Pregnancy is a contraindication for intraoperative autotransfusion because of concern that final, postprocessed, packed red blood cells (RBCs) might be contaminated with amniotic fluid elements that could cause amniotic fluid embolism syndrome. Despite a lack of proven safety in obstetrics, there have been anecdotal reports in the literature of least 122 cases1,2 in which blood salvaged intraoperatively was used for autotransfusion in women who had cesareans. It is likely that many more instances were not reported. In in vivo case reports in which the cell-saver was used during cesarean, when timing of salvage was described, the blood was collected after placental removal in all cases, and retransfusion resulted in no untoward complications. In the present study, our aim was to determine which fetal products remained in blood salvaged and processed after removal of the placenta during cesarean delivery, and to analyze the packed RBCs for immunoreactivity and compatibility with maternal serum.
With Human Rights Committee approval and informed consent obtained, we salvaged blood shed from 27 term gravidas with intact membranes who had elective cesarean deliveries, using the COBE BRAT-2™ (COBE Cardiovascular, Arvada, CO) intraoperative autotransfusion apparatus. Women with known hemoglobinopathies or other potential contraindications to autologous blood transfusions were excluded. Subjects were a consecutive series of women who met inclusion criteria and came under the care of the investigator (JF) on days when she was the attending anesthesiologist in the labor and delivery department during the 8 months from August 1996 to April 1997. Based on changes in preprocessing and postprocessing concentrations of alpha-fetoprotein (AFP), cellular material, and tissue factor in earlier abstracts and studies of blood salvage during cesarean delivery, we calculated that ten subjects would detect a significant change between preprocessing and postprocessing specimens, with a power of 99% at significance of 0.05, assuming independent measurements and no effect on treatment.
According to protocol, salvaging began 4 minutes after removal of placenta to allow the surgeons enough time to suction the peritoneal gutters and remainder of the operative field, then change to a clean, unused suction cannula and canister system for collection of blood from the surgical site. The salvaged blood was mixed in the cannula with 30 U/mL heparinized saline at a concentration of approximately 1:7. Suction pressure was 100 mmHg. Salvaged blood was processed in a 135-mL centrifuge bowl with a fill speed of 200 mL/minute, a wash speed of 400 mL/minute, and an empty speed of 200 mL/minute. Washing was done with 1500 mL 0.9% normal saline solution. Figure 1 is a schematic diagram of the autotransfusion apparatus. Salvaged blood was used for study purposes only; none was retransfused to any women.
Maternal peripheral blood samples were collected preoperatively and analyzed for AFP concentration, hemoglobin concentration, hematocrit, and standard blood type and antibody screen. Cord blood was collected at delivery and analyzed for blood type and antibody screen. Separate preprocessing and postprocessing samples of salvaged blood were taken from the suction canister and blood bag, respectively. Tests done on the preprocess and postprocess scavenged blood were AFP concentration, using the Abbott AxSYM System, a microparticle enzyme immunoassay technology (Abbott Laboratories, Abbott Park, IL) that has a lower limit of detection of 7.5 ng/mL; Papanicolaou smears, using three 5-mL specimens each from preprocess and postprocess specimens to detect lanugo, vernix, and anucleate squamous cells that might be fetal3; hematocrit, hemoglobin, and plasma-free hemoglobin concentrations; and Ouchterlony tests of antigen-antibody reaction4 run against maternal serum from the preoperative sample, in which samples were placed on an agar medium at room temperature and analyzed for precipitate reactions at 18, 24, and 48 hours postincubation, a qualitative technique that can detect protein interactions at microgram concentrations. All postprocessing-salvaged packed RBC samples were subjected to Kleihauer-Bethke analyses, typed, and crossmatched against maternal serum, including mixed fields.
Comparisons of laboratory values between preprocessing and postprocessing samples were made using paired t tests for normally distributed data and Wilcoxon signed-rank tests for non-normally distributed data. The Friedman repeated-measures analysis of variance on ranks test and the Dunnett method for all pairwise multiple comparisons were used to analyze differences among preoperative maternal, preprocessing, and postprocessing AFP values. Demographic comparisons between evaluated subjects and those who could not be evaluated were made using t tests or χ2 tests for normally distributed data, and Mann-Whitney rank-sum tests for non-normally distributed data. P < .05 was significant for all tests.
A total of 27 subjects were enrolled to find ten who shed enough blood after removal of placenta for a postprocessing product. There was no difference between the eligible group and the ineligible group with respect to age, gravidity, parity, prepregnant weight, fetal gestational age, neonatal weight, race, Apgar scores, duration of cesarean, surgeon's estimated blood loss, or reason for cesarean. When compared with the eligible women, the ineligible group had significantly lower pregnant weights (mean ± standard deviation [SD]), 82.7 ± 14.5 kg versus 70.6 ± 10.8 kg; weight gain, 16.7 ± 5.3 kg versus 12.3 ± 4.2 kg; and height, 157.9 ± 9.4 cm versus 150.2 ± 7.2 cm. The indications for cesarean were previous uterine surgery, elective repeat cesarean, or breech presentation. Only one subject had placenta previa; she was in the ineligible group.
The mean (± SD) volume of blood salvaged from the ten eligible women before processing was 485.3 ± 122.4 mL, with a mean (± SD) hematocrit of 23.10 ± 5.8% and mean (± SD) plasma-free hemoglobin concentration of 337.7 ± 98.0 mg/dL. Processing yielded a mean (± SD) volume of packed RBCs of 186.2 ± 36.6 mL, with a mean (± SD) hematocrit of 48.95 ± 14.2% and mean (± SD) plasma-free hemoglobin concentration of 159.6 ± 67.2 mg/dL. All differences between preprocessing and postprocessing values were statistically significant (P < .001). When we applied the calculation of free hemoglobin mass reported by Kindscher,5 we found that the median free-hemoglobin mass of the preprocessing specimens was 1130 mg (interquartile range [IQR] 1055–1608 mg), whereas in the final products, it was 103 mg (IQR 83–139 mg), representing a median percent reduction in free-hemoglobin mass from preprocessing to postprocessing specimens of 91.6% (IQR: 89.5–92.2%), which was statistically significant (P = .004).
Table 1 lists the test results for the ten eligible cases. The median AFP concentration in preoperative maternal serum was 58 ng/mL (IQR: 52 ng/mL–75 ng/mL), not significantly different from the median AFP concentration in the preprocessed salvaged blood of 53 ng/mL (IQR: 31 ng/mL–73 ng/mL). However, AFP was cleared to below lower limits of detection by our immunoassay (less than 7.5 ng/mL) in all postprocessing specimens, which by pairwise comparisons were significantly different from the preoperative and preprocessing specimens (P < .05). No antigen-antibody reaction between any pairings of maternal serum with preprocessed or postprocessed salvaged blood proteins was found by the Ouchterlony method at 18, 24, or 48 hours postincubation. Crossmatching was successful in all pairings of maternal serum with postprocessing salvaged blood, with negative mixed fields in all cases, even when maternal and fetal blood types were different. All Kleihauer-Bethke analyses on postprocessing salvaged blood were positive, with a mean of 0.4 ± 0.133% (IQR: 0.2–1.0%). All women had negative Kleihauer-Bethke analyses preoperatively, and none had hematologic disorders that increased the amount of fetal hemoglobin present in circulation. Anucleate squamous cells were detected on Papanicoalau smear in four of ten preprocessed specimens, one of which was cleared by processing. Lanugo and vernix were not detected in any specimens.
Major obstetric hemorrhage causes 13.4% of maternal mortality in the United States.6 Although fatal obstetric hemorrhage is rare, obstetric patients receive 4% of blood components used annually in the United States.7,8 According to the February 1997 report from the United States General Accounting Office,9 the risk of developing a chronic disease or dying from homologous blood transfusion is four in 10,000, making autologous blood transfusion for pregnant women appealing to women and physicians.
A literature search using PubMED and a review of anesthesiology and obstetric abstracts and letters from January 1985 to September 1998 found at least 1221,2 women who had intraoperative cesarean cell salvage despite the lack of proven safety. The ability of the cell-saver apparatus to remove amniotic fluid elements from salvaged blood was the subject of a few in vitro analyses. Bernstein et al10 salvaged all blood and amniotic fluid shed at cesarean, analyzed preprocessing and postprocessing samples, and found no tissue after processing. Using Kleihauer-Bethke analysis, Durand et al11 detected close to 1% fetal red blood cells in 20% of specimens collected during 15 cesareans and processed by the autotransfusion apparatus. Rainaldi et al2 transfused cell-salvaged, packed RBCs into 34 women who had cesareans without complications. In a separate analysis, 15 specimens salvaged during cesarean but not retransfused showed no microbacterial contamination. Two of fifteen had procoagulant activity, and four had fetal RBCs. Subject selection criteria, transfusion indication, and interval between placenta removal and blood salvage were not specified.
The cell salvage technology is being used clinically, so the present study was conducted to assess the quality and nature of the product collected during intraoperative blood salvage after removal of placenta, simulating in vivo case reports. Only 37% of our subjects shed enough blood for a postprocessing specimen after removal of the placenta. It is unlikely that the difference in weight gain and height between the eligible and ineligible groups affected the quantity of blood salvaged because the surgeon's estimated blood loss and operating times were similar between groups. Despite a smaller centrifuge bowl (135 mL), the mean hematocrit of the final packed RBCs was 48.95 ± 14.2%. We found a mean reduction in free-hemoglobin mass of 90.89 ± 4.14%, which, according to Kindscher,5 is a more accurate measure of washing efficacy because reliance upon plasma free-hemoglobin concentration is confounded by the differences in hematocrit of the preprocessing and postprocessing specimens. Although of potentially less quantity, the quality of the product collected from intraoperative blood salvage after placenta removal was quite good.
Alpha-fetoprotein was cleared from the final processed packed RBCs. The results of our Kleihauer-Bethke analyses indicated fetal RBCs were present in salvaged blood and could not be eliminated by processing. The range of Kleihauer-Bethke results was 0.2% to 1% fetal cells in the postprocessing samples, translating to 1–5 mL of fetal blood in postprocessing packed RBCs,12 which raises concern over maternal alloimmunization after intraoperative autotransfusion, not just for Rhesus factor,13 but also other antigens such as Kell, Kidd, or Duffy. Post-transfusion maternal blood specimens should be sent for accurate RhoGAM dosing. The risk also exists with exposure to homologous blood products, so the fetal antigens to which women are sensitized during intraoperative autotransfusion show the father's antigenicity, putting subsequent pregnancies at proportionately greater risk.
In our study, despite the inability of cell-saver processing to clear cellular debris, the product was compatible with maternal blood by crossmatch, which was consistent despite positive Kleihauer-Bethke results, due to low fetal RBC load in the processed packed RBCs. The product's supernate did not react immunologically with maternal serum. The negative Ouchterlony testing before and after processing and the lack of lanugo or vernix, showed nearly complete elimination of amniotic fluid from the operative field before blood salvage.
Anucleate squamous cells that might be fetal were in our postprocessed packed RBCs. Based on animal and human data, those cells might be innocuous. Stolte et al14 administered to monkeys large amounts of IV primate and human amniotic fluid, with and without meconium, and none of the animals developed amniotic fluid embolism syndrome. Several studies of pulmonary-artery, catheter-blood specimens collected from normal, asymptomatic, pregnant women had possibly fetal squames and anucleate squamous cells during pregnancy.15–17
The cause of amniotic fluid embolism syndrome is unknown, so it is unclear if tissue18 or other humoral factors such as leukotrienes19,20 are causes, or whether fetal cells are significant risks. Clark et al21,22 postulated that an unknown component of amniotic fluid causes anaphylactoid reaction in susceptible individuals, with an incidence of 1:8,000–1:80,000. Assuming that this unknown component is not cleared by autotransfusion processing and Clark was correct, the risk from auto-transfusion is 1:8,000–1:80,000, less than the 1:2,500 risk9 of chronic morbidity or mortality from homologous transfusion. We caution that amniotic fluid embolism syndrome has an 86% rate of severe, irreparable neurologic damage21 or death.
Intraoperative autotransfusion as an alternative to homologous transfusion might be valid, despite lack of proven safety, as long as risks are explained to women. We cannot prove safety, but we did not find a reason to not use cell-salvaged packed RBCs in women who had cesareans. Until a marker for amniotic fluid embolism syndrome is found, clinical investigations should focus on decreasing cellular elements in salvaged products. Waiting longer than 4 minutes after removing placenta, or leukocyte depletion filters might eliminate squamous cells,23,24 white cells, or fetal RBCs.
1. Grimes DA. A simplified device for intraoperative autotransfusion. Obstet Gynecol 1988;72:947–50.
2. Rainaldi MP, Tazzari PL, Scagliarini G, Borghi B, Conte R. Blood salvage during caesarean section. Br J Anaesth 1998;80:195–8.
3. Covone AE, Mutton D, Johnson PM, Adinolfi M. Trophoblast cells in peripheral blood from pregnant women. Lancet 1984;2:841–3.
4. Ouchterlony O, Nilsson LA. Immunodiffusion and immunoelectrophoresis. In: Weir DM, ed. Handbook of experimental immunology. Oxford, England: Blackwell Scientific, 1978;1:19.1–19.4.
5. Kindscher J. Hemolysis, plasma free hemoglobin, and washing efficiency during red blood cell scavenging. Anesth Analg 1995;80:SCA1–SCA141.
6. Kaunitz AM, Hughes JM, Grimes DA, Smith JC, Rochat RW, Kafrissen ME. Causes of maternal mortality in the United States. Obstet Gynecol 1985;65:605–12.
7. Klapholz HL. Blood transfusion in contemporary obstetric practice. Obstet Gynecol 1990;75:940–3.
8. Menitove JE. How safe is blood transfusion? In: Smith DM, Dodd RY, eds. Transfusion transmitted infections. Chicago: ASCP Press, 1991:7.
9. United States General Accounting Office. Blood supply: Transfusion-associated risks. Report to the ranking minority member, Committee on Commerce, House of Representatives. GAO/PEMD-97-2. Washington, DC: February 1997.
10. Bernstein HH, Rosenblatt MA, Gettes M, Lockwood C. The ability of the Haemonetics® cell saver system to remove tissue factor from blood contaminated with amniotic fluid. Anesth Analg 1997;85:831–3.
11. Durand F, Duchesne-Gueguen M, LeBervet JY, Marcorelles P, Tardivel R, Vovan JM, et al. Rheologic and cytologic study of autologous blood collected with Cell Saver 4 during cesarean. Rev Fr Transfus Hemobiol 1989;32:179–91.
12. Walker RH. Fetal-maternal hemorrhage. Lab Med 1978;9:22–5.
13. Zipursky A, Pollock J, Neelands P, Chown B, Israels LG. The transplacental passage of fetal red blood cells and the pathogenesis of Rh immunisation during pregnancy. Lancet 1963;6:489–94.
14. Stolte L, Kessel H, van Seelen J, Eskes T, Wagatsuma T. Failure to produce the syndrome of amniotic fluid embolism by infusion of amniotic fluid and meconium into monkeys. Am J Obstet Gynecol 1967;98:694–7.
15. Kuhlman K, Hidvegi D, Tamura R, Depp R. Is amniotic fluid material in the central circulation of peripartum patients pathologic? Am J Perinatol 1985;2:295–9.
16. Clark SL, Pavlova Z, Greenspoon J, Horenstein J, Phelan J. Squamous cells in the maternal pulmonary circulation. Am J Obstet Gynecol 1986;154:104–6.
17. Lee W, Ginsburg K, Cotton DB, Kaufman RH. Squamous and trophoblastic cells in the maternal pulmonary circulation identified by invasive hemodynamic monitoring during the peripartum period. Am J Obstet Gynecol 1986;155:999–1001.
18. Lockwood CJ, Bach R, Guha A, Zhou X, Miller WA, Nemerson Y. Amniotic fluid contains tissue factor, a potent initiator of coagulation. Am J Obstet Gynecol 1991;165:1335–41.
19. Clark SL. Arachidonic acid metabolites and the pathophysiology of amniotic fluid embolism. Semin Reprod Endocrinol 1985;3:253–7.
20. Azegami M, Mori N. Amniotic fluid embolism and leukotrienes. Am J Obstet Gynecol 1986;155:1119–24.
21. Clark SL, Hankins GD, Dudley DA, Dildy GA, Porter TF. Amniotic fluid embolism: Analysis of the national registry. Am J Obstet Gynecol 1995;172:1158–69.
22. Clark SL. New concepts of amniotic fluid embolism: A review. Obstet Gynecol Surv 1990;45:360–8.
23. Miller GV, Ramsden CW, Primrose JN. Autologous transfusion: An alternative to transfusion with banked blood during surgery for cancer. Br J Surg 1991;78:713–5.
24. Torre GC, Ferrari M, Favre A, Razzetta F, Borgonovo G. A new technique for intraoperative blood recovery in the cancer patient. Eur J Surg Oncol 1994;20:565–70.