Category Archives: Analytical Sensitivity

Analytical Sensitivity

Can a personalized approach improve IVF success rates?


Test Tube Baby

This post was written by Robert D. Nerenz, PhD, an assistant professor of pathology and laboratory medicine at the University of Kentucky, in Lexington.

In the United States, an estimated one in seven couples experience infertility and for many of these couples, in vitro fertilization (IVF) represents their best chance of achieving pregnancy. However, IVF cycles constitute a significant expense (approximately $12,500 per cycle), disrupt patients’ daily lives and only result in a healthy, live birth 30% of the time! Furthermore, the majority of IVF cycles performed in the United States involve the transfer of multiple embryos. This is of particular concern because multiple embryo transfer carries a finite risk of a multiple gestation pregnancy. Bringing multiple infants to term is associated with an increased risk of poor fetal and maternal outcomes including decreased birth weight, increased rate of fetal death, preeclampsia, gestational diabetes and preterm labor. Clearly, there is a significant need to improve IVF success rates while also minimizing the likelihood of multiple gestation pregnancies.

One strategy that may accomplish both of these goals is to perform “single embryo transfer” by implanting one embryo that has a high likelihood of producing pregnancy and, ultimately, a live birth. This is the focus of an upcoming symposium at the AACC meeting to be held July 29th at 10:30 am in Atlanta, Georgia. Fertility clinics around the world currently attempt to do this by observing embryos under a microscope and choosing the best embryo on the basis of its physical appearance. Unfortunately, this approach does not provide any information about the embryo’s genetic status. This is an important limitation because aneuploidy (the gain or loss of a chromosome) is the most common cause of pregnancy loss. It is also estimated to occur in ≥10% of clinical pregnancies and becomes more frequent with increasing maternal age.

To ensure that aneuploid embryos are not selected for transfer, several research groups have developed methods collectively known as comprehensive chromosome screening (CCS). CCS involves culturing embryos for 5-6 days, removing a few cells from the trophectoderm (the outer cell layer that develops into the placenta), isolating the DNA from those cells and assessing the copy number of each chromosome using techniques such as quantitative PCR, comparative genomic hybridization, or single nucleotide polymorphism arrays. Following determination of the embryos’ genetic status, only embryos with the normal number of chromosomes are chosen for transfer. In multiple prospective, randomized controlled trials described here and here, CCS has been shown to increase the pregnancy rate and decrease the frequency of multiple gestation pregnancies. As a result, CCS is beginning to make the transition from the research setting to use with patients.

The ability to transfer only euploid embryos represents the most promising application of novel technologies to IVF but ongoing research is focused on other ways to improve the IVF success rate. Many different groups are analyzing the culture medium that embryos are grown in prior to implantation. It is hoped that this will provide information about the embryos’ metabolic health and might help identify which embryos are most likely to result in pregnancy and live birth. Other groups are evaluating endometrial gene expression profiles to assess endometrial receptivity and ultimately determine the best time to perform embryo transfer. While both of these approaches have technical limitations and are not quite ready for primetime, they have the potential to greatly improve our current standard of care and may be ready for clinical use in the near future.

Detecting hCG in urine: how low is low enough?


A recent post on this blog described the inability of qualitative point-of-care (POC) hCG tests to detect hCG when it was present in urine or serum at a concentration that should, according to the test manufacturer, always be detected. The inability of these devices to detect hCG is a serious concern.

A false-negative result from a home pregnancy test can be initially disappointing if a pregnancy is desired or a temporary relief if it is not. By contrast, a false-negative result in the health care setting can result in serious harm to the fetus if a patient who is assumed to not be preganant undergo interventions that are potentially harmul to the pregnancy.

A recent case report has been published that, like other reports, emphasizes the limitations of qualitative urine hCG testing. The case describes a young woman who required radioactive iodine therapy for Grave's disease. Importantly, this young woman was also recently pregnant. That fact would likely not have been discovered had the physician relied on a qualitative urine hCG test. Fortunately, the laboratory had performed a quantiative urine hCG test (note that quantitative hCG tests using urine may be performed by the lab but the results are reported as qualitative (e.g. yes or no) and not the actual hCG concentration) which was interpreted as "positive" because the measured hCG concentration was greater than the lab's cutoff of less than or equal to 5 IU/L (it was 12 IU/L). A serum hCG test performed the same day produced a result 15 IU/L (not pregnant=less than or equal to 5 IU/L). Two days later a repeat serum hCG test produced a result of 147 IU/L confirming that she was in the very early stages of pregnancy. 

As has been noted in this blog in the past (here and here), urine hCG testing is commonly performed in the health care setting because it is convenient. However, the problems with urine hCG tests are so numerous (see here and here) that urine hCG testing should not be relied upon to determine a patient's pregnancy status.

The authors of the case described above correctly point out that the detection thresholds of most qualitative urine hCG tests are stated to be 20–50 IU/L (recent evidence suggest these cutoffs are not always accurate). Further, they call for more sensitive qualitative urine hCG tests in order to decrease the number of false-negative hCG results in the health care setting and suggest that a detection threshold of 5 IU/L (which is the same threshold used for interpreting quantitative serum hCG tests) should be used. Interestingly, this conclusion is similar to the one my group suggested in regards to qualitative serum hCG testing.

I am in complete agreement that when it comes to the detection of early pregnancy, hCG tests that are capable of accurately detecting and/or measuring hCG are required. Currently, this means that serum hCG tests should be used exclusively, in the health care setting, for this purpose. To rely on less sensitive tests and less accurate urine hCG tests is a disservice to our patients.

Wanted! A sensitive qualitative hCG test.


Today’s post is by a guest author, Dina N. Greene, Ph.D. Dr. Greene is a Scientific Director at Northern California Kaiser Permanente Regional Laboratories in Berkeley, CA. She discovered that qualitative hCG tests may not be as analytically sensitive as we all have come to believe and she shares her observations here. A report of her work has been published in Clinica Chimica Acta.
Neg pregnancy test
The assessment of very early pregnancy (from conception until about two weeks following the
expected menses) is dependent on the detection of hCG in serum or urine. In health care settings a urine sample is often the specimen of choice because it is convenient and usually easy to obtain.

When urine samples are tested for hCG they are most frequently tested using qualitative (yes/no) point-of-care (POC) devices. This type of testing is attractive because it is performed close to the patient and the test results can be obtained within minutes. In general, when challenged with urine or serum containing hCG these devices work well. However, what was not known was how sensitive these devices are for detecting very early pregnancy. That is, could pregnancy be ruled out if a qualitative POC test was negative?

To answer that question we completed a study that took a systematic approach to this question by testing urine and serum specimens collected from patients that spanned a wide range of hCG concentrations with two commonly used POC devices.

While many concentrations of hCG were represented in these samples, we purposefully skewed the specimens so that a large percentage (~30%) had concentrations of hCG expected to be seen only in very early pregnancy. The results were surprising.

We found was that the devices did not always detect hCG at the lowest detectable concentration claimed by the manufacturer (20 IU/L for urine and 10 IU/L for serum). In fact, we had many false-negative results when the urine concentration of hCG was as high as 200 IU/L or the serum hCG concentration was as high as 50 IU/L. We further showed that the urine specimens were collected from patients that were at approximately 4 weeks’ of gestation which, if calculated from the day of the last menstrual period, is close to the day of expected menses.

Anecdotally, medical providers at some institutions have recognized this phenomenon. If a sexually active woman is unsure of her pregnancy status, and the POC urine hCG test result is negative, the provider may encourage the patient to return for retesting in a few days. Alternatively, if the patient’s pregnancy status must be known urgently, the provider may collect a blood sample for quantitative serum hCG testing performed in the laboratory to confirm the negative POC test result.

Interestingly, the package insert of one qualitative hCG POC device used in our study states “If a negative result is obtained, but pregnancy is suspected, another sample should be collected and tested 48-72 hours following.” Most other hCG POC devices provide a similar disclaimer. Although it is empirically recognized that false-negative results are possible in early pregnancy, most individuals (health care professionals and consumers alike) assume that this corresponds to the period of gestation that precedes hCG production. What our study showed is that hCG is present in the urine and serum of these women, but the concentration is too low for the POC devices to always detect reliably.