Three-dimensional growth dynamics of the liver in the human fetus

The present study describes a cross-sectional interpretation of the three dimensions (length, transverse and sagittal diameters) of the growing liver in human fetuses. In total, 69 autopsied formalin-fixed fetuses at the age of 18–30 weeks were enrolled for final analysis. So, the sample was relatively numerous, comparable and representative of fetal age. In fact, the fetuses studied might potentially be smaller for gestational age. However, the assumption that fetuses could suffer from intrauterine growth retardation was unambiguously disapproved, because the correlation coefficient between the gestational age based on the crown-rump length and that calculated by either the amenorrhea age or ultrasound age reached the value R = 0.99 (P < 0.001) for the whole sample. In clinical practice, a reliable estimate of gestational age up to 14 weeks is confirmed by ultrasound measurement of fetal crump-rump length. As recommended by the Fetal Growth Longitudinal Study, part of the International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21st) Project, in women commencing antenatal care at less than 14 weeks and 0 days of gestation until birth, a combination of known values of the five fetal ultrasonic measurements: head circumference, bi-parietal diameter, occipitofrontal diameter, abdominal circumference, and femur length behooves estimation of accurate fetal ages [4, 18, 23‐28, 31]. The Fetal Growth Longitudinal Study prospectively assessed fetal growth from 14 to 42 weeks of gestation in 4,321 mothers who were at low risk of intrauterine growth restriction (optimal health and nutritional maternal wellbeing, satisfactory antenatal care, singletons without congenital malformations) in eight geographically diverse urban populations [18]. Of note, the ultrasound measurements were taken specifically and the method was rigorous and implemented across all study sites with the same specially adapted ultrasound equipment to allow blinding of measurements. As a result, the 3rd, 5th, 10th, 50th, 90th, and 97th centile curves vs. gestational age for the five primary ultrasound growth measures were produced, as the international standards for fetal growth.

As far as formalin-fixed fetuses are concerned, it should be emphasized that formalin fixation little (0.5–1.0 %) influences both size and shape of parenchymal viscera preserved in situ in the sealed thoracic and abdominal cavities [4, 11, 16, 23‐28]. On the contrary, any isolated organ subjected to formalin fixation is prone to artifacts as a result of tissue shrinkage. We decided not to assess the liver weight, because the 12–24 months formalin preservation could certainly gain the weight and density of fixed organs, even by 20–25 % for the liver [8]. Our results were based on plausible direct measurements, instead of deduced, extrapolated through a series of indirect measurements. All the parameters measured were exact and clearly definable. We opine that our empirical results could be considered as veracious because both the material studied and the methods used were apposite.

In the present study, no significant male–female differences concerning all the three morphometric parameters of the fetal liver were found, in accordance with other reports [1, 6, 13, 21]. Thus, we did not attempt to further separate our findings with relation to sex.

It is explicitly apparent that the fetal liver growth is three-dimensional with simultaneous evolution of length, transverse and sagittal diameters. To date, however, except for one anatomical study by Albay et al. [1] and one ultrasonographic research by Chang et al. [6], reference data for liver dimensions in the human fetus have focused on the liver length only. Albay et al. [1] basing on 121 autopsied fetuses aged 9–40 weeks examined the height, width, and thickness of the liver, that were tantamount to its length, transverse and sagittal diameters, respectively. The liver length gradually increased from 13 ± 2 mm in fetuses aged 9–12 weeks (1st trimester) through 25 ± 8 mm in fetuses aged 13–25 weeks (2nd trimester) and 41 ± 7 mm in fetuses aged 26–37 weeks (3rd trimester) to 47 ± 8 mm in fetuses aged 38–40 weeks (full term). Some authors [12, 17, 22, 29, 30] 2-D ultrasonically studied the liver length to present its normal range for every week of pregnancy. Vintzileos et al. [30] ascertained that the mean liver length incrementally increased from 27.3 to 59.0 mm in fetuses aged 20 and 40 weeks, respectively. Roberts et al. [22] showed that in 350 normal pregnancies, the fetal liver length crept up in a probably parabolic manner, from 12.5 mm in fetuses aged 13 weeks to 51 mm in fetuses aged 40 weeks. In the ultrasonographic group of Mirghani et al. [12], consisting of 31 fetuses aged 21–24 weeks, the liver length ranged from 30 to 33 mm. Phatihattakorn et al. [17] presented sonographically derived results from liver length measurements of 750 fetuses at the age of 13–40 weeks, ranging from 12.59 ± 2.08 to 60.06 ± 3.49 mm. These findings coincide with those reported by Tongprasert et al. [29], who published the results on normal liver length in 640 fetuses from 14 to 40 weeks of gestation. During that time, the liver length increased from 9.9 to 51.9 mm for the 5th percentile, from 15.8 to 57.7 mm for the 50th percentile, and from 21.7 to 63.6 mm for the 95th percentile [29]. As reported by Kuno et al. [9], the liver length increased from 25 mm for the 21-week fetus to 58 mm for the 38-week fetus. In the present study, we found the liver length to increase from 19.51 ± 1.02 mm in the 18-week fetus to 39.65 ± 7.05 mm in the 30-week fetus. The meticulous comparative analysis of numerical data revealed that our anatomical results concerning the liver length were in secure consensus with the ultrasonic findings reported by Vintzileos et al. [30], Roberts et al. [19, 22] and Kuno et al. [9]. On the contrary, the means for liver length obtained by Mirghani et al. [12] were smaller by 1–3 %, and those reported by Phatihattakorn et al. [17] and Tongprasert et al. [29] were greater by 4–7 % when compared to our findings. We suggest that the possible differences may be attributed to either anatomical varieties or racial reasons.

In the anatomical material of Albay et al. [1], the transverse diameter of the fetal liver attained the following values: 19 ± 3, 39 ± 12, 67 ± 11, and 82 ± 9 mm for the three successive trimesters of gestation and full term fetuses, respectively. The liver sagittal diameter in the fetus grew from 11 ± 1 mm in the 1st trimester, through 18 ± 5 mm in the 2nd trimester and 26 ± 5 mm in the 3rd trimester to 31 ± 8 mm at full-term fetuses. Regrettably, these authors did not present the means of transverse and sagittal diameters for particular weeks. Moreover, the ultrasonic study reported by Chang et al. [6] showed only two scattergrams for transverse and sagittal diameters, without any numerical data. Finally, our results present the liver transverse and sagittal diameters, changing from 29.44 ± 3.73 to 53.13 ± 5.31 mm, and from 22.97 ± 3.79 to 43.22 ± 5.49 mm, respectively. So, for the forenamed reasons, we could not compare our age-specific intervals for transverse and sagittal diameters of the fetal liver with the findings obtained by Albay et al. [1] and Chang et al. [6]. This has considerably limited our discussion on this subject.

To date, assessment of precise growth dynamics for the growing liver has featured in only a few publications [6, 9, 17, 29, 30]. The growth models divulged by these authors are all alike, presenting a linear relationship between parameters studied and fetal age. In a linear function the rate of growth remains constant throughout the study period. Its value (a) precisely corresponds with the coefficient accompanying the fetal age (age), according to the linear formula y = a × (age) + b. According to some authors [2, 11, 22, 30], the fetal liver length grew at a rate of 1.0–1.2 mm per week up to 28 weeks of gestation, and at 1.70–1.76 mm per week after 28 weeks. Phatihattakorn et al. [17] demonstrated a proportionate increase in liver length throughout gestation, that followed in accordance with the linear function: liver length (mm) = 1.528 × (age) − 5.676. Similarly, Tongprasert et al. [29] found the best regression model for the liver length as liver length (mm) = 1.61 × (age) − 6.750 (R ² = 0.94; P < 0.05). In the 3-D sonographic study of 14 appropriate-for-gestational-age fetuses, reported by Kuno et al. [9], the optimal model for liver length was as follows: liver length (mm) = 1.74 × (age) − 9.47 (R ² = 0.858; P < 0.001). Therefore, having fastidiously performed the comparative analysis of growth dynamics for liver length, constructed by Phatihattakorn et al. [17], Tongprasert et al. [29] and Kuno et al. [9], we emphasize the affinity between the three growth rates, i.e. 1.528 mm per week, 1.61 mm per week, and 1.74 mm per week, respectively. To date, only Chang et al. [6] with the use of three-dimensional ultrasound studied the growth dynamics of the liver in 55 normal singleton fetuses aged 20–30 weeks with relation to liver length, transverse and sagittal diameters. The scattergrams reflected the following linear functions: y = 1.191 × (age) + 0.658 (R = 0.58; P < 0.001) for liver length, y = 2.370 × (age) − 15.897 (R = 0.78; P < 0.001) for transverse diameter, and y = 1.520 × (age) + 4.593 (R = 0.56; P < 0.001) for sagittal diameter. This means that the growth rate for liver length was much slower (1.191 mm per week) than those reported by the three aforementioned authors. An increase in transverse and sagittal diameters ensued more energetically, when compared to that of liver length, with the following growth rates: 2.370 mm per week and 1.520 mm per week, respectively. In the present study, the numerical data have been presented in a similar manner as the INTERGROWTH-21st Project data [18], including the fitted 3rd, 10th, 50th, 90th, and 97th smoothed centile curves. The length, transverse and sagittal diameters of the fetal liver did not generate linear functions on nomograms. In fact, we substantiated that the best-fit growth models were the following natural logarithmic functions: y = −82.778 + 35.752 × ln(age) ± Z × (−2.778 + 0.308 × age) for liver length, y = −123.06 + 52.668 × ln(age) ± Z × (3.156 + 0.049 × age) for liver transverse diameter, and y = −108.94 + 46.052 × ln(age) ± Z × (−0.541 + 0.188 × age) for liver sagittal diameter. It is noteworthy that the natural logarithmic function y = ln(x) is one-to-one (for each y there is one and only one x), continuous, and increasing. Besides, it indicates a declining rate of change, expressed precisely by a concave down graph, more and more deviating from the linear function y = x with advanced fetal age. That is because in the material under examination, the growth rates per week for the 50th centile were gradually decreasing from 1.93 to 1.21 mm for length, from 2.85 to 1.79 mm for transverse diameter, and from 2.49 to 1.56 mm for sagittal diameter of the liver (P < 0.05) in fetuses aged 18 weeks and 30 weeks, respectively. Therefore, the nomograms presented in the material under examination emphasize a much more extensive growth rate for liver length only in younger fetuses than that presented by all other authors. On the contrary, in older fetuses, the growth rate of liver length was affinitive to that found by Chang et al. [6], and much slower to those presented by Phatihattakorn et al. [17], Tongprasert et al. [29], and Kuno et al. [9]. As far as the transverse and sagittal diameters of the fetal liver are concerned, the only nomograms constructed by Chang et al. [6] support a regular increase in transverse diameter by 2.370 mm per week, and in sagittal diameter by 1.520 mm per week. This growth rate for liver transverse diameter (2.370 mm per week) was comparable to our findings (from 2.85 mm in the 18-week fetus to 1.79 mm in the 30-week fetus). On the contrary, the growth rate for liver sagittal diameter (1.520 mm per week) presented by Chang et al. [6] turned out to be slower to our results (from 2.49 mm in the 18-week fetus to 1.56 mm in the 30-week fetus).

We extrapolated that notwithstanding an absolute increase in the three values studied, the spatial growth of the liver ensued in a relatively proportionate manner. The three-dimensional evolution in the mean length, transverse and sagittal diameters of the fetal liver followed proportionately as 0.71:1:0.87.

Having elucidated the normative growth of liver length, transverse and sagittal diameters in the human fetus, we would like to accentuate the importance of the measurements in the material under examination, because the reader should be endowed with relevant data that are useful for distinguishing abnormal from normal fetal development. Due to the best-fit growth models for the mean and SD (Table 3) for each measure (liver length, transverse and sagittal diameters), readers can calculate any desired centiles according to gestational age. Of note, the value of Z always equals −1.88 for the 3rd centile, −1.28 for the 10th centile, 0 for the 50th centile, +1.28 for the 90th centile, and +1.88 for the 97th centile. Murao et al. [14] showed that the fetal liver was considerably small compared to other organs in intrauterine growth retardation. The fetal liver length was found to be below average in 18 % [30] or even 30 % [9, 12, 19] of small-for-gestational-age fetuses. Vintzileos et al. [30] reported on 8 fetuses with severe erythroblastosis fetalis correlated with excessive increases in fetal liver length by 5 mm and over per week. As reported by Roberts et al. [22], enlargement of the fetal liver in isoimmunized pregnancies substantively correlated with increased liver hemopoiesis, and liver length proved to be a conducive indicator of the degree of fetal anemia. Moreover, fetal transfusion and correction of fetal anemia influenced a decrease in fetal liver length. Besides, a considerable increase in liver dimensions in the fetus is a good indicator of the severity, the clinical outcome in pregnancies with Rh isoimmunization, Hb Bart’s disease, congestive heart failures, certain metabolic diseases, tumors, and even intrauterine infections [1, 3, 13, 14, 20]. Roberts et al. [21] ascertained that in fetuses at the age of 18 weeks and over the liver length was 12 % greater in diabetic pregnancies, with relation to normal pregnancies. These authors substantiated that knowledge of the normal range for the fetal liver length is beneficial to diagnose and follow-up both gestational diabetes and hepatomegaly. Mirghani et al. [12] found the mid-trimester fetal liver length to be significantly greater (36 vs. 31 mm) in pregnant women with gestational diabetes mellitus. The measurement of fetal liver length in the diabetic pregnancy is reproducible and of outmost relevance as a parameter for monitoring the effectiveness of treatment in diabetic pregnancies [2, 10].

In our opinion, the main advantages of the present study result in a unique fetal material, objective digital-image analysis, the appropriate statistical analysis, and completely novel growth patterns for the growing liver. The main limitation of the present study seems to be a relatively narrow fetal age, ranging from 18 to 30 weeks of gestation. Had we been able to collect a larger fetal sample size with fetuses younger than 18 weeks and older than 30 weeks, this would considerably hone the growth curves obtained. In nature, all autopsy examinations include retrospective design without prospective ultrasound quality control. Because of this, a disadvantage may be that our findings have been presented as if describing a developmental sequence in one fetus. Therefore, it is rather a populational perspective than a true representation of growth in itself. Furthermore, in our research, measurements were conducted by a single observer in a blind fashion. Besides, our results lack inter-observer variability, because all numerical data have been controlled by one researcher.

In summary, the present paper attempts to cogently extend the existing literature relating to development of the fetal liver in human fetuses. We confer that the crucial findings in our study are that the growth dynamics for the liver length, transverse and sagittal diameters do not follow proportionately, but indicate a disparate fashion, according to logarithmic regressions. The three novel nomograms constructed in the present study may potentially be useful in monitoring normal fetal development and in the diagnosis, monitoring and treatment of severe fetal pathologies that affect the liver in the intrauterine period.