Re-defining the classification of heterotopic ossification of the elbow: A retrospective study

doi:10.21203/rs.3.rs-1444306/v1

Download PDF

Research Article

Re-defining the classification of heterotopic ossification of the elbow: A retrospective study

https://doi.org/10.21203/rs.3.rs-1444306/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

None of the existing classifications of heterotopic ossification (HO) of the elbow is based on anatomy, therefore they cannot accurately guide or predict excision. This study was to evaluate the anatomical morphology of HO of the elbow resulting from different etiologies and to propose a new HO classification system based on three-dimensional (3D) mapping technology.

Methods

We retrospectively included 76 cases of HO of the elbow from 2014 to 2021. All their demographic and etiologic information were collected and imported into medical imaging software. The distribution of the HO of the elbows and the 3D heat map was then created to evaluate the location and pattern of the HO. The volume of each HO based on mapping model and the length of the HO were measured to better analyze the characteristics of different types of HO.

Results

From the aggregated 3D image of the 76 HO cases, 85.53% have affected the dorsal side of the elbow, especially on the humeroulnar side (81.58%). And 26.32% have involved the anterior part of elbow, few cases were distributed among the radial head (10.53%) and olecranon fossa (1.32%). Subsequently, in accordance with the anatomical and morphological characteristics, a new HO classification of the elbow was proposed. The kappa score for inter- and intra-observer agreement of this classification system was 0.86 (95% CI, 0.72 -1.00) and 0.90 (95% CI, 0.78 -1.00), respectively.

Conclusion

This new HO classification system of the elbow may help surgeons to speculate the possible pathological anatomy and to design the best preoperative plan for excision.

heterotopic ossification

classification

pattern

three-dimensional CT

elbow

fracture

Heterotopic ossification (HO) refers to pathological bone formation in soft tissues that develops after trauma such as bone fracture and dislocation, joint replacement, muscle injury, traumatic brain injury, spinal cord injury, combat injury and explosion injury. HO is associated with localized pain and limited joint mobility[1]. Elbow is the most common joint that for HO formation. Its prevalence is relatively high[2], with the radiologic incidence of 37% among surgically treated elbow fractures[3]. Over 20% of HO patients report varying degrees of dysfunction such as chronic pain, skin ulcers, vascular nerve compression, joint contractures and stiffness[4]. However, the specific pathogenesis, mechanisms, morphology, and the distribution of trauma induced HO remains poorly understood. For instance, neurogenic HO, surgical HO, burn HO, and fracture-dislocation may not share a common etiology. Furthermore, there is a paucity of studies reporting the characteristics of HO of the elbow including its specific spatial distribution and morphology.

Although there are several classification systems for HO of the elbow, such as the Ilahi, Hastings and the Graham system, and the Brooker classification[5–7], none of them is based on anatomic distribution; hence they cannot improve surgical planning or guide excision operation.

Fracture mapping was initially used to describe the morphology of bone structures[8], which can reveal the frequency of fracture lines and the comminution zones. Several studies have utilized these techniques for characterizing the different types of fractures[9, 10], however, no studies have been published on its application in measuring HO of the elbow joint. Through 3D reconstruction of CT images and the use of fracture mapping analysis technology, the anatomical and morphological characteristics of HO can be accurately described leading to accurate HO classification and treatment. In this study, we set out to use to establish the anatomical morphology of HO of the elbow using a 3D mapping technology and to propose a new classification that will guide surgical resection of HO.

Subjects

In this retrospective study, a total of 82 patients with clinically symptomatic elbow HO were treated from 2014 to 2021 in [removed for blind peer review]. We excluded those younger than 18 years of age (5 patients) and those without CT scan reports (2 patients). Finally, we enrolled 75 patients (28 female, 47 male) with a mean age of 42 years (range, 18–57 years). Among which, 36 patients had right HO, 40 with left HO, and 1 patient had bilateral HO of the elbow. All patients had no CT artifacts. And all patients signed an informed consent before entering the study.

Ho Mapping Models

All patients had their CT scans performed using a standardized protocol in our hospital, and the DICOM (Digital Imaging and Communications in Medicine) raw data of the elbow imported into Mimics Software (Materialise NV, Leuven, Belgium). And then we divided each model into 4 parts: humerus, radius, ulnar and HO. Specifically, HO was divided as per the difference of bone mineral density (BMD) between ossification and normal bone tissue. The angle of longitudinal axis of the humerus and the forearm of each HO elbow was adjusted to 135°, based on a normal standard elbow model. We mirrored the left elbow models to the right and matched each model with the normal standard elbow model. We imported all the adjusted models into the E3D medical imaging software (E3D, Central South University, Changsha, China) and evaluated the pattern and location of the HO. We aggregated all 76 HO 3D images and stacked them on a normal elbow template, then we marked each HO part with a translucent red color, so that overlapping parts was conspicuous (Fig. 1). Subsequently, we created 3D heat maps for the relative frequency of HO formation, in which blue indicated lower incidence and red indicated higher incidence (Fig. 2).

To better analyze the characteristics of different types of HO, we measured the volume of each HO based on mapping model and the length (L) of the HO when it is on the posterior part of the humerus. Specifically, we measured the length (L) of HO by drawing a perpendicular line on the dorsal side of the axis of the humerus and choose the intersection of this vertical line and HO as a; subsequently, we chose the intersection point of the ulnar shaft axis and olecranon cortex as b (olecranon apex). Finally, we defined the distance between ab as the length (L) of the HO (Fig. 3).

Statistical analysis

To validate our classification system, two independent orthopedic surgeons and one radiologist who were blinded to the surgical procedures evaluated all the HO CT scan data and classified them into different types. Inter and intra-observer variations were assessed by calculating the kappa correlation coefficient.

A total of 75 patients with 76 HO elbows were enrolled (28 women, 47 men), including 40 (52.63%) left elbows, 36 (47.37%) right elbows. From the aggregated 3D image (Fig. 1), the most overlapping parts were the dorsal side of the elbow (65 cases, 85.53%), especially on the humeroulnar side (62 cases, 81.58%), which is the most common site of HO. 20 cases (26.32%) affected anterior part of elbow, and few cases were distributed among the radial head (8 cases, 10.53%) and olecranon fossa (1 case, 1.32%). On 3D heat mapping (Figs. 2), correspondingly, the hot zones were mainly concentrated around the dorsal medial condyle of the humerus and olecranon apex, and the cold zones were scattered in the anterior and the radial side of the elbow.

Thus, in accordance with the anatomical and morphological characteristics of HO, we propose a new HO classification of the elbow by means of 3D mapping technology. These include: anterior ossification (AO), posterior ossification (PO), peri-radial head ossification (PRHO), mixed pattern ossification (MO), and circumferential ossification (CO) (Table 1, Fig. 4). Each HO pattern was further subdivided according to its specific anatomical location (Table 2). The difference between MO and CO was that MO affected not only the anterior and posterior side of elbow, but also lateral and/or medial side of elbow (i.e., encircling the elbow).

Table 1

Anatomical classification based on 3D-CT reconstructed HO of elbow
Classification of HO	N (%)	Volume (mm³)	Surface (mm²)	Length (mm)
Anterior ossification (AO)	4 (5.26)	6078.61	3654.45	-
Posterior ossification (PO)	45 (59.21)	7522.43	3728.50	28.70
Peri-radial head ossification (PRHO)	4 (5.26)	7622.49	3602.13	-
Mixed pattern ossification (MO)	16 (21.06)	19766.54	8922.08	40.87
Circumferential ossification (CO)	7 (9.21)	9994.94	9027.70	13.85

Table 2

Specific anatomical location of each HO classification
Classification of HO		Specific location of HO	N	Volume (mm³)	Surface (mm²)	Length (mm)
AO		Anterior humerus	3	4233.51	2853.85	-
AO		Anterior humerus and ulnar	1	11613.90	6056.25	-
PO		Posterior ulnar	23	4657.01	2806.66	26.44
		Posterior radius	1	963.76	738.56	0
		Posterior ulnar and radius	7	2909.10	2227.74	21.43
		Posterior ulnar and middle	2	7201.07	4260.37	27.7
		Posterior middle	2	9316.26	4643.75	48.53
		Posterior full	10	17703.58	6908.82	41.50
PRHO		Around radial head	4	7622.49	3602.13	-
MO	AO + PO	Anterior radius + Posterior ulnar	1	5380.78	4034.55	37.26
		Anterior ulnar + Posterior ulnar	2	5193.41	10291.45	32.32
		Anterior ulnar + Posterior full	3	7593.99	15789.79	55.14
		Anterior humerus + Posterior ulnar and middle	3	2559.99	3112.21	15.94
		Anterior humerus + Posterior full	3	7633.72	13472.47	30.16
		Olecranon fossa + Coronoid fossa	1	3826.09	3754.96	23.61
	PO + PRHO	Posterior full + Peri-radial head	3	20779.73	57272.87	69.21
CO		Circumferential	7	9127.05	8646.50	15.07

On the etiology of HO, 46 patients had different types of fractures, 7 patients were injured by simple elbow dislocation, 11 patients had explosive burn, 5 patients had crushing injury of the elbow, and 4 patients (5 elbows) had a craniocerebral injury (Table 3). Among these 46 fracture-inducing patients, the specific pathogenesis were quite different. In patients with the PRHO, radial head was always injured. For instance, both the Monteggia fracture (2 cases) and radial head fracture (2 cases) had a varying degree of radial head injuries. CO was strongly associated with elbow dislocation. For example, among the 7 cases of CO, 5 had elbow dislocation, and the other 2 had coronoid fracture and humeral medial epicondyle fracture. The volume of HO induced by craniocerebral injury was the highest among all pathogenesis (26302.09 mm³).

Table 3

Etiology of different patterns of HO
Damage reason		AO	PO	PRHO	CO	MO	Total (%)	Volume (mm³)		Surface (mm²)
Fractures	Anterior transolecranon fracture-dislocations	1	2		1		46 (60.53)	4873.83	8908.64	3516.99	4447.90
	Avulsion fracture of lateral collateral ligament		1					963.76		738.56
	Coronoid fracture		2		1			2375.57		2196.90
	Distal humerus fracture		3			1		4922.94		2840.90
	Humeral medial epicondyle fracture		2		1	3		23958.24		10185.27
	Humeral medial epicondyle fracture-dislocations					1		10291.45		5193.41
	Monteggia fracture		3	2				15667.16		6248.78
	Olecranon fracture		5					9941.64		4914.77
	Radial head fracture		4	2
	Supracondylar fracture	2						7978.72		4083.40
	Terrible elbow triad		7		1	1		4538.07		2555.11
Simple elbow dislocation		1	2		3	3	9 (11.84)	7732.44		5791.37
Explosion burn			8			3	11 (14.47)	7961.88		4357.44
Craniocerebral injury			1			4	5 (65.79)	26302.09		12508.92
Crushing injury of the elbow			5				5 (6.58)	7261.07		4111.77
AO, anterior ossification; PO, posterior ossification; PRHO, peri-radial head ossification; CO, circumferential ossification; MO, mixed pattern ossification.

Compared to previous studies, we found that most of our patients (97.37%) did not have an olecranon fossa etiology. Olecranon fossa was rarely involved no matter the type of underlying HO etiology (Fig. 5). However, only two cases had an olecranon fossa involvement: one with olecranon and coronoid fossa ossification due to burn, and another one with posterior middle ossification resulting from distal humeral fracture.

The kappa score and the related 95% confidence interval (CI) for inter-observer agreement was 0.86 (0.72-1.00), whereas those for intra-observer agreement was 0.90 (95% CI, 0.78-1.00). The highest percentage of differences in the new classification system was for circumferential and mixed patterns HO.

In previous studies, the most commonly reported site of HO is the posterolateral aspect of the elbow, where a bridge of bone typically spans from the lateral humeral condyle to the posterolateral olecranon process, filling the olecranon fossa[1, 6]. However, in our study, the olecranon fossa was not involved in most of cases (97.37%), making the posteromedial aspect of the elbow the most frequent HO site, specifically along the triceps. Similarly, Foruria et al.[3] reported that ectopic bone was particularly common around the posterior aspect of the ulna and the neck of the radius, yet in our study we found many cases developing on the ulnar side of the distal humerus.

This may be due to recent advancement in imaging technology as compared to the data used in previous studies since it was mainly based on plain radiographs with poor anatomical delineation as compared to 3D-CT. Compared to plain radiographs, CT scans can detect the subtle focal mineralization with superior spatial and contrast resolution. On the other hand, the 3D-CT images can be viewed at a 360° without overlapping on the radiograph[11], therefore, HO grade classification based on CT imaging is more accurate. To our knowledge, this is the first study that utilized 3D fracture mapping technique to detect the morphology and the anatomy of HO.

Another plausible reason on the higher incidence of HO on the ulnar side of the distal humerus could be the differences in anatomical characteristics of the ulnar and radial sides. On the ulnar side, there are no obvious muscle attachment in the ulnar nerve groove, making it suitable for the development of HO. As commonly seen during surgery, HO rarely invloves the ulnar nerve, instead it grows in the groove thereby displacing the ulnar nerve. On the radial side, however, the anconeus muscle is attached to the dorsal side of the humeral head, thereby depriving viable space for HO formation unless fracture or dislocation occurs in this area.

There was no HO involving the olecranon fossa in our study, and we hypothesize that it may due to the presence of a fat pad in the olecranon which acts as a barrier between the bone and the muscles. But when we reviewed the cases of double plate fixation of comminuted intercondylar fracture of the humerus, in which fat pad in the olecranon fossa was missing, we found no obvious HO formation in their olecranon fossa, either. In addition, during HO excision, we found that HO was not directly involving the muscles nor the tendons, instead, it was always adhered to the bony surface. Thus, our hypothesis is that HO formation could be induced by the interaction between muscles and bone. Furthermore, due to elbow instability and dislocation, involved muscles and tendons are passively and forcibly stretched leading to re-injuries, thereby causing HO formation. Intriguingly, PRHO secondary to Monteggia fractures is rare in children[12], but was paradoxically prevalent in our adult cohort manifesting as radial head fracture with or without dislocation. Moreover, the volume of HO induced by craniocerebral injury was far greater than the volume induced by fractures, therefore further studies should be carried out to detect the true incidence of HO in relation to the different underlying etiologies.

Kamineni et al.[13, 14] reported the absolute safe zone for pin placement into the lateral distal humerus to be the area lying within the 70% of the caudad line, equivalent in length to the patient’s trans-epicondylar distance (57.47mm). In our study, the average and maximum length of HO was 28.40mm and 69.21mm, respectively, thus avoiding radial nerve injury during HO excision is very important.

Previous traditional classifications of HO were unbale to delineate the anatomic distribution and provide guidance for excision, due to various reasons. The Ilahi classification of HO includes four grades of HO, with grade I indicating the largest with an angle of < 30° as measured from the center of the capitellum; grade II 30°-60°; grade III an angle > 60° without bony ankylosis, and grade IV indicating frank bony ankyloses (humeroulnar or humeroradial) on any radiographic view[5]. This grading system is only useful in identifying anterior HO of the elbow, whereas in our study we found HO is mostly located posteriorly. As for the Hastings and the Graham HO classification, grade I indicates ectopic bone without functional limitation, grade II includes HO patients with limited range of motion which is further subdivided into three categories depending on different affected planes of motion, and grade III includes patients with both HO and ankylosis[6]. This classification system is mainly based on the degree of functional limitation, which has minimal association with anatomical characteristics and does not guide surgical approach. The widespread used Brooker classification is mainly used to stratify the degree of hip HO and not that of the elbow[7]. Another HO classification system of the elbow was defined according to the anatomic location and severity such as the degree of articular damage, range of motion and clinical and radiographic characteristics[3]. But the inter-and intra-observer reliability of this system is high, with minimal flexibility in interpretation.

Our new classification system which is based on the anatomic distribution of HO, can guide surgical approach and excision with high precision. The surgical approach of PO is relatively simple, and it involves HO excision by a posterior approach. As for the CO, single surgical approach, without considering the joint mobility, may not be effective. After the excision of the HO, ligament reconstruction needs to be performed if the elbow is unstable, and an internal joint stabilizer or articulated hinge external fixation of the elbow is needed for temporary protection. In addition, measurements of length, volume and surface area of HO are also helpful for precise excision and avoidance of neurovascular injuries.

Limitations of this study include relatively small sample size that could lead to higher variability and bias. Also, since the patients we retrospectively included in our study were severely injured and needed surgical intervention, there could be a selection bias by missing to enroll patients with mild HO who did not require surgery, therefore, our findings should be interpreted with caution.

Based on the 3D-CT technology, anatomical and morphological characteristics of HO can be accurately described. Our new HO classification system of the elbow may help surgeons to speculate the possible pathological anatomy and to design the best preoperative plan for excision.

anterior ossification

circumferential ossification

computed tomography

heterotopic ossification

mixed pattern ossification

posterior ossification

PRHO

peri-radial head ossification

three-dimensional.

Ethics approval and consent to participate

The research was approved by the Ethics Committee on Research of Xiangya Hospital prior to implementation (No.2019030383).

Consent for publication

Not applicable.

Availability of data and materials

All the data are available if qualified authors apply for them.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding

This study was supported by grants from the National Natural Science Foundation of China (No. 81902277, 81874006), the Natural Science Foundation of Hunan Province (CN) (No. 2017JJ3477), and the Xiangya Clinical Big Data Project(xyyydsj9).

Authors’ Contributions

Conceptualization, L.C. and H.W.; methodology, H.W. and R.Z.; software, L.C. and H.T.; validation, Y.Z. and Z.L.; writing-original draft preparation, H.W.; writing-review and editing, L.C. All authors have read and agreed to the published version of the manuscript.

Acknowledgments

The author thank Prof. Shenghui Liao’s team for their technical support of E-3D software. And we would like to thank all patients for participating in this study. And we would like to thank Dr. Omar Abdihamid for proofreading and editing our paper.

Casavant AM, Hastings H: Heterotopic Ossification about the Elbow: A Therapist's Guide to Evaluation and Management. Journal of Hand Therapy 2006, 19(2):255–267. https://doi.org/10.1197/j.jht.2006.02.009
McAfee PC, Cunningham BW, Devine J, Williams E, Yu-Yahiro J: Classification of Heterotopic Ossification (HO) in Artificial Disk Replacement. Journal of Spinal Disorders & Techniques 2003, 16(4):384–389. https://doi.org/10.1097/00024720-200308000-00010
Foruria AM, Augustin S, Morrey BF, Sanchez-Sotelo J: Heterotopic ossification after surgery for fractures and fracture-dislocations involving the proximal aspect of the radius or ulna. J Bone Joint Surg Am 2013, 95(10):e66. https://doi.org/10.2106/JBJS.K.015331684782
Ranganathan K, Loder S, Agarwal S, Wong VW, Forsberg J, Davis TA, Wang S, James AW, Levi B: Heterotopic Ossification: Basic-Science Principles and Clinical Correlates. Journal of Bone and Joint Surgery 2015, 97(13):1101–1111. https://doi.org/10.2106/jbjs.N.01056
Ilahi OA, Bennett JB, Gabel GT, Mehlhoff TL, Kohl HW, 3rd: Classification of heterotopic ossification about the elbow. Orthopedics 2001, 24(11):1075–1077; discussion 1077–1078.
Hastings H, 2nd, Graham TJ: The classification and treatment of heterotopic ossification about the elbow and forearm. Hand Clin 1994, 10(3):417–437.
Brooker AF, Bowerman JW, Robinson RA, Riley LH, Jr.: Ectopic ossification following total hip replacement. Incidence and a method of classification. J Bone Joint Surg Am 1973, 55(8):1629–1632.
Armitage BM, Wijdicks CA, Tarkin IS, Schroder LK, Marek DJ, Zlowodzki M, Cole PA: Mapping of Scapular Fractures with Three-Dimensional Computed Tomography. The Journal of Bone and Joint Surgery-American Volume 2009, 91(9):2222–2228. https://doi.org/10.2106/jbjs.H.00881
Molenaars RJ, Mellema JJ, Doornberg JN, Kloen P: Tibial Plateau Fracture Characteristics: Computed Tomography Mapping of Lateral, Medial, and Bicondylar Fractures. Journal of Bone and Joint Surgery 2015, 97(18):1512–1520. https://doi.org/10.2106/jbjs.N.00866
Zhang Y, Sun Y, Liao S, Chang S: Three-Dimensional Mapping of Medial Wall in Unstable Pertrochanteric Fractures. BioMed Research International 2020, 2020:1–8. https://doi.org/10.1155/2020/8428407
Mary Jiayi T, Linda P, Michael P, Hans K, Markku N, Richard J, Bo Angela W, May T, Elizabeth B, Edward C: Potential discrepancy between plain films and CT scans in Brooker classification of heterotopic ossification. Br J Radiol 2017, 90(1080):20170263. https://doi.org/10.1259/bjr.20170263
Goyal T, Arora SS, Banerjee S, Kandwal P: Neglected Monteggia fracture dislocations in children: a systematic review. J Pediatr Orthop B 2015, 24(3):191–199. https://doi.org/10.1097/BPB.0000000000000147
Kamineni S, Ankem H, Patten DK: Anatomic relationship of the radial nerve to the elbow joint: Clinical implications of safe pin placement. Clinical Anatomy 2009, 22(6):684–688. https://doi.org/10.1002/ca.20831
Crosio A, Mattei L, Blonna D, Castoldi F: Anatomy and Specimens. In: The Elbow: Principles of Surgical Treatment and Rehabilitation. edn. Edited by Porcellini G, Rotini R, Stignani Kantar S, Di Giacomo S. Cham: Springer International Publishing; 2018: 19–28.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Re-defining the classification of heterotopic ossification of the elbow: A retrospective study

Status:

Version 1

Abstract

Figures

Background

Methods

Subjects

Ho Mapping Models

Statistical analysis

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1