Fractures in infants and toddlers with rickets

Pediatr Radiol (2010) 40:1184–1189 DOI 10.1007/s00247-009-1470-8 ORIGINAL ARTICLE Fractures in infants and toddlers with rickets Teresa Chapman & Naomi Sugar & Stephen Done & Joanne Marasigan & Nicolle Wambold & Kenneth Feldman Received: 7 September 2009 / Revised: 1 October 2009 / Accepted: 19 October 2009 / Published online: 9 December 2009 # Springer-Verlag 2009 Abstract Background Rickets affects young infants and toddlers. However, there is a paucity of literature regarding the types of fractures that occur in rachitic patients. Objective To evaluate the age of patients at which radiographically evident rickets occurs, and to characterize the age incidence and fractures that are observed in infants and toddlers with radiographically evident rickets. Materials and methods A retrospective study of children younger than 24 months was performed. Clinical data and radiographs were reviewed. Radiographs obtained within 1 month of the diagnosis were evaluated for the presence or absence of osteopenia, presence or absence of fraying– cupping, and presence and characterization of fractures. Results After exclusion criteria were applied, 45 children were included in the study. Children with rickets evident by radiograph were in the age range of 2–24 months. Fractures were present in 17.5% of the study group, exclusively in mobile infants and toddlers. Fracture types included transverse long bone fractures, anterior and anterior-lateral rib fractures, and metaphyseal fractures. All fractures occurred exclusively in patients with severe, overtly evident rickets. T. Chapman (*) : S. Done Department of Radiology, Seattle Children’s Hospital, MS R-5417, 4800 Sand Point Way NE, Seattle, WA 98105, USA e-mail: Teresa.chapman@seattlechildrens.org N. Sugar : K. Feldman Children’s Protection Program, Seattle Children’s Hospital, Seattle, WA, USA J. Marasigan : N. Wambold College of Arts & Sciences, University of Washington, Seattle, WA, USA Conclusion Fractures occur in older infants and toddlers with overt rickets and can be seen by radiograph. Fractures do not resemble high-risk non-accidental trauma fractures. Keywords Rickets . Non-accidental trauma . Infant . Fracture Introduction A recent controversy has been generated by the assertion that multiple fractures occur spontaneously in young infants with subclinical vitamin D deficiency and insufficiency, in the absence of radiographic findings of rickets [1]. Rickets is a deficiency of bone mineralization, most recognizable at the growth plate, that results in radiographically evident abnormalities, including metaphyseal flaring and cupping, physeal widening, and both focal and generalized osteomalacia. Fractures occur in both normal and abnormal bone, and the forces required to fracture bone of varying strength differ. Kleinman’s [2, 3] extensive work regarding fracture patterns in non-accidental trauma is based on scientific method and pathologic correlation. Kleinman has also addressed the differential diagnosis of the classic metaphyseal lesion, including rickets [4]. The radiologist might suspect and at times make the diagnosis of non-accidental trauma based on recognition of highly specific fracture types and on multiple fractures of different ages. Multiple fractures also occur in the setting of bone disease, such as rickets. Therefore, an understanding of what fractures are expected in the setting of non-accidental trauma versus rickets is important. This study aims to evaluate infants and toddlers with radiographically evident rickets and to document the Pediatr Radiol (2010) 40:1184–1189 1185 Fig. 1 Age of included children, as a percentage of the total number of patients frequency of injury, age of injury and types of fractures that occur within this specific population. We hypothesize that fractures in children with rickets occur in abnormal bone and can be explained by a failure of normal load-bearing of abnormal bone. Materials and methods IRB approval was obtained from the Seattle Children’s Hospital. This retrospective study was limited to children younger than 24 months. Children were identified through two means: Group 1 patients were identified by their radiographic diagnosis, using a radiology report word search for cases of rickets imaged from Jan. 1, 2000, through Dec. 31, 2007; group 2 patients were identified by a laboratory report search for elevated alkaline phosphatase at least 1.5 times normal as a possible indicator of metabolic bone disease. These laboratory values were collected between Jan. 1, 2007, and June 30, 2008. These search methods resulted in identification of 62 children. Children were excluded for the following reasons: endstage renal or liver disease, history of renal or liver transplant, short-gut with prolonged TPN use, prematurity <34 weeks, complicated cancer, and no diagnostic radiographs available for review. Table 1 Ethnicity of children in the nutritional rickets patients versus the metabolic and secondary rickets patients Age at diagnosis White African American African Hispanic Pacific Islander A clinical chart review was performed by two undergraduate students (JM and NW) with two child abuse pediatricians (NS and KF). Radiographic review was performed by two pediatric radiologists (TC and SD). Clinical information collected includes patient age at diagnosis, ethnicity, nutrition, stage of gross motor development, and medical history. Laboratory values collected include calcium, free ionized calcium, phosphate, alkaline phosphatase, 25-hydroxyvitamin D, 1-25-hydroxyvitamin D, and parathyroid hormone. Only radiographs obtained within 1 month of diagnosis were evaluated. They were reviewed for the presence or absence of osteopenia, presence or absence of fraying– cupping, and presence and characterization of fractures. Discrepancies between the two radiologists were rectified by consensus. Results The final number of children included was 45. Children included in the study were ages 2 to 24 months (mean age, 13 months). Only four children were younger than 7 months (Fig. 1). The majority of children (n=32) had nutritional rickets. The other types of rickets were as follows: congenital Nutritional Metabolic and secondary 13 months (+/− 6) (Range 2–24) 1/27 (3%) 13/27 (41%) 9/27 (28%) 3/27 (9%) 1/27 (3%) 12 4/8 1/8 0/8 3/8 0/8 months (+/− 6) (Range 4–23) (50%) (12.5%) (0%) (37.5%) (0%) 1186 Table 2 Laboratory values available for nutritional rickets patients (normal range provided in parentheses) a Normal levels for 1-25-OH Vitamin D levels are not available for infants and toddlers at this institution’s laboratory; therefore, values provided are for children ages 3–17 Pediatr Radiol (2010) 40:1184–1189 Test N Abnormal Mean SD Range Calcium (8.7–10.7 mg/dL) Ca++ (1.17–1.35 mmol/L) PO4 (3.9–6.5 mg/dL) Alk phos (95–380 U/L) 25-OH-vit-D (>30 ng/mL) 34 15 30 32 27 8.8 1.15 3.45 938 20.4 1.4 0.22 1.07 526 18.8 5.0–10.7 0.64–1.4 1.2–5.9 233–2,339 4.0–74 1-25-OH (27–71 pg/mL)a 23 99.2 60.0 23–231 PTH (9–59 pg/mL) 24 14 low (41%) 5 low (33%) 21 low (70%) 31 high (97%) 17 (< 20; 63%) 21 (< 30; 78%) 1 low (4%) 12 high (52%) 20 (83%) 263.7 185 9–699 metabolic disease (n=4; three cases were x-linked hypophosphatemic rickets, one case was partial vitamin D-resistant rickets), secondary (n=4; cases in this group were secondary to Fanconi syndrome, enteritis, hepatitis, and cystinosis), and unknown (n=2). An additional three patients were identified by an elevated alkaline phosphatase level but had no radiographic evidence of rickets. The two cases with unknown causes of rickets and the three cases with elevated alkaline phosphatase levels but no radiographic evidence of rickets were not included in the data analysis. However, none of them had fractures. Therefore 40 children were included in the data analysis. Ethnicity information was available for 35 children and is summarized in Table 1. The majority of nutritional rickets patients were either African American (41%) or were the children of recent African immigrants (28%), whereas half of the metabolic and secondary rickets patients were Caucasian (50%). Laboratory values are summarized in Table 2. The following percentages of children had these abnormal laboratory values: calcium (41%, ranging from 5.0 to 10.7 mg/dL), free ionized calcium (33%, from 0.64 to 1.4 mmol/L), phosphate (70%, from 1.2 to 5.9 mg/dL), alkaline phosphatase (97%, from 233 to 2,339 U/L), Fig. 2 Number of fractures per child 25-OH-vitamin D (78%, from 4.0 to 74 ng/mL), 1-25OH-vitamin D (57%, from 23 to 231 ng/mL), parathyroid hormone (83%, from 9 to 699 ng/mL). Of the 40 children with nutritional, congenital or secondary rickets, seven were found to have at least one fracture (17.5%). All seven children with fractures had nutritional rickets. Twenty-five-OH-Vitamin D levels were available for five of these seven children, and levels were abnormally low in all but one (the highest vitamin D level in the fractured group measured 44.0 ng/mL). Thirty-eight of the 40 children (95%) were judged to have radiological evidence of osteopenia. The seven children with fractures had obvious rachitic changes with widespread metaphyseal fraying/cupping. The number of fractures per child is represented (Fig. 2). The fractures observed were of three types: transverse diaphyseal, metaphyseal, and lateral or anterior-lateral rib fractures. The types of bones fractured and types of fractures all resemble structural insufficiency fractures and are summarized in Table 3. Eleven bone surveys allowed for 22 hand and foot films to be reviewed, revealing one metatarsal fracture. Examples of each fracture type can be seen (Fig. 3). The designation of metaphyseal fracture was given to long bones with profoundly irregular metaphyses. They appeared different from the classic metaphyseal lesion (CML). The usual location of CMLs, nearly adjacent to the physis, was lost in the rachitic fraying and cupping; the lower extremity CMLs we saw were located further toward the diaphysis. Their morphology indicated that either the metaphysis had collapsed under an axial load or had sheared off in a direction transverse to the diaphysis (Fig. 3). None of the lowerextremity metaphyseal fractures we saw had the appearances of solid “chips” of the distal medial or lateral cortex or “bucket handles” of the distal metaphysis. Only one child had proximal humeral metaphyseal fractures. The appearance of these was closer to the usual CML appearance (Fig. 3), suggesting they, too, might have resulted from limb traction, but these humeral lesions demonstrate a greater degree of irregularity and physeal widening, typical of rachitic changes, than is seen with the usual healing CMLs. Pediatr Radiol (2010) 40:1184–1189 1187 Table 3 Fracture types and locations. The number of radiographs available for the particular bone described in the entire group of rickets patients studied is provided # Fractures Rib Spine Pelvis Humerus Radius Ulna Femur Tibia Fibula Metatarsal 4 0 0 3 3 1 0 3 1 1 # Patients 2 n/a n/a 2 2 1 n/a 1 1 1 # of radiographs reviewed 23 19 25 49 51 51 73 59 59 22 The age distribution of the children with and without fractures is summarized in Table 4. Rickets patients without fractures spanned ages 2–24 months (mean 12.9 months; median 12.0; standard deviation 6.1). Rickets patients with fractures were ages 8 months to 19 months (mean 13.3 months; median 13; standard deviation 4.0). The seven fractured infants were all mobile (two crawling, four cruising, one walking). Discussion This study demonstrates that fractures occur in mobile infants and toddlers with radiographically overt nutritional rickets. Although a wide spectrum of severity of rickets was observed in the overall group of study patients, none of the fractures observed in the 40 children included was seen against a background of radiographically normal bone or bones with subtle rachitic abnormalities. Further, the fractures we observed resemble structural insufficiency fractures. None of the fractures observed included classic metaphyseal fractures, posterior medial rib fractures, or vertebral fractures. None of the children had skull fractures. Our data do not support the claim that multiple fractures occur in infants with radiographically subclinical vitamin D insufficiency or deficiency [1]. To the best of our knowledge, the supposition that multiple fractures suspicious for abuse in infants with normal-appearing bones might be a result of vitamin D insufficiency or deficiency rather than non-accidental trauma has not been supported. The findings we report here indicate that severe rickets predisposes to fractures in mobile infants and toddlers, but the fractures acquired do not resemble those of nonaccidental trauma. There are data that reduced mineralization is associated with easy fracturing in elderly people and older children Description of fracture Lateral (3); anterior-lateral (1) – – Proximal metaphyseal Distal metaphyseal (2); Distal transverse (1) Distal transverse – Bilateral distal metaphyseal (2); Proximal transverse (1) Proximal transverse Proximal transverse [5]. However, there are not comparable data that poor mineralization in infants results in easier fracturing. Likewise, there is no information about what degree of reduced mineralization might be required to increase fracture susceptibility. We have found that, in children with much more severe under-mineralization than the infants reported by Keller and Barnes [1], when fractures occurred, they involved older infants who had gained some independent mobility. The fractures we saw included primarily long bone fractures typical of normally active children and metaphyseal fractures caused by failure of load-bearing. We found a few rib fractures, but no skull fractures, no vertebral fractures, and no classic metaphyseal lesions (CML). In our study, we also did not find infants who sustained fractures in the first 6 months of life. Few of our infants had multiple fractures and none had fractures typical of common abusive fractures like CMLs or posterior-medial rib fractures. The fractures we saw were either present in the face of florid rachitic changes (the metaphyseal injuries) or were accompanied by moderate radiological abnormality and histories consistent with the fracture types and appropriate to their injury mechanics (diaphyseal injuries). As such, we are confident that radiologists familiar with the infant skeleton would not have difficulty recognizing the underlying metabolic bone disease. Our study is hampered by not identifying rachitic children in the same age range as the infants reported by Keller and Barnes [1]. We only identified four children younger than 7 months with rickets. However, our search strategy included searching for younger infants with elevated levels of alkaline phosphatase levels. In a review by Weisberg et al. [6] of 166 cases of literature-reported rickets, an elevated alkaline phosphatase level was present in 99%, while a low vitamin D level only identified 68%. As such, we should have recognized subjects with functional vitamin D insufficiency in early infancy, suffi- 1188 Pediatr Radiol (2010) 40:1184–1189 Fig. 3 Examples of fractures and healing seen in rickets patients. a PA and b lateral right wrist films of a 9-month-old with nutritional rickets presenting with distal radius and distal ulna transverse fractures. The bones are osteopenic with metaphyseal irregularity and flaring. c Chest radiograph of an 11-month-old with nutritional rickets shows anterior rib end widening (rachitic rosary), in addition to right anterior lateral rib fractures (arrows) and bilateral proximal humeral severe irregularity and disruption judged to represent metaphyseal fractures. d AP and e lateral right tibia/fibula films of a Table 4 Age of children with fractures Range Mean Median Std. deviation No fracture Fracture 2–24 months 12.9 12.0 6.1 8–19 months 13.3 13 4.0 15-month-old with nutritional rickets shows severe osteomalacia, a proximal right transverse fracture (arrow), and distal tibia and fibula metaphyseal fraying and cupping. The distal tibia was interpreted as fractured. The distortion at the distal femur was not interpreted as a fracture but as severe metaphyseal irregularity. Findings were symmetric (left side not shown). f Same child as shown in d and e, 1 year after treatment for vitamin D deficiency. Metaphyses are now well mineralized, but bowing remains cient to affect bone strength, if they were other than extremely rare. The assertion that a premobile infant’s fractures are secondary to the infant’s normal-appearing but fragile bones based on Vitamin D levels alone implies postulation of a new entity not related to the radiologic diagnosis of rickets. Congenital rickets has been reported, but all of those infants had overt radiological changes [7]. Our study included four cases of x-linked hypophosphatemic rickets, which is a congenital form of rickets; however, we did not observe fractures in this group of children. Pediatr Radiol (2010) 40:1184–1189 Like any proposed new disease entity, one of temporary brittle bones caused by vitamin D deficiency should be subjected to scientific scrutiny. In the absence of a good measurement of bone mineral density in infants, other useful investigations might include comparing surrogate markers for underlying bone abnormalities such as PTH, free calcium, alkaline phosphatase, and phosphate between infants with non-accidental trauma and those with accidental trauma, or infants with multiple injuries including head injuries versus infants with fractures alone. Conclusion This study demonstrates that fractures occur in mobile infants and toddlers with radiologically overt nutritional rickets. None of the fractures we observed was seen against a background of normal bone or mild rickets on radiograph. Further, the fractures observed were consistent with a failure of load-bearing and could be distinguished from the highly specific fractures seen in non-accidental trauma. 1189 This study does not support the assertion that low vitamin D levels are a significant cause of fractures in pre-mobile infants. References 1. Keller KA, Barnes PD (2008) Rickets vs. abuse: a national and international epidemic. Pediatr Radiol 38:1210–1216 2. Kleinman PK, Marks SC, Blackbourne B (1986) The metaphyseal lesion in abused infants: a radiologic–histopathologic study. AJR 146:895–905 3. Kleinman PK (1998) Diagnostic imaging of child abuse, 2nd edn. Mosby, St. Louis 4. Kleinman PK (2008) Problems in the diagnosis of metaphyseal fractures. Pediatr Radiol 38(Suppl 3):S388–394 5. Bianchi ML (2007) Osteoporosis in children and adolescents. Bone 41:486–495 6. Weisberg P, Scanlon KS, Li R et al (2004) Nutritional rickets among children in the United States: review of cases reported between 1986 and 2003. Am J Clin Nutr 80(suppl):1697S–1705S 7. Slovis TL, Chapman S (2008) Evaluating the data concerning vitamin D insufficiency/deficiency and child abuse. Pediatr Radiol 38:1221–1224