Achondroplasia natural history study (CLARITY): 60-year experience in orthopedic surgery from four skeletal dysplasia centers

One thousand three hundred and seventy-four subjects with achondroplasia, constituting the Primary Achondroplasia Cohort (PAC), comprise the CLARITY population, with a mean age at the last encounter of 15.4 ± 13.9 years and a median age at the last encounter of 11.9 years (5.9–19.7) (Table 1) [17]. Overall, 408 patients (29.7%) had at least one orthopedic surgery during their lifetime. At the time of the first procedure, the mean age was 13.7 ± 12.7 years with a median age of 9.9 years (0.1–62.7). One hundred seventy-five individuals underwent one or more spine procedures alone. Their mean age was 22.4 ± 15.3 years with a median age of 16.7 years (0.1–62.7). 291 individuals underwent one or more lower extremity procedures alone at a mean age of 9.9 ± 8.3 years and a median age of 8.2 years (0.2–57.8). Fifty-eight individuals had both a spinal and lower extremity procedure. In this group, the mean age for the second procedure was 13.8 ± 11.4 years with a median age of 12.8 years (1.1–57.8). The most common spinal procedure was decompression, while the most common lower extremity procedure was osteotomy.

To account for the age of the subject at the last observation, Kaplan Meier analysis of the 1365 patients with complete surgical history is shown in Fig. 1. At 10 years of age, 120 of 635 (19.0%) had either type of orthopedic procedure, with 103 of 635 (16.2%) being lower extremity procedures. At 20 years of age, this increased to 41.8% (87 of 212) patients with the majority of these procedures (68 of 212; 32%) involving the lower extremities. After the age of 20, further increases were predominately spinal procedures, and by 60 years of age, 85.3% of 8 patients who were known to reach that age had an orthopedic procedure with a shift to spine procedures being more common.

One hundred and seventy-five patients (12.7%) underwent 425 spine procedures (Table 1). Spine procedures were classified as fusion and/or laminectomy of the cervical (below C2), thoracic or lumbosacral levels. Laminectomy of C1 and C2 are excluded here but have been published [29]. The most common spine surgery was decompression with 152 patients undergoing 271 laminectomy procedures in the cervical, thoracic, and lumbosacral spine. Lumbar laminectomies were most frequent with L3 being the most common level. Thoracic and cervical regions followed in frequency, with T11 and T12 being the most common thoracic levels. The mean age for all laminectomies was 25.5 ± 15.5 years with a median age of 20.6 (0.1–67.6). One hundred and six patients underwent 142 fusion procedures with a mean age of 21.0 ± 15.1 years and a median age of 15.6 years (0.8–65.8). Laminectomy was performed in patients with spinal stenosis, and fusion was performed in those patients who had a risk of progressive kyphosis after decompression, as determined by the surgeon. Most surgeries classified as ‘other’ spine procedures are implant/hardware revision or removal. Spine surgery continued throughout the lifespan (Fig. 2).

Two hundred and ninety-one patients (21.2%) underwent 684 lower extremity procedures (Table 1). Lower extremity surgical procedures included osteotomy, guided growth, epiphysiodesis, fibular shortening, limb lengthening, and total hip arthroplasty. The most common lower extremity surgery was osteotomy with 200 patients undergoing 434 lower extremity osteotomies. The mean age for osteotomy was 9.1 ± 5.8 years with a median age of 7.9 years (1.1–36.9). Twenty-three patients underwent 27 guided growth tension band plating. The mean age for guided growth tension band plating was 8.6 ± 2.9 years with a median age of 8.9 years (4.0–12.3). Of the 27 procedures, none were performed in the ‘before 1980s’ birth cohort. One was performed in the 1980s birth cohort with the numbers increasing to 6, 16 and 4 respectively in the 1990s, 2000s and 2010s birth cohorts. The other category of lower extremity procedures included arthroscopic discoid meniscus saucerization, fixator removal, and arthrograms. All patients underwent lower extremity surgery before 60 years of age (Fig. 1).

Risk factors associated with having any spine, lower extremity or both types of surgery are shown in Table 2. There was no difference in the odds of these surgeries given the sex of the subject. There was a 1.85 times greater odds of spine surgery in this cohort if a cervicomedullary decompression (CMD) had been performed previously (p =  < 0.01) but there was no increased risk for any lower extremity surgery (p = 0.8). Hydrocephalus requiring shunting increased the risk for both a spine surgery (OR 1.97; p = 0.011), and for both a spine and lower extremity procedure (OR 1.94; p =  < 0.01), but not a lower extremity procedure alone (OR1.46; p = 0.118). Of all the risk factors evaluated, a history of both shunting and CMD was associated with the highest risk for subsequent additional spinal surgery (OR 2.268; p = 0.017). The risk for a lower extremity procedure (1.93, p = 0.031) or both types of surgery (2.38, p < 0.01) were also increased. The use of CPAP to treat obstructive sleep apnea (OSA) increased the odds of any orthopedic surgery combined (OR 1.5, p = 0.02), but not individually for spine procedures or lower extremity procedures. If an individual underwent a surgery in the spine or lower extremity, the odds of them needing a subsequent procedure in the other domain was increased 2.05 times (p =  < 0.01).

The relationship between patient height, weight, and occipital frontal circumference (OFC) and the need for orthopedic surgery was also evaluated (Table 2). Modest statistically significant increases in the odds ratios (OR 1.03, p =  < 0.001 and OR 1.01, p = 0.027) were found for increasing height related to spine or spine and lower extremity surgery, respectively (the taller the more likely). No associated risk was seen between height and lower extremity surgery. Weight also showed a modest significant increase in the odds ratios for spine or spine and lower extremity surgery, respectively (OR 1.04, p =  < 0.001 and OR 1.01, p = 0.034). Increasing weight however, slightly decreased the likelihood of requiring a lower extremity surgery (OR 0.98, p = 0.023). In this group of risk factors, OFC increased the risk for spine surgery (OR 1.06, p =  < 0.001) but not for lower extremity procedures or both lower extremity and spine surgery.

CLARITY is the largest study reporting the results of a multicenter historical cohort of achondroplasia. This manuscript focuses on the orthopedic aspects of achondroplasia within the PAC. Orthopedic surgery was a common event in the care of achondroplasia with 408 patients (29.7%) undergoing at least one orthopedic intervention during their lifetime at a mean age of 13.7 years. This information is valuable for practicing orthopedic surgeons, as the new American Academy of Pediatrics care guidelines recommend orthopedic referrals for all children with achondroplasia [30]. Our data present the most detailed natural history to date, and should be used to facilitate improved care.

In the next largest study to report orthopedic procedures in patient with achondroplasia, Hunter et al. presented data on 193 patients with achondroplasia from one US center and five other centers across the world [16]. They reported that 21.6% of patients 20 years of age and older had undergone tibial osteotomy, similar to the frequency reported in our study. However, the frequency of spine surgery for stenosis in the prior study was 24.1% in patients 40 years of age and older. This rate of spine surgery was nearly twice the overall rate reported in our study (12.7%). In part, this difference may be due to a difference in the age distribution between the studies since our study population was younger, with only 7% of patients older than 40 years of age at the time of their last clinical contact.

However, in our subjects who reached 40 years of age, 39.6% had undergone a spinal surgery, which is 1.6 fold higher than the Hunter study.

For the purposes of this analysis, we only included in this category, subjects who underwent CMD and shunt placement of different dates. This is due to the fact that during the 1980s and 1990s these procedures were often performed simultaneously. The subjects who underwent both procedures on the same date were included in the CMD group, as that was the primary surgical indication. The single largest risk factor associated with spinal, a lower extremity or or spinal and lower extremity surgery was previously having both a CMD for foramen magnum stenosis and a shunt placed for the treatment of hydrocephalus. The odds ratios were 2.26 (p =  < 0.017)1.93 (p-0.031) and 2.38 (p =  < 0.01), respectively.

In patients that had undergone a shunt placement for the treatment of hydrocephalus, the odds of requiring a spinal surgery, or both a spine and lower extremity procedure was increased, but not as great as when a shunt and CMD were performed. In patients that had undergone CMD alone, only the risk of spine surgery was increased (1.85, p =  < 0.01). The mechanism for narrowing of the foramen magnum and jugular foramina in FGFR3-related disorders is likely similar to that in the rest of the spine and may explain the increased odds of spinal decompression surgery in those with a prior shunt in the PAC [11, 19, 31]. These findings support the previous observations of the association between spinal stenosis and CMD in a much smaller achondroplasia population [11]. In that study of 44 patients undergoing surgery for spinal stenosis, Sciubba et al. found that more than 61% (27/44) also had CMD. Of these patients, 93% (25/27) had the CMD first and then additional spinal decompression later in life [11]. Although the proportion of patients in CLARITY with laminectomy who also underwent CMD was less (30.9%, 47/152), most patients had the CMD prior to their laminectomy (87.2%, 41/47).

OSA requiring CPAP was a risk factor for both types of orthopedic surgery (OR 1.50, 95% CI 1.06–2.12) but not spine or lower extremity surgery individually. OSA, in part, is the manifestation of abnormal cartilaginous development of the midface structures, but also associated with other factors such as increased weight. These pathologic processes also occur in the axial and appendicular spine and may interact with weight to accountfor this association.Future studies will be necessary to clarify this overall association and the temporaliy (i.e. did the requirement precede or follow the surgical procedures)..

Compared to the general population, patients with achondroplasia undergo spine surgery much more frequently. In an analysis of the Nationwide Inpatient Sample of average stature individuals, the rate of lumbar fusion, thoracic fusion, and cervical fusion was estimated at 69.1, 7.9 and 51.9 per 100,000 patients [32]. In contrast, 7,700 per 100,000 patients (7.7%) with achondroplasia in our cohort underwent spine fusion which is over 100 times more common than the general population. This finding is likely related to the altered pedicular anatomy in achondroplasia, which predisposes to symptomatic spinal stenosis [18]. In average statured patients, stenosis remains a common indication for surgery but is more likely caused by other factors, including spondylosis [33]. Although spinal stenosis is treated with decompression, fusion is frequently necessary in patients with achondroplasia to prevent or address additional sagittal spine deformity [34].

The second most common indication for spine surgery was TLK (n = 34). TLK is present in most infants and young children with achondroplasia and usually resolves spontaneously with standing and walking (20). In those individuals who needed spine surgery for persistent TLK, the spine surgery occurred at a younger age (13.2 years) as compared to those who had all other indications (22.7 years). Unfortunately, CLARITY was unable to provide data regarding progression of TLK or the impact of bracing.

Laminectomies were most commonly performed in the lumbar spine, most frequently at the L3 level. These findings are consistent with previous study of spinal stenosis in achondroplasia, which reported that the L2-3 level was most commonly decompressed [35]. In a study of pediatric patients, Scubbia et al. found that the most common level for decompression was at the thoracolumbar region (65%) followed by lumbar (20%) spine [11]. In our study, a significant number of laminectomies were performed in the cervical and thoracic spine, highlighting the importance of looking for stenosis beyond the lumbar spine and consideration of more than one focal area of stenosis in the same patient.

Genu varum is common among patients with achondroplasia and is related to a complex combination of multiple factors including lateral, dynamic instability of the knee; distal femur, proximal and distal tibial varus; internal tibial torsion; fibular overgrowth and tibial recurvatum [5, 12, 25]. It is frequently corrected with surgery [16]. Among patients with achondroplasia, lower extremity deformity involves the coronal, sagittal and transverse planes [12]. Historically, patients with genu varum were treated with corrective lower extremity osteotomies [36]. Although osteotomies are generally effective for addressing genu varum, these procedures are invasive. More recently, less invasive guided growth techniques utilizing tension band plates emerged to correct lower extremity deformity [37]. In a retrospective review of tension band plates among patients with skeletal dysplasia, Yilmaz et al. demonstrated correction in 34 of 38 valgus knees and 7 of 12 varus knees and concluded that this procedure was relatively safe and effective, even in young patients [38]. In another case series, McClure et al. reported the use of guided growth techniques in four patients with achondroplasia [28]. They found improvement in alignment in all patients; however, one patient did require subsequent osteotomies. They concluded that guided growth should be initiated at a younger age compared to patients without achondroplasia. Because guided growth was utilized in patients with skeletal dysplasia more recently, our data included only 23 patients undergoing 27 guided growth procedures. In our follow up of patients undergoing tension band plating for guided growth, none had a subsequent osteotomy. We anticipate that guided growth will be used with greater frequency, and future analysis should determine the impact of guided growth on the timing and frequency of lower extremity osteotomy as well as long-term physical function. Nevertheless, internal tibial torsion, which is common in achondroplasia, is difficult to address with guided growth, and guided growth cannot be used in patients with closed physes. In these situations, osteotomy will be necessary.

Lower extremity procedures were performed at a younger age in four different geographic locations. Similarly, lower extremity osteotomies, the most common lower extremity surgery, was also performed at a younger age in all four centers. Interestingly, a smaller second peak in lower extremity procedures was detected between 40 and 60 years of age at three centers. One of the four centers included in this study was a pediatric hospital; therefore, a second peak in lower extremity procedures was not expected at this particular center. Lower extremity osteotomies were predominantly performed at a younger age and did not contribute to the second peak in the bimodal distribution of lower extremity procedures. This analysis did not try to determine which type or types of fixation were used following osteotomies.

The second peak in lower extremity procedures between 40 and 60 years of age was related to surgery performed in nine patients. The Kaplan Meier analysis revealed this second peak due to the relatively small number of patients in this age group: the cohort born prior to 1980 had 17% (234/1374) of the entire sample. Although three of these patients in the 40-to-60-year age group received total hip arthroplasty and one patient received total knee arthroplasty, the overall frequency of hip and knee arthroplasty in the achondroplasia cohort was much smaller than the general population. In an analysis of the Healthcare Cost and Utilization Project State Inpatient Databases and the National Hospital Discharge Survey of the general population, Kremers et al. found a prevalence of 0.83% for total hip arthroplasty and 1.52% for total knee arthroplasty in the United States [39]. Once again, this finding may relate to the relatively young ages of patients within our cohort. This finding may also reflect the smaller mechanical forces at the hip and knee in patients with achondroplasia. In addition, mouse studies demonstrate a protective mechanism against osteoarthritis for the FGFR3 mutation [40, 41].

The consistency of the timing of lower extremity procedures, including lower extremity osteotomies, across centers in different parts of the country stands in contrast to previous study of orthopedic procedures demonstrating geographic variation. To examine geographic trends in revision and primary hip and knee arthroplasty, Hilibrand et al. utilized the Nationwide Inpatient Sample and found a 2.2-fold variation and 2.1-fold variation in the revision rate ratio by state for revision total knee arthroplasty and revision total hip arthroplasty [42]. The authors concluded that significant variation exists in the performance of revision total knee and hip arthroplasty procedures from state to state. In another study examining primary total hip and knee arthroplasty performed in England, Judge et al. found that significant geographic variation existed across districts, controlling for distance measures [43]. Demographic variables contributed to variation in some districts while other districts were not influence by these factors. They concluded that the geographic variation in joint replacement surgery needs to be further delineated to facilitate access for all patients.

The consistency in the timing of lower extremity surgery in patients with achondroplasia may be due to the small number of specialists performing these procedures. In this relatively small dysplasia community, consensus likely exists in performing lower extremity surgery, particularly lower extremity osteotomies, in younger patients. In contrast, arthroplasty surgery is performed by a large number of surgeons who may have diverging views on indications and timing of surgery.

The highest incidence of lower extremity procedures, including lower extremity osteotomies, was performed in the cohort born between 1990 and 1999. Interestingly, the oldest patients born prior to 1980 had the lowest incidence of lower extremity procedures, suggesting that lower extremity interventions were performed more frequently in the last three decades with an increase from 14.21 surgeries per 1,000 person years in the cohort born prior to 1980 to 69.96 surgeries per 1,000 person years in the cohort born between 1990 and 1999.

The increase in the number of lower extremity procedures performed more recently in patients with achondroplasia is consistent with a general trend in orthopedics for overall increase in number of procedures. In a study of arthroscopic knee surgery performed in England between 1997 and 2017, Abram et al. found that the incidence of arthroscopic partial meniscectomy increased from 51/100,000 in 1997–1998 to 120/100,000 in 2016–2017 [44]. In addition, they found that the incidence of arthroscopic chondroplasty increased from 3.2/100,000 in 1997–1998 to 51/100,000 in 2016–2017. The trend toward more frequent procedures in orthopedics is predicted to continue to increase. In a study of the number of hip arthroscopies performed by the National Health Service in England, Palmer et al. predicted a 1388% increase in the number of hip arthroscopies performed in 2023 as compared to 2002 [45].

Risk factors for repeat spine surgery included increased weight and height. The theme of increased patient size leading to worse outcomes and/or repeat surgical intervention is prevalent in orthopedic surgery [46-49]. In a study of patients undergoing orthopedic trauma procedures, the complication rate in obese patients was 38% compared to 28% (p = 0.03) in non-obese patients [50]. The authors concluded that obese orthopedic trauma patients are at higher risk for in hospital complications with further study required to optimize results. In another analysis of failed total hip arthroplasty, Goodnough et al. found that the rate of aseptic loosening leading to failure of primary total hip arthroplasty was 30% in obese patients and 18% in non-obese patients [51]. The authors found a similar increase in infected failed total hip arthroplasty among obese patients.

This analysis demonstrates consistency in the timing of lower extremity procedures across all four centers and a trend towards more frequent lower extremity procedures in recent decades. Finally, malalignment is a consistent indication for surgery for first time as well as repeat lower extremity surgery. Future study focusing on decreasing the number of repeat lower extremity surgery should be performed. Nevertheless, several meaningful observations are drawn in this study for orthopedic surgeons managing the complex lower extremity deformities observed in patients with achondroplasia.

The major limitation of this study is its clinic-based and retrospective nature. The number of individuals is skewed very heavily towards the younger decades. The data which could be extracted into the database is limited by what was available in the charts reviewed. Specific details such as the precise indications and outcomes were limited by what was documented and there was no specific predefined criteria. The main strengths of our study findings are the size of this natural history cohort, the follow-up of over four decades, and the uniform data collection by medical providers with extensive achondroplasia care experience.