated cirrhosis, portal hypertension, and hepatocellular carcinoma (2, 27, 28). The occurrence of late chronic hepatitis C should be assumed as a failure of the health system to timely diagnose infected individuals. Overcoming this issue will require to promote early screening using a strategy as effective and simple as possible. Since the reflex testing strategy is driven by—and depends on—a central laboratory, patient treatment start (including the assessment of liver fibrosis) can be achieved without implementing sophisticated and expensive screening programs.

Our analysis has the typical limitations of an observational study in which all data have been obtained from a record designed with purposes of care and laboratory operations management. This design has allowed us to bring together data from 128,721 patients, providing an accurate perspective of the population reality; however, no patient social-demographic and clinical information could be collected, which could have improved the study. Furthermore, we must bear in mind that the scope of the results may be conditioned by the organization of the health care system and the laboratory itself, so, depending on the context, the contribution of the one-step protocol to the diagnosis of new cases of HCV could be lower or even higher.

In summary, our study shows the potential that implementing a diagnostic protocol based on reflex testing in a central laboratory may have to identify new cases of HCV infection significantly faster (2 days instead of weeks or months) and with much higher effectiveness than the traditional six-step process. Although results must be interpreted within the context of laboratory practice and the health care system of each area, our experience encourages the implementation of this type of analyses, which may notably contribute to the elimination of HCV infections. In this regard, it seems evident that universal screening programs for HCV infection represent the optimal solution to detect hidden HCV cases and start treatment, thus meeting the WHO aims. Given that the general population is mostly managed in the primary care setting, one-step diagnosis programs can effectively cover a high proportion of the general population at a low cost. Overall, the cost-effectiveness of implementing hepatitis C screening programs in the general adult population has been demonstrated (29). The availability of high-performance analytical platforms like Cobas 8000 and Cobas 6800, used in our laboratory and able to process thousands of samples in a normal working day, strongly facilitates the implementation of such screening programs. However, additional strategies for groups at high risk, very frequently not linked to the health system, might be necessary to reach the entire population of HCV-infected individuals. In this regard, the accurate and simple point-of-care testing platforms, specially the HCV finger-stick assay in the GenXpert analyzer (Cepheid), represent a very interesting alternative for these special populations.