Screening for Babesia microti in the U.S. Blood Supply

Transcription

1 The new england journal of medicine Original Article Screening for Babesia microti in the U.S. Blood Supply Erin D. Moritz, Ph.D., Colleen S. Winton, S.B.B. (A.S.C.P.), Laura Tonnetti, Ph.D., Rebecca L. Townsend, B.A., Victor P. Berardi, Mary Ellen Hewins, B.S., Karen E. Weeks, B.S., Roger Y. Dodd, Ph.D., and Susan L. Stramer, Ph.D. From Scientific Affairs, American Red Cross, Gaithersburg, MD (E.D.M., C.S.W., L.T., R.L.T., R.Y.D., S.L.S.); and the Research Division, Imugen, Norwood, MA (V.P.B., M.-E.H., K.E.W.). Address reprint requests to Dr. Stramer at Scientific Affairs, American Red Cross, 935 Gaither Rd., Gaithersburg, MD 877, or at susan. redcross. org. N Engl J Med 6;375:36-5. DOI:.56/NEJMoa6897 Copyright 6 Massachusetts Medical Society. ABSTRACT BACKGROUND Babesia microti, a tickborne intraerythrocytic parasite that can be transmitted by means of blood transfusion, is responsible for the majority of cases of transfusiontransmitted babesiosis in the United States. However, no licensed test exists for screening for B. microti in donated blood. We assessed data from a large-scale, investigational product-release screening and donor follow-up program. METHODS From June through September, we performed arrayed fluorescence immunoassays (AFIAs) for B. microti antibodies and real-time polymerase-chain-reaction (PCR) assays for B. microti DNA on blood-donation samples obtained in Connecticut, Massachusetts, Minnesota, and Wisconsin. We determined parasite loads with the use of quantitative PCR testing and assessed infectivity by means of the inoculation of hamsters and the subsequent examination for parasitemia. Donors with test-reactive samples were followed. Using data on cases of transfusiontransmitted babesiosis, we compared the proportions of screened versus unscreened donations that were infectious. RESULTS Of 89,53 blood-donation samples tested, 335 (.38%) were confirmed to be positive, of which 67 (%) were PCR-positive; 9 samples were antibody-negative (i.e., antibody-negative sample per 996 screened samples), representing 3% of all PCR-positive samples. PCR-positive samples were identified all through the year; antibody-negative infections occurred from June through September. Approximately one third of the red-cell samples from PCR-positive or high-titer AFIA-positive donations infected hamsters. Follow-up showed DNA clearance in 86% of the donors but antibody seroreversion in 8% after year. In Connecticut and Massachusetts, no reported cases of transfusion-transmitted babesiosis were associated with screened donations (i.e., cases per 75,33 screened donations), as compared with cases per 53,3 unscreened donations (i.e., case per 8,7 unscreened donations) (odds ratio, 8.6; 95% confidence interval,.5 to ; P =.5). Overall, 9 cases of transfusion-transmitted babesiosis were linked to blood from infected donors, including blood obtained from donors whose samples tested positive on the PCR assay to 7 months after the implicated donation. CONCLUSIONS Blood-donation screening for antibodies to and DNA from B. microti was associated with a decrease in the risk of transfusion-transmitted babesiosis. (Funded by the American Red Cross and Imugen; ClinicalTrials.gov number, NCT589.) 36 n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

2 Screening for Babesia microti in U.S. Blood Supply Babesia microti is an intraerythrocytic parasite that causes babesiosis. The severity of babesia infection ranges from asymptomatic, most commonly in healthy persons, to fatal, most frequently in persons older than 5 years of age, those who have no spleen (or no functional spleen), and those who are immunocompromised. In the United States, B. microti is transmitted to humans primarily by means of the bite of Ixodes scapularis (also called the deer tick). 3 Babesiosis became a nationally notifiable disease (as defined by the Centers for Disease Control and Prevention [CDC]) in and was reportable (i.e., reportable to the state, which then notifies the CDC) in 7 states in 3, with 95% of 796 cases reported by 7 states: Connecticut, Massachusetts, Minnesota, New Jersey, New York, Rhode Island, and Wisconsin. The occurrence of babesiosis resulting from the bloodborne transmission of parasites from infected donors to recipients through transfusion of red-cell units or platelets that have been contaminated with red cells is of increasing concern. 5 From 979 through 9, a total of 6 cases of transfusion-transmitted babesiosis were reported in the United States, 59 of which were caused by B. microti. These numbers probably underestimate the incidence of transfusiontransmitted babesiosis, because infections may be overlooked or misdiagnosed and because investigations to determine the source of babesia infection in the blood recipient may not be initiated. 6 No blood-donation screening tests for B. microti have been licensed by the Food and Drug Administration (FDA). However, donors are asked whether they have ever had babesiosis, and those who respond yes are indefinitely deferred; this method is considered to be ineffective. 5,7 In June, the American Red Cross began screening blood obtained at selected blood drives using an investigational product-release testing protocol that consisted of an IgG arrayed fluorescence immunoassay (AFIA) to detect B. microti antibodies and a polymerase-chain-reaction (PCR) assay to detect B. microti DNA. We used data collected during screening, donor follow-up, and investigations of cases of transfusion-transmitted babesiosis to assess the natural history of infection in blood donors and the effect of screening on blood safety. Methods Assays and Screening Protocol The characteristics and performance of the screening assays (AFIA and PCR assay) and supplemental assays for the confirmation of reactive samples (in-house IgG and IgM Western blots, enhanced-sensitivity digital PCR assay, and quantitative PCR assay; all from Imugen) have been described previously (see the Supplementary Appendix, available with the full text of this article at NEJM.org). 8 In brief, the AFIA used B. microti piroplasms as an antigen substrate, and for the purposes of screening, a titer of :8 or more was considered to indicate reactivity of the index donation sample; titers that were less than :8 were considered to indicate a negative result. DNA was extracted with the use of an automated membrane-based isolation system (Taigen Bioscience). PCR primers and probes targeted the B. microti 8S ribosomal RNA gene; the 95% lower limit of detection was 66 piroplasms per milliliter. Samples that were reactive on AFIA but negative on PCR assay were retested by enhancedsensitivity PCR assay (95% lower limit of detection, piroplasms per milliliter). A quantitative PCR assay was performed on samples that were reactive on PCR assay and enhanced-sensitivity PCR assay to determine the parasite load. 8 A donation sample was considered to be reactive if the result was initially positive or inconclusive on the screening AFIA, PCR assay, or both. A sample was considered to be confirmed as positive if the index donation sample tested positive on supplemental testing or if it infected hamsters (see below and the Supplementary Appendix). In targeted areas of Connecticut, Massachusetts, Minnesota, and Wisconsin that were selected on the basis of the number of babesiosis cases reported to health departments and on a previously published study of prevalence, 9 donors who met current donor-suitability criteria were given information regarding the study and were prospectively screened unless the donor opted out. Screening was conducted from June through September on Mondays through Thursdays at selected drives so that the number of screened donation samples was approximately 5 to per day. This procedure was approved by the institutional review board of the American Red Cross and was considered to be an acceptable form of A Quick Take is available at NEJM.org n engl j med 375;3 nejm.org December 8, 6 37 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

3 The new england journal of medicine providing informed consent. Whole-blood and plasma samples were collected in 5-ml EDTA tubes and were tested for B. microti DNA by PCR assay and for B. microti antibodies by AFIA before the release of the product (product release was defined as release for transfusion, on the basis of negative test results). Results were available within hours after receipt of the sample. Blood donations that were identified by screening as AFIA-reactive or PCR-reactive, which were considered to be the index donations, were removed from the blood supply. Recipient tracing was performed for a -month period for donors of reactive samples (see the Supplementary Appendix). Donors were notified of their reactive results and were encouraged to seek medical advice, were indefinitely deferred from donating blood, and were invited to participate in followup. Index donation samples that tested reactive by at least one supplemental assay (IgM or IgG Western blot or enhanced-sensitivity PCR assay) or by repeat AFIA or PCR assay on residual samples from the index unit were considered to be confirmed positive samples. Positive predictive values (PPVs) were calculated. Data from this study were combined with data from a repository validation study to assess the results of donor followup (see the Supplementary Appendix). 8 Donors who participated in follow-up provided additional samples every 6 to 8 weeks until they had AFIA-negative test results (titer of <:6). Study Oversight This study was a collaboration between the American Red Cross and Imugen and included a signed confidentiality agreement. The study was supported by internal funds of the American Red Cross. All testing was conducted at Imugen with the use of FDA-approved cost recovery (the FDA allows investigators to recover the actual costs of the investigational testing program). The protocol was approved by the institutional review board of the American Red Cross and the FDA (Investigational New Drug Application number, 53). All the authors contributed to the study design and data collection and vouch for the integrity and completeness of the data and analyses presented. Assessment of Infectivity Red-cell samples from index donations that were PCR-positive or that were PCR-negative and had a high-titer antibody level (titer of :5) were injected into hamsters. - For PCR-positive samples, two Syrian hamsters (Mesocricetus auratus, Envigo) that had not been previously used for research were injected intraperitoneally with.5 ml of red cells (range, to red cells per milliliter) on consecutive days. For PCR-negative, high-titer donation samples, three hamsters were injected. Blood was obtained weekly up to 8 weeks after the initial injection and examined for parasites by means of acridine orange staining and microscopy. To confirm B. microti infection, an in-house PCR assay was performed at weeks and 8 weeks after the initial injection. Genomic DNA was extracted from hamster whole blood with the use of the Gentra Puregene Blood Kit (Qiagen). A real-time PCR assay was performed as described previously. 3 Animals were killed when the development of parasitemia was observed or after 8 weeks, whichever occurred first. A redcell donation was considered to be infectious in hamsters if infection developed in at least one animal. Procedures were approved by the institutional animal care and use committee of the American Red Cross. Cases of Transfusion-Transmitted Babesiosis Suspected cases of transfusion-transmitted babesiosis in patients who had received transfusions were identified by hospitals on the basis of clinician suspicion and were reported to the American Red Cross. Cases in which an implicated donation was confirmed to be positive (i.e., when a donor had an AFIA-positive result at a titer of :6 or had a PCR-positive result on a follow-up sample that was obtained as part of a transfusion-transmission investigation) were included in this analysis. Statistical Analysis Kaplan Meier product-limit survival estimates that modeled the time to the first PCR-negative and AFIA-negative result were generated with the use of SAS software, version 9. (SAS Institute). Comparisons of red-cell storage age with regard to infectivity in hamsters were made by means of Student s t-test with a two-sided P value, with the use of InStat software. The proportions of infectious donations among unscreened and screened donations that were collected in Connecticut and Massachusetts over the 38 n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

4 Screening for Babesia microti in U.S. Blood Supply study period were compared with the use of Fisher s exact test. Results Screening of Donations From June through September, a total of 89,53 blood donations from 6,5 donors were screened. The test results for 337 donation samples (.38%; 95% confidence interval [CI],.3 to.) showed reactivity for B. microti antibodies or DNA. A total of 335 samples, each representing a unique donor, were confirmed to be positive. Of these 335 samples, 67 (%) were positive on PCR assay, including 9 that were AFIA-negative (titer of <:6 and IgM-negative), for a yield of antibody-negative sample per 996 donation samples screened. This yield represents 3% of all PCR-positive results. Of the 9 donors who had PCR-positive, AFIA-negative results, 8 underwent seroconversion; the ninth donor had a repeated PCR-positive result in an independent red-cell sample. One sample that was not confirmed to be positive on AFIA and one that had an inconclusive result on AFIA (because of nonspecific fluorescence) were excluded from further analysis (Fig., and Table S in the Supplementary Appendix). Figure shows the county selection and the results for Connecticut and Massachusetts, which represented 93% of the positive samples in this study. All but one donor resided in the seven states in which babesia is endemic plus New Hampshire (Table S in the Supplementary Appendix). PCR-positive donations were identified in all months of the year except April, and PCRpositive, AFIA-negative donations were identified in June through September (Fig. 3). Among the samples that were positive on PCR assay or enhanced-sensitivity PCR assay, the estimated parasite loads ranged from 5 parasites to 3 million parasites per milliliter (median, 8 parasites per milliliter). As compared with donors of nonreactive samples, donors of reactive samples were more likely to be men, white, and at least 5 years old (Table S in the Supplementary Appendix; Fig. SA and SB in the Supplementary Appendix provide information on donor risk factors and symptom history). The PPV was % among PCR-positive, AFIAnegative samples (9 of 9), among PCR-positive, AFIA-positive samples (67 of 67), and among,66,8 Donations were obtained in Connecticut, Massachusetts, Minnesota, and Wisconsin 89,53 Donation samples were screened 6,5 Unique donors were screened 5,956 Were screened more than once,556 Were screened once 6 Had reactive result 9 Had PCR-positive and AFIA-negative result 337 Had reactive result 335 Had confirmed positive result 67 Had PCR-positive and AFIA-positive result 75 Had reactive result Had false reactive result Had inconclusive result 59 Had PCR-negative and AFIA-positive result Figure. Screening of Blood and Donors for Babesia microti. Shown is the screening process for B. microti in donated blood samples and in the donors, including testing results that were reactive and those that were confirmed to be positive. One donor with an inconclusive result had a negative polymerase-chain-reaction (PCR) assay and inconclusive results on arrayed fluorescence immunoassay (AFIA) owing to nonspecific fluorescence on the index donation sample and seven follow-up samples. The donor had PCR-negative, AFIA-negative results approximately years after the index donation. Among the nine donors with PCR-positive, AFIAnegative results, eight underwent seroconversion on follow-up and had results that were considered to be confirmed as positive on PCR assay and negative on AFIA. One donor did not undergo seroconversion; PCR reactivity on the index donation sample was confirmed by means of a detectable parasite load and PCR reactivity on an independent red-cell sample from the same index donation. The 67 PCR-positive, AFIA-positive samples include samples that were negative on the initial screening PCR assay but positive on the enhanced-sensitivity PCR assay. n engl j med 375;3 nejm.org December 8, 6 39 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

5 The new england journal of medicine A Counties Targeted for Screening Fairfield Connecticut Massachusetts Windham New London Middlesex Middlesex Norfolk Essex Plymouth Dukes B Areas in Connecticut with Donation Samples Positive for B. microti Barnstable PCR-positive and AFIA-negative (N=5) PCR-positive and AFIA-positive (N=) PCR-negative and AFIA-positive (N=9) C Areas in Massachusetts with Donation Samples Positive for B. microti PCR-positive and AFIA-negative (N=3) PCR-positive and AFIA-positive (N=7) PCR-negative and AFIA-positive (N=5) Figure. Maps of Connecticut and Massachusetts Showing Areas of Screening and Positive Results for B. microti. Shaded areas in Panel A represent counties that were targeted for screening. Screening was conducted from June through September on Mondays through Thursdays at selected drives so that the number of screened donation samples was approximately 5 to per day. High-risk counties were targeted on the basis of the number of babesiosis cases that had been reported to health departments and on a previously published study of prevalence. 9 The map in Panel B indicates the residential ZIP Codes of blood donors in Connecticut whose screened samples were positive for B. microti. In Connecticut, 39,65 donations were screened between June,, and September 3,. The percentages of positive samples per total tested were.% (95% CI,. to.3) for PCR-positive, AFIA-negative samples,.% (95% CI,.8 to.5) for PCR-positive, AFIA-positive samples, and.9% (95% CI,. to.56) for PCR-negative, AFIA-positive samples. One false positive donation sample and one sample that was inconclusive on AFIA and PCR-negative were excluded. Four PCR-positive, AFIA-positive samples and four PCR-negative, AFIA-positive samples that were obtained from donors who did not live in Connecticut were excluded. Eight donation samples that were negative on the initial PCR assay but positive on an enhanced-sensitivity PCR assay are included in the PCR-positive, AFIA-positive group (titer range, :8 to :). The map in Panel C indicates the residential postal codes of donors in Massachusetts whose screened samples were positive for B. microti. In Massachusetts, 36,66 donations were screened between July 3,, and September 3,. The percentages of positive samples per total tested were.% (95% CI, <. to.) for PCR-positive, AFIA-negative samples,.5% (95% CI,.3 to.75) for PCR-positive, AFIApositive samples, and.5% (95% CI,. to.9) for PCR-negative, AFIA-positive samples. One PCR-negative, AFIA-positive sample was excluded because it was obtained from a donor who did not live in Massachusetts. Six donors whose results were negative on the initial PCR assay but positive on an enhanced-sensitivity PCR assay are included in the PCR-positive, AFIA-positive group (titer range, :8 to :). PCR-negative, AFIA-positive samples with titers of :5 or more (68 of 68). The PPV was 99.5% among PCR-negative, AFIA-positive donations with titers of :8 or :56 (9 of 9 donations). The overall PPV was 99.7% (335 of 336 donations) (Table S3 in the Supplementary Appendix). High-titer antibody-positive samples were more likely than low-titer ones to be PCR-positive, n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

6 Screening for Babesia microti in U.S. Blood Supply 6 No. of Positive Donations PCR-positive and AFIA-negative (N=9) PCR-positive and AFIA-positive (N=67) PCR-negative and AFIA-positive (N=59) January February March April May June July August September October November December Figure 3. Positive Donation Samples from June,, to September 3,, According to Month. A total of donation samples were negative on the initial PCR assay but were positive on the enhanced-sensitivity PCR assay and are included in the PCR-positive, AFIA-positive count. The analysis excluded false reactive sample and inconclusive sample (owing to nonspecific fluorescence on AFIA). but most antibody-positive samples had a low titer and were PCR-negative (Fig. ). Follow-up Samples Follow-up samples were provided by 5 donors, including 5 who were identified in a repository validation study 8 ; the overall participation rate through September 5, 5, was 58% (5 of 386 donors). A total of 56 donors with PCRpositive results and 69 with PCR-negative results participated, providing to 3 samples per donor. The median time to the first follow-up sample was.7 months (interquartile range,. to.7), and the median time to the last follow-up sample was 7. months (interquartile range,. to 7.6) after the index donation. Participants were more likely than nonparticipants to be white and at least 5 years old (Table S in the Supplementary Appendix). The median time to a PCR-negative sample was.7 months (interquartile range,. to 7.); 86% of the donors with PCR-positive results (8 of 56 donors) had resolution of reactivity by year (Fig. 5A). The median time to seroreversion, defined as an AFIA titer of less than :6, in pooled data from donors with PCRpositive results and those with PCR-negative results, was 7. months (interquartile range, 9.9 to 7.6). A total of 9 of donors (8%) had resolution of antibody reactivity by year (Fig. 5B and 5C). One of 9 donors with PCR-positive, AFIA-negative results never had seroconversion ( parasites per milliliter). The PCR status of samples from 5 donors fluctuated between posi- No. of Index Donations PCR-negative (N=59) 75 PCR-positive (N=67) 3 :8 :56 :5 : Antibody Titer of Index Donation Figure. Relationship between AFIA Titer and Result on PCR Assay in Index Donation Samples That Tested Positive. A total of donation samples were negative on the initial PCR assay but were positive on the enhanced-sensitivity PCR assay; these samples are included in the PCR-positive count, including at an AFIA titer of :8, at a titer of :56, at a titer of :5, and 5 at a titer of : or more. The analysis excluded false reactive sample, inconclusive sample (owing to nonspecific fluorescence on AFIA), and 9 PCR-positive, AFIA-negative (titer of <:6) samples. 6 n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

7 The new england journal of medicine A PCR-Negative Follow-up Sample..9.8 Yr Yr AFIA-negative AFIA-positive Months Probability No. at Risk AFIA-positive 7 AFIA-negative 9 7 B Antibody-Negative Follow-up Sample in PCR-Positive Donors <:6 :5. :.3. :56. Yr Yr Yr 3 Yr. 3 5 Months Probability No. at Risk <:6 :56 :5 : Probability C Antibody-Negative Follow-up Sample in PCR-Negative Donors :. : : :8.. Yr Yr Yr 3 Yr Yr 5. 6 Months No. at Risk :8 :56 :5 : Figure 5. Kaplan Meier Estimates of the Survival Function. Panel A shows the Kaplan Meier estimate of the survival function modeling the time to the first PCR-negative follow-up sample in blood donors whose samples were PCR-positive for B. microti on the index donation sample, stratified according to the presence or absence of antibodies (according to the AFIA result on the index donation sample). Circles indicate the last follow-up for donors who did not have reversion to PCR-negative status during the study period ( donors whose index donation sample was PCR-positive, AFIA-negative and whose index donation sample was PCR-positive, AFIApositive). For example, the first circle on the AFIA-positive plot represents a donor whose last follow-up occurred.7 months after the index donation, at which point the sample remained PCR-positive. Dashed lines mark year and years since the index donation. Panel B shows the Kaplan Meier estimate of the survival function modeling the time to the first antibody-negative (titer of <:6) follow-up sample in blood donors whose index donation samples were PCR-positive for B. microti, stratified according to index titer. One donor with a PCR-positive, AFIA-negative result who did not undergo seroconversion was excluded from the analysis. Circles indicate the time of last follow-up for 36 donors (65% of those followed) who did not undergo seroreversion to AFIA-negative status during the study period (3 donors whose index donation sample had a titer of <:6, with an index titer of :56, 9 with an index titer of :5, and with an index titer of :). Dashed lines indicate the number of years since the index donation. Panel C shows the Kaplan Meier estimate of the survival function modeling time to first antibody-negative follow-up sample in donors whose index donation samples were PCR-negative for B. microti, stratified according to index titer. Circles indicate the time of last follow-up for 35 donors (8% of those followed) who did not undergo seroreversion to AFIA-negative status during the study (6 donors whose index donation samples had a titer of :8, 36 whose index samples had a titer of :56, 3 whose index samples had a titer of :5, and 5 whose index samples had a titer :). tive and negative during follow-up (Table S5 in the Supplementary Appendix). Infectivity in Hamsters Among 38 donation samples that were either PCR-positive or PCR-negative with a high titer ( :5) on the screening AFIA, red-cell samples for injection into hamsters were available from 93 (67%) 6 PCR-positive samples and 7 PCRnegative, AFIA high-titer samples. Overall, 7 samples (9%) were infectious. Of the 6 PCRpositive samples that were injected into hamsters, 5 (5%) were infectious. Samples from of the 7 PCR-negative, AFIA high-titer units (%) were infectious, of which was positive on enhanced- n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

8 Screening for Babesia microti in U.S. Blood Supply sensitivity PCR assay (AFIA titer of :; parasites per milliliter), leaving infectious redcell sample (AFIA titer of :5) that had PCRnegative test results according to the methods used in this study. Infectious red-cell samples were generally held for shorter periods of time before injection than were noninfectious red-cell samples (mean, 7 days [range, 7 to 3] vs. days [range, to 55]; P =.), which indicates that infectivity was probably underestimated. Cases Associated with Unscreened Donations During the study period, 9 cases of transfusiontransmitted babesiosis from unscreened donations were confirmed by linkage to a transfused cellular component from donors who later had a positive test result. Donors of implicated infectious samples resided in eight states: Connecticut (eight donors), Massachusetts (seven), New Jersey (five), New Hampshire (three), Maine (three), Minnesota (one), New York (one), and California (one travel-associated case linked to a donor infected in Rhode Island). Table S6 in the Supplementary Appendix provides data on the reported cases of transfusion-transmitted babesiosis to the American Red Cross each year during the period from January,, through August 3, 6. The estimated risk of transfusion-transmitted babesiosis in these states in which babesia is endemic, including New Hampshire and Maine, was 8 cases per,85, donations or case per, donations, whereas the incidence outside these states that was attributable to travel was case per,359, donations. The proportions of infectious donations among the unscreened and screened donations were compared. Among 75,33 screened donations, zero cases of transfusion-transmitted babesiosis were identified from screened donation samples, which implied that none of the donations were infectious. Among 53,3 unscreened donations, (i.e., in 8,7) were implicated in cases of transfusion-transmitted babesiosis, which suggests that donations were infectious and 53,7 were not infectious, although we cannot exclude the possibility that the disease was not associated with transfusion. These proportions ( in 75,33 and in 53,3) were compared, and the odds of an unscreened donation being infectious were 8.6 times (95% CI,.5 to ) as high as the odds of a screened donation being infectious (P =.5). Among implicated donors who had follow-up results that were found to be positive for babesia, had PCR-positive, AFIA-negative results 89 days after the implicated unit was obtained, 9 had PCR-positive, AFIA-positive results (median, 9 days; interquartile range, 76 to 98), and 9 had PCR-negative, AFIA-positive results (median, days; interquartile range, 88 to 57). A total of donors with PCR-positive results were identified at to 7 months after their implicated donation. All 9 presumed cases of transfusion-transmitted babesiosis were attributed to red cells; the mean time from donation to transfusion was days (range, 8 to ). Discussion From June through September, the American Red Cross screened 89,53 blood donations for B. microti DNA and antibodies to B. microti and removed 335 positive donations (.38%) from the blood supply, including 9 that were positive only on PCR testing. Hamster-infectivity data, which probably underestimated risk, and the continuing occurrence of cases of transfusiontransmitted babesiosis from unscreened donations, for which the risk is also likely to be underestimated, suggest that this investigational screening protocol removed infectious donations from the blood supply. Most asymptomatic donors who are deferred from donating blood retain PCR-positive status for less than year, whereas antibody reactivity (titer of >:6) may be retained for several years (Fig. 5A, 5B, and 5C). Screening has continued for hospitals that have specifically requested tested blood. An additional 367 confirmed positive donation samples from 3,36 screened samples (.8%) were identified by the American Red Cross between October,, and August 3, 6. This study has limitations. The screening protocol is specific to B. microti and does not detect other babesia species, including B. duncani, which was implicated in 3 of 6 cases of transfusiontransmitted babesiosis from 979 through 9. 6 Moreover, index donation samples had screening results that were PCR-negative but had detectable levels of DNA when tested by means of enhanced-sensitivity PCR assay, which shows the need for continued improvements in the sensitivity of the PCR assay. Despite the size of the study, these data may not be representative of n engl j med 375;3 nejm.org December 8, 6 3 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

9 The new england journal of medicine B. microti infection in the general population and should be interpreted with caution. Blood donors are healthy persons whose clinical course probably differs from persons in whom severe disease develops. Furthermore, we cannot exclude some degree of bias related to the decision not to participate in screening or follow-up (Tables S and S in the Supplementary Appendix). Continual donor follow-up and additional well-controlled studies are necessary to assess further the natural history of infection in healthy and susceptible persons. Data on infection and disease in donors (e.g., the number of donors who opted out of screening or who sought treatment after a positive screening test) and recipients are limited and are subject to the constraints of passive reporting. Our data were collected in areas that were geographically limited (Connecticut, Massachusetts, Minnesota, and Wisconsin), and any discussion of risk estimates, effect of screening, and interpretation of the data should take this into consideration. Finally, although we report a marginally significant association between the odds of infectious samples in unscreened versus screened donations, the statistical inference, which is based on zero counts, small proportions, and absent covariates, must be viewed with caution. Babesiosis is currently an infectious risk to the U.S. blood supply. It was responsible for 7% of the deaths ( of 5 deaths) in blood-transfusion recipients that were reported to the FDA from through. The American Red Cross identified 6 probable cases of transfusiontransmitted babesiosis in which a B. microti positive donor was identified in the period from January,, through August 3, 6, and 9 of these occurred during this study period (Table S6 in the Supplementary Appendix). A previous pilot study that used similar assays also showed that screening may be efficacious. 5 Our data imply that screening with the use of a PCR-based and AFIA-based protocol removes infectious donation samples from the blood supply and reduces the incidence of transfusiontransmitted babesiosis. Our geographically specific per-donation risk of transfusion-transmitted babesiosis with unscreened samples of approximately case per 8, donations is higher than the current residual risks of other serious transfusion-related adverse events, including sepsis from contaminated platelets ( case per 7, donations), transfusion-associated acute lung injury ( case per 38, donations), and infection with human immunodeficiency virus or hepatitis B or C virus ( case per,, donations for each type of infection). 6-9 Although reducing the frequency of transfusion-transmitted babesiosis is important, there are a number of issues that need to be resolved. Our data suggest that the greatest efficacy would be achieved by testing for both antibodies and DNA, because studies in animals have shown infectivity from single-marker positive samples, particularly those that test positive with the use of the PCR assay. Furthermore, the geographic areas affected by B. microti are restricted but expanding. Our data, combined with reports of cases of babesiosis nationwide, suggest that a screening strategy that is limited to states in which babesia is endemic, including antibody and nucleic acid testing year-round, would balance risk reduction and the judicious use of resources. Periodic reevaluation will be necessary to address the identification of cases in states, such as Pennsylvania, that were previously considered to be at low risk for babesiosis (Table S6 in the Supplementary Appendix). Risk-targeted screening (i.e., transfusing units screened for babesia only into high-risk patients) is initially appealing but introduces many potential challenges and has not been favorable in cost-effectiveness modeling.,- Seasonal testing has likewise been proposed as a means of decreasing the financial burden; however, our data show that PCR-positive donation samples have been identified year-round (Fig. 3). In May 5, the Blood Products Advisory Committee of the FDA voted in favor of antibody screening in all 5 U.S. states. 5 Although the testing algorithm that was used in this study performed better than other individual assays that have been used to ensure a safe blood supply, with a combined specificity of 99.98% in retrospective studies and more than % in prospective studies and an overall PPV of 99.7% in prospective studies (Table S3 in the Supplementary Appendix), 8 it is possible that when the antibody test is implemented widely in areas of low prevalence, the PPV will decrease (the lowest PPVs occurred in the category of AFIA-positive low- n engl j med 375;3 nejm.org December 8, 6 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.

10 Screening for Babesia microti in U.S. Blood Supply titer, PCR-negative samples). If universal screening is adopted, it may result in the unnecessary deferral of donors with false reactive results and of those with resolved infection, while also imposing a substantial financial burden on hospitals and the blood industry. Of the,79 total donations that were prospectively screened from June,, through August 3, 6, only four false antibody-reactive results occurred (one was described in this study). The high degree of specificity ensures the ongoing availability of donors and minimizes incorrect messages to donors. Moreover, our data show that, although PCR-reactivity resolves within year in most donors, antibodies are retained much longer. Algorithms for the reentry of persons into the pool of eligible donors are crucial for maintaining the engagement of donors who have reactive results, particularly those with false reactive results and those with remote previous infections. In conclusion, transfusion-transmitted babesiosis is of increasing concern and requires effective interventions. Our experience with prospective blood-donation screening provides a potential testing strategy. Supported by the American Red Cross and Imugen. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank the many employees of the American Red Cross who contributed to this study, including Gregory Foster and David Krysztof in Scientific Affairs, Stephanie Johnson (formerly with the Connecticut Region), the Babesia Team and case investigators at the Donor and Client Support Center, and numerous persons in the Connecticut, New England, North Central, and Badger Hawkeye Blood Collection Regions who assisted in the implementation of the screening; Edward Notari IV and Jaye Brodsky for advice on statistical analysis; Paula Saa for assistance with earlier versions of figures; and Evan Caten of the Massachusetts Department of Health for mapping assistance. References. Western KA, Benson GD, Gleason NN, Healy GR, Schultz MG. Babesiosis in a Massachusetts resident. N Engl J Med 97; 83: Vannier EG, Diuk-Wasser MA, Ben Mamoun C, Krause PJ. Babesiosis. Infect Dis Clin North Am 5; 9: Homer MJ, Aguilar-Delfin I, Telford SR III, Krause PJ, Persing DH. Babesiosis. Clin Microbiol Rev ; 3: Adams D, Fullerton K, Jajosky R, et al. Summary of notifiable infectious diseases and conditions United States, 3. MMWR Morb Mortal Wkly Rep 5; 6(53): Leiby DA. Transfusion-transmitted Babesia spp.: bull s-eye on Babesia microti. Clin Microbiol Rev ; : Herwaldt BL, Linden JV, Bosserman E, Young C, Olkowska D, Wilson M. Transfusion-associated babesiosis in the United States: a description of cases. Ann Intern Med ; 55: Kessler DA, Erwin JL, Busch MP, Levin AE, Shaz BH. Serological findings in donors who are deferred for a history of babesiosis. Transfusion ; 5: Suppl: A. abstract. 8. Moritz ED, Winton CS, Johnson ST, et al. Investigational screening for Babesia microti in a large repository of blood donor samples from nonendemic and endemic areas of the United States. Transfusion ; 5: Johnson ST, Cable RG, Tonnetti L, Spencer B, Rios J, Leiby DA. Seroprevalence of Babesia microti in blood donors from Babesia-endemic areas of the northeastern United States: through 7. Transfusion 9; 9: Cullen JM, Levine JF. Pathology of experimental Babesia microti infection in the Syrian hamster. Lab Anim Sci 987; 37: Brandt F, Healy GR, Welch M. Human babesiosis: the isolation of Babesia microti in golden hamsters. J Parasitol 977; 63: Gleason NN, Healy GR, Western KA, Benson GD, Schultz MG. The Gray strain of Babesia microti from a human case established in laboratory animals. J Parasitol 97; 56: Tonnetti L, Proctor MC, Reddy HL, Goodrich RP, Leiby DA. Evaluation of the Mirasol pathogen [corrected] reduction technology system against Babesia microti in apheresis platelets and plasma. Transfusion ; 5: Fatalities reported to FDA following blood collection and transfusion: annual summary for fiscal year. Silver Spring, MD: Food and Drug Administration ( downloads/ BiologicsBlood Vaccines/ SafetyAvailability/ ReportaProblem/ TransfusionDonationFatalities/ UCM596.pdf). 5. Young C, Chawla A, Berardi V, Padbury J, Skowron G, Krause PJ. Preventing transfusion-transmitted babesiosis: preliminary experience of the first laboratorybased blood donor screening program. Transfusion ; 5: Eder AF, Meena-Leist CE, Hapip CA, Dy BA, Benjamin RJ, Wagner SJ. Clostridium perfringens in apheresis platelets: an unusual contaminant underscores the importance of clinical vigilance for septic transfusion reactions (CME). Transfusion ; 5: Eder AF, Dy BA, O Neill EM. Predicted effect of selectively testing female donors for HLA antibodies to mitigate transfusionrelated acute lung injury risk from apheresis platelets. Transfusion 6; 56: Zou S, Dorsey KA, Notari EP, et al. Prevalence, incidence, and residual risk of human immunodeficiency virus and hepatitis C virus infections among United States blood donors since the introduction of nucleic acid testing. Transfusion ; 5: Stramer SL, Notari EP, Krysztof DE, Dodd RY. Hepatitis B virus testing by minipool nucleic acid testing: does it improve blood safety? Transfusion 3; 53: Bish EK, Moritz ED, El-Amine H, Bish DR, Stramer SL. Cost-effectiveness of Babesia microti antibody and nucleic acid blood donation screening using results from prospective investigational studies. Transfusion 5; 55: Acosta ME, Ender PT, Smith EM, Jahre JA. Babesia microti infection, eastern Pennsylvania, USA. Emerg Infect Dis 3; 9: Fang DC, McCullough J. Transfusiontransmitted Babesia microti. Transfus Med Rev 6; 3: Ness PM, Lipton KS. Selective transfusion protocols: errors and accidents waiting to happen. Transfusion ; : Simon MS, Leff JA, Pandya A, et al. Cost-effectiveness of blood donor screening for Babesia microti in endemic regions of the United States. Transfusion ; 5: Transcript of the meeting of the Blood Products Advisory Committee, May 3, 5. Silver Spring, MD: Food and Drug Administration ( down loads/ AdvisoryCommittees/ Committees MeetingMaterials/ BloodVaccinesandOther Biologics/ BloodProductsAdvisory Committee/ UCM953.pdf). Copyright 6 Massachusetts Medical Society. n engl j med 375;3 nejm.org December 8, 6 5 Downloaded from nejm.org on January, 8. For personal use only. No other uses without permission. Copyright 6 Massachusetts Medical Society. All rights reserved.