8
Detection and Diagnosis
The term "detection" is used here to denote identification of the virus in the environment, while "diagnosis" refers to determination that the virus or pathogen has infected a human host. The need to detect variola virus could arise as a result of experimentation with the virus under BSL-4 conditions in well controlled laboratories, but is perhaps more likely to occur as a result of experimentation with unregistered variola virus under less optimal conditions. Detection technology could provide additional safety by offering proof of containment if research were to be conducted on live variola virus. Exposure to variola virus could also result from the intended or unintended action of a terrorist individual or group or the planned action of a rogue state. Should such an event occur, timely environmental detection and early diagnosis of human infection would be extremely valuable.
The development of sensitive and specific detection and diagnostic strategies would probably involve identification of variola virus nucleic acid or protein. Such approaches would be dependent on knowledge of the range of variability in natural variola sequences and/or the sequences of their encoded proteins. It is likely that from current knowledge of the sequences of individual orthopoxvirus genes and from the complete sequences of three variola major virus isolates that are available, polymerase chain reaction (PCR) primer pairs capable of differentially amplifying DNA segments from all previously sequenced orthopoxviruses could be selected.
Such primers would probably be valid for the sensitive detection of any variola virus. The sequencing of additional strains of variola virus DNA from dispersed geographic origins, however, would provide an additional margin of validity with minimal additional risk or cost. Sequences that are unique to vari-
ola virus have greater potential for variability among different strains than sequences that are common to the orthopoxvirus family. Therefore, additional sequencing of variola virus samples is critically important for development of the nucleic acid detection technology needed to identify variola virus, as opposed to other orthopoxviruses. There is no guarantee that an emerging strain would be represented in the archive, but understanding of the variation would assist in determining the relation of the new strain to known isolates. Moreover, the specific identification of variola virus would be a necessary feature of detection or diagnosis of variola virus infection should the precise source of the infection be unknown. The two variola major isolates that have been sequenced come from India and Bangladesh, within or near the Indian subcontinent, and one variola minor isolate comes from Brazil. The degree of similarity between these variola virus sequences and those of strains of variola from other parts of the world is unknown. Research on this issue—including sequencing of the entire genome or selected genome segments of additional isolates, or extended PCR and restriction fragment length polymorphism (RFLP) assay of entire genome DNA—should move forward as rapidly as possible.
It may be noted that additional sequencing and PCR/RFLP assay of other orthopoxvirus DNA sequences may be important for specificity issues. This is especially true for monkeypox virus DNA. Little is known about the range of DNA sequence variability among isolates of monkeypox virus.
Concerns analogous to those regarding DNA sequence conservation and variability would hold for protein- or antibody-based detection or diagnostic strategies. Some work on validating target sequences for sensitive or specific detection or diagnosis has been done at CDC and at VECTOR. The current state of that work needs to be critically assessed.
Environmental Detection
Environmental detection would involve instrumentation for sampling the environment. For example, a vacuum system with a contained filter trap could be developed for air sampling, and various adsorbents could be prepared for surface sampling. For the most part, any orthopoxvirus could be used in place of variola virus to devise suitable sampling techniques, assuming similar biophysical characteristics and stability (as appears to be the case). However, variola virus or possibly another live orthopoxvirus expressing variola virus surface protein(s) or containing variola virus DNA might be essential for final testing of a detection technology and strategy. For example, a filter coated with a variola-specific monoclonal antibody might trap variola virus. Identification of the virus might then rely on PCR for DNA testing or a second antibody for antigen testing. A solely protein-based detection assay, such as those using monoclonal antibodies in antigen capture, would be relatively insensitive, but might be faster
and more portable. Another live orthopoxvirus containing one or a few variola virus genes would be preferable to live variola virus for final test validation because of safety concerns. However, current policies prohibit the making of recombinant poxviruses containing any variola virus gene(s), and it is conceivable that variola proteins expressed in such a recombinant might not exhibit authentic structural and functional properties.
Furthermore, little is known about the surface proteins of variola. Detection is based solely on surface components of other orthopoxviruses that are extrapolated to variola components. This extrapolation is based on predictions that depend on the homology of the corresponding DNA of these surface proteins and those of variola virus. Limited biochemical studies of variola surface proteins would therefore be valuable. These studies could be conducted prior to destruction of the virus stocks.
One alternative might be to construct a variola virus recombinant that was incapable of replicating in normal human cells and to use such a recombinant for test validation under highly restricted conditions. Altering the regulations to permit limited insertion of part or all of a variola virus gene into vaccinia virus would be preferable to undertaking work with a host-range-restricted variola virus recombinant. Nonetheless, there would still be considerable security and safety concerns associated with using an almost complete variola virus. Moreover, the rigorous testing needed to prove the authenticity of the protein structure and function of a host-range mutant might make such an approach time-consuming and excessively expensive.
Diagnosis of Infection
Sensitive diagnosis of variola virus infection at the earliest stages would most likely be accomplished using some form of nucleic acid amplification technology to detect variola in saliva, sputum, blood, or lymphatic aspiration. However, initial epidemiological or medical interventions might be indicated solely upon finding orthopoxvirus-specific nucleic acid in human material in a setting in which variola virus infection was a possibility. Although technology for detecting antibodies to variola virus would be of limited or no utility in the first few days after exposure, robust antibody detection schemes might be highly useful for epidemiological surveillance following an outbreak.
Use of host-specific components—for example, in blood, saliva, or urine—to screen for variola virus upon early indications of infection might be feasible were the pathogenesis of variola better understood. IgM (immunoglobulin M) antibodies might be detectable within 2 weeks of exposure and IgG antibodies within 3 to 4 weeks of exposure. Detection of antibodies may be less expensive, more portable, and more stable than detection of nucleic acid or antigen. Rapid
and safe type-specific bedside tests would be useful in the event of potential variola infection.
Diagnostic strategies could be verified using mouse or subhuman primate models of orthopoxvirus infection. A recombinant vaccinia virus carrying all or part of a variola virus gene would be useful for a step in the validation, subject to the concerns expressed above with regard to detection strategies. Research with a vaccinia virus construct carrying a single variola gene would not be constrained by the need to work in BSL-4 facilities. Some parallel work with monkeypox virus diagnostics in monkey models would be desirable to validate ease of detection in the context of an analogous primate virus-host model.
Alternatives to Live Virus
The development of detection and diagnostic strategies would not require live variola virus per se, beyond the need for additional recombinant DNA stocks and sequencing. Much of the developmental in vitro research could be done with isolated recombinant variola virus DNA and recombinant produced protein. The requirement for virus for field epidemiological detection, for particle stability verification, and for diagnostic test validation in experimental models could be bypassed by using vaccinia virus as a representative of the orthopoxvirus family. A vaccinia virus recombinant containing all or part of the relevant variola virus DNA segment would be useful for the validation of tests in appropriate epidemiological or experimental animal infection model systems, although assessment of the authenticity of the recombinant as compared with live variola virus would be needed. Parallel work with monkeypox virus might add a small margin of additional validation.