The Table shows the geographic origin, age range, sex, and clinical score of each of the 21 patients. Patients traveled to Albuquerque, New Mexico, as funds permitted from the University of New Mexico’s Research Allocation Committee. The patients were subjected to a battery of clinical tests, including studies examining pulmonary, renal, and hepatic function. The results of the clinical studies will be reported separately.

Serum samples were collected for virologic studies and examined for residual viral RNA by RT-PCR and for neutralizing antibodies. Although positive control RNA preparations produced amplification products as expected, no viral RNA was detected in any of the patients’ serum samples (data not shown). However, the serum samples examined continued to exhibit neutralizing antibodies at titers ranging from 100 to 3,200 (Figure). No statistically significant relationship existed between neutralizing antibody titers after the patient’s recovery from SNV infection and the severity of the previous illness, nor did a relationship exist between antibody titers and the use of extracorporeal membrane oxygenation (ECMO) in our patient group.

Although case-fatality ratios for HCPS have declined since the initial description of the disease, probably reflecting improved recognition of mild illness, therapeutic options remain limited. The efficacy, if any, of antiviral drugs such as ribavirin remains to be demonstrated. Intensive care management of the manifestations of the disease constitutes the primary method of clinical intervention. At some centers, ECMO has been utilized in those patients who exhibit hemodynamic or pulmonary instability that is predictive of fatal outcome (9). (Six of our patients underwent this procedure.)

Although treating patients with viral hemorrhagic fevers with the plasma of patients who have recovered from the disease is often proposed or discussed, rigorous demonstration of its efficacy in humans or animals is often lacking. Argentine hemorrhagic fever (AHF) due to Junín virus and other arenaviral diseases are among the exceptions (10–13). The efficacy of convalescent-phase plasma in treating AHF was demonstrated before the nature of the protective substance contained therein was positively identified. With the advent of a test for neutralizing activity, Argentine investigators retrospectively demonstrated a correlation between the quantity of the neutralizing activity of plasma and the degree of protection afforded by its infusion (11).

The neutralizing activity of a dose of plasma is defined by multiplying the reciprocal of the highest dilution that would reduce the number of viral foci by 80% by the volume of the inoculum, in milliliters. Overall, the infusion of convalescent-phase plasma containing antibodies to Junín virus reduced deaths by approximately 90%. In a retrospective analysis, investigators showed that those patients who received >2,000 U/kg of neutralizing antibodies appeared to respond better than did those patients who received 1,000–2,000 U (11).

Because HCPS is rare, the number of treatments that can be independently examined through clinical trials without delaying the identification of an effective intervention is inherently limited. To determine whether a clinical rationale exists for attempting passive immunotherapy for HCPS with neutralizing antibodies, we first studied the relationship between the titer of such antibodies on a patients’ admission to the hospital and the clinical outcome. That study showed a profound and inverse relationship between antibody titer and disease severity (6).

It then became important to determine whether the small number of patients who have recovered from SNV infection might constitute a pool of plasma donors of sufficient size to provide a stable source of plasma to treat future patients. The study reported herein was initiated to determine whether neutralizing antibodies persist for substantial periods after recovery from the disease. Results indicate that the answer is affirmative but that not all recovered patients are equally rich sources of neutralizing antibodies.

Modern techniques for collecting plasma allow for convenient and low-risk preparation of 500–600 mL or more of plasma from healthy donors; this can be repeated on multiple occasions over several weeks. For that and other reasons, each of those patients whose titers remained at >800 represents a rather rich source of antibodies. If one extrapolates from the experience of the Argentine investigators in their treatment of AHF, a single 600-mL plasma donation by a person with a neutralizing antibody titer of 800 can be used to treat two 80-kg (~175-lb) patients. A patient such as the person whose titer was 3,200 could be identified as a particularly valuable repeat donor, one whose every donation could be divided and used to treat eight new patients with HCPS.

Thus far, little or no overlap occurs among the behavioral characteristics associated with HCPS and those associated with infection by bloodborne pathogens such as HIV or hepatitis C virus. Potential donors would be subjected to the same battery of questions and tests for bloodborne pathogens as would any other directed blood donors, with the exception that donors would be allowed to donate despite the history of hospitalization for HCPS and receipt of blood products during ECMO therapy. Although we therefore expect that most potential donations of convalescent-phase plasma would be free of bloodborne pathogens, ensuring that SNV has been completely cleared from the blood is more difficult. Thus far, no evidence for a persistent state of SNV infection in humans has been shown. In fact, studies published to date show that in all cases that could be evaluated, SNV is undergoing rapid clearance at the time of clinical presentation (14).

Although our RT-PCR data are reassuringly negative, we cannot know whether a larger dose of convalescent-phase plasma would contain infectious SNV. In fact, SNV has never been cultivated from any human source, despite attempts by multiple laboratories.

We believe that passive immunotherapy may be efficaceous if administered very early in the course of illness, before the severity of a particular patient’s disease course is known. To reduce the risk for accidental inoculation of potential recipients of anti-SNV convalescent-phase plasma, at least two strategies are possible. First, one could impose a delay of 6 to 36 hours by requiring that potential recipients have a positive specific test for SNV antibodies before the plasma is administered. Alternatively, one could impose either purification or viral inactivation procedures for the antibodies or await the development of human or humanized monoclonal antibodies before instituting this therapy. We presume in our risk analysis that no new risk is associated with the inadvertent inoculation of a second strain of SNV in a patient who is undergoing active infection with a different genotype of the same virus, since there is no a priori evidence to suggest otherwise.

No current evidence suggests that infusing anti-SNV antibodies would enhance the pathogenesis of SNV. The data of Bharadwaj et al. (6), in fact, support the presumption that treatments that lead to higher titers of neutralizing and even nonneutralizing anti-SNV antibodies are more likely to lead to less severe disease. We believe that convalescent-phase plasma therapy may represent a practical intervention that is feasible, given current resources and funding levels for potential antihantavirus therapies.