October 2009 Clinical Laboratory News: New Paradigms in Fetal Lung Maturity Testing

CLN Banner Logo

October 2009: Volume 35, Number 10


New Paradigms in Fetal Lung Maturity Testing
What Are the Options for Labs?
By Genna Rollins

Since the early 1970s, fetal lung maturity (FLM) testing has played an important role in identifying neonates at risk of respiratory distress syndrome. The various assays, developed in the 1970s and 1980s, maintained their popularity into the 1990s. However, FLM testing has been on the wane in recent years as physicians have found the results less informative to their decision making about when to time delivery. Even so, information about FLM still has a place in certain situations, and the potential loss of one of the assays will require laboratorians and clinicians to collaborate about available options and develop continuity plans.

“Labs can really play an important role in this transition time,” said David Grenache, PhD, DABCC, assistant professor of pathology at University of Utah School of Medicine and medical director of special chemistry at ARUP Laboratories in Salt Lake City. “It’s a good opportunity for them to work in tandem with doctors, not only to adapt to the changing environment, but also to have discussions about FLM test utilization and how and why it’s being used.”

A Look at Test Utilization

Abbott is one of the primary manufacturers of FLM tests and its TDx FLMII assay is one of the more popular testing options. However, the potential discontinuation of this platform has prompted researchers to review the utilization of and indications for FLM testing. Although the company has not made a formal announcement about its TDx FLM II test for surfactant/albumin ratio (S/A ratio) by fluorescent polarization, an Abbott spokesperson indicated that it is decommissioning certain legacy systems. “Abbott is retiring many of the assays that run on the company’s older diagnostic instrument platforms —IMx, TDx and TDxFLx. These platforms are more than 20 years old and use older equipment and technology that is difficult to support and service as parts and raw materials are scarce. The decision to retire these older products will enable Abbott to invest in state-of-the art technologies,” said Ann Fahey-Widman, director of global diagnostic public affairs. “We have not made an announcement regarding the Abbott FLM assay on TDx as we're currently evaluating this and other products internally, [but] what I can tell you is that we are committed to communicating any additional assay retirements to our customers at least 12 months in advance and will continue to help customers through any transitions.”

Should the S/A ratio platform become unavailable, it appears that many labs will be affected, according to a survey of AACC members conducted this spring by Grenache and Ann Gronowski, PhD, DABCC, FACB, associate professor of pathology, immunology, and obstetrics and gynecology at Washington University School of Medicine in St. Louis. Their results indicated that the S/A ratio was by far the most frequently offered test, followed by phosphatidylglycerol (PG) by agglutination, lecithin/sphingomyelin ratio, lamellar body count, PG by thin layer chromatography, and (rarely) foam stability. CAP survey data for 2007–2008 indicated roughly the same utilization pattern, except it did not have data on lamellar body count and foam stability tests.

Changing Volumes, Practice Patterns

The potential retirement of the TDx platform may pose a dilemma for some labs, but perhaps not as much as it might have in the past. Grenache’s and Gronowski’s survey showed a 10-year trend of decreasing FLM test utilization, and in a separate analysis, University of Mississippi Medical Center researchers reported an overall decline in FLM test volume in their lab from 1,090 in 1994 to 389 in 2004 (J Perinatol 2008;28:20–23). The authors also surveyed other maternal-fetal medicine practices and found that there was a net decrease in test utilization for three of four FLM tests, ranging from 33% for foam stability to 56% for PG. The exception was Abbott’s TDx FLM II, for which respondents reported a net increase of 29%. This spring, Grenache and Gronowski also surveyed fellows of the Society for Maternal-Fetal Medicine. Although the response rate was quite modest (<5%), the results indicated that one-quarter of participants were ordering the tests less often.

The decline in FLM testing stems from several factors, according to Thomas McElrath, MD, PhD, assistant professor of obstetrics and gynecology at Harvard Medical School and Brigham and Women’s Hospital in Boston. “Multiple trends have each contributed to the former testing paradigm that had been dominant,” he observed. “We were hoping at one point that fetal lung maturity was a marker for overall maturity. We now realize that it’s less that and more an issue specific to the lungs. Fetal lung maturity is still an important issue, but we were probably crediting it with providing us too much information at an earlier time and today have put it in a more proper context.”

Trends McElrath cited as contributing to the decline included studies indicating that “the window of utility for FLM testing is more narrow than most of us expected it would be,” and medical practices that have enabled safer delivery of preterm babies. “NICU technology has improved dramatically and a neonatologist can now support a baby at 34 weeks much more safely than he or she could have done two decades ago,” he said. In addition, research has illuminated the overall dysmaturity of near-term premature babies. “We realize that there’s more to maturity than just lungs. The use of surfactant has increased so dramatically, and it’s been a revolution in treating these kids. But there are so many other problems that aren’t being tested for like necrotizing enterocolitis, intravascular hemorrhages, and behavior problems, that FLM is still informative but it’s perhaps less useful than we had thought it was.”

FLM testing not being needed for patient care was by far the most common reason cited for the decline in test utilization reported in the Journal of Perinatology study. “Clinicians are going more on clinical judgment, and they’re not doing amniocentesis as often is the bottom line,” observed senior author John Morrison, MD, professor of obstetrics and gynecology and pediatrics at University of Mississippi Medical Center in Jackson.

Updated Guidelines Clarify Indications

Updated American College of Obstetricians and Gynecologists practice guidelines note that while FLM status can assist clinicians in determining when delivery should occur, FLM testing “should not be performed, and is contraindicated, when delivery is mandated for fetal or maternal indications” (ACOG Practice Bulletin, 97, September 2008). Thomas Peng, MD, a maternal fetal medicine specialist at Bon Secours Richmond Health System in Richmond, Va. agrees. “The way I put it is, if your strategy for management is going to be to deliver regardless of the test results, then there’s no point in doing the test. You use the test when you have the option to wait to deliver. For example, a mom with diabetes with a macrosomic baby where you’re concerned about fetal loss, you’d do an amniocentesis at 37 weeks, and if the test indicated low risk of neonatal respiratory distress, you could proceed and deliver the baby,” he said. Peng served as consultant to ACOG in updating the practice bulletin.

One of the most common circumstances in which McElrath uses FLM testing is when the mother has ruptured membranes and her pregnancy is at least 32 weeks’ gestation. “It’s been demonstrated in the medical literature that if a baby has mature lungs, after 32 weeks the overall risk of morbidity to the baby is lower to go ahead and deliver than to continue to manage it in utero,” he explained. “That scenario was a small percentage of the FLM testing we used to do but has now grown to approximately 50 percent of all of the FLM tests we order.”

The ACOG guidelines call for FLM testing in scheduled deliveries at less than 39 weeks unless fetal maturity can be deduced from other tests performed earlier in the pregnancy, including ultrasound, Doppler ultrasonography to detect fetal heart tones, and positive serum or urine HCG pregnancy test results. Likewise, NACB’s Laboratory Management Practice Guideline on Maternal-Fetal Risk Assessment and Reference Values in Pregnancy states that “FLM testing is not indicated in normal pregnancies if the gestational age is accurately known to be at least 36 weeks.”

Another trend in FLM testing is that repeat testing is less common now. The updated ACOG practice bulletin notes a lack of consensus about the benefits of repeat testing when the first results are either immature or indeterminate, a change from ACOG’s prior recommendation in 1996. “That change was made for two practical reasons. First, most practitioners in general don’t do sequential testing,” Peng explained. “Also, one study evaluated the rate of increase in TDx-FLM II results and found it to be a constant of approximately 15 milligrams per gram per week. So depending on the initial value, one could extrapolate to determine if by waiting another week on average the FLM test will be mature.”

FLM Test Options

For labs that offer the S/A ratio test, the key is to begin reviewing FLM testing options now, according to Gronowski. “Without a doubt, labs need to start thinking about this and working together now. Most of us don’t have a huge volume of this testing, and over time it has declined, so those of us who need to validate a test will need to help each other and share samples,” she said. Including clinicians early-on is a must, advised Peng. “If you spring a new test on them at the last minute, they will have no faith in the particular test that you’re offering because they will never have used it before. In order to foster their trust, you have to get them together and develop a plan,” he noted.
Accuracy and timeliness are the two most important factors in physician preference for FLM tests, according to Peng. “The test has to be reproducible and reasonably accurate in terms of predicting respiratory distress syndrome,” he said. “We’re also looking for a short turnaround time.” Most respondents to Grenache and Gronowski’s survey of Society of Maternal-Fetal Medicine fellows indicated they would like FLM results within 12 hours. NACB LMPG recommends that FLM testing be available in hospitals that deliver infants, with results available routinely once per day and in urgent settings within an hour of collecting the specimen.

According to NACB LMPG, all FLM tests have a high—at least 95%—sensitivity, but mediocre specificity. Taking the prevalence of respiratory distress syndrome into account, the tests have a high negative predictive value, meaning the probability of the neonate having respiratory distress syndrome is low when results indicate the lungs are mature. The positive predictive value, or likelihood of respiratory distress syndrome when the FLM test indicates the lungs are immature, ranges from 13% to 61%, according to ACOG. “Very few babies with mature results go on to develop respiratory distress syndrome,” Grenache explained. “But the tests’ immature, or positive, predictive value is not good at all. There are lots of babies born who had an immature result and never develop respiratory distress syndrome.” Meconium and blood contamination can affect the performance of the tests to varying degrees, depending on the assay.

In their survey of Society of Fetal-Maternal Medicine fellows, Grenache and Gronowski asked respondents to give their impression of the clinical accuracy of the various FLM tests. Respondents rated S/A ratio and PG by TLC equally high, with >90% reporting the tests to be excellent or good. Nearly 90% also ranked the L/S ratio as either excellent or good, whereas about two-thirds rated PG by agglutination and lamellar body count as either good or excellent. Foam stability was a distant sixth. Grenache and Gronowski have also published a review article about FLM testing (Clin Biochem 2006;39:1–10); see table, below, for a summary of the various tests.

Table 1
Fetal Lung Maturity Tests

Test

Platform

Pros

Cons

Mature Predictive Value (%)

Immature
Predictive Value (%)

Foam Stability Index

Lab-developed

Rapid

Blood and meconium affect results

Qualitative

Technically challenging

≈95

≈15

Lamellar Body Count

Lab-developed

Rapid, simple

Instrumentation widely available

Blood and meconium affect results

QC/PT not available until 2010

97–98

29–35

L/S Ratio

TLC (lab-developed or commercially available)

Quantitative

QC/PT available

Blood and meconium affect results

Technically challenging

Time consuming

86–100

19–83

PG

AmnioStat-FLM (Agglutination)

Rapid, simple

QC/PT available

Blood and meconium do not affect results

PG last surfactant to rise

Subjective test interpretation

Single vendor

97–100

9–50

PG

TLC (lab-developed or commercially available)

QC/PT available

Quantitative

Blood and meconium do not affect results

PG last surfactant to rise

Technically challenging

95–100

23–53

S/A Ratio

TDx

Rapid, precise

Quantitative

Standardized QC/PT available

Blood and meconium affect results

Single vendor

99–100

14–36

Legend: S/A Ratio = surfactant/albumin ratio; PG = phosphatidylglycerol; L/S Ratio = lecithin/sphingomyelin ratio

All but three FLM tests are lab-developed, so another consideration for labs in choosing a test is how labor-intensive and technique-dependent the methods are. L/S ratio by TLC can be performed alone or in combination with PG by TLC. Both the tests “are old stand-bys but they also require an experienced person to know how to prepare and read the gels,” Peng observed. A lab he worked with stopped offering the test “because it lost the one person who had expertise in doing the gels, and when he was gone, that was it.” In general, the test is time-consuming and many labs do not have volumes recommended to maintain proficiency with the assay, according to Gronowski. A commercial version of the L/S ratio, Fetal-Tek 200, is available from Helena Laboratories.

Another lab-developed test that a few labs offer is the foam stability index test. It has the advantage of rapid results, but is only semi-quantitative and can pose technical challenges owing to the use of hygroscopic ethanol. The assay also is highly operator-dependent. Physicians at a major teaching hospital informed Peng that they preferred the assay but that there “was only one person in the lab they trusted to do the test.” Relying on a test that is so technologist-dependent is a recipe for disaster, according to Gronowski. “With our automated tests we try not to have a single instrument that does a test, so why would a lab have a single employee that can do a test?” she asked. By the same token, labs can be equally vulnerable if they offer only one of the three commercial FLM tests.

Irvine Scientific’s AmnioStat-FLM PG by agglutination test gives rapid results and is easy to perform. However PG is the last surfactant to increase, at about 36 weeks, and makes up a relatively modest proportion of overall pulmonary surfactant, so a mature result is closely tied to timing of the test. In addition, the manufacturer twice has experienced production challenges, once during much of 2004, and for a period of about three months this year. The problems were unrelated and have been fully resolved, according to Janie Kim, senior product manager.

The Case for Lamellar Body Count

Aware of the limitations of the various FLM tests, as well as the possibility that Abbott might one day retire the TDX-FLM II platform, Grenache and Gronowski began exploring other options some time ago. Their research has lead both to conclude that the lamellar body count assay is the best available alternative. Lamellar bodies are hydrophobic structures made of surfactants and synthesized in type II pneumocytes.

Although Gronowski is bullish on the test now, that wasn’t always the case. “I had an FDA-approved test that had external proficiency testing, standardized methods and quality control, so why would I want to substitute that for something I was going to have to validate in-house, that I didn’t have external proficiency testing for and there wasn’t a lot of data on?” she recalled. “However, overtime, by compiling data I’ve convinced myself, and hopefully others, that clinically it performs as well as the S/A ratio.”

The test, which involves counting lamellar bodies using standard hematology counters, is quick and easy to perform, and can be done with small sample volumes. However, it is a lab-developed test that will require internal validation with hard-to-come-by amniotic fluid samples. In addition, both laboratorians and clinicians are not as familiar with it as other FLM tests, according to Grenache and Gronowski’s surveys.


Lamellar body count performed on hematology counters
uses an amniotic fluid sample. CAP will offer proficiency testing
for this analyte next year.
Photo courtesy of Ann Gronowski, PhD

The two main hurdles to implementing the test are overcoming the perception that it is in some way injurious to cell counters and accumulating enough amniotic fluid samples to validate the test. “Lamellar body count is quantitative and easy to do, and I’d be willing to bet that every lab in the country has a cell counter. But people are afraid of running samples through cell counters that are not whole blood,” Grenache observed. “However, amniotic fluid is not going to harm the cell counter.”

It may take a while for labs to accumulate sufficient amniotic fluid samples for validation, but the good news is “the fluid is very stable, it can be refrigerated and you don’t have to rush it off anywhere or do any special handling, so you can just hold it until you have enough samples,” explained Gronowski. She recommends that labs interested in implementing the test pool samples to support their validation requirements. CAP has announced that it will begin offering proficiency testing for the assay next year.

Regardless of which FLM test(s) labs offer in the future, it is imperative that they start the review process now, according to Gronowski. “Labs have to open a dialogue with their Ob-Gyns. Whether their decision is to send FLM tests out or develop them internally, honestly, it could take a while to get everything worked out and ready.”

Drs. Grenache and Gronowski have received consultant fees from Abbott.

Page Access: