None of the authors have any conflicts of interest to disclose.
How to Cite this Article: Bober MB, Niiler T, Duker AL, Murray JE, Ketterer T, Harley ME, Alvi S, Flora C, Rustad C, Bongers EMHF, Bicknell LS, Wise C, Jackson AP. 2012. Growth in individuals with Majewski osteodysplastic primordial dwarfism type II caused by pericentrin mutations. Am J Med Genet Part A 158A: 2719–2725.
Microcephalic primordial dwarfism (MPD) is a group of related disorders characterized by severe pre- and postnatal growth failure together with microcephaly [Klingseisen and Jackson, 2011]. Majewski osteodysplastic primordial dwarfism type II (MOPDII; OMIM #210720) is a distinctive diagnostic entity within this group and one of the most common conditions encountered [Hall et al., 2004; Rauch, 2011]. Aside from the classic features of MPD, individuals with MOPDII have a characteristic skeletal dysplasia [Hall et al., 2004; Willems et al., 2009], abnormal dentition [Kantaputra et al., 2011], an increased risk for cerebrovascular disease [Brancati et al., 2005; Waldron et al., 2009; Bober et al., 2010], and insulin resistance [Huang-Doran et al., 2011]. MOPDII is caused by mutations in the pericentrin (PCNT) gene and is inherited in an autosomal recessive manner [Rauch et al., 2008]. Typically, infants with intrauterine growth restriction (IUGR) are fed with increased caloric density formulas and more aggressive feeding regimes in an attempt to improve growth [Hall et al., 2004]. However, given the underlying genetic nature of MOPDII, typical growth velocities cannot be attained. Although longitudinal growth data for a number of cases was presented in the natural history study by Hall et al. , the population in that study was genetically heterogeneous. No detailed growth data exist for individuals with MOPDII caused by PCNT mutations. As poor growth is such an important characteristic of MOPDII, establishing normative curves would provide an important tool to aid in the diagnosis as well as management of children with MOPDII. Here we present the first detailed growth curves for height, weight, and head circumference (OFC) for 26 individuals with MOPDII with PCNT mutations or demonstrated absence of PCNT protein. Moreover, we describe the effect of growth hormone therapy on this population.
To systematically ascertain the medical problems associated with MOPD II and other forms of MPD, a Primordial Dwarfism Registry has been established at the Alfred I. duPont Hospital for Children (Institutional Review Board approved). The registry collects retrospective medical records. To date, 62 patients with various forms of MPD have been enrolled via a written consent process. Thirty-four of these patients have a clinical diagnosis of MOPDII. Biallelic PCNT mutations have been identified by sequencing in 24 of these patients (see Supplemental eTable I in Supporting Information online), PCNT status has not been established in eight patients. For the remaining two patients, biallelic mutations in PCNT could not be identified through capillary sequencing of coding exons and intron–exon boundaries [Griffith et al., 2008]. However, absence of PCNT protein was confirmed through both Western immunoblotting and immunofluorescence against PCNT (see Supplementary eFig. 1 in Supporting Information online) using patient-derived dermal fibroblasts, in keeping with the diagnosis of MOPD II [Griffith et al., 2008]. Absence of PCNT suggests that cryptic mutations in PCNT are present in these patients. Sex, gestational age, height, weight, and OFC measurements were obtained from the medical records of those 26 individuals with a clinical diagnosis of MOPDII and either biallelic PCNT mutations or the absence of PCNT protein. In addition, growth hormone treatment history was recorded.
Standard Deviation Calculation
Standard deviations (SDs) for height, weight, and OFC normalized for age and sex were calculated using Cole's LMS method using UK 1990 cohort data [Freeman et al., 1995]. Age at examination was corrected for prematurity where birth was before 37 weeks gestation and age <2 years. Age was then rounded to the nearest month prior to SD calculation.
Growth Chart Calculation
Data were analyzed using R software for statistical computing (R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org). After separating data into human growth hormone (hGH) treated and non-hGH treated groups, height, weight, and OFC for males and females were all plotted against age over the ranges of 0–24 months and 2–18 years. For children under the age of two a correction for gestation age under 37 weeks was made. Data were modeled using cubic spline fits to obtain representative averages of the data as a function of age. SDS as a function of age were automatically calculated by the software and then added to the graphs. Weight–height curves for subjects in the non-hGH group were generated for subjects between ages 0 and 24 months. Height velocity curves were generated for non-hGH exposed measurements in subjects from ages 2 to 15 years by fitting a cubic spline model to the height data and calculating the derivative using the five point stencil method. These results were then plotted along with the CDC height–velocity norms for girls matched in age.
In total, we studied 26 patients with molecularly confirmed MOPDII ranging in age at time of final data collection from 35 months to 33 years. Twelve are male and 14 are female. The overall average gestational age at birth was 34.8 ± 2.8 weeks. For males the average gestational age was 34.8 ± 3.0 weeks and for females the average gestational age was 34.9 ± 2.8 weeks. Eleven of the patients were treated with growth hormone. Eleven of the patients reached skeletal maturity as defined by either the same consecutive height measurements or documented closed growth plates at the time of a height measurement. The total number data points collected for height was 490, for weight was 608, and for OFC was 240. Prior to the initiation of growth hormone there were 266 data points for height, 360 for weight, and 152 for OFC.
Growth in MOPD II
Marked growth failure was evident from birth in MOPD II patients. Nineteen of the patients had birth lengths notated which, when corrected for gestational age were on average −7.0 ± 2.4 SD below the population mean (see Table I). Twenty-six of the patients had BW measurements which when corrected for gestational age were on average −3.9 ± 1.1 SD below the mean. Fifteen of the patients had a birth OFC measurement which when corrected for gestational age were on average −4.6 ± 1.7 SD below the mean. For skeletally mature patients, final height was −10.3 ± 2.3 SD (n = 11), weight −14.3 ± 7.7 SD (n = 10), and OFC −8.5 ± 2.1 SD (n = 9) below the mean. Sex-specific growth parameters are shown in Table I. Table II shows the skeletally mature patients stratified by history of hGH treatment or non-treatment. No statistically significant differences were seen when comparing the average SDs of height, weight, or OFC in hGH treated versus non-treated patients.
Table I. Growth Parameters as Measured by Standard Deviations Compared to the General Population at Birth and Skeletal Maturity
At skeletal maturity
All participants (n = 26)
All mature (n = 11)
Males (n = 6)
Females (n = 5)
Birth measurements were corrected for gestational age if birth was prior to 37 weeks. n, number of measurements; ASD, average standard deviation; SD, standard deviation of ASD.
Table II. Growth Parameters as Measured by Standard Deviations Compared to the General Population at Skeletal Maturity by Sub-Groups
hGH treatment (n = 6)
No hGH treatment (n = 5)
P values comparing hGH treatment to no treatment are derived from two tailed unpaired Student's t-tests and are shown in the rightmost column. n, number of measurements; hGH, human growth hormone; ASD, average standard deviation; SD, standard deviation of ASD.
P = 0.18
P = 0.91
P = 0.58
Growth Curves for MOPD II
Combined charts for males and females for height (Figs. 1 and 2), weight (Figs. 3 and 4), and OFC (Figs. 5 and 6) were constructed. For children under the age of two a correction for gestation age under 37 weeks was made. The Center for Disease Control (CDC) growth charts for girls [Kuczmarski et al., 2000] with the 5th, 50th, and 95th centiles are superimposed for comparison on all of the curves. For OFC over the age of two, the Rollins [Rollins et al., 2010] head growth curve for girls was used. Given the extreme small size of individuals with MOPD II, female curves were used as in general, girls are smaller than boys and therefore provided the most conservative comparison. In addition, a weight for height chart was constructed for boys and girls together <2 years of age (Fig. 7).
Hall et al.  were the first to publish graphs of various growth parameters drawn from growth data in 58 (27 historical and 31 unpublished) individuals with a clinical diagnosis of MOPD II. These graphs were not specific growth charts, but rather aggregate plots of growth data. There are several factors including exposure to hGH, correction for gestational age and clinical versus molecular diagnosis which make direct comparisons between these graphs and ours difficult. There is however a reasonable degree of similarity in the 2–18-year growth patterns for height and OFC. There is an appearance that our patient's weights may be slightly higher. The causative gene for MOPD II [Griffith et al., 2008; Rauch et al., 2008; Willems et al., 2009] was elucidated 4 years after the publication of the natural history by Hall et al. . With molecular testing now available, there are individuals previously assigned a diagnosis of MOPD II that are now thought to have a different form of MPD, given negative PCNT gene testing results. In addition there are individuals who had a previous clinical diagnosis of Seckel syndrome in whom mutations in PCNT have been identified [Griffith et al., 2008], and subsequent evaluations have demonstrated that these patients' phenotypes fit most closely with MOPD II [Willems et al., 2009]. In our cohort of 26 patients with a clinical diagnosis of MOPD II, biallelic PCNT mutations have been identified by sequencing in 24 of these patients and in 2 patients, absence of PCNT protein was confirmed through both Western immunoblotting and immunofluorescence against PCNT (see Supplemental eTable I in Supporting Information online). Absent PCNT suggests that cryptic mutations in PCNT are present in these patients. Thus, data presented here are more comprehensive than that of Hall et al.  and benefit from molecular validation of the clinical diagnosis.
The ages of our patients ranged from 35 months of age to 33 years, with 11 having reached skeletal maturity. Prematurity was common and the overall gestational age at birth was 34.8 ± 2.8 weeks. This is entirely consistent with the 35 weeks described by Hall et al. . It is possible that prematurity is over-estimated as the revision of due dates in this population is common given the degree of IUGR identified during prenatal ultrasounds. When possible, gestational ages have been calculated based upon last menstrual period dating. The average size at birth for length, weight, and OFC respectively was −7.0, −3.9, and −4.6 SDs from the population mean (after correcting for gestational age <37 weeks). This means that on average, at term, length of an infant with MOPD II is that of a 28–29-week neonate, weight that of a 31–32-week neonate and the OFC that of a 30–31-week neonate. These estimations differ only slightly from those of Hall et al.  who suggested that at term, these children were the size of a 28-week preemie for height, weight, and OFC. When taking the height SDs at skeletal maturity and determining the age which that height was the 50th centile, it could be determined that the mature male is the height of an average 3-year-10-month-old boy, the weight of an average 5-year-2-month-old boy with the OFC of an almost 5-month-old male infant. For mature females, the height is that of an average 3-year-11-month-old girl, the weight of an average 5-year-3-month-old girl with the OFC of a 6-month-old female infant.
Given the distinct clinical courses and associated morbidities of various forms of MPD, it is important to make a precise clinical diagnosis with subsequent molecular confirmation where possible. Growth parameters taken in conjunction with other clinical features can have diagnostic utility. For example, MOPDII is often morphometrically distinguishable from Meier–Gorlin syndrome (MGS), another microcephalic dwarfism disorder caused by mutations in five genes encoding components of the pre-replication complex [Bicknell et al., 2011a, b; Guernsey et al., 2011]. Detailed analyses of children with MGS by de Munnik et al. (submitted) demonstrate at birth, mean length was −3.9 SD, with a mean weight of −3.4 SD, and OFC of −2.14 SD (compared to −7, −3.9, and −4.6 SD, respectively for MOPDII). Thus these children are much larger at birth, particularly in respect to birth length. Although minor ear anomalies such as simple ears and attached lobes have been described in both MOPDII and MGS [Hall et al., 2004; Munnik et al., 2012] a much shorter child with a smaller OFC would suggest MOPDII rather than MGS. Of course, other features such as patella aplasia would also point to MGS as the diagnosis as well. While the precise size of infants with MOPD I is not well delineated, they would likely be smaller than an infant with MOPDII. In a MOPD I cohort of 14 patients, the mean BW was −5.8 SD below the mean and the mean OFC was −7.0 SD below the mean [Nagy et al., 2011]. Significant brain malformations, skeletal dysplasia and recognizable physical features such as proptosis and sparse hair suggest that these disorders will be readily distinguishable on clinical features alone [Sigaudy et al., 1998; Juric-Sekhar et al., 2010; Edery et al., 2011; Nagy et al., 2011].
The final height of 11 patients who completed their growth were −10.3 SDs below the mean, their weights −14.3 SDs below the mean, and their OFCs were −8.3 SDs below the mean. P values comparing hGH treatment to no treatment are derived from two tailed unpaired Student's t-tests. No statistical differences were seen in the height (P = 0.18), weight (P = 0.91), or OFC (P = 0.58) SDs of patients who were treated with hGH and those who were not (see Table II). This confirms the general belief of treating physicians that hGH has been ineffective in their individual MOPD II patients. In Figure 8, average linear growth is plotted for those who have had hGH treatment in comparison to those who have not. Though there is a small initial gain in growth, by adulthood there is no clinically meaningful gain in height with maximum male height at skeletal maturity being 111.8 cm (n = 1), while untreated height was 111.0 cm (n = 1). For females, the maximum height at skeletal maturity in an hGH treated patient was 104.0 cm. The only untreated female had a height of 69.0 cm, but is likely an outlier, with other confounding clinical issues, including severe feeding issues with reflux and aspiration as well as subglottic stenosis leading to tracheal reconstructive surgery requiring permanent tracheostomy. Even with her inclusion, the difference between untreated and treated MOPD II patients (both sexes) remains non-significant (P = 0.18). Furthermore, hGH therapy may be deleterious to MOPD II patients, given that a patient in this cohort developed an elevated total insulin level (without evidence of diabetes), which became apparent during growth hormone therapy, and returned to normal following cessation of hGH supplementation. Insulin resistance also became apparent during hGH therapy in two patients in a previous study [Huang-Doran et al., 2011], which also reported that hGH therapy is ineffective in promoting linear growth. In summary, there is no evidence to support using hGH treatment to improve linear growth, and there is at least anecdotal evidence that it can contribute to the development of insulin resistance in this population.
Cessation of meaningful head growth in MOPD II occurs at approximately 18 months of age although some limited growth does continue to occur, resulting in a final OFC of −8.5 ± 2.1 SD. Individuals with MOPD II have cognitive abilities which far exceed those of a 6-month-old infant, and therefore demonstrate that extreme microcephaly need not necessarily be associated with cognitive impairment. Adults with MOPD II, not affected by a cerebrovascular accident, have average to slightly below average intelligence [Hall et al., 2004] and in our experience can hold jobs, drive cars, and be productive members of society.
As evidenced by Figure 9, the height velocity is much reduced in MOPD II, and effectively plateaus by 2 years of age and decreases starting at approximately 7 years of age, with no pubertal growth spurt. The height curve therefore drops further away from the typical stature curve as age increases (Figs. 1 and 2), and the skeletally mature growth parameters have SDs farther below the mean than the birth parameters. The widening of the SD curves in Figures 3 and 4 is likely explained by the relative paucity of growth points in the older ages, and the fact that one individual with severe medical problems in early childhood and subsequent severely reduced growth (38% smaller than the tallest individual) is included in the points influencing the spread.
More points are included in the weight analysis in the younger ages predominately because a majority of these infants (without older affected siblings) were being monitored closely to combat a perceived failure to thrive, so had frequent measurements in the younger ages, or alternatively they were hospitalized in early infancy because of their extreme small size, and weights were recorded from their admission notes. The curve seems to dip downward in later infancy (Fig. 3) which can be explained by our experience that clinicians eventually become aware that they are overfeeding these children, and back off of their aggressive feeding regimes to allow the child to return to a better homeostasis. The spread becomes larger around age 10 (Fig. 4) as truncal obesity becomes more commonplace in this population, which is a known risk with MOPDII [Hall et al., 2004; Huang-Doran et al., 2011]. Again, the spread is skewed downward due to the one individual that is much shorter than the rest of the group, and therefore weighs much less. The most important contribution of these growth curves is therefore likely to be in the early management of nutrition of these children and availability of these charts will hopefully reduce morbidity that previously resulted from aggressive feeding regimes.
The major strengths of these data are that they are from a genetically homogenous group. However, because this is a smaller more defined group of an already rare group of conditions, with genetic testing only very recently becoming clinically available, the number of individuals eligible to be a part of this study is small. The cubic spline method statistically weights data according to the reciprocal variance of its measurement error and thereby ensures that uncertainty estimates are rather conservative. Where there is less data, the uncertainties tend to be rather larger. Also because of the small number of participants, any one outlying individual can more dramatically affect the end result. For this reason, the plotted SD confidence intervals are large, particularly at older ages. Therefore though they are provided to assist clinicians in judging if there is significant deviation from expected growth in future patients, caution is required in interpretation. Actual data points are also included as we find these helpful and may be a more useful guide to inform clinical judgment on appropriate growth in this disorder.
In summary, despite these limitations, we feel these growth curves will be valuable tools to help clinicians, both in assessing if MOPDII seems an appropriate diagnosis, and once diagnosed, to set better goals in height and weight gains over time. In particularly we hope that these growth parameters will reduce unnecessary nutritional interventions from well-intentioned clinicians trying to boost weight when it is not a realistic goal in this genetically determined disorder of globally reduced growth.
This work is supported in part by the Potentials Foundation, the Walking with Giants Foundation, and the Texas Scottish Rite Hospital for Children Research Fund. We thank all of the families who participate in the Primordial Registry at the Alfred I. duPont Hospital for Children. This work would not be possible without their contributions. APJ's lab is funded by the MRC and Lister Institute for Preventative Medicine. We also thank Tim Cole and Laurens Holmes, Jr. for their assistance in the early stages of this work.