Conventionalmanagement ofmaternal red cell alloimmunization

compared with management by Doppler assessment of middle

cerebral artery peak systolic velocity

Leonardo Pereira, MD,a Thomas M. Jenkins, MD,b and Vincenzo Berghella, MDa

Philadelphia, Pa, and Madison, Wis

OBJECTIVE: This study was undertaken to compare management of red blood cell alloimmunization by

Doppler measurement of middle cerebral artery peak systolic velocity (MCA-PSV) to conventional

management with amniocentesis.

STUDY DESIGN: A historical cohort of 28 fetuses at risk for anemia caused by red blood cell alloimmunization

was followed between 1999 and 2002 at a single institution. The decision to perform percutaneous umbilical

cord blood sampling (PUBS) was based on conventional management. MCA-PSV Doppler was measured

before amniocentesis or PUBS but not used clinically.

RESULTS: Twenty-eight fetuses were followed up: 4 had severe anemia, 1 had moderate anemia, 3 had mild

anemia, and 20 were nonanemic. Conventional management had a sensitivity and positive predictive value

for moderate-to-severe anemia of 80% and 44%, with a false-positive rate of 56%. In the same patients,

MCA-PSV Doppler had a sensitivity and positive predictive value for moderate-to-severe anemia of 100%

and 71%, with a false-positive rate of 28%.

CONCLUSION: Compared with conventional management, MCA-PSV Doppler may have a better predictive

accuracy for moderate-to-severe fetal anemia in red blood cell alloimmunization. Management by MCA-PSV

Doppler may eliminate the need for amniocentesis and reduce the number of PUBS performed in red blood

cell-alloimmunized pregnancies. (Am J Obstet Gynecol 2003;189:1002-6.)

Key words: Fetal anemia, hemolytic disease of the newborn infant, middle cerebral artery peaksystolic velocity, Doppler, red blood cell alloimmunization

Several prospective studies have found that elevated

middle cerebral artery peak systolic velocity (MCA-PSV)

Doppler measurements accurately predict fetal anemia in

red blood cell–alloimmunized pregnancies.1-6 Currently

in the United States, the management of alloimmunized

pregnancies varies among institutions. Some practitioners

rely on conventional management (maternal blood titers,

amniocentesis, and ultrasound evaluation for fetal

hydrops), whereas others have adopted management

strategies centered solely on noninvasive techniques

(MCA-PSV Doppler assessment and ultrasound evalua-

tion for fetal hydrops).

Conventional management has been validated by an

abundance of scientific and clinical evidence7 but has the

inherent complication rate associated with amniocen-

tesis, including transplacental hemorrhage, premature

From the Departments of Obstetrics and Gynecology, Jefferson MedicalCollege of Thomas Jefferson University,a and University of Wisconsin-Madison.b

Presented at the Twenty-Third Annual Meeting of the Society forMaternal-Fetal Medicine, San Francisco, Calif, February 3-8, 2003.Reprints not available from the authors.� 2003, Mosby, Inc. All rights reserved.0002-9378/2003 $30.00 + 0doi:10.1067/S0002-9378(03)00771-3

1002

rupture of membranes, premature labor, infection, and

fetal loss.8-11

Management by MCA-PSV Doppler does not carry

the risks associated with amniocentesis but involves

multiple ultrasound examinations, which must be per-

formed with strict adherence to proper technique.

Published studies have reported impressive sensitivities

(88%-100%),3,5 although clinical experience with MCA-

PSV Doppler is far less than with conventional manage-

ment.

To date, there has been only 1 article comparing the 2

management strategies in the same patient population,

and this series was limited to Rh disease.12 The purpose of

our study was to compare conventional management of

maternal alloimmunization with antierythrocyte antigens

with management by MCA-PSV Doppler. Our focus was on

the sensitivity and false-positive rates of each management

strategy for the detection of moderate-to-severe fetal

anemia.

Material and methods

This was a retrospective study of fetuses at risk for

anemia caused by maternal erythrocyte alloimmunization

followed between 1999 and 2002 at Thomas Jefferson

Volume 189, Number 4Am J Obstet Gynecol

Pereira, Jenkins, and Berghella 1003

University, Philadelphia, Pa. The institutional review

board approved the study before data collection. No cases

presenting during the study period were excluded. All

fetuses at risk for anemia secondary to maternal red cell

alloimmunization were monitored with the use of our

conventional management. This consisted of serial ultra-

sound evaluations to rule out fetal hydrops and amnio-

centeses performed once the maternal serum antibody

titer crossed a critical threshold level (16), or earlier if the

patient had a prior affected pregnancy. Once performed,

amniocenteses were repeated at 2- to 4-week intervals, and

ultrasound examinations were performed every 2 weeks.

The decision to perform percutaneous umbilical cord

blood sampling (PUBS) was based on abnormal amnio-

centesis results or sonographic evidence of fetal hydrops,

with the exception of Kell sensitization. Patients with anti-

Kell alloimmunization had serial PUBS performed for

worsening titers secondary to lack of sensitivity of DOD450

levels.13-14

Blood transfusions were performed for a fetal hemo-

globin level below 9 g/dL, which corresponded to either

moderate or severe fetal anemia as defined by Mari et al1

in all cases. After transfusion, PUBS with possible repeat

transfusion was performed until 34 weeks’ gestation, and

ultrasounds were performed weekly.

During the study period, MCA-PSV Doppler mea-

surements were obtained between 22 and 34 weeks’

gestation during each sonographic assessment for fetal

hydrops. Ultrasound examinations were not scheduled

expressly to measure MCA-PSVs. Physicians were not

blinded to MCA-PSV values; however, these were not used

for clinical management. The sonographic diagnosis of

fetal hydrops was made if two or more of the following

findings were present: ascites, skin edema, pleural

effusion, or pericardial effusion.

Doppler measurements of the MCA were performed

with color Doppler imaging (Acuson XP, Sequoia, and

Aspen, Acuson, Mountain View, Calif). All patients had

MCA-PSV Doppler measurements done by one of six

experienced sonographers and reviewed by a perina-

tologist, using the technique described by Mari et al in

2000.3 This measurement was repeated multiple times

during periods of fetal apnea, and the highest MCA-PSV

Doppler measurement was recorded. When performed in

this fashion, multiple sources have previously reported

low intraobserver and interobserver variability rates

between 2.3% and 4.0%.1,15,16

MCA-PSV Doppler values were expressed as multiples

of the median (MoM) by using previously established

reference ranges,3 and defined as abnormal if greater

than 1.5 MoM. Anemia was defined as a function of

gestational age by using previously established reference

ranges: mild anemia less than 0.84 MoM; moderate

anemia less than 0.65 MoM; and severe anemia less than

0.55 MoM.3

Data were entered into an SPSS data document.

Statistical analysis was performed with the SPSS statistical

package (Statistical Package for Social Sciences, SPSS, Inc,

Chicago, Ill). Fisher exact test was used to calculate P

values. A two-sided P value < .05 was used to define

statistical significance.

The predictive accuracy of our conventional manage-

ment for moderate-to-severe fetal anemia was deter-

mined. The number of invasive procedures, which

would have been attempted in our patient population if

abnormal MCA-PSV Doppler had been the indication for

PUBS was then determined. The predictive accuracy of

MCA-PSV Doppler imaging for moderate-to-severe fetal

anemia was determined, and the predictive accuracy

between management strategies compared.

Results

Twenty-eight fetuses were monitored during the study

period. Nine pregnancies were complicated by RhD

disease, 7 by RhD with additional erythrocyte antigens, 2

K1, 2 c, 3 E, and 5 by miscellaneous erythrocyte antigens

(Table 1). Patients were monitored from time of initial

presentation (range 22 to 30 weeks’ gestation) until

delivery. Five fetuses had moderate-to-severe anemia

develop, and of these, 2 were hydropic.

With the use of our conventional management, 24 of 28

fetuses (86%) underwent at least 1 amniocentesis; and

a total of 49 amniocenteses were performed. Of 28 fetuses,

9 (32%) underwent at least 1 PUBS; and a total of 14 PUBS

were performed. In the 9 fetuses that underwent PUBS, 3

had severe anemia, 1 had moderate anemia, 3 had mild

anemia, and 2 had normal hemoglobin levels (Table 2).

Two nonanemic fetuses underwent PUBS. In the first

case, serial amniocenteses were performed because of an

anti-c titer of 32 in a patient with hemolytic disease of the

newborn infant in a previous gestation. Amniocentesis at

25 weeks’ gestation revealed a DOD450 of 0.073 (Liley

Table I. Distribution of erythrocyte antibody specificities

Antibody specificity n

D 9D and C 3D and M 2D, C, E, Jka 1D, C, e, Jka 1K1 1K1 and Jka 1c 2E 3C and Jka 1u 1Anticoltan 1Jka 1Fya 1

1003

October 2003Am J Obstet Gynecol

1004 Pereira, Jenkins, and Berghella

Table II. Antenatal course and test results up to first PUBS

nAntibodyspecificity

Prioraffectedfetus

Indication foramniocentesis

Amniocentesisresults

Ultrasoundfindings

MCA-PSVDoppler(m/s)

Indication forPUBS

Fetal hgb(g/dL) Anemia

1 D No Anti-D titer 128 0.239 at 29 wk Fetal hydrops 0.80 Zone 3DOD450;fetal hydrops

5.3 at 29w Severe

2 D Yes Anti-D titer 1024 0.165 at 24 wk Normal 0.74 High zone2DOD450

5.4 at 31w Severe

0.169 at 27 wk0.137 at 31 wk

3 D No Anti-D titer 512 0.330 at 31 wk Fetal hydrops 0.68 Zone 3 DOD450;fetal hydrops

6.9 at 31w Severe

4 D No Anti-D titer 512 0.286 at 31 wk Ascites 0.71 Zone 3 DOD450 7.2 at 31w Moderate5 D and C No Anti-D titer 512 0.118 at 30 wk

0.111 at 32 wkNormal 0.41 High zone

2 DOD45010.8 at 32w Mild

6 K1 No — — Normal 0.37 Kell titer 1:64 10.6 at 29w Mild7 K1 and JKa No — — Normal 0.47 Kell titer 1:128 9.1 at 30w Mild8 c Yes Anti-c titer 32 0.073 at 25 wk Normal 0.40 Zone 3 DOD450 13.9 at 30w None

0.355 at 30 wk9 D, C, E, Jka No Anti-D titer 8192 0.363 at 28 wk Normal 0.31 Zone 3 DOD450 14.1 at 28w None

Hgb; Hemoglobin.

zone 1); however, a subsequent amniocentesis at 30 weeks’

gestation revealed dark-brownish fluid and a DOD450

of 0.355 (Liley zone 3). The second case occurred in

a pregnancy complicated by anti-D, anti-C, anti-E, anti-Jka

antibodies. Amniocentesis was performed at 28 weeks’

gestation because of an anti-D titer of 8192 and revealed

hazy fluid, which appeared meconium stained, and

a DOD450 of 0.363 (Liley zone 3). Ultrasound ex-

aminations were normal in both cases, as were MCA-PSV

Doppler measurements.

One case of severe fetal anemia, in a pregnancy

complicated by anti-D, anti-C, anti-e, and anti-Jka

alloimmunization, was missed antenatally. Serial amnio-

centeses were persistently in mid-Liley zone 2. There was

no evidence of fetal hydrops on serial ultrasounds;

however, MCA-PSV Doppler measurements were consis-

tently abnormal. The neonate had an initial hemoglobin

of 6.3 g/dL after an uncomplicated vaginal delivery at 37

weeks’ gestation.

Conventional management had a sensitivity for mod-

erate-to-severe anemia of 80%, specificity 78%, positive

Table III. Comparison of management strategies for

detection of moderate-to-severe anemia

Conventionalmanagement

MCA-PSV Dopplerimaging P-value

Sensitivity (cases) 80 (4/5) 100 (5/5) >.999Specificity (cases) 78 (18/23) 91 (21/23) .41PPV (cases) 44 (4/9) 71 (5/7) .36NPV cases 95 (18/19) 100 (21/21) .48FP rate (cases) 56 (5/9) 28 (2/7) —FN rate (cases) 5 (1/19) 0 (0/21) —RR (95% CI) 3.7 (1.5–9.0) 11.5 (3.1–43.2) —

FP; False positive; FN; false negative.

1004

predictive value (PPV) 44%, and negative predictive value

(NPV) 95%. Five PUBS were performed on fetuses with

either mild anemia or normal hemoglobin levels (false-

positive results), for a false-positive rate of 56%. If con-

ventional management indicated PUBS, the fetus had

a relative risk (RR) for moderate-to-severe anemia of 3.7

(95% CI 1.5-9.0) compared with fetuses in which a PUBS

was not indicated (Table 3).

In the course of ultrasound screening for fetal hydrops,

104 MCA-PSV Doppler measurements were obtained

before 35 weeks’ gestation. MCA-PSV Doppler mea-

surements greater than 1.5 MoM were considered ab-

normal. By using this cutoff value, if the decision to

perform PUBS had been based solely on MCA-PSV

Doppler measurements, 7 patients rather than 9 would

have been offered sampling. If these procedures had been

successfully performed, all 5 cases of moderate-to-severe

anemia would have been detected, with 2 nonanemic

fetuses being sampled.

MCA-PSV for the prediction of moderate-to-severe

anemia in our population had a sensitivity of 100%,

specificity of 91%, PPV 71%, and NPV 100%. Two PUBS

would have been performed on nonanemic fetuses for

a false-positive rate of 28% (2/7). Fetuses with abnormal

MCA-PSV measurements had a RR for moderate-to-severe

anemia of 11.5 (95% CI 3.1-43.2) compared with fetuses

with MCA-PSV less than 1.5 MoM. There were no

statistically significant differences between management

strategies (Table 3).

Comment

The discovery that Rh sensitization could be quantified

by assessing the spectral absorption curve of amniotic

Volume 189, Number 4Am J Obstet Gynecol

Pereira, Jenkins, and Berghella 1005

fluid at 450 nm (DOD450) was described by Liley in 1961.17

From this important observation, successful antenatal

management strategies for maternal red blood cell

alloimmunization were developed.

Despite its widespread use and accepted efficacy, con-

ventional management of red blood cell alloimmuni-

zation has limitations. Amniocentesis DOD450 values may

be falsely elevated in the presence of meconium or

blood and may provide misleading low values after inad-

vertent exposure to light or in cases of Kell alloimmu-

nization.13-14 Transplacental fetal hemorrhage, which

may worsen sensitization, occurs after 2% to 11% of

amniocenteses.8-10 Another 1% to 2% of amniocenteses

are complicated by rupture of amniotic membranes,

premature labor, vaginal bleeding, or infection, whereas

fetal loss occurs in approximately 0.5% of cases.11

In 1995, Mari et al1 reported that MCA-PSV Doppler

measurements could accurately predict fetal anemia in

a prospective series of 16 pregnancies complicated by

maternal red blood cell alloimmunization. Since that

initial report, several other studies, including a large

prospective, intent-to-treat trial of 125 cases by

Zimmermann et al2-6 have confirmed these findings.

Although these studies have generated enthusiasm

for noninvasive management of erythrocyte immunized

pregnancies, management by MCA-PSV Doppler has

limitations. Serial MCA-PSV Doppler measurements

must be conducted in strict adherence with proper

technique to maintain diagnostic accuracy. Furthermore,

the accuracy of MCA-PSV Doppler appears to diminish

after 35 weeks’ gestation.5 The reliability of MCA-PSV

Doppler to predict fetal anemia after multiple in-

trauterine transfusions has not been tested prospec-

tively, and false-negative cases have been reported.5,18

Intrauterine transfusions increase fetal blood viscos-

ity,18,19 which may alter the predictive accuracy of MCA-

PSV Doppler.

On the other hand, years of experience have shown

amniocentesis to be a reliable predictor of fetal anemia

in most cases of red blood cell alloimmunization.

Although MCA-PSV Doppler measurements do not carry

the risks associated with amniocentesis, this argument

may not be sufficient to justify replacing conventional

management, given the paucity of studies directly

comparing the two management strategies. However, if

trials comparing both strategies conclude that MCA-PSV

Doppler is at least as effective as conventional manage-

ment, then the safety of noninvasive management makes

it preferable.

A recently published study of 28 nonhydropic fetuses by

Nishie et al12 found that conventional management and

MCA-PSV Doppler were both accurate predictors of fetal

anemia in Rh disease. The authors suggested in their

conclusion that management with MCA-PSV Doppler

could decrease the number of invasive procedures per-

formed in this population.

The results of our study suggest that management by

MCA-PSV Doppler, compared with conventional manage-

ment, may have better predictive accuracy for moderate-

to-severe fetal anemia. The sensitivity, specificity, PPV,

NPV, and RR were all higher for MCA-PSV Doppler

measurements than conventional management, whereas

the false-positive and false-negative rates were lower. With

the use of MCA-PSV Doppler measurements, severely

anemic fetuses were more likely to be identified, whereas

nonanemic fetuses were less likely to meet criteria for

PUBS. As previously stated, however, there were no

statistically significant differences between management

strategies.

Although both management strategies had false-posi-

tive cases, management by MCA-PSV Doppler reduced the

number of false-positive cases in half (from 4 to 2).

Furthermore, management by MCA-PSV Doppler did not

miss any fetuses with moderate-to-severe anemia com-

pared with conventional management, which missed 1

case.

A limitation of this study was inherent in its retrospec-

tive design, precluding intent-to-treat randomization.

Furthermore, amniocentesis results were used for man-

agement, whereas MCA-PSV Doppler measurements were

not. Nevertheless, the study design did permit a com-

parison of both management strategies in the same

population.

The study had a power of 0.02 to detect the observed

difference in sensitivity between management strategies.

Of 28 fetuses, only 5 had moderate-to-severe anemia. With

the observed effect size, 11 additional fetuses with

moderate-to-severe anemia would be needed for the study

to reach a power of 0.80.

Furthermore, when applied to a larger cohort, the

sensitivity, specificity, PPV, and NPV of MCA-PSV Doppler

may decrease. This must be assessed as management of

alloimmunization by MCA-PSV Doppler becomes more

widespread. Large randomized studies comparing man-

agement strategies will be helpful in determining optimal

clinical management of red blood cell-alloimmunized

pregnancies.

In conclusion, compared with conventional manage-

ment, management by MCA-PSV Doppler imaging may

have a better predictive value and sensitivity for moderate-

to-severe anemia in red blood cell alloimmunization.

Management by MCA-PSV Doppler may ultimately elim-

inate the need for amniocentesis in pregnancies compli-

cated by red cell blood alloimmunization and reduce the

number of PUBS performed on nonanemic fetuses.

We thank Stuart Weiner, MD, Anthony Sciscione, DO,Natasha Pereira, MD, Kelly McCollum MPH, Dennis

1005

October 2003Am J Obstet Gynecol

1006 Pereira, Jenkins, and Berghella

Wood Jr, RDMS, Gjergji Bega, MD, and Marion Kauf-mann, RN, BSN, for helping us complete this project.

REFERENCES

1. Mari G, Adrignolo A, Abuhamad AZ, Pirhonen J, Jones DC,Ludomirsky A, et al. Diagnosis of fetal anemia with Dopplerultrasound in the pregnancy complicated by maternal blood groupimmunization. Ultrasound Obstet Gynecol 1995;5:400-5.

2. Teixeira JMA, Duncan K, Letsky E, Fisk NM. Middle cerebral arterypeak systolic velocity in the prediction of fetal anemia. UltrasoundObstet Gynecol 2000;15:205-8.

3. Mari G, Deter RL, Carpenter RL, Rahman F, Zimmermann R, MoiseKJ Jr., et al. Noninvasive diagnosis by Doppler ultrasonography offetal anemia due to maternal red-cell alloimmunization. N Engl JMed 2000;342:9-14.

4. Delle Chiaie L, Buck G, Grab D, Terinde R. Prediction of fetalanemia with Doppler measurements of the middle cerebral arterypeak systolic velocity in pregnancies complicated by maternal bloodgroup alloimmunization or parvovirus B19 infection. UltrasoundObstet Gynecol 2001;18:232-6.

5. Zimmermann R, Durig P, Carpenter RJ Jr., Mari G. Longitudinalmeasurement of peak systolic velocity in the fetal middle cerebralartery for monitoring pregnancies complicated by red cell alloimmu-nisation: a prospective multicentre trial with intention-to-treat. BJOG2002;109:746-52.

6. Mari G, Detti L, Oz U, Zimmermann R, Durig P, Stefos T. Accurateprediction of fetal hemoglobin by Doppler ultrasonography. Ultra-sound Obstet Gynecol 2002;99:589-93.

7. Bowman J. Hemolytic disease (Erythroblastosis fetalis). In: CreasyRK, Resnik R, editors. Maternal-fetal medicine. Philadelphia: WBSaunders; 1999. p. 736-67.

8. Bowman JM, Pollock JM. Transplacental fetal hemorrhage afteramniocentesis. Obstet Gynecol 1985;66:749-54.

1006

9. Woo Wang MYF, McCutcheon E, Desforges JF. Fetomaternal

hemorrhage from diagnostic transabdominal amniocentesis. Am J

Obstet Gynecol 1967;97:1123-8.10. Peddle LJ. Increase of antibody titer following amniocentesis. Am J

Obstet Gynecol 1968;100:567-9.11. Simpson JL. Incidence and timing of pregnancy losses: relevance to

evaluating safety of early prenatal diagnosis. Am J Med Genet

1990;35:165-73.12. Nishie EN, Brizot ML, Liao AW, Carvalho MHB, Toma O, Zugaib M.

A comparison between middle cerebral artery peak systolic velocity

and amniotic fluid optical density at 450 nm in the prediction of fetal

anemia. Am J Obstet Gynecol 2003;188:214-9.13. Caine ME, Mueller-Heubach E. Kell sensitization in pregnancy. Am J

Obstet Gynecol 1986;154:85-90.14. Vaughan JI, Warwick R, Letsky E, Nicolini U, Rodeck CH, Fisk NM.

Erythropoietic suppression in fetal anemia because of Kell alloim-

munization. Am J Obstet Gynecol 1994;171:247-52.15. Akiyama M, Detti L, Abuhamad A, Bahado-Singh R, Mari G. Is the

middle cerebral artery peak systolic velocity measurement affected by

the site of vessel sampling?. Am J Obstet Gynecol 2002;187(Suppl):

S164.16. Mari G, Abuhamad A, Brumfield J, Ferguson JE II. Doppler

ultrasonography of the middle cerebral artery peak systolic velocity

in the fetus: reproducibility of measurement. Am J Obstet Gynecol

2001;185(Suppl):S261.17. Liley AW. Liquor amnii analysis in the management of the pregnancy

complicated by rhesus sensitization. Am J Obstet Gynecol 1961;82:

1359-70.18. Detti L, Oz U, Guney I, Ferguson JE, Bahado-Singh RO, Mari G.

Doppler ultrasound velocimetry for timing the second intrauterine

transfusion in fetuses with anemia from red cell alloimmunization.

Am J Obstet Gynecol 2001;185:1048-51.19. Stefos T, Cosmi E, Detti L, Mari G. Correction of fetal anemia on the

middle cerebral artery peak systolic velocity. Obstet Gynecol

2002;99:211-5.