| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Departments of *Obstetrics and Gynecology, Brigham & Women's Hospital, Boston, MA; and Obstetrics and Gynecology and Laboratory Medicine, University of Washington, Seattle, Washington.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Amniotic fluid was collected by transabdominal amniocentesis from women at 20–41 weeks of gestation and contaminated with blood, meconium, and vaginal secretions to concentrations of 0.5, 1, 2, 5, and 10%. An additional 20% concentration was performed with meconium and vaginal secretions. Fluorescence polarization was determined by a TDx Analyzer with the NBD-PC fluorescent probe. Results were compared for each contaminant by concentration level using repeated-measures analysis of variance.
RESULTS: Forty-eight samples from women at a mean gestational age of 35 weeks (range 20–41.5 weeks) were evaluated. Before contamination, 16 (33%) samples had fluorescence polarization values greater than 290 mPol (immature), 10 (21%) were 260- 289 mPol (transitional), and 22 (46%) were less than 260 mPol (mature). Contamination with blood significantly altered fluorescence polarization values in AF samples with baseline values in the immature and mature categories such that values trended toward the transitional range. Contamination of baseline immature samples with vaginal secretions at 20% contamination level resulted in more mature fluorescence polarization values. Contamination with meconium more than 2% in the baseline immature category or more than 20% in the baseline transitional category also resulted in significantly more mature fluorescence polarization values.
CONCLUSION: Amniotic fluid contamination with blood can result in more transitional range fluorescence polarization values, whereas contamination with meconium and vaginal secretions can result in more mature fluorescence polarization values.
LEVEL OF EVIDENCE: II-2
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Blood was obtained from umbilical cords after delivery on the labor and delivery suite. First-pass meconium was obtained from infants in the special care nursery, and vaginal secretions were obtained from women undergoing routine speculum examinations on the labor and delivery suite or maternity clinic. The blood, meconium, and vaginal secretions used to contaminate the AF were not necessarily from the same patient.
For performing standardized AF contamination, the density of meconium and vaginal secretions was assumed to be equivalent to water. Based on this assumption, a 20% contamination was performed by adding 0.240 mg of meconium or vaginal secretions to individual test tubes for a final volume of 1.2 mL of amniotic fluid. These contaminated specimens were vortexed to allow complete suspension of the contaminants. These 20% contamination stocks and remaining uncontaminated AF samples were then centrifuged for 20 minutes at 500 rpm in a Sorval RT-6000 (Sorval Norwalk, Norwalk, CT) centrifuge at 4°C per standard fluorescence polarization procedure protocol.6 The supernatants of all spun material were removed and appropriately labeled as pure AF, 20% meconium stock, or 20% vaginal secretions stock. A minimum volume of 0.33 mL per cuvette was required to process each specimen; 0.33 mL of uncontaminated AF was mixed with 0.6 mL of TDx Buffer (Abbott Laboratories, Abbott Park, IL). A similar procedure was performed with the 20% contaminated stock specimens (ie, 0.33 mL of 20% stock plus 0.6 mL TDx Buffer). The subsequent meconium and vaginal secretion stock specimens were diluted with the pure AF to obtain final concentrations of 0.5, 1, 2, 5, and 10%. For example, to constitute a 10% contaminant, 0.165 mL of the 20% stock solution was mixed with 0.165 mL of pure AF before adding the standard 0.6 mL TDx Buffer. Collected blood was centrifuged under the same protocol. The serum supernatant was removed, labeled serum, and then used to contaminate the pure AF for final concentrations of 0.5, 1, 2, 5, and 10%.
The fluorescence polarization values were determined on the uncontaminated and contaminated samples as previously outlined by Tait et al7 with the NBD-PC fluorophore (Avanti Polar Lipids, Birmingham, AL). Fluorescence polarization values were categorized as mature (less than 260 mPol), transitional (260–289 mPol), or immature (more than 290 mPol) based on prior studies which stratified results based on risk of development of respiratory distress syndrome (RDS).7,12,13
Fluorescence polarization results from uncontaminated AF were compared with fluorescence polarization results from the same fluid after the addition of each contaminant at the various concentration levels using repeated-measures analysis of variance. Separate models for each type of contaminant were generated using the general linear modeling function of SPSS X (SPSS Inc., Chicago, IL) with contaminant concentration as within subject and baseline lung maturity category as between subject factors.
A power calculation determined that AF from 8 pregnancies were needed to detect a 5 mPol difference in fluorescence polarization values between the uncontaminated and contaminated AF with power of 90% (alpha of 0.05), assuming the fluorescence polarization standard deviation of ± 3 mPol.11 To determine whether the effect of AF contamination varied by baseline AF fluorescence polarization values, we planned to enroll 8 women in each of the fluorescence polarization categories (mature, immature, transitional). Samples with baseline transitional fluorescence polarization values were less frequent; thus more women in other categories were enrolled to meet the sample size estimates needed in the transitional range.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There was a significant linear association between blood contamination concentration and fluorescence polarization values (P < .001, Fig. 1). However, the direction of the effect of blood contamination varied significantly by the baseline fluorescence polarization lung maturity category (P < .001, Fig. 2A). In AF samples that had a mature baseline lung maturity category, blood contamination resulted in more immature fluorescence polarization values. Conversely, in AF samples that were immature at baseline, contamination led to more mature fluorescence polarization values. The magnitude of these mean differences was statistically significant for most concentrations of blood contamination in the AF samples that were immature or mature at baseline (Fig. 2A). Sixteen percent of the mature samples became transitional, and 9% of the immature samples became transitional (Table 1).
|
|
|
There was a significant linear association between AF fluorescence polarization values and increasing meconium contamination (Fig. 1, P < .001). Furthermore, the effect of meconium contamination on AF fluorescence polarization values varied by the baseline fluorescence polarization lung maturity category (Fig. 2B, P < .001). For immature samples, meconium contamination to more than 2% concentration resulted in significantly more mature fluorescence polarization values; for transitional samples, only at 20% contamination did fluorescence polarization values become significantly more mature; and for mature samples, no significant effect of meconium contamination occurred (Fig. 2B). These changes in the baseline immature and transitional AF samples were large enough that crossover into the mature range occurred for 44% of the immature samples and 31% of the transitional samples (Table 1).
Amniotic fluid fluorescence polarization values were linearly associated with vaginal secretions contamination level (Fig. 1, P = .001). The effect of vaginal secretions contamination on fluorescence polarization results did vary significantly by baseline lung maturity categories (Fig. 2C, P < .001). Repeated-measures pair-wise comparisons of mean differences between baseline and contaminated fluorescence polarization results were significant at 20% concentration for the baseline immature AF samples and at less than 5% for the baseline mature AF samples (Fig. 2C). Eight percent of the immature samples and 7% of the transitional AF samples became mature with vaginal secretions contamination (Table 1). However, no mature AF samples became immature and only 4% became transitional.
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The effect of blood contamination on fluorescence polarization results is consistent with the known fluorescence polarization value of pure blood at approximately 255 mPol. With increasing blood contamination, the AF fluorescence polarization value trended toward blood's inherent value, which is in the transitional range. Clinically identifiable contamination with blood was first evident at the 2% contamination level. Unfortunately, the effect of blood was significant at all contamination levels, including the lowest concentration tested (0.5%), when contamination would not be clinically evident. For those samples contaminated at 0.5 or 1%, 6/84 (7%) crossed over or changed lung maturity categories (data not shown).
Meconium contamination also affected the fluorescence polarization results differently depending on the baseline lung maturity. There is a significant phospholipid composition in the sloughed cells and mucous elements of meconium. It is likely that the effect of this increasing phospholipid composition is more noticeable in the immature, or phospholipid-poor samples, than in the transitional or mature samples. Clinically, lightly stained meconium amniotic fluid resembles our 0.5% and 1% contaminations, whereas thick particulate meconium-stained amniotic fluid resembles our 2% contamination. Unlike with blood contamination, crossover from the immature and transitional categories into the mature category did occur, making the possibility of false positive mature results a clinical concern (Table 1).
The effect of vaginal secretions was more complex in that low amounts (0.5–2%) of contamination caused falsely higher (more immature) fluorescence polarization values, but 20% contamination resulted in lower (more mature) fluorescence polarization values. From previous work done with vaginal secretions and other lung maturity assays, it has been established that vaginal flora add a phospholipid–rich component to contaminated amniotic fluid samples.12,13 Clinically, a 20% contamination could easily be attained, especially if rupture of membranes occurred well before collection of the vaginal pool sample. However, unlike with blood or meconium, contamination with vaginal secretions was not visible at any concentration. Again, as crossover into the mature category did occur, the reliability of a mature result would have to be questioned if a vaginal pool specimen was analyzed.
The effect of contaminants on fluorescence polarization has not been widely studied. Tait suggested in his initial work6 that blood and meconium would interfere with analysis of AF specimens; however, the significance of the effect and amount of contamination were not defined. In later studies Tait further characterized the effect of blood contamination, suggesting an effect modification based on level of baseline fluorescence polarization values. This conclusion was drawn after visibly contaminated samples were processed, thus the amount of contamination necessary to produce this effect was not identified.7 Carlan et al14 supported an effect modification of blood contamination on fluorescence polarization values suggested by Tait, with increasing amounts of blood contamination causing mature values to become more immature and immature values to become more mature. Our work confirms the effect modification suggested by Tait and Carlan, but is in contrast to those findings of Carlan in that none of our mature specimens crossed over into the immature category. However, in both our study and Carlan's study, no transitional or immature specimen crossed over into the mature category after contamination with blood. Thus the clinical interpretation of a mature result in both studies remains reliable.
Our study suggests that the presence of vaginal secretions at significant concentrations might effect fluorescence polarization values obtained from vaginal pool specimens. These findings challenge the studies by Edwards15 and Runowicz16 that assessed the reliability of fluorescence polarization generated from vaginal pool collections. In contrast to our study, these studies concluded that vaginal pool fluorescence polarization values could serve as a reliable predictor of fetal lung maturity.15,16 However, vaginal secretions are extremely heterogeneous in their composition and without a reliable test to ascertain the presence or concentration of phospholipids attributable to vaginal flora in the vaginal pool sample, our study suggests that the results from a vaginal pool sample need to be interpreted with caution.
Although contamination with meconium and vaginal secretions at high enough concentrations led to falsely mature fluorescence polarization results, our study was an in vitro study and did not evaluate the clinical endpoint of RDS in neonates with falsely mature postcontamination results. Tait previously showed that the sensitivity of RDS was approximately the same at all gestational ages when using a cutoff of 260 mPol for fetal lung maturity; however, there was a trend toward lower specificity at gestational ages less than 36 weeks. He also showed that there was an association between fluorescence polarization and gestational age.9 Although we cannot comment on the rate of RDS using the prior work of Tait, our data may aid the clinician in interpreting fluorescence polarization results.
Although in vitro contamination was performed in our study, it is unlikely that amniotic fluid contaminated in vivo would behave differently. In vitro contamination allowed for analysis at various levels on contamination that would be clinically relevant. We did not perform an evaluation for baseline contamination on collected AF samples other than gross evaluation. However, each sample was analyzed and a baseline fluorescence polarization value was generated. All subsequent concentrations of contamination were compared with that baseline value. This likely negated the effect of any unseen contamination.
Previous studies have questioned the accuracy and clinical reliability of several other fetal lung maturity assays in the setting of contaminated AF specimens.2–5 Our study further defines and demonstrates the significant effects of blood, meconium, and vaginal secretions contamination on AF fluorescence polarization results. Amniotic fluid samples contaminated with blood significantly differed from baseline fluorescence polarization values, such that the level trended toward the transitional zone. This becomes clinically important because immature specimens contaminated with blood may yield lower (more mature) fluorescence polarization values. However, when a mature fluorescence polarization result is obtained in the presence of blood contamination, management decisions based on this information can be made safely. In contrast to blood contamination, the ability of high concentrations of meconium and vaginal secretions to change immature to mature fluorescence polarization values raises the possibility of falsely reassuring test results. The effect of meconium and vaginal secretions contamination on AF fluorescence polarization results should be considered when making clinical decisions based on lung maturity testing.
|
Footnotes |
|---|
Received April 1, 2004. Received in revised form July 27, 2004. Accepted September 1, 2004.
doi:10.1097/01.AOG.0000146637.96281.a5
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Cohen B, Ludmir J, Towers C, Alvarez JG. Performance of the dipalmitoyl phosphatidylcholine test in predicting respiratory distress syndrome in contaminated samples of amniotic fluid. Obstet Gynecol 1997;89:719–22.[Abstract]
3. Buhi WC, Spellacy WN. Effects of blood or meconium on the determination of the amniotic fluid lecithin/sphingomyelin ratio. Am J Obstet Gynecol 1975;121:321–3.[Medline]
4. Keniston RC, Noland GL, Pernoll ML. The effect of blood, meconium, and temperature on the rapid surfactant test. Obstet Gynecol 1976;48:442–6.
5. Dombroski RA, Mackenna J, Brame RG. Comparison of amniotic fluid lung maturity profiles in paired vaginal and amniocentesis specimens. Am J Obstet Gynecol 1981;140:461–4.[Medline]
6. Tait JF, Franklin AC, Simpson JB, Ashwood ER. Improved fluorescence polarization assay for use in evaluating fetal lung maturity. I. Development of the assay procedure. Clin Chem 1986;32:248–54.
7. Tait JF, Foerder AC, Ashwood ER, Benedetti TJ. Prospective clinical evaluation of an improved fluorescence polarization assay for predicting fetal lung maturity. Clin Chem 1987;33:554–8.
8. Chen C, Roby PV, Weiss NS, Wilson JA, Benedetti TJ, Tait JF. Clinical evaluation of the NBD-PC fluorescence polarization assay for prediction of fetal lung maturity. Obstet Gynecol 1992;80:688–92.
9. Ron M, Cohen T, Yaffe H, Beyth Y. The clinical significance of blood contaminated midtrimester amniocentesis. Acta Obstet Gynecol Scand 1982;61:43–5.[Medline]
10. Tabsh KM, Brinkman CR, Bashore R. Effect of meconium contamination on amniotic fluid lecithin/sphingomyelin ratio. Obstet Gynecol 1981;58:605–8.
11. Foerder CA, Tait JF, Franklin RW, Ashwood ER. Improved fluorescence polarization assay for use in evaluating fetal lung maturity. II Analytical evaluation and comparison with lecithin/sphingomyelin ratio. Clin Chem 1986;32:255–9.
12. Lambers DS, Brady K, Leist PA, Setser C, Helmchen R. Ability of normal vaginal flora to produce detectable phosphatidylglycerol in amniotic fluid. Obstet Gynecol 1995;85:651–5.[Abstract]
13. Schumaker RE, Parisi VM, Steady HM, Tsao FHC. Bacteria causing false positive test for phosphatidylglycerol in amniotic fluid. Am J Obstet Gynecol 1985;151:1067–8.[Medline]
14. Carlan SJ, Gearity D, O'Brien WF. The effect of maternal blood contamination on the TDx-FLM II assay. Am J Perinatol 1997;14:491–4.[Medline]
15. Edwards RK, Duff P, Ross KC. Amniotic fluid indices of fetal pulmonary maturity with preterm premature rupture of membranes. Obstet Gynecol 2000;96:102–5.
16. Runowicz CD, Beers PC. The assessment of fetal pulmonary maturity by fluorescence polarization (FP-Value) of pooled vagina amniotic fluid associated with ruptured membranes. Int J Gynaecol Obstet 1985;23:481–8.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
