Background: Middle cerebral artery peak systolic velocity (MCA-PSV) measurements are spectral Doppler waveforms determined from the middle cerebral artery, indicating the blood flow velocity within these vessels. MCA-PSV values were correlated to hemoglobin (Hb) concentration/deficit, all as a function of gestational age (GA), and the observed inverse relationship between MCA-PSV values and Hb concentration inferred a significant role in assessing fetal anemia, particularly in the context of Rhesus disease. MCA-PSV values were previously investigated and identified as a non-invasive tool for assessing fetal anemia. The objective of the study is to compare the methodology and results from the two major studies concerning their MCA-PSV curves.
Method: The data from the two studies were reviewed for sample size and composition, demographic information, statistical methods, and significance, and analyzed using computer software microsoft excel 2003 and graph version 4.4.2 (Build 543) for graphical representation.
Results: The two studies appear to be similar in many respects. Both exhibited comparable MCA-PSV values across the same range of gestational ages, with average curves displaying the calculated arithmetical mean MCA-PSV value for the two given curves. Two average curves were created. Improved linear slopes, with lower gradients for a linear model, were suggested, indicating better detection of fetuses with mild anemia.
Conclusion: The equations provided below can be integrated into ultrasound software packages to improve diagnostic capabilities.

Middle cerebral artery peak systolic velocity (MCA-PSV), Fetal anemia, Ultrasound, Fetuses, Doppler.

MCA-PSV measurements are spectral doppler waveforms determined from the middle cerebral artery, indicating the blood flow velocity within these vessels. MCA-PSV values were correlated to Hb concentration/deficit, all as a function of GA, and the observed inverse relationship between MCA-PSV values and Hb concentration inferred a significant role in assessing fetal anemia, particularly in the context of Rhesus disease.[1,2] These MCA-PSV measurements were shown to successfully identify fetuses with Rhesus disease anemia in two unrelated but paramount earlier studies.[1,2] Mari et al. were the first to perform this type of study in 2000 [1], followed by Scheier et al. in 2004.[2] Both studies aimed at using the MCA-PSV value as an intervention or indicator of fetal anemia, and both have independently shown a relationship between MCA-PSV values and fetal haemoglobin deficit.[1,2] Mari et al. used the median as a measurement of central tendency and Multiples of the Median (MoM) as a standard deviation approximation.[1] Scheier et al., on the other hand, used the mean as a measurement of central tendency and the standard deviation (SD) ‘proper’ as an indication of variance.[2] Although the outcomes of these two studies.[1,2] were similar, the quantification of MCA-PSV values may be confounded. The determination of MCA-PSV values, from two different curves, could introduce ambiguity, making it difficult and possibly inconclusive to assign a diagnosis of fetal anemia when the MCA-PSV values are at, or on, the threshold curve eg, is a 26-week-old fetus anemic at an MCA-PSV of 50.5 cm/s or 46.6 cm/s? Hence, necessitating the need for this review and comparison of the MCA-PSV curves from both studies.[1,2]

The study aims to compare the methodologies and relate the outcomes of the two major studies.[1,2] regarding their MCA-PSV curves, and implications for clinical practice. To directly compare the standard (median and mean) curves and the (1.5 standard deviation (SD) and the 1.5 multiples of the median (MoM)) anemia threshold curves from each study to each other.[1,2] To reassess the linear MCA-PSV model suggested for fetal anemia detection 3.

The data from the two studies[1,2] were reviewed for sample size and composition, demographic information, methodology, statistical methods, and significance, and analyzed using office excel 2003 and graph version 4.4.2 (build 543) for graphical representation and display purposes.

The equation for the standard curve correlating MCA-PSV to GA was given by Mari et al.[1] as MCA-PSV=exp(2.31+0.0464x). Similarly, the equation relating MCA-PSV to GA as given by Scheier et al.[2], was MCA-PSV =10^(0.0223x+0.963). Data points from each study were entered into Microsoft Office Excel 2003 to regenerate the curves from each study. The curves were also reproduced with graph version 4.4.2 (build 543). Tangent-derived linear slopes (dy/dx) were calculated at 0.01 increments, summed, and divided by the total number of slopes (n), for each curve between 15- and 40-weeks GA.

The two studies exhibited comparable MCA-PSV values across the same range of gestational ages, with average curves displaying the calculated arithmetical mean MCA-PSV value for the two given curves. Improved slopes for a linear model were suggested, indicating better detection of fetuses with mild anemia.

The 1.5MoM threshold curve used by Mari et al.[1], was calculated to be MCA-PSV =15.104exp (0.0464x). This curve was generally higher (GA>20 weeks) than the 1.5SD curve used by Scheier et al.[2] which in turn was calculated to be MCA-PSV=12.233exp (0.0514x). Conversely, the median standard curve[1] was lower than that of the mean standard curve[2] for GA<40 weeks. Thus, the difference between normal and affected fetuses was less pronounced for the 1.5SD-mean pair[2] than for the 1.5MoM-median pair[1] (Figure 1).

Figure 1 Thick solid curved line indicates the MCA-PSV average threshold, and the thin solid curved line indicates the MCA-PSV average standard curve, best-fitted (regression) with their respective equations. Dashed curved lines indicate the 1.5MoM and 1.5SD curves (short dashes), and the MCA-PSV mean and median standard curves (short-long dashes), and their respective equations. Double arrowhead vertical lines depict the ‘graphical gap’ between the 1.5MoM-median pair and the 1.5SD-mean pair, respectively. The vertical thick solid line indicates the difference between the MCA-PSV Mean and Median standard curves at approximately 37 weeks of GA. Rectangles, diamonds, crosses, asterisks, and dots represent actual data points, including data points from earlier studies.[1,2]

We calculated and generated an average MCA-PSV curve for the normal fetal population as MCA-PSV=9.5975exp (0.0489x), with a corresponding average MCA-PSV threshold curve for fetal anemia as MCA-PSV=13.635exp (0.0488x).

The first normal range for MCA-PSV values was published in 1995 by Mari et al.[3] This exponential standard curve[4] was identical to the normal median curve depicted by Mari et al.[1], despite using only 135 normal fetuses. However, it should be noted that some of the anemic-affected fetuses were in their study.[4] were not purely of Rhesus etiology. Furthermore, each of the normal curves mentioned in this study (median, mean, and average standard curve) was within the 95% confidence interval predicted by Mari et al.[3]

Subsequently, Detti et al.[4] proposed a linear model for predicting fetal anemia.[4,5] Their best-fitted linear slope[4,5] of 1.9 cm/s/wk, suggested for normal healthy fetuses, compared favorably to the average tangent-calculated slopes for the Median, Mean, and average standard curves, which were 1.77, 2.0,7 and 1.92 cm/s/wk, respectively (this study). The average tangent-calculated slopes of the 1,5MoM, 1.5SD, and average threshold curves were 2.65, 2.07, and 2.71 cm/s/wk, respectively (this study). These tangent-derived gradients appear to be markedly lower than the best-fitted slopes of 4.7 and 5.98 cm/s/wk, which were previously reported for mild and moderate/severe anemic fetuses, respectively.[4,5]

The differences between the results of the two major studies[1,2] could possibly be attributed to subtle differences in the population composition and participant selection. This holds true, particularly if subsets of small-for-gestational-age and intrauterine growth-restricted (IUGR) fetuses have not been effectively accounted for or excluded from the singleton population. Inclusion of significant numbers of IUGR fetuses in either the normal (control) or the anemic populations would consequently alter the true correlation between the determined MCA-PSV value and its given corresponding GA. Furthermore, and interestingly, Stagnati et al. reported on intertwin MCA-PSV discrepancies associated with IUGR occurring within the twin fetal populations, albeit not necessarily directly related to fetal anemia.[6]

In terms of the absolute numerical MCA-PSV values, the 1.5MoM curve is more stringent than the 1.5SD curve in identifying fetal anemia (Figure 1). However, despite the 1.5MoM curve being generally steeper than the 1.5SD curve, the spread or spectrum (graphical gap) between normal and affected fetuses was generally greater between the 1.5MoM-median pair than for the 1.5SD-mean curve pair (shown by double arrowhead vertical lines in Figure 1). The visible difference between the mean standard curve[2] and the median standard curve[1] (shown by vertical thick solid line in Figure 1) suggested that the normal fetal population MCA-PSV values (given by Hb counts and/or corresponding GA) of either study may be biased. To this end, 3 times more normal fetuses were included in the normal fetal population selected by Scheier et al.[2] than Mari et al.[1], 813 vs. 265 normal fetuses.

Methodological bias could have been introduced due to minor differences in sampling and measuring technique of the MCA-PSV (index standard), ultrasound machine, and transducer type, and operator-dependent variability.[7]
The Hb counts/concentrations (reference standard) per se could have been minimally biased according to Hb assay types, blood sampling methods, and/or laboratory specifications, etc.[7]

The flow and timing of each of the two studies did not appear to add any additional bias [7], ie, no overlapping samples and an acceptable time between index standards and reference standards were conformed to, unlike the initial study of Detti L et al.[4], where some fetuses were said to be studied on more than one occasion.

During their later study, Mari et al.[1] commented that the MCA-PSV test was not particularly good in identifying fetuses with only mild anemia. Therefore, a reduction of their stringent anemia threshold, through using the calculated average threshold (and linear slopes from this study) instead, should, in retrospect, putatively accommodate and identify more fetuses with mild anemia from within Mari’s cohort of participants.[1]

A statistical model, using modal values as a measurement of central tendency, would be ideal for completeness. However, the average curves generated in this investigation, as a combination of the two major previous studies[1,2], may be more appropriate. These average regression curves for MCA-PSV values could enhance the accuracy of ultrasound assessments in the clinical setting, facilitating timely interventions for fetal anemia. The equations provided below can be integrated into ultrasound software packages to improve diagnostic capabilities and performance. MCA-PSV=9.5975exp (0.0489x), with a corresponding average MCA-PSV threshold curve for fetal anemia as MCA-PSV=13.635exp (0.0488x).

Thank you to the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) for accepting the content as a preliminary poster for the 33rd ISUOG World Congress. The Fetal Medicine Group at Tygerberg Hospital, second floor, under the guidance of Prof L. Geerts for assistance and time there, and the Cape peninsula university of technology (CPUT) team.

No funding was provided for this study.

Corresponding Authors:
Elton C. Nelson
Department of Radiology
Cape Peninsula University of Technology, Cape Town, South Africa
Email: eltonnelson12345@gmail.com

Co-Author:
Mark Marais
Department of Medical Imaging, Senior Lecturer
Cape Peninsula University of Technology, Cape Town, South Africa

All authors contributed to the conceptualization, investigation, and data curation by acquiring and critically reviewing the selected articles. They were collectively involved in the writing – original draft preparation, and writing – review & editing to refine the manuscript. Additionally, all authors participated in the supervision of the work, ensuring accuracy and completeness. The final manuscript was approved by all named authors for submission to the journal.

Not applicable

No conflict of interest

Not reported

https://doi.org/10.63096/medtigo30622414

Elton CN, Mark M. MCA-PSV Determination in Fetal Anemia. medtigo J Med. 2024;2(4):e30622414. doi:10.63096/medtigo30622414 Crossref