Importantly, when we divided the hamate into two parts with a sagittal plane through the proximal pole and central ridge between the 4th and 5th carpometacarpal joint surface, most arterial entrances (including all trunk arteries) and the intra-osseous vascularity were located in the radial part [Figure 2A–D].
Classification and outcomes of hamate fractures
Based on the radiograph and/or CT data from all 127 cases of hamate fractures reviewed in this study, they were divided into four types: fractures of transversal/proximal pole, medial tuberosity, dorsal coronal of the hamate body and fractures of the hamate hook. The coronal fractures of the hamate body were sub-divided into two types: dorsal oblique fractures and splitting fractures [Figure 3].
The mean follow-up time for the 94 cases was 12.4 months (range: 4.0–37.0 months). Sixty-one were male and 33 were female, with an average age of 34.4 years (from 16.0 to 67.0 years). Although not all patients underwent surgical intervention, the union rate of hamate fracture was still up to 92.6% (87 of 94 cases). The non-union rate was only 5.3% (5 of 94 cases) and the remaining two cases ended up with malunion.
From the view of classification, the fracture of the hamate hook has a significantly high rate of non-union (P = 0.031). But the patients were generally satisfied with the results on the last follow-up. The classification results of hamate fractures are shown in Table 2.
Intra-osseous vascularity in the hamate hook
In most specimens (5/6), the major blood supply of the hook was from the volar non-articular surface. Some tiny arteries around the distal and ulnar base of the hook could be identified. In the other specimen there was a trunk artery entering the hook from its tip and running to the base of the hook. The distribution features of intra-osseous arteries in the hamate hook can be divided into three patterns.
Pattern 1: An artery, which was slightly thinner than trunk arteries, entered the volar non-articular surface and then ran to the tip underneath the radial cortex of the hook. Some other small arteries were found at the distal hook and the ulnar side of the hook base. This type was the most common and was found in four of the six specimens [Figure 4A and 4B].
Pattern 2: Two branches from the volar trunk arteries were identified. They ran to the tip underneath the radial cortex of the hook. Some small arteries were also found. This type was found in one specimen [Figure 4C and 4D].
Pattern 3: The arteries from the volar non-articular surface were very small. There was one trunk artery entering the hook from the ulnar tip and running to the base of the hook. Some other small arteries were found around the hook. This pattern was only found in one specimen [Figure 4E and 4F].
In the two specimens of sample 1 (a 55-year-old woman), the contrast agents were found accumulated inside the hook of the hamate. But, this situation did not exist in other specimens. The accumulated contrast agents appeared as continuous small lumps, like grapes [Figures 2B and 4B].
Intra-osseous vascularity of hamate body
There were 18 arteries from the volar non-articular surface. The median quantity and diameter of all 18 arteries were 3.00 mm (IQR: 2.80–3.30 mm) and 0.13 mm (IQR: 0.09–0.16 mm), respectively. There were 55 arteries from the dorsal side and the median quantity and diameter of them were 8.50 mm (IQR: 6.00–13.50 mm) and 0.11 mm (IQR: 0.08–0.14 mm), respectively. The difference of the arterial diameter between the volar and dorsal sides was not significant based on the Mann-Whitney U test (Z = –0.783, P = 0.434). However, the quantity of arteries from the dorsal side was significantly larger than that from the volar side based on the Wilcoxon signed-rank test (Z = –2.032, P = 0.042).
There were seven trunk arteries on the volar side and ten trunk arteries on the dorsal side. The median quantity was 1.00 mm (IQR: 1.00–1.30 mm) for the volar side and 1.50 mm (IQR: 1.00–2.30 mm) for the dorsal side, while the median diameter was 0.16 mm (IQR: 0.14–0.20 mm) for the volar side and 0.18 mm (IQR: 0.17–0.21 mm) for the dorsal side. However, there were no significant differences for both the quantity (Z = –1.732, P = 0.083,) and diameter (Z = –1.430, P = 0.153) between the two sides.
The mean location value of trunk and non-trunk arteries (including the arteries around the radial and ulnar surfaces) was 67% ± 11% and 80% ± 17%, respectively. The difference was significant based on the independent sample t test (t = –3.091, P = 0.003). Most location values of the trunk (94%, 16/17) and non-trunk (96%, 67/70) arteries were larger than 50%.
We recorded the location value of the most proximal artery and the most proximal trunk artery in each specimen. The mean location value of the most proximal arteries was 47% ± 11%, whereas the mean location value of the most proximal trunk arteries was 58% ± 13%. But the difference was not significant (t = 2.254, P = 0.074) [Table 3].
In the specimens of transversal/proximal pole fractures of the hamate body, the mean location value of fracture lines was 45% ± 5%. This value was significantly smaller than the mean value of most proximal trunk arteries in our specimens with angiography (t = –2.478, P = 0.031).
New findings of intra-hamate vascularity
After Lee first studied the intra-osseous vascularity of the lunate with the Spalteholz method, Panagis et al systematically investigated the intra-osseous vascularity of all carpal bones and established the anatomical basis of the wrist arterial system. Failla's study concerned the distribution of vessel entrances around the hamate hook, but the inadequacies of this study were obvious. The research only focused on the foramina distribution instead of intra-osseous vessels. According to our observations, almost 28% of foramina contained no artery in the hamates. So the results of the foramina distribution cannot accurately reflect the artery distribution. Recently the latest studies about the intra-osseous arterial system in scaphoid, lunate, and capitate bones with micro-CT angiography demonstrated many valuable results and confirmed the feasibility and reliability of the new angiography method.[9,11,12,14]
Panagis's research on the vascularity of the carpal bone is still fundamental. But their descriptions about intra-hamate vascularity were quite simple as follows. Dorsally, three to five vessels entered the non-articular surface and branched in all directions. These branches supplied the dorsal 30% to 40% of the bone. Palmarly, the hamate's pre-dominant blood supply was supplied by one large artery entering through the radial base of the hook. Within the bone, it branched and formed significant anastomoses with dorsal vessels in 50% of specimens. The vascularity of the hook was consistently provided by one to two small vessels entering through the medial base and the tip of the hook. These two vessels anastomosed with each other in each specimen but, in the majority of specimens, failed to anastomose with the blood vessels of the body of the hamate.
However, our research has obtained some important and different results: (1) The blood supply from the dorsal side may be dominant because the number of dorsal arteries is significantly larger than it on the volar side; (2) The rate (five of six specimens) of existence of anastomoses between the trunk arteries from dorsal and volar sides is much greater than Panagis's observation (50%); (3) Most of the arterial entrances and intra-osseous vascularity are located in the radial part and the blood supply of the ulnar part and proximal part is relatively poor, which is never reported before; (4) There are many thin arteries entering the hamate from the radial non-articular surface and the ulnar surface. But they are usually small and inconstant; (5) All arterial entrances are located around the ligaments, such as the capitohamate ligament, deep capitohamate ligament, carpometacarpal ligament, and transverse carpal ligament. This arterial distribution pattern can also be seen in other carpal bones.[9,11] In this way, the arteries are well protected by the rigid ligaments.
The accumulation of the contrast agent in the hamate hook was not reported in previous studies. The contrast agents demonstrated grapes-like continuous small lumps, which were more likely physiological blood sinuses rather than pathological bleeding. Especially, this situation was found bilaterally in one sample. The intra-osseous blood sinus is quite salient in the red marrow, which is very rare in a normal adult. But some studies have proved that the reconversion of yellow marrow to red marrow could occur in adults under many conditions, such as obese women who smoke heavily,[15,16] non-Hodgkin's lymphoma, or hematopoietic growth factors. Unfortunately, the life history of this donor is unclear. Further research, if there are complete life histories of this donor, will help us to explain this phenomenon.
Possible relationship between the avascular risk of hamate fractures and intra-hamate vascularity
Xiong et al concluded the clinical classification of hamate hook fractures. For hamate body fractures, the classification methods are not unique.[19,20] Some authoritative works[1,10] sub-divide the hamate body fractures into four types: fractures of proximal pole, medial tuberosity, sagittal oblique, and dorsal coronal. The sagittal oblique fractures were image-based classification. We reviewed the literature but found no reported typical sagittal hamate fracture cases. All the literatures just quoted the description from Milch. According to our observation of hamate fractures, we believe that the sagittal oblique type should be classified as a transversal fracture, which mainly involves the proximal part of the hamate. So we divided the hamate fractures into four types: fractures of transversal/proximal pole, medial tuberosity, dorsal coronal of the hamate body, and fractures of the hamate hook.
In the past, plaster immobilization might be used to deal with the acute cases of hook fractures, but the healing rate was not ideal. The feature of vascularity of the hook was considered as a possible key factor of hook fracture non-union.[1,8] Our study has demonstrated that the blood supply from the volar non-articular surface is more important for the hamate hook than in the previous concept, and only in a few specimens, the trunk artery on the tip of hook was dominant with some small arteries from the base of the hook supplying the basal part. This feature may have different influences on different locations of hook fractures. Obviously both the basal part and the tip of the hook have more blood supply than the middle part. We believe that this is a very important factor which led to a significant higher fracture non-union rate for middle part fractures compared with the basal fractures.[2,4] The fractures of tip of the hook were generally avulsion fractures and the symptoms were always mild. But the middle part fractures of the hook may result in the complications of tear or disruption of flexor tendons of the ring and little fingers. The statistical analysis has shown the hook fracture has a statistically significant high non-union rate than other fracture types of hamate, which was consistent with the feature of vascularity. In our study, we observed that the trunk arteries of the hook were generally located on the radial side and the tip of the hook, so using a dorsal approach or medial approach for open reduction and internal fixation is recommended. The intra-osseous arteries of the hook will be protected if the screw is centrally placed.
Coronal fractures of the hamate body are often the results of axial forces from metacarpi, so this type is usually combined with sub-luxation or dislocation of the 4/5th carpometacarpal joint. Ebraheim et al proposed a classification for coronal body fractures. However, according to the intra-hamate vascularity, it is more appropriate to sub-divide this fracture into two sub-types: splitting fractures involving injuries of intra-osseous trunk arteries, and dorsal oblique fractures not involving injuries of intra-osseous trunk arteries. This classification method is similar to Hirano and Inoue's work. Although the splitting fracture line can massively disrupt the anastomoses of the trunk arteries, the volar and dorsal fragments can be well supplied by the volar and dorsal trunk arteries, respectively. If the fracture line is limited to a small dorsal part such as the dorsal oblique fracture, it can hardly influence the main blood supply of the hamate body, but for the small fragment, some non-trunk arteries can also provide the blood supply [Figure 5]. This might be the reason that most of coronal fractures of the hamate body generally have good outcomes after surgical intervention.[25,26] However, the stable fixation of the fragment is still pivotal. The internal fixed screws should be centrally placed but slightly ulnar to prevent damage of the main intra-osseous arteries.
The fracture lines of transversal/proximal pole fractures are usually located more proximal than the trunk arteries according to our result [Figure 6A and 6B]. The blood supply of the proximal part is retrograde and poor. So there should be a high avascular risk for transversal/proximal pole fractures. But studies on isolated transversal and proximal pole fractures of the hamate are so rare that the clinical healing rate of these fractures is unclear. However, according to our follow-up results and the published researches of complex carpal injuries with transhamate fractures, the outcomes were generally good.[27,28,29] It was assumed that the tough ligament connecting the capatite will provide rigid stability of hamate fractures. In addition, the axial load helps to maintain the stability of transversal fractures. Also, unlike lunate fractures and proximal fractures of the scaphoid, the proximal part of the hamate is not a stress concentrating location of the wrist. It is less likely to cause a collapse of the fragment.[12,30]
Finally, the medial tuberosity fracture of the hamate body is a rare type of hamate injury. This type is usually caused by a direct blow to the ulnar side of the wrist [Figure 6C]. Although there are some non-trunk arteries located around this area, they are usually inconstant and small. We believe that the influence to the intra-osseous vascularity of this fracture type is limited. Generally the fracture fragments are small and the outcomes are good. Conservative immobilizations are recommended for this type of fracture.
At present, it is impossible to illustrate intra-osseous arteries for the hamate fracture patients in vivo. So it is hardly to obtain the direct evidence of the relationship between the fractures and intra-hamate vessels. But we found there were one to three constant trunk arteries from both volar and dorsal sides in each hamate sample. The abundant blood supply obviously contributes to the high healing rate of fracture of the hamate body. As to the fracture of hamate hook, the feature of intra-osseous arteries of hamate hook is highly correlated with the non-union rate of different type of the fracture.
The new feature of intra-hamate vascularity we obtained is more abundant than the previous studies. However, it is a pity that our sample size is still not big enough to well establish a classification for the intra-hamate arterial system. But we have obtained many new and interesting results, and this can lead to some significant instructive clinical and treatment recommendations. A larger sample size is needed to confirm our classification and extend our understanding for the relationship of arterial patterns between different sides, genders, and ages. Otherwise, some spurious significance may be obtained. We believe that our research can help hand surgeons take the features of intra-hamate vascularity into full consideration to properly assess and avoid damage of the intra-osseous arterial system when they deal with a hamate fracture in their practices.
The authors thanks to Dr. Edward C. Mignot from Shandong University, for linguistic advice.
This study was supported by the grant from the Peking University Medicine Information Foundation (No. BMU20160600).
Conflicts of interest
1. Suh N, Ek ET, Wolfe SW. Carpal fractures. J Hand Surg Am
2014; 39:785–791. doi: 10.1016/j.jhsa.2013.10.030.
2. Xiong G, Dai L, Zheng W, Sun Y, Tian G. Clinical classification and treatment strategy of hamate hook fracture. J Huazhong Univ Sci Technolog Med Sci
2010; 30:762–766. doi: 10.1007/s11596-010-0654-7.
3. Bernstein RA. Are hamate fractures
common? Letter regarding “carpal fractures” article. J Hand Surg Am
2014; 39:2344doi: 10.1016/j.jhsa.2014.08.026.
4. Xiong G. Hook of hamate fractures
: location and tendon rupture. J Hand Surg Am
2014; 39:175–176. doi: 10.1016/j.jhsa.2013.11.020.
5. Yamazaki H, Kato H, Nakatsuchi Y, Murakami N, Hata Y. Closed rupture of the flexor tendons of the little finger secondary to non-union of fractures of the hook of the hamate. J Hand Surg Br
2006; 31:337–341. doi: 10.1016/j.jhsb.2005.12.015.
6. Gelberman RH, Panagis JS, Taleisnik J, Baumgaertner M. The arterial anatomy of the human carpus. Part I: the extraosseous vascularity. J Hand Surg Am
1983; 8:367–375. doi: 10.1016/s0363-5023(83)80194-4.
7. Panagis JS, Gelberman RH, Taleisnik J, Baumgaertner M. The arterial anatomy of the human carpus. Part II: the intraosseous vascularity. J Hand Surg Am
1983; 8:375–382. doi: 10.1016/s0363-5023(83)80195-6.
8. Failla JM. Hook of hamate vascularity: vulnerability to osteonecrosis and nonunion. J Hand Surg Am
1993; 18:1075–1079. doi: 10.1016/0363-5023(93)90405-R.
9. Xiao ZR, Zhang WG, Xiong G, Zhang YL. Three-dimensional intralunate arteries visualization with red lead (Pb3O4) angiography. Chin Med J
2017; 130:2575–2578. doi: 10.4103/0366-6999.213909.
10. Wolfe SC, Hotehkiss RN, Pederson WC, Kozin SH. Greens's Operative Hand Surgery. 6th Edn2010; Philadelphia: Churchill Livingstone, 1814.
11. Xiong G, Xiao ZR, Zhang WG. Vascular anatomy of the capitate determined by micro-computed tomography angiography
. J Hand Surg Eur Vol
2017; 42:966–967. doi: 10.1177/1753193417714400.
12. Xiao Z, Xiong G, Zhang W. New findings about the intrascaphoid arterial system. J Hand Surg Eur Vol
2018; 43:1059–1065. doi: 10.1177/1753193418758890.
13. Lee ML. The intraosseus arterial pattern of the carpal lunate bone and its relation to avascular necrosis. Acta Orthop Scand
1963; 33:43–55. doi: 10.3109/17453676308999833.
14. Kadar A, Morsy M, Sur YJ, Laungani AT, Akdag O, Moran SL. The vascular anatomy of the capitate: new discoveries using micro-computed tomography imaging. J Hand Surg Am
2017; 42:78–86. doi: 10.1016/j.jhsa.2016.12.002.
15. Gonzalez FM, Mitchell J, Monfred E, Anguh T, Mulligan M. Knee MRI patterns of bone marrow reconversion and relationship to anemia. Acta Radiol
2016; 57:964–970. doi: 10.1177/0284185115610932.
16. Poulton TB, Murphy WD, Duerk JL, Chapek CC, Feiglin DH. Bone marrow reconversion in adults who are smokers: MR imaging findings. AJR Am J Roentgenol
1993; 161:1217–1221. doi: 10.2214/ajr.161.6.8249729.
17. Manceron V, Guignard S, de Broucker F, Paycha F, Pouchot J, Vinceneux P. Bone marrow reconversion and magnetic resonance imaging: case report. Rev Med Interne
2003; 24:830–834. doi: 10.1016/j.revmed.2003.08.003.
18. Saadate-Arab M, Troufléau P, Stines J, Verhaeghe JL, Rios M, Molé D. MR imaging findings of bone marrow reconversion induced by growth factors in 3 patients. J Radiol
2002; 83 (2 Pt 1):147–152. doi: 10.1097/00004424-200202000-00007.
19. Hirano K, Inoue G. Classification and treatment of hamate fractures
. Hand Surg
2005; 10:151–157. doi: 10.1142/S0218810405002747.
20. Milch H. Fracture of the hamate bone. J Bone Joint Surg
21. Carroll RE, Lakin JF. Fracture of the hook of the hamate: acute treatment. J Trauma
1993; 34:803–805. doi: 10.1097/00005373-199306000-00009.
22. Takeda S, Tatebe M, Ishii H, Morita A, Wakai K, Hirata H. Computerized tomographic prediction of flexor tendon injuries complicating hamate hook fractures. J Hand Surg Eur Vol
2019; 44:367–371. doi: 10.1177/1753193418823503.
23. Nanno M, Sawaizumi T, Ito H. Simplified dorsal approach to fracture of the hamate hook with percutaneous fixation with screws. J Plast Surg Hand Surg
2010; 44:214–218. doi: 10.3109/02844310801956714.
24. Cain JE Jr, Shepler TR, Wilson MR. Hamatometacarpal fracture-dislocation: classification and treatment. J Hand Surg Am
1987; 12:762–767. doi: 10.1016/s0363-5023(87)80064-3.
25. Ebraheim NA, Skie MC, Savolaine ER, Jackson WT. Coronal fracture of the body of the hamate. J Trauma
1995; 38:169–174. doi: 10.1097/00005373-199502000-00004.
26. Wharton DM, Casaletto JA, Choa R, Brown DJ. Outcome following coronal fractures of the hamate. J Hand Surg Eur Vol
2010; 35:146–149. doi: 10.1177/1753193408098907.
27. Nunez FA Jr, Luo TD, Jupiter JB, Nunez FA Sr. Scaphocapitate syndrome with associated trans-scaphoid, trans-hamate perilunate dislocation: a case report and description of surgical fixation. Hand (N Y)
2017; 12:N27–N31. doi: 10.1177/1558944716668837.
28. Sabat D, Dabas V, Suri T, Wangchuk T, Sural S, Dhal A. Trans-scaphoid transcapitate transhamate fracture of the wrist: case report. J Hand Surg Am
2010; 35:1093–1096. doi: 10.1016/j.jhsa.2010.04.023.
29. Schweizer A, Kammer E. Transhamate periscaphoid axial radial fracture dislocation of the carpus. J Hand Surg Am
2008; 33:210–212. doi: 10.1016/j.jhsa.2007.11.003.
30. Xiong G, Xiao Z, Wang H, Guo S, Tao J. Microstructural study of the lunate in stage III Kienböck's disease with micro-computed tomography imaging. J Hand Surg Eur Vol
2017; 42:71–77. doi: 10.1177/1753193416664502.
31. Sauerland S, Lefering R, Bayer-Sandow T, Brüser P, Neugebauer EA. Fingers, hands or patients? The concept of independent observations. J Hand Surg Br
2003; 28:102–105. doi: 10.1016/s0266-7681(02)00360-1.
Keywords:© 2019 by Lippincott Williams & Wilkins, Inc.
Micro-computed tomography angiography; Intra-hamate arteries; Hamate fractures; Avascular risk