The n e w e ng l a n d j o u r na l of
Original Article
From the American Red Cross, Scientific
Affairs, Gaithersburg, MD (P.S., M.P.,
G.F., D.K., C.W., R.Y.D., S.L.S.); Grifols Di-
agnostic Solutions, San Diego, CA (J.M.L.,
K.G.); Quality Analytics, Riverwoods, IL
(J.P.B.); and Wadsworth Center, New York
State Department of Health, Albany
(R.J.L.). Address reprint requests to Dr.
Stramer at the American Red Cross, 9315
Gaither Rd., Gaithersburg, MD 20877, or
at ­susan​.­stramer@​­redcross​.­org.
N Engl J Med 2018;378:1778-88.
DOI: 10.1056/NEJMoa1714977
Copyright © 2018 Massachusetts Medical Society.
1778
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
m e dic i n e
Investigational Testing for Zika Virus
among U.S. Blood Donors
Paula Saá, Ph.D., Melanie Proctor, B.S., Gregory Foster, B.A.,
David Krysztof, M.B.A., Colleen Winton, B.A., Jeffrey M. Linnen, Ph.D.,
Kui Gao, Ph.D., Jaye P. Brodsky, B.S., Ronald J. Limberger, Ph.D.,
Roger Y. Dodd, Ph.D., and Susan L. Stramer, Ph.D.​​
A BS T R AC T
BACKGROUND
Because of the potential severe clinical consequences of Zika virus (ZIKV) infection,
the large numbers of asymptomatic travelers returning from ZIKV-active areas, the
detection of ZIKV nucleic acid in blood, and reports of transmission of ZIKV through
transfusion, in 2016 the Food and Drug Administration released recommendations
for individual-unit nucleic acid testing to minimize the risk of transmission of ZIKV
through blood transfusions.
METHODS
The American Red Cross implemented investigational screening of donated blood for
ZIKV RNA by means of transcription-mediated amplification (TMA). Confirmatory
testing of reactive donations involved repeat TMA, TMA testing in exploratory mini-
pools, real-time reverse-transcriptase polymerase chain reaction, IgM serologic testing,
and red-cell TMA. Viral loads in plasma and red cells were estimated by means of end-
point TMA. The costs of interdicting a donation that was confirmed to be positive
were calculated for the 15-month period between June 2016 and September 2017.
RESULTS
Of the 4,325,889 donations that were screened, 393,713 (9%) were initially tested
in 24,611 minipools, and no reactive donations were found. Of the 3,932,176 donations
that were subsequently tested individually, 160 were initially reactive and 9 were
confirmed positive (a 1:480,654 confirmed-positive rate overall; positive predictive
value, 5.6%; specificity, 99.997%). Six (67%) of the confirmed-positive donations
were reactive on repeat TMA, of which 4 were IgM-negative; of these 4, all 3 that
could be tested were reactive on minipool TMA. Two confirmed-positive donors
had infections that had been transmitted locally (in Florida), 6 had traveled to
ZIKV-active areas, and 1 had received an experimental ZIKV vaccine. ZIKV RNA
levels in red cells ranged from 40 to 800,000 copies per milliliter and were detected
up to 154 days after donation, as compared with 80 days of detection in plasma
at levels of 12 to 20,000 copies per milliliter. On the basis of industry-reported
costs of testing and the yield of the tests in our study, the cost of identifying
8 mosquito-borne ZIKV infections through individual-unit nucleic acid testing was
$5.3 million per ZIKV RNA–positive donation.
CONCLUSIONS
Screening of U.S. blood donations for ZIKV by individual-donation TMA was costly and
had a low yield. Among the 9 confirmed ZIKV-positive donations, only 4 were IgM-
negative; of these donations, all 3 that were tested were reactive on minipool TMA.
(Funded by the American Red Cross and Grifols Diagnostic Solutions.)
n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
 Testing for Zik a Virus among U.S. Blood Donors
Z
ika virus (ZIKV), an arthropod-borne
virus (arbovirus) of the genus flavivirus,
was identified in a sentinel rhesus monkey
in the Zika Forest of Uganda in 1947 and was
isolated from mosquitoes (Aedes africanus) in
1948.1 Outbreaks of ZIKV infection were observed
in Yap Island, Federated States of Micronesia, in
20072 and in French Polynesia in 2013.3 ZIKV is
transmitted through the bite of infected female
A. aegypti mosquitoes4 and potentially A. albopictus
mosquitoes,5 as well as through nonvector routes,
including vertical (i.e., mother-to-infant),6-9 trans-
fusion,10-12 and sexual transmission.13,14 Acute ZIKV
infection is often asymptomatic (in 50 to 80% of
cases), but it can manifest as a self-limited dis-
ease that usually resolves within 1 week.2,3,15
Severe complications related to ZIKV infection
have been identified, including congenital defects,
such as microcephaly,7-9,16 and noncongenital neu-
rologic disorders — notably encephalitis,17-19
myelitis,20 and the Guillain–Barré syndrome.21-23
On the basis of the large proportion of asymp-
tomatic cases, the severe clinical consequences
to an infected fetus, the detection of ZIKV RNA
in asymptomatic blood donors,12,24 and the re-
ports of suspected cases of transfusion trans-
mission of ZIKV in Brazil,10,11 ZIKV infections
represent a threat to blood safety. Accordingly,
the Food and Drug Administration (FDA) issued
recommendations to minimize the risk of ZIKV
transmission through blood components.25,26 On
June 20, 2016, the American Red Cross initiated
screening for ZIKV RNA using the Procleix Zika
Virus Assay (Grifols Diagnostic Solutions) under
an FDA-approved investigational-new-drug pro-
tocol. Data collected during investigational-new-
drug testing and donor follow-up were used to
assess the performance of the assay and the ef-
fect of donor screening on blood safety.
Me thods
Screening Protocol
We informed donors who met current donor-
suitability criteria about the study and included
them in prospective screening unless they opted
out. Investigational nucleic acid testing was ini-
tiated on June 20, 2016, at the American Red
Cross National Testing Laboratories; nucleic acid
testing was conducted in 16-member minipools
with the Procleix Zika Virus Assay on the Panther
System (Grifols Diagnostic Solutions). Testing was
performed on collections of blood donations
n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
from five southeastern U.S. states that were pre-
sumed to be at high risk (Fig. 1). The assay uses
A Quick Take
magnetic-based target capture, transcription- is available at
mediated amplification (TMA), and chemilumi- NEJM.org
nescent detection, with an estimated 95% limit
of detection of 3.9 copies per milliliter (95% fi-
ducial limits, 3.2 to 4.8), as determined by probit
analysis.27,28 In accordance with the FDA revised
guidance from August 26, 2016, investigational
nucleic acid testing was extended to all U.S.
blood donations, including conversion from mini-
pool TMA to individual-donation TMA. Phased
implementation of individual-donation TMA was
completed on December 12, 2016 (Fig. 1). Dona-
tions that were initially found to be reactive on
individual-donation TMA were retested in tripli-
cate with the use of the same testing platform;
donations with reactivity in any of the triplicate
retests were considered to be repeat reactive. The
screening protocol remains active, with 15 months
of the results presented here.
Confirmatory Testing
Surplus sample tubes from all initially reactive
donations were obtained; corresponding plasma
units and red-cell components were quarantined
and retrieved from collection regions. Samples
from retrieved plasma units were each retested
by TMA in 20 replicates. In addition, samples of
plasma were sent to the Wadsworth Center (New
York State Department of Health) for confirma-
tory testing, including alternate nucleic acid test-
ing by in-house real-time, reverse-transcriptase
polymerase chain reaction (RT-PCR), which had
an estimated limit of detection of 70 copies per
milliliter,29 and testing for anti-ZIKV IgM with
an IgM antibody-capture enzyme-linked immuno-
sorbent assay (MAC-ELISA) (performed under
emergency-use authorization).30 IgM-positive do-
nations (considered presumptively positive) were
tested with an arbovirus plaque-reduction neutral-
ization test to determine the presence of ZIKV,
dengue viruses 1 through 4, or West Nile virus
(WNV) neutralizing antibodies. When available,
samples of red cells or residual whole blood and
samples of plasma were sent to Grifols Diagnos-
tic Solutions for further assessment of reactivity
and estimation of viral loads by end-point TMA.27
Donors were classified as confirmed positive if
their plasma, red cells, or whole blood was re-
peat reactive on TMA or if they were found to be
positive by subsequent serologic or molecular
testing.
1779
 The n e w e ng l a n d j o u r na l of m e dic i n e

February 16, 2016

FDA issues recommendations
for industry to reduce the risk
of transmission of ZIKV through
whole blood and blood components

March 14, 2016
Introduction of ZIKV
travel question in the
donor-history question-
naire, and associated
deferral
July 21, 2016 September 6, 2016 October 3, 2016 December 12, 2016
Conversion from MP-TMA Implementation of MP-TMA Conversion from MP-TMA Conversion to ID-TMA
to ID-TMA for FL collections for CA, OR, and WA to ID-TMA for AL, AZ, CA, in all remaining states
collections GA, MS, NY, OK, SC, and
TX collections

June 20, 2016
Implementation of IND
testing for the presence
of ZIKV RNA by MP-
TMA in AL, FL, GA,
MS, and SC
August 29, 2016 September 26, 2016 November 14, 2016
Implementation of Implementation of Implementation of MP-TMA
MP-TMA for AZ, MS, MP-TMA for CA in all remaining states;
(additional areas), OK, (additional areas) conversion to ID-TMA for
and TX collections and NY collections ID, MT, NV, OR, UT, and
WA collections

August 26, 2016 January 23, 2017

FDA issues revised Removal of ZIKV travel
guidance recommending question from donor-
screening of all donated blood history questionnaire
in the United States by ID-NAT

Figure 1. Timeline of Zika Virus (ZIKV) Investigational Testing at the American Red Cross (ARC).
In accordance with Food and Drug Administration (FDA) guidance, the ARC modified the donor-history questionnaire on March 14,
2016, and implemented donor deferral that was based on whether the donor had traveled to an area with active transmission of ZIKV
or had had sexual contact with a man who had ZIKV infection or who resided in or had traveled to an area with active transmission of
ZIKV. On June 20, 2016, the ARC voluntarily implemented ZIKV testing under an FDA-approved investigational-new-drug (IND) protocol,
using the Procleix Zika Virus Assay on the Panther System (Grifols Diagnostic Solutions) with minipool transcription-mediated amplifi-
cation (MP-TMA) in five states that were presumed to be at risk for ZIKV transmission — Alabama, Florida, Georgia, Mississippi, and
South Carolina. On July 21, 2016, after the identification of the first autochthonous cases of ZIKV infection in Florida, the ARC started
testing all Florida collections by individual-donation transcription-mediated amplification (ID-TMA). On August 26, 2016, the FDA issued a
revised guidance recommending the individual screening by nucleic acid testing (ID-NAT) of all blood donations collected in the United
States and its territories. On August 29, 2016, ZIKV testing by MP-TMA was initiated for selected states and regions: all donations from
Arizona, Oklahoma, and Texas, as well as from additional areas of Mississippi. On September 6, 2016, MP-TMA was implemented for
California, Oregon, and Washington collections. On September 26, 2016, MP-TMA started for New York collections and additional re-
gions in California. On October 3, 2016, conversion from MP-TMA to ID-TMA took place for collections from Alabama, Arizona, California,
Georgia, Mississippi, New York, Oklahoma, South Carolina, and Texas. On November 14, 2016, the ARC implemented MP-TMA in all
­remaining states and converted from MP-TMA to ID-TMA for donations collected in Idaho, Montana, Nevada, Oregon, Utah, and Wash-
ington. On December 12, 2016, the ARC finished converting all remaining states to ID-TMA. The ZIKV-related travel question was removed
from the donor-history questionnaire on January 23, 2017.

To assess the sensitivity of minipool TMA, 3 months before sexual contact. All donors of
donations that were confirmed to be positive reactive samples were notified of their reactive
after individual-donation TMA were further test- results, temporarily deferred from future blood
ed in triplicate in exploratory 16-sample mini- donation, and invited to participate in the Zika
pools that were created by combining the reac- Virus Follow-up Study, for which the partici-
tive donation with 15 donations that had been pants provided additional blood samples begin-
found nonreactive on TMA. ning 1 week after the index reactive donation.
Deferred donors were reinstated for ongoing dona-
Donor Follow-up tion when a follow-up sample tested nonreactive
All donors whose samples tested reactive by TMA and 120 days had elapsed after the last reactive
were contacted within 24 hours of the reactive donation.
test result to inquire about potential risk factors,
including travel in the previous 28 days or sexual Study Oversight
contact with a person who received a diagnosis This study was a collaboration between the
of ZIKV infection or traveled to or resided in American Red Cross and Grifols Diagnostic So-
an area with active ZIKV transmission in the lutions and included a signed confidentiality

1780 n engl j med 378;19 nejm.org  May 10, 2018

The New England Journal of Medicine

Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
 Testing for Zik a Virus among U.S. Blood Donors
agreement. The study was supported by internal
funds of the American Red Cross with FDA-
approved cost recovery (the FDA allows investi-
gators to recover the actual costs of the investi-
gational testing program). The costs of both
minipool and individual-donation TMA were cal-
culated on the basis of the mean of a range of
cost-recovery figures reported by the industry.31
The protocol was approved by the institutional
review board of the American Red Cross and by
the FDA (Investigational New Drug application
17003). The authors vouch for the accuracy and
completeness of the data and analyses presented.
R e sult s
ZIKV RNA Detection and Assay Performance
Between June 20, 2016, and September 9, 2017,
a total of 4,325,889 donations were tested; 393,713
(9%) of the donations were tested in 24,611 mini-
pools without a reactive result (Fig. 2). Of the
other 3,932,176 donations (91%) that underwent
subsequent testing by individual-donation TMA,
160 were initially reactive, of which 9 (5.6%)
were confirmed positive by repeat TMA or ad-
ditional testing (Fig. 2), for a rate of confirmed
ZIKV-positive donations of 1 per 480,654. The
remaining 151 initially reactive donations were
found to be nonreactive on alternate nucleic acid
testing, serologic testing, or additional TMA
testing (or a combination of the three tests) of
plasma and red cells or whole blood and were
classified as false positive. The positive predic-
tive value was 9 confirmed-positive donations
among 160 reactive donations, or 5.6%. The
specificity was the percentage of donations with
nonreactive test results that were truly negative:
4,325,729 of 4,325,880, or 99.997% (Table 1).
Of the 9 confirmed-positive donations, 6 (67%;
from Donors 1, 2, 3, 5, 8, and 9) were repeat
reactive on individual-donation TMA; 3 were
initially reactive on individual-donation TMA,
but ZIKV infection was confirmed by the detec-
tion of ZIKV IgM in plasma and ZIKV RNA in
red cells or whole blood (Donors 4, 6, and 7)
(Fig. 2). Among the 6 donors of repeat-reactive
blood, the index donations of 4 (67%; Donors 1,
3, 5, and 9) were IgM-negative — that is, they
were window-period donations (collected before
seroconversion had occurred). The results of al-
ternate nucleic acid testing were positive in 2 of
4 donors (Donors 1 and 5) and equivocal in the
other 2 window-period donors (Donors 3 and 9).
n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
All repeat-reactive donations were confirmed
positive, for a positive predictive value of 100%
(Table 1). Detection of ZIKV RNA at index by
alternate nucleic acid testing in the 5 IgM-posi-
tive donors (Donors 2, 4, 6, 7, and 8) was unsuc-
cessful. Plasma from confirmed-positive dona-
tions had an estimated 12 or fewer to 20,000
copies per milliliter at donation, as compared
with red cells, which had 40 or fewer to approxi-
mately 800,000 copies per milliliter, with the
exception of Donor 9, from whom index samples
were not available (Table S1 in the Supplemen-
tary Appendix, available with the full text of this
article at NEJM.org).
The confirmed-positive donors resided in Texas,
California, Florida (3 donors), Massachusetts
(2 donors), New York, and West Virginia (Table 2,
and Fig. S1 in the Supplementary Appendix).
Four (80%) of the 6 donors of repeat-reactive
blood reported travel to a ZIKV-active area (Donors
1, 2, 5, and 8), with a mean of 14 days (range,
2 to 31) elapsing from travel return to blood
donation (Table 2). Two of 3 donors of non–
repeat-reactive confirmed-positive blood — Do-
nors 4 and 7 — returned to the United States
from a ZIKV-active area 73 and 59 days before
donation, respectively, and had “tail-end” infec-
tions (i.e., residual, low-level infection), as re-
ported previously.27 In addition, Donor 7 report-
ed sexual contact with a person who traveled to
an area with active ZIKV transmission (Table 2).
Two of 9 confirmed-positive donors (Donors 4
and 7) became symptomatic after travel (Table 2),
for a rate of symptomatic infection among do-
nors of 22% — a finding similar to that previ-
ously reported for this virus,32 although the rate
of symptomatic infection is probably higher.15
Each confirmed-positive donor is described in
detail in the Supplementary Appendix.
Exploratory Minipool Testing
Confirmed-positive index donations were tested
by TMA in simulated minipools of 16 donations.
The results of minipool TMA were 100% reactive
for all tested window-period donations (from
Donors 1, 5, and 9) (Table S1 in the Supplemen-
tary Appendix). Conversely, ZIKV IgM–positive
donations were found to be nonreactive in TMA
minipool testing (Donors 4, 7, and 8); the excep-
tion was that of Donor 2, whose blood was found
to be reactive in one of three tests. Donors 3
(IgM-negative at index) and 6 (IgM-positive at
index) were not included in minipool testing
1781
 1782
4,325,889 Donations were tested between
June 20, 2016 and September 9, 2017
393,713 Were tested by MP-TMA
3,932,176 Were tested by ID-TMA

160 Were initially reactive 4,325,729 Were nonreactive

6 Were repeat reactive

154 Were non–repeat reactive
The

(Donors 1, 2, 3, 5, 8, and 9)

2 Had positive 2 Had equivocal 2 Had negative 3 Did not undergo 151 Had negative
result for alternate result for alternate result for alternate alternate result for alternate
NAT testing NAT NAT NAT NAT
(Donors 1 and 5) (Donors 3 and 9) (Donors 2 and 8)

0 Had positive 0 Had positive 0 Had negative

IgM test result IgM test result IgM test result
n e w e ng l a n d j o u r na l

The New England Journal of Medicine

2 Had negative 2 Had negative 2 Had positive 3 Did not 3 Had positive 148 Had nega-
of

IgM test result IgM test result IgM test result undergo IgM test result tive IgM test
(Donors 1 and 5) (Donors 3 and 9) (Donors 2 and 8) IgM testing (Donors 4, 6, result
and 7)

n engl j med 378;19 nejm.org  May 10, 2018

Copyright © 2018 Massachusetts Medical Society. All rights reserved.

m e dic i n e

2 Were reactive 0 Were not 1 Was reactive 1 Did not 2 Were reactive 0 Were not 3 Were not 3 Were reactive 0 Were not 3 Did not 145 Were not
in whole-blood reactive in whole-blood undergo in whole-blood reactive reactive in in whole-blood reactive undergo reactive
test (Donors 1 in whole- test (Donor 3) whole-blood test (Donors in whole- whole-blood test (Donors 4, in whole- whole- in whole-
and 5) blood test testing 2 and 8) blood test test 6, and 7) blood test blood testing blood test

Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
(Donor 9)

9 Were confirmed ZIKV-positive donations

 Testing for Zik a Virus among U.S. Blood Donors

Figure 2 (facing page). Flow Chart of ZIKV Investigational

Screening Results for Donations Tested between June 20,
2016, and September 9, 2017.
A total of 4,325,889 donations were tested with the
Procleix Zika Virus Assay during the study period,
393,713 of which were tested by MP-TMA in 24,611
minipools. The remaining 3,932,176 donations were
tested by ID-TMA; of these, 160 were reactive. Six dona-
tions were repeat reactive (Donors 1, 2, 3, 5, 8, and 9),
and 154 were non–repeat reactive. Of the 6 repeat-­
reactive donations, the results of alternate nucleic acid
testing (NAT) (reverse-transcriptase polymerase chain
reaction [RT-PCR]) were positive in 2 (Donors 1 and 5),
equivocal in 2 (Donors 3 and 9), and negative in 2
(Donors 2 and 8). All donations that were positive or
equivocal by alternate NAT were ZIKV IgM negative by
antibody-capture enzyme-linked immunosorbent assay
(MAC-ELISA); of these, 3 were reactive in whole blood
(Donors 1, 5, and 3), and 1 did not have whole blood
tested (Donor 9). The 2 alternate NAT–negative donations
were ZIKV IgM–positive and reactive in whole blood
(Donors 2 and 8). Of the 154 non–repeat-reactive
­donations, 3 were not tested by either alternate NAT
or MAC-ELISA, but they were TMA-negative in whole
blood. Of the 151 alternate NAT–negative donations,
3 tested positive for ZIKV IgM by MAC-ELISA and were
reactive in whole blood (Donors 4, 6, and 7), which
brought the total number of confirmed ZIKV-positive
donations to 9.
provided their first follow-up sample between
4 and 38 days (mean, 19.5 days) after their index
donation (Table S1 in the Supplementary Appen-
dix). Confirmed-positive donors provided between
1 and 25 follow-up samples (mean, 8). In 3 of
4 window-period donors (75%), seroconversion
occurred at between 7 and 17 days of follow-up
(mean, 11 days). The fourth window-period do-
nor was the recipient of an experimental ZIKV
vaccine (ZIKV purified inactivated vaccine [ZPIV]);
the donor provided three additional samples at
14, 23, and 38 days, and none of the samples
were reactive on ZIKV IgM testing (Table S1 in
the Supplementary Appendix). ZIKV IgM detec-
tion continued in all seropositive donors for the
duration of the follow-up study (range, 4 to 176
days; mean, 82) (Fig. 3). When the estimated time
of ZIKV exposure of imported cases was included,
IgM detection continued for a mean of 113 days
after exposure (range, 18 to 235). All follow-up
samples were nonreactive on alternate nucleic
acid testing but were TMA-reactive in plasma or
red cells. The longest duration of detection of
ZIKV RNA in red cells was 154 days (Fig. 3B), as
compared with 80 days in plasma (Fig. 3G).

because of inadequate sample volume. TMA sig- Cost of Testing

nal levels were higher in IgM-negative donations. On the basis of a per-donation cost of $6 for
minipool nucleic acid testing (range, $3 to $9),
Follow-up of Confirmed ZIKV-Positive Donors which is similar to that for other transfusion-
Each confirmed-positive donor consented to par- transmitted viruses for which minipool nucleic
ticipate in the Zika Virus Follow-up Study and acid testing is currently used; $10 for individual-

Table 1. Performance Characteristics of the Procleix Zika Virus (ZIKV) Assay Based on the Number of Reactive Donations,
June 20, 2016, to September 9, 2017.*

Positive
TMA Confirmed False Predictive
Test Tested Reactive Positive Positive Specificity Value
number of donations percent
TMA
Minipool 393,713 0 0 0 100.000 —
Individual donation 3,932,176 160 9 151 99.996 5.63
All 4,325,889 160 9 151 99.997 5.63
Repeat reactive 4,325,889 6 6 0 100.000 100.00

* Of the 160 donations tested by individual-donation testing that were ZIKV RNA–reactive, 9 (5.63%) were confirmed
positive by repeat transcription-mediated amplification (TMA) or subsequent testing, for an overall confirmed-positive
rate of 1 confirmed-positive donor per 480,654 tested donations. The assay specificity for minipool TMA was greater
than that for individual-donation TMA (100.000% vs. 99.996%). The overall TMA specificity was 99.997%, and the posi-
tive predictive value was 5.63%. No reactive donations were identified by minipool TMA (100% specificity), which pre-
vented the comparison of the positive predictive value of minipool versus individual-donation TMA. All repeat-reactive
donations were confirmed positive by subsequent testing, yielding a positive predictive value and specificity of repeat
reactivity of 100%.

n engl j med 378;19 nejm.org  May 10, 2018 1783

The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Confirmed ZIKV-Positive Donor Demographic Characteristics and Risk Factors.*

Days from
Age in Return to Sexual
Donor State Collection Date Sex Yr Travel† Donation Symptoms‡ Contact§
1 Texas Nov. 2, 2016 Male 62 Yes 8 No No
2 California Nov. 18, 2016 Male 61 Yes 14 No No
3 Florida¶ Dec. 5, 2016 Male 68 No — No Yes
4 Massachusetts‖ Dec. 27, 2016 Female 58 Yes 73 Yes No
5 Florida Jan. 10, 2017 Female 20 Yes 2 No No
6 Florida¶ Jan. 12, 2017 Male 22 No — No No
7 New York‖ Jan. 31, 2017 Female 26 Yes 59 Yes Yes
8 West Virginia Mar. 13, 2017 Male 67 Yes 31 No No
9 Massachusetts** May 16, 2017 Female 19 No — No No

* Of 160 donors identified by individual-donation transcription-mediated amplification (ID-TMA), 9 were confirmed as
ZIKV-positive by supplemental testing. Five confirmed-positive donors were male, and 4 were female. The age distribu-
tion was 19 to 68 years (mean, 45 years). The donors resided in California, Florida (3 donors), Massachusetts (2 donors),
New York, Texas, and West Virginia.
† Six confirmed-positive donors reported traveling to a ZIKV-active area; donors traveled to the following destinations:
Donor 1, Puerto Rico; Donor 2, St. Maarten; Donor 4, St. Thomas; Donor 5, Curaçao; Donor 7, St. Kitts and Nevis;
and Donor 8, Aruba.
‡ Donor 4 reported symptoms consistent with ZIKV infection, including low-grade fever, aches, and rash and vasculitis
in the hands. Donor 7 reported clinical symptoms between December 5 and December 8, 2016, after her return to the
United States, including a rash and a head cold.
§ Donors 3 and 7 reported sexual contact with someone who traveled or resided in an area of active ZIKV transmission.
The sex partner of Donor 3, also a resident of Miami-Dade County with no travel risk, was invited to participate in the
follow-up study. The partner provided five samples that were TMA-reactive in whole blood and seropositive by IgM
antibody-capture enzyme-linked immunosorbent assay and plaque-reduction neutralization testing analyses.
¶ Donors 3 and 6 were residents of Miami-Dade County, an area of active ZIKV transmission at the time of donation;
these donors did not report a travel risk and therefore had probable cases of autochthonous ZIKV transmission.
‖ Donors 4 and 7 had tail-end ZIKV infections; these donors provided TMA-reactive samples 73 and 59 days after their
return from a ZIKV-active area, respectively.
** Donor 9 did not report traveling to a ZIKV-active area and is a resident of Massachusetts; the donor is the recipient
of an experimental ZIKV vaccine.

Figure 3 (facing page). Long-Term Follow-up of Selected Confirmed-Positive

Donors.
All confirmed-positive donors consented to participate in the Zika Virus
Follow-up Study. Confirmed-positive donors provided between 1 and 25
follow-up samples, which were tested by TMA with the Procleix Zika Virus
Assay and by RT-PCR to determine the presence of ZIKV RNA in plasma and
red cells or whole blood, as well as by MAC-ELISA and plaque-reduction
neutralization testing (PRNT) to determine the presence of ZIKV antibodies
in serum. PRNT results are expressed as antibody dilution titers. Samples
with neutralizing ability at a dilution greater than 1:10 are considered posi-
tive and represented in the figure. ZIKV RNA loads in plasma and red cells
or whole blood were determined by limiting-dilution TMA from samples
that were 100% TMA-reactive; these are expressed as RNA copies per milli-
liter. RT-PCR results are expressed as cycle threshold (Ct) values. Ct values of
less than 40 are considered positive and are shown in the figure. MAC-ELISA
results (IgM) are represented as the ratio of the mean optical density (OD)
of the test specimen reacted with Zika viral antigen and the mean OD of
the normal human serum reacted with Zika viral antigen. Results greater
than 2 are considered presumptively positive and are shown in the figure.
All results are expressed in logarithmic scale. Detailed descriptions of each
donor’s test results can be found in the Supplementary Appendix.

1784 n engl j med 378;19 nejm.org  May 10, 2018

The New England Journal of Medicine

Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
 Testing for Zik a Virus among U.S. Blood Donors

ZIKV PRNT (log10 antibody dilution titer) DENV PRNT (log10 antibody dilution titer) ZIKV RNA red cells/whole blood (log10 copies/ml)
ZIKV RNA plasma (log10 copies/ml) ZIKV RT-PCR plasma (log10 Ct) ZIKV IgM (log10 OD ratio)

A Donor 1 B Donor 2
104

104
103

103
102
102

101
101

100 100
0 8 15 21 0 4
Days of Follow-up Days of Follow-up

C Donor 3 D Partner of Donor 3

105 104

104
103

103
102
102

101
101

100 100
0 15 20 25 33 43
0
17
23
30
39
45
49
52
56
59
63
66
70
79
85
91
109
115
122
122
157
151
164
162
178
6
9

Days of Follow-up Days of Follow-up

E Donor 4 F Donor 5
104 106

105
103
104

102 103

102
101
101

100 100
0 10 0 7 10 14 16 21 23 28 37 44 58 76 78 83 85 92 94 97 133
Days of Follow-up Days of Follow-up

G Donor 6 H Donor 7 I Donor 8

104 104 104

103 103 103

102 102 102

101 101 101

100 100 100

0 18 21 27 32 35 41 43 60 63 0 38 121 176 0 36 80 147
Days of Follow-up Days of Follow-up Days of Follow-up

n engl j med 378;19 nejm.org  May 10, 2018 1785

The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l
donation nucleic acid testing (range, $7 to $13),
including all expenses related to testing31; and
the number of collections tested by the Ameri-
can Red Cross between June 20, 2016, and Sep-
tember 9, 2017, the cost of ZIKV testing was
approximately $41.7 million over the course of
the 15-month study.
Discussion
The ZIKV outbreak in the Americas in 2015 and
2016 revealed unprecedented characteristics for
an arbovirus, including causal association with
a severe congenital syndrome9 and sexual trans-
mission.13,14 These findings, together with the
large proportion of asymptomatic cases and re-
ports of probable transfusion-transmitted ZIKV
infections,10,11 triggered a rapid response to mini-
mize the risk of ZIKV transmission to pregnant
women and blood-transfusion recipients, includ-
ing the rapid development of ZIKV nucleic acid
testing assays for donor screening.27,33 The Amer-
ican Red Cross implemented the Procleix Zika
Virus Assay in minipool TMA testing on June 20,
2016. After the identification of the first cases of
local mosquito transmission in Florida, the Amer-
ican Red Cross converted from minipool to indi-
vidual-donation TMA for donations collected in
Florida, subsequently extending this to all U.S.
collections.
Of 4,325,889 donations, 160 were initially re-
active, including 9 (5.6%) confirmed-positive
samples, for a rate of 1 per 480,654 donations
screened. The specificity of minipool TMA was
higher than that of individual-donation TMA
(100% vs. 99.996%), for an overall TMA specific-
ity of 99.997% and a positive predictive value of
5.6%. All repeat-reactive donations were con-
firmed positive by subsequent testing, yielding a
positive predictive value of 100%. The calculated
specificity of the Procleix Zika Virus Assay in
individual-donation TMA was similar to that
previously published for ZIKV nucleic acid test-
ing.27,33 The overall performance characteristics
of ZIKV TMA are similar to those of another
assay, the Procleix WNV Assay (Table 1, and
Table S2 in the Supplementary Appendix). How-
ever, the Procleix Zika Virus Assay had a sub-
stantially lower positive predictive value because
individual-donation TMA for ZIKV is performed
throughout the year in areas in which the virus
is not endemic, as opposed to WNV, for which
1786 n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
of m e dic i n e
individual-donation TMA occurs only during the
active WNV season of June through November,
with minipool TMA used during other times.34
Minipool TMA has the advantage of an addi-
tional cycle of reactive minipool resolution test-
ing before the identification of an individual
TMA-reactive donation (i.e., each donation in a
reactive minipool is tested individually in a sec-
ond round of testing).
Four window-period donations were identified
by ZIKV screening; seroconversion had occurred
in all window-period donors with the exception
of the vaccine recipient, Donor 9, at the time of
the first follow-up. The absence of seroconver-
sion in Donor 9 may have been due to receipt of
only a single vaccine dose of two planned by the
protocol.35 Alternatively, 38 days of follow-up
may have been insufficient to detect seroconver-
sion. It is possible that receipt of only a single
vaccine dose plus subsequent plateletpheresis in
Donor 9, which occurred 6 hours 20 minutes
after intramuscular administration of the vaccine,
together contributed to the absence of serocon-
version. Nonetheless, this study shows that ZIKV
RNA can be detected in plasma by a sensitive
TMA method after ZPIV administration. In all
other confirmed-positive donors, IgM detection
persisted for at least 113 days after exposure.
Like WNV, ZIKV associates with red cells in
the blood.36 The presence of ZIKV RNA in red
cells has been reported for as long as 101 days
when tested by RT-PCR with a limit of detection
exceeding 100 copies per milliliter,37 as com-
pared with 54 days in serum as determined by
Trioplex RT-PCR with an estimated limit of de-
tection of 40 copies per milliliter.38 The longest
period of ZIKV RNA detection in red cells in our
study was 154 days, as compared with 66 days
in plasma from the same donor, whose sample
was repeat-reactive on TMA, a finding consistent
with the use of a more sensitive assay with an
estimated limit of detection of 3.9 copies per
milliliter. Viral RNA loads in reactive samples
were generally low, ranging from approximately
12 to 200 copies per milliliter in plasma and ap-
proximately 40 to 800 copies per milliliter in red
cells. The sole exception was Donor 5, a window-
period donor whose index plasma and red-cell
samples contained 20,000 and 800,000 copies
per milliliter, respectively; 7 days later, the ZIKV
RNA levels in this donor underwent a several-log
decrease, concomitant with seroconversion. Red
 Testing for Zik a Virus among U.S. Blood Donors
cells may be the source of ZIKV RNA detected in
plasma by TMA (see the Supplementary Appendix).
The correlation between ZIKV genome copies
and viral infectivity is currently unknown, but it
has been estimated to be within the range of
200 to 500 genome copies per infectious viral
particle in culture.2 On the basis of this range,
the number of infectious units in all but one of
the confirmed-positive donations in this study
would be approximately 1 infectious unit per mil-
liliter in plasma and approximately 1.6 infectious
units per milliliter in red cells (the exception
again being Donor 5, whose sample was repeat-
reactive on minipool TMA and who had returned
from travel just 2 days before blood donation).
Donors 3 (repeat reactive) and 6 (initially re-
active) did not report a travel risk and are resi-
dents of Miami-Dade County, an area previously
designated as ZIKV active. ZIKV infection in
Donor 3’s partner was confirmed by RNA reac-
tivity in plasma and whole blood and by sero-
positivity at the same time point as the donor.
The partner also resided in Miami-Dade County
and had not traveled. In the absence of addi-
tional epidemiologic data, Donor 3 and his part-
ner represent two potential cases of autochtho-
nous ZIKV infection or one case of autochthonous
ZIKV infection with associated sexual transmis-
sion. Similarly, Donor 6 is another patient with
a probable case of local ZIKV infection.
Given that most cases of ZIKV infection are
asymptomatic and that there are unusual non-
vector routes of ZIKV transmission, donor-defer-
ral policies that are based on travel and symp-
tom reporting may have limited sensitivity for
blood safety. Consequently, the FDA has recom-
mended individual-donation nucleic acid testing
of all U.S. blood donations. Such an implemen-
tation came at a substantial expense: $42 mil-
lion over a period of 15 months. If ZIKV screen-
ing after the investigational-new-drug assessment
follows current FDA guidance, and given that
the American Red Cross collects 42% of the U.S.
blood supply, the projected annual cost for na-
tional screening is estimated to be $137 million
(range, $109 million to $167 million),31 which
will pose an additional strain on the blood in-
dustry.39 Our data for 8 donations that were
confirmed positive for mosquito-borne ZIKV
translates to a cost of $5.3 million for each RNA-
positive donor identified per year by individual-
donation TMA, whether infectious or not, or $21
n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
million for each of 2 donors without travel risk.
Note that these costs do not reflect increases
attributable to commercial pricing. In addition,
further testing beyond the period covered here
(6.32 million donations through February 3,
2018) has yielded only one additional confirmed-
positive donation: a window-period donation
from a traveler who donated blood 4 days after
return from Puerto Vallarta, Mexico. The esti-
mated viral loads in this donor were 1,000,000
and 20,000 copies per milliliter in plasma and
red cells, respectively. There is currently no con-
clusive evidence that low-level, antibody-positive
donations, as in Donors 2, 4, 6, 7, and 8, are in-
fectious, whereas previous experience with WNV
indicates that seronegative donations can result
in transfusion transmission,34,40 which suggests
that the donations from Donors 1, 3, and 5 were
potentially infectious.
We also explored the detection of ZIKV RNA
in confirmed-positive donations in simulated
pools of 16 samples. Minipool TMA results were
reactive for all window-period donors who could
be tested (including the most recently identified
donor), whereas seropositive donations that were
tested in pools were found to be nonreactive.
The exception was the repeat-reactive donation
from Donor 2, for whom one of three pool re-
sults was found to be reactive.
In conclusion, screening of U.S. blood dona-
tions for ZIKV by means of individual-donation
TMA had a low yield and a high cost. Among the
nine donors whose blood was confirmed as
ZIKV-positive, only four had acute infection (i.e.,
were IgM-negative); among these four donors,
all three with adequate sample volumes for test-
ing also tested positive on minipool TMA. Red
cells had higher viral loads over longer periods
than plasma in the confirmed-positive donors
who underwent repeat testing.
Supported by the American Red Cross and Grifols Diagnostic
Solutions.
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
(ARC) who contributed to this study, including Sandy S. Dick-
son, Joua Yang, and Erin Sash of the ARC National Testing Labo-
ratories, Dawn Roth and Bryan Blumanhourst of ARC Testing
Support, and the case investigators at the Donor and Client Sup-
port Center; Kathryn Stephenson and Diane Ananos of Beth
Israel Deaconess Medical Center for assisting with follow-up
sample collection from the vaccine-recipient donor; and Kirsten
St. George, Amy Dean, Susan Wong, Karen Kulas, Laura Kramer,
and Alan Dupuis of the Wadsworth Center for their assistance
with confirmatory testing.
1787
 Testing for Zik a Virus among U.S. Blood Donors
References
1. Dick GW, Kitchen SF, Haddow AJ. Zika
virus. I. Isolations and serological speci-
ficity. Trans R Soc Trop Med Hyg 1952;​46:​
509-20.
2. Lanciotti RS, Kosoy OL, Laven JJ, et al.
Genetic and serologic properties of Zika
virus associated with an epidemic, Yap
State, Micronesia, 2007. Emerg Infect Dis
2008;​14:​1232-9.
3. Cao-Lormeau VM, Roche C, Teissier A,
et al. Zika virus, French Polynesia, South
Pacific, 2013. Emerg Infect Dis 2014;​20:​
1085-6.
4. Marchette NJ, Garcia R, Rudnick A.
Isolation of Zika virus from Aedes aegypti
mosquitoes in Malaysia. Am J Trop Med
Hyg 1969;​18:​411-5.
5. Grard G, Caron M, Mombo IM, et al.
Zika virus in Gabon (Central Africa) —
2007: a new threat from Aedes albopictus?
PLoS Negl Trop Dis 2014;​8(2):​e2681.
6. Besnard M, Lastere S, Teissier A, Cao-
Lormeau V, Musso D. Evidence of perina-
tal transmission of Zika virus, French
Polynesia, December 2013 and February
2014. Euro Surveill 2014;​19:​20751.
7. Calvet G, Aguiar RS, Melo ASO, et al.
Detection and sequencing of Zika virus
from amniotic fluid of fetuses with micro-
cephaly in Brazil: a case study. Lancet In-
fect Dis 2016;​16:​653-60.
8. Mlakar J, Korva M, Tul N, et al. Zika
virus associated with microcephaly. N Engl
J Med 2016;​374:​951-8.
9. Rasmussen SA, Jamieson DJ, Honein
MA, Petersen LR. Zika virus and birth de-
fects — reviewing the evidence for causal-
ity. N Engl J Med 2016;​374:​1981-7.
10. Barjas-Castro ML, Angerami RN,
Cunha MS, et al. Probable transfusion-
transmitted Zika virus in Brazil. Transfu-
sion 2016;​56:​1684-8.
11. Motta IJF, Spencer BR, Cordeiro da
Silva SG, et al. Evidence for transmission
of Zika virus by platelet transfusion. N Engl
J Med 2016;​375:​1101-3.
12. Musso D, Nhan T, Robin E, et al. Poten-
tial for Zika virus transmission through
blood transfusion demonstrated during
an outbreak in French Polynesia, Novem-
ber 2013 to February 2014. Euro Surveill
2014;​19:​20761.
13. Foy BD, Kobylinski KC, Chilson Foy
JL, et al. Probable non-vector-borne trans-
mission of Zika virus, Colorado, USA.
Emerg Infect Dis 2011;​17:​880-2.
14. Davidson A, Slavinski S, Komoto K,
Rakeman J, Weiss D. Suspected female-
to-male sexual transmission of Zika virus
— New York City, 2016. MMWR Morb
Mortal Wkly Rep 2016;​65:​716-7.
15. Gallian P, Cabié A, Richard P, et al.
Zika virus in asymptomatic blood donors
in Martinique. Blood 2017;​129:​263-6.
16. Kleber de Oliveira W, Cortez-Escalante
1788 n engl j med 378;19 nejm.org  May 10, 2018
The New England Journal of Medicine
Downloaded from nejm.org on May 9, 2018. For personal use only. No other uses without permission.
Copyright © 2018 Massachusetts Medical Society. All rights reserved.
J, De Oliveira WT, et al. Increase in report-
ed prevalence of microcephaly in infants
born to women living in areas with con-
firmed Zika virus transmission during
the first trimester of pregnancy — Brazil,
2015. MMWR Morb Mortal Wkly Rep 2016;​
65:​242-7.
17. Soares CN, Brasil P, Carrera RM, et al.
Fatal encephalitis associated with Zika
virus infection in an adult. J Clin Virol
2016;​83:​63-5.
18. Carteaux G, Maquart M, Bedet A,
et al. Zika virus associated with meningo­
encephalitis. N Engl J Med 2016;​ 374:​
1595-6.
19. Rozé B, Najioullah F, Signate A, et al.
Zika virus detection in cerebrospinal fluid
from two patients with encephalopathy,
Martinique, February 2016. Euro Surveill
2016;​21:​30205.
20. Mécharles S, Herrmann C, Poullain P,
et al. Acute myelitis due to Zika virus in-
fection. Lancet 2016;​387:​1481.
21. Oehler E, Watrin L, Larre P, et al. Zika
virus infection complicated by Guillain-
Barre syndrome — case report, French
Polynesia, December 2013. Euro Surveill
2014;​19:​20720.
22. Paploski IA, Prates AP, Cardoso CW,
et al. Time lags between exanthematous
illness attributed to Zika virus, Guillain-
Barre Syndrome, and microcephaly, Sal-
vador, Brazil. Emerg Infect Dis 2016;​22:​
1438-44.
23. Cao-Lormeau VM, Blake A, Mons S,
et al. Guillain-Barré Syndrome outbreak
associated with Zika virus infection in
French Polynesia: a case-control study.
Lancet 2016;​387:​1531-9.
24. Chevalier MS, Biggerstaff BJ, Basavaraju
SV, et al. Use of blood donor screening
data to estimate Zika virus incidence,
Puerto Rico, April-August 2016. Emerg
Infect Dis 2017;​23:​790-5.
25. Food and Drug Administration. Donor
screening recommendations to reduce the
risk of transmission of Zika virus by hu-
man cells, tissues, and cellular and tissue-
based products. 2016 (https:/​/​w ww​.fda​
.gov/​downloads/​biologicsbloodvaccines/​
guidancecomplianceregulatoryinformation/​
guidances/​t issue/​ucm488582​.pdf).
26. Food and Drug Administration. Re-
vised recommendations for reducing the
risk of Zika virus transmission by blood
and blood components. 2016 (https:/​/​w ww​
.fda​.gov/​downloads/​biologicsbloodvaccines/​
guidancecomplianceregulatoryinformation/​
guidances/​blood/​ucm518213​.pdf).
27. Williamson PC, Linnen JM, Kessler
DA, et al. First cases of Zika virus-infected
US blood donors outside states with areas
of active transmission. Transfusion 2017;​
57:​770-8.
28. Gao K, James A, Green T, et al. Perfor-
mance and evaluation of Procleix Zika vi-
rus assay on the fully automated Procleix
Panther System. Presented at the 27th Re-
gional Congress of the ISBT, Copenhagen,
June 17–21, 2017. abstract.
29. St George K, Sohi IS, Dufort EM, et al.
Zika virus testing considerations: lessons
learned from the first 80 real-time reverse
transcription-PCR-positive cases diagnosed
in New York State. J Clin Microbiol 2017;​
55:​535-44.
30. Zika MAC-ELISA. Atlanta:​Centers for
Disease Control and Prevention, 2017
(https:/​/​w ww​.fda​.gov/​downloads/​medical
devices/​safety/​emergencysituations/​
UCM488044​.pdf).
31. Ellingson KD, Sapiano MRP, Haass
KA, et al. Cost projections for implemen-
tation of safety interventions to prevent
transfusion-transmitted Zika virus infec-
tion in the United States. Transfusion
2017;​57:​Suppl 2:​1625-33.
32. Duffy MR, Chen TH, Hancock WT,
et al. Zika virus outbreak on Yap Island,
Federated States of Micronesia. N Engl J
Med 2009;​360:​2536-43.
33. Galel SA, Williamson PC, Busch MP,
et al. First Zika-positive donations in the
continental United States. Transfusion
2017;​57:​762-9.
34. Groves JA, Shafi H, Nomura JH, et al.
A probable case of West Nile virus trans-
fusion transmission. Transfusion 2017;​
57:​850-6.
35. Modjarrad K, Lin L, George SL, et al.
Preliminary aggregate safety and immu-
nogenicity results from three trials of a
purified inactivated Zika virus vaccine
candidate: phase 1, randomised, double-
blind, placebo-controlled clinical trials.
Lancet 2017;​391:​563-71.
36. Rios M, Daniel S, Chancey C, Hewlett
IK, Stramer SL. West Nile virus adheres to
human red blood cells in whole blood.
Clin Infect Dis 2007;​45:​181-6.
37. Froeschl G, Huber K, von Sonnenburg
F, et al. Long-term kinetics of Zika virus
RNA and antibodies in body fluids of a
vasectomized traveller returning from
Martinique: a case report. BMC Infect Dis
2017;​17:​55.
38. Paz-Bailey G, Rosenberg ES, Doyle K,
et al. Persistence of Zika virus in body
fluids — preliminary report. N Engl J Med
2017. DOI:​10.1056/NEJMoa1613108.
39. Klein HG, Hrouda JC, Epstein JS.
­Crisis in the sustainability of the U.S.
blood system. N Engl J Med 2017;​377:​
1485-8.
40. Dodd RY, Foster GA, Stramer SL.
Keeping blood transfusion safe from West
Nile virus: American Red Cross experience,
2003 to 2012. Transfus Med Rev 2015;​29:​
153-61.
Copyright © 2018 Massachusetts Medical Society.