HIV Detection in Cuba : Role and Results of the National Laboratory Network

This article explains how laboratory diagnosis based on an extensive laboratory network effective confirmatory algorithms and quality control with external evaluation is an essential component of Cubas National HIV/AIDS Program. It describes the role of laboratory diagnosis in the program and discusses how to develop an appropriate algorithm for the laboratory. In addition it touches on diagnostic quality control and building laboratory capacity for the future.


Policy & Practice
Timely and accurate detection of HIV infection is critical due to the seriousness of AIDS and its impact on individuals and families, the long symptom-free incubation period, the benefi ts of enrolling patients as early as possible in therapy and follow-up, and as a pillar of population-based prevention strategies aimed at curbing the epidemic.Thus, an essential component of Cuba's National HIV/AIDS Program (full title, National Program for Prevention and Control of Sexually-Transmitted Diseases and HIV/AIDS) is laboratory diagnosis, [1,2] based on an extensive laboratory network, effective confi rmatory algorithms and quality control with external evaluation.

Role of Laboratory Diagnosis in a National HIV Screening Strategy
The National HIV/AIDS Program was established months after the fi rst AIDS cases were diagnosed in Cuba in 1985.The extremely low HIV prevalence at the time offered a window of opportunity to establish early diagnosis as a central strategy for prevention and containment.Thus, a national screening program was developed for all blood donors, and HIV tests were indicated for vulnerable, at-risk groups including: Cubans returning from international service (among whom the fi rst cases had been diagnosed); pregnant women; contacts of seropositive individuals; persons with other sexually-transmitted diseases; persons who sought testing on their own; patients whose clinical evaluation required ruling out HIV; hospitalized patients, and individuals identifi ed through routine surveillance.[1,2] From the start, the National Program envisioned incorporation of strong diagnostic capabilities.Accordingly, in 1986, the Ministry of Public Health (MINSAP, its Spanish acronym) established a nationwide laboratory network to provide the necessary technical backbone for national HIV screening.The network, extended throughout Cuba, began providing services for blood banks, Provincial Hygiene and Epidemiology Centers, hospitals and primary care facilities.The number of laboratories in the network rose from 33 in 1986 to 64 a decade later and to 115 by 2009 (Figure 1).
The network required a highly effective diagnostic protocol, assuring both high sensitivity and specifi city (MINSAP, National HIV/AIDS Program, Cuba 1985), and diagnostic strategies lacking confi rmatory testing were rejected from the start.[2] Throughout the years, the network's methodology and organization have been guided by a national reference center, the National Retrovirus Diagnostic Reference Center (LNR, its Spanish acronym, located in the AIDS Research Laboratory, LISIDA), which also carries out the confi rmatory testing; procedures developed fully comply with standards set by the national regulatory agency: the Government Center for Drug Quality Control (CECMED, its Spanish acronym); [3] LNR maintains CECMED certifi cation.

Developing an Appropriate Algorithm for the National Laboratory Network
The diagnostic algorithm displayed in Figure 2 was established on the basis of WHO recommendations and consultation of other expert opinions, [4][5][6][7] using tests from various suppliers previously registered for use in Cuba [8] to enable clarifi cation of diffi cultto-interpret reactivity patterns, follow-up of indeterminate Western blot results for HIV-1, and confi rmation of HIV-2 infections.[9] HIV

Policy & Practice
Nevertheless, while the WHO recommends algorithms that do not rely on a single commercial supplier, we have found domestic test production to be an essential pillar of the sustainability of the vital diagnostic capability of the National Program.Some imported tests from reliable suppliers are still used, but most specimens are examined with domestic tests, stabilizing the diagnostic process.These tests include primarily UMELISA \ HIV 1+2 RECOMBINANT (TecnoSuma, Cuba) [10] for screening and DAVIH-blot (trademark of the Western blot test used by DAVIH Laboratories, Cuba) for confi rmation at LNR.Samples received at LNR from network laboratories are analyzed with a commercial ELISA system with a different principle and antigenic composition than that used for screening.(VironostikaHIV Ag/Ab, Biomériux).[11,12] Repeatedly non-reactive (RNR) samples are reported as negative, while repeatedly reactive (RR) samples are then tested with the DAVIH-blot system, with three possible results: positive, negative, and indeterminate.Positive DAVIH-blot samples are reported to the National HIV/AIDS Program so that, after counseling and informed consent, a second confi rmatory specimen can be taken.Indeterminate or negative results are tested with a commercial ELISA preparation specifi c to HIV-2 and those that are RR are analyzed with Western blot for HIV-2, which also has the same three possible results: positive, negative and indeterminate.A second sample is also requested from individuals with positive Western blot for HIV-2 (DAVIH-blot HIV-2, DAVIH Laboratories, Cuba).Persons with indeterminate results for any of the Western blot tests remain in serological and epidemiological follow-up for at least one year.[1,3] While several international organizations require reactivity to only one of the three envelope glycoproteins to defi ne reactivity in Western blot analysis, [7] the Cuban algorithm complies with the WHO criteria, considered the most rigorous.[13] WHO criteria specify reactivity to at least two viral envelope glycoproteins for a result to be considered positive.A negative result is defi ned by non-reactivity; any other reactivity pattern not including two envelope glycoproteins is considered indeterminate.[13] We chose these criteria because they are more specifi c and include careful follow-up of individuals with indeterminate results, allowing us to catch later seroconversion and minimizing false positive diagnoses.
We rejected the WHO option of dispensing with supplementary Western blot testing and instead using one or more different ELISA systems with different principles and antigens depending on specifi c diagnostic objectives and infection prevalence in the population from which the specimen is taken.The situation in Cuba, with a very low prevalence of HIV infection in the general population and a national program of medical and social attention to all those affected, requires an algorithm with very high specifi city, including confi rmatory tests to avoid false positives.[14] Defi nitive HIV infection diagnosis is made by serologic confi rmation with supplementary DAVIH-blot in a second specimen, complemented by epidemiologic criteria for exposure and clinical fi ndings specifi c to each case.
Infants born to seropositive mothers are followed until age 18 months using a special diagnostic algorithm, including Western blot, PCR detection of maternal viral genetic material and detection of circulating antigens using ELISA for protein 24 Kd (p24) capture.[15,16] PCR and p24 antigen detection are also used in follow-up of persons with indeterminate Western blot results and to clarify samples with diffi cult-to-interpret reactivity patterns.[1]

Diagnostic Quality Control
In 1989, LNR's responsibilities were expanded to include evaluation and certifi cation of all diagnostic tests used for screening and confi rmation; scientifi c research mainly aimed at developing new diagnostic tools; and training of personnel for the national laboratory network.It was also charged with establishing a National Program for External Quality Control (PEEC, its Spanish acronym) of network laboratories.
Tests included in the diagnostic algorithm are evaluated according to established regulatory standards, [8] using serum or plasma panel rounds previously characterized by similar sys-* Policy & Practice tems.From 2003 through 2009, 15 serum panel rounds were distributed across the network, with over 80% laboratory participation and a quality index of >95%.Sensitivity and specifi city of the diagnostics evaluated at that time were >97%, comparable to internationally accepted results.[17][18][19][20][21] Simultaneously, registries were designed not only to preserve patient confi dentiality, but to ensure traceability of the diagnostic process.Specimen storage and the transport chain from health services to network laboratories and LNR were also strengthened.
PEEC has pinpointed potential vulnerabilities in the functioning of network laboratories, contributed to improving good laboratory practices and to assessing overall performance of the diagnostic component of the National HIV/ AIDS Program.For example, laboratories with poorer performance in the panel rounds were identifi ed and plans developed to assist them to meet quality standards.

Results through 2009
Table 1 gives an overview of the laboratory network's activity from 1986 through 2009.Over 40 million tests have been performed, with an annual average of nearly 1.7 million.Given that the mean number of new HIV infections detected annually was 509-persons averaging two tests-then the total number of persons screened annually can be estimated at over 1.5 million, or about 13.6% of the Cuban population.Note that LNR confi rmatory testing volume more than doubled between the 2000 through 2004 and 2005 through 2009 periods (mean number of tests annually 17,230 vs. 8181), a statistically-signifi cant difference (p < 0.001); while increase in the laboratory network's testing volume between the same periods was not statistically signifi cant.
The mean annual number of confi rmed positive specimens also increased between the same two fi ve-year periods, from 962 to 2162 (p = 0.008), as did overall proportion of seropositive tests (12.5% vs. 11.8%,p < 0.001).
Table 2 shows the number of tests performed, seropositive cases detected, number of tests per case detected, and annual seropositivity rate in the population screened.Cumulatively, a mean of 509 individuals were found HIV-positive annually from 1986 through 2009; however, the mean numbers of new diagnoses annually during the 2000 through 2004 and 2005 through 2009 periods were both statistically signifi cantly greater than the cumulative historical mean (668 and 1238, respectively, p = 0.001 in both cases).
The mean number of tests done for every seropositive case decreased signifi cantly over time (p < 0.001).Nonetheless, despite the rising trend in detection of HIV in specimens received, cumulative population seropositivity by the end of 2009 was still only 0.03%.In 2009, more than 2 million tests were done, of which 0.071% were positive.
Annual numbers of HIV-positive persons detected annually are shown in Table 3, differentiated by HIV virus type.Unlike HIV-1, HIV-2 showed a declining trend in new diagnoses, with fewer than one new case annually from 2005 through 2009.This is not statistically signifi cant from the cumulative annual mean (0.4 vs. 0.75 per year, p = 0.47), not surprising with such small numbers.
Since the increase in confi rmatory testing volume at LNR was accompanied by an increase in seropositivity rates in specimens received, we conclude that it was infl uenced by both increased seroprevalence and active case-fi nding among groups at risk.The analysis of seropositivity reported here is by specimens received for confi rmation, not by population group, but it is interesting that the data are consistent with epidemiologic reports on HIV prevalence in different study populations in Cuba.Prevalence of HIV-1 in those aged 15 to 49 years was 0.1% in 2009, among the lowest prevalences in the Caribbean.[22] Knowing that HIV-2 is less transmissible than HIV-1 [23] and that the fi rst HIV-2-positive individuals in Cuba were infected in West Africa [3] does not explain why we have been unable to detect a source or transmission to the most recent individuals diagnosed.Epidemiologic follow-up of their sexual partners would help understand the true HIV-2 situation in Cuba.

Creating Knowledge and Building Laboratory Capacity for the Future
We are currently evaluating new strategies to decentralize the initial confi rmatory process to reduce the number of reactive specimens sent to LNR for defi nitive diagnosis.Such a strategy requires building greater capacity in regional and provincial laboratories, part and parcel of LNR's ongoing work.Besides its reference function and its roles in quality assessment

Policy & Practice
and assurance, and scientifi c research, LNR also builds human resource capacity through training and continuing education of professional and technical staff in network laboratories.Personnel in the laboratory network include an array of technicians in various biomedical sciences and medical specialists in microbiology and laboratory medicine; it is a highly-qualifi ed workforce that is predominantly female.
The Immunoassay Center (CIE, its Cuban acronym) courses on SUMA technology, offered by TecnoSuma regional centers, are a prerequisite to working in HIV detection and diagnosis.Staff also follow a systematic training program that includes the latest diagnostic techniques, good laboratory practices, and related topics.To this end, LNR has held 11 national workshops, with participation by staff from all network laboratories.
Cuba benefi ts from a broad network of laboratories providing coverage for blood bank safety, epidemiologic surveillance in vulnerable groups, and services with high social impact, such as early diagnosis in pregnancy, diagnosis of perinatal infection, and anonymous counseling.This strategy helps ensure timely access to medical attention under the National HIV/AIDS Program and to contain the spread of infection within the population.

Figure 2 :
Figure 2: Diagnostic algorithm for Cuba's National HIV/AIDS Program