Chirurgie de la main 26 (2007) 13–20

http://france.elsevier.com/direct/CHIMAI/

Original article

The anatomy and vascularity of the lunate:

considerations applied to Kienböck’s disease
L’anatomie et vascularisation du semilunaire : considérations appliquées
à la maladie de Kienböck
C. Lamasa,*, A. Carrerab, I. Proubastaa, M. Llusàb,c, J. Majóa, X. Mirc
a
Hand Unit and Upper Extremity, Department of Orthopedic Surgery, Hospital de la Santa Creu i Sant Pau, Autonomous University of Barcelona, Barcelona,
Spain
b
Department of Human Anatomy, Faculty of Medicine, University of Barcelona, Barcelona, Spain
c
Department of Orthopedic Surgery, Hospital Valle Hebrón, Autonomous University of Barcelona, Barcelona, Spain

Received 9 November 2006; accepted 14 January 2007

Abstract
Purpose. – The purpose of this study was to assess the anatomy and vascularity of the lunate. The genesis of lunatomalacia requires some
combination of vascular risk and mechanical predisposition. The findings will be correlated with the major existing theories of the cause of
Kienböck’s disease.
Methods. – We studied 27 cadaver upper limbs using latex injection and the Spalteholz technique. We investigated the blood supply to the
lunate. In 24 wrists we evaluated the incidence and distribution of anatomic features, arthrosis, and soft tissue lesions. We correlated the lunate
morphology and ligaments disruptions with the arthritic changes.
Results. – The lunate morphology results as classified by Antuña-Zapico were five type I (20.8%), 18 type II (75%) and one type III (4.2%).
The lunate was found to have a separate facet for the hamate in 11 cases (45.8%). The most common size of the facet was found to be 3 mm
(range, 3–6 mm). Arthrosis was identified with most frequency in the radius (88.2%) and lunate (94.1%). The triangular fibrocartilage complex
was found torn in 58.3%, the lunotriquetral interosseous ligament was torn in 20.8% and the scapholunate interosseous ligament (SLIL) was torn
in 54.2% of the wrists. There was a correlation between the presence of arthrosis at the hamate and the presence of a lunate facet (P = 0.027) and
a correlation between the presence of a tear in the SLIL and arthrosis in the scaphoid (P = 0.002). The nutrient vessels entered the lunate through
the dorsal and volar poles in all the specimens. The dorsal intercarpal and radiocarpal arches supply blood to the lunate from a plexus of vessels
located directly over the lunate’s dorsal pole. Vessels entered the dorsal aspect of the lunate through one to three foramina. One to five nutrient
vessels were observed entering the volar pole through various ligament insertions, including the ligament of Testut-Kuentz (radio-scapho-lunate
(RSL) ligament) and the radiolunate triquetrum ligament (or dorsoradial carpal ligament) and ulnar lunate triquetral ligament.
Conclusions. – The lunate had consistent dorsal and palmar arteries entering the bone in all the specimens. The blood supply and foramina
number is greater in the volar pole of the lunate than the dorsal pole. The lunate blood supply comes from different ligaments. In the etiopatho-
geny of Kienböck’s disease it is possible that an acute or chronic, traumatic or non-traumatic injury of the vessel bearing ligaments, particularly
because of their structure and the location of the RSL ligament, may have an important role in the appearance of lunate necrosis.
© 2007 Elsevier Masson SAS. All rights reserved.
Résumé
Objectif. – L’objectif de cette étude était de décrire l’anatomie et la vascularisation du semi-lunaire. La genèse de la nécrose du semi-lunaire
peut être la combinaison du risque vasculaire et d’une prédisposition mécanique. Les conclusions seront comparées aux principales théories
existantes pour la maladie de Kienböck.

* Correspondingauthor.
E-mail address: clamasg@hsp.santpau.es (C. Lamas).

1297-3203/$ - see front matter © 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.main.2007.01.001
14 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20

Méthodes. – Nous avons étudié la vascularisation de 27 poignets de cadavres en utilisant une injection de latex et la technique de Spalteholz.
Nous avons évalué dans 24 poignets la fréquence et la distribution des aspects anatomiques, de l’arthrose et des arrachements ligamentaires. Nous
avons mis en corrélation la morphologie du semi-lunaire et les arrachements de ligaments avec les changements arthrosiques.
Résultats. – La morphologie du semi-lunaire résultant de la classification d’Antuňa-Zapico était cinq types I (20,8 %), 18 types II (75 %) et un
type III (4,2 %). Le semilunaire avait une facette séparée pour l’hamatum dans 11 cas (45,8 %). La taille la plus habituelle de la facette a été de
3 mm (3–6 mm). L’arthrose a été identifiée le plus souvent sur le radius (88,2 %) et le semilunaire (94,1 %). Le fibrocartilage triangulaire
complexe était déchiré dans 58,3 % des cas, le ligament interosseux semi-lunaire-pyramidal dans 20,8 % et le ligament interosseux scapholunaire
dans 54,2 % des poignets. Il y avait une corrélation entre la présence d’arthrose à la face proximale de l’hamatum et la présence d’une facette du
semilunaire (p = 0,027) et une corrélation entre la présence d’une rupture dans le ligament interosseux scapholunaire et une arthrose du scaphoïde
(p = 0,002). Les vaisseaux nutritifs entrent dans le semilunaire à travers des artères palmaires et dorsales dans tous les spécimens. Les arches
radiocarpiennes et intercarpiennes dorsales paraissent se diviser en branches et converger vers un plexus de vaisseaux localisé directement à la
face dorsale du semilunaire. Les vaisseaux entrent à la face dorsale du semilunaire, à travers un à trois foramina. Un à cinq vaisseaux nutritifs
entrent à la face palmaire du semilunaire et à travers plusieurs ligaments : le ligament de Testut-Kuentz (ligament radioscapholunaire), le ligament
radius–semilunaire–pyramidal et le ligament semilunaire–pyramidal–ulna.
Conclusions. – Les semilunaires présentaient des artères dorsales et palmaires qui pénétraient l’os dans tous les spécimens. L’apport sanguin
et le nombre des foramina sont plus grands à la face palmaire qu’à la face dorsale. La vascularisation du semi-lunaire vient de ligaments diffé-
rents. Dans la pathogénie de la maladie de Kienböck, il est possible qu’une blessure aiguë ou chronique, traumatique ou non traumatique :
arrachement de ligament porte-vaisseaux, et en particulier le ligament radioscapholunaire, par leur structure et leur emplacement, peuvent avoir
un rôle important dans l’apparition d’une nécrose du semilunaire.
© 2007 Elsevier Masson SAS. All rights reserved.

Keywords: Anatomy of the lunate; Vascularity of the lunate; Kienböck’s disease; Avascular necrosis; Spalteholz technique

Mots clés : Anatomie du semilunaire ; Vascularisation du semilunaire ; Maladie de Kienböck ; Nécrose avasculaire ; Technique de Spalteholz

1. Introduction ease, and different types of corrective osteotomies have been

Kienböck theorized that lunate malacia was the result of a
traumatic disruption of the blood supply to the lunate and sub-
sequent disturbance of the bony nutrition [1–3]. The etiology
of lunatomalacia has remained a source of controversy [4].
Theories regarding the mechanism for the development of
osteonecrosis of the lunate include primary compression frac-
ture, traumatic disruption of the extraosseous blood supply of
the lunate, repetitive loading of the lunate and emboli [5–7].
Stahl [8] found in 31 specimens only one or two narrow
arteries entering the lunate from the volar side, and he was
able to demonstrate only one dorsal artery. Lee [9] and Gelber-
man et al.’s studies [10,11] showed that in 7–26% of speci-
mens there is only a single volar or dorsal blood supply to
the lunate. It is possible that this single blood supply may
lead to lunate necrosis.
Several causes of Kienböck’s disease have been proposed
based on anatomical studies. Antuña-Zapico [12,13] observed
a relationship between lunate shape and ulnar length. He
described three types of lunate morphology with the trabecular
pattern in type I the weakest with a greater potential for bone
fatigue and stress fracture under loads. Fragmentation was
more frequent in ulna minus variants and Hultén [14] postu-
lated that cubitus minus was an etiological factor in the disease.
Persson [15] performed the first corrective osteotomies for this
variant. More recently, other mechanical predisposing factors
have been investigated, including the size of the bone (smaller
lunates would be more prone to fail under high load) and an
increased ulnar inclination of the distal articular surface of the
radius [16]. The inclination of the articular surface of the
radius, measured as the radial inclination angle, has proven to
be an anatomical factor influencing the appearance of the dis-
proposed [5,6,17–20]. Nakamura et al. [18] advocated radial
wedge osteotomy to increase the radiolunate contact surface
producing a more uniform distribution of forces and thus redu-
cing load on the lunate. In Kienböck’s disease vascularized
bone grafts is an alternative to load altering procedures [21].
The purpose of this cadaveric study was to assess the anat-
omy and vascularity of the lunate to try to support the hypoth-
esis that avascular necrosis of the lunate may be due to poor
vascularity. We evaluated the incidence and distribution of
anatomic features, arthrosis and soft tissue lesions. We corre-
lated the lunate morphology, ligament and TFCC disruptions
and the presence of arthritic changes. We correlated the fora-
mina number in the lunate with the lunate morphology of
Antuña-Zapico.
2. Materials and methods
We studied 27 cadaver upper limbs using latex injection and
the Spalteholz technique. The specimens ranged from 18 to
90 years old (average age 59). The cadavers were men in 16
cases and women in 11 cases. We studied 18 right and nine left
wrists. The adult cadaver arms were amputated proximal to the
elbow joint. The brachial artery was cannulated with a catheter
and irrigated with normal saline. Transverse incisions were
made over the distal phalanges of all digits. Latex solution
(40–50 cm3) was then injected into the brachial artery under
firm manual pressure. The adequacy of injection was deter-
mined by the appearance of the latex through the transverse
incisions in the fingertips. The injected specimens were refri-
gerated for 12 hours to ensure consolidation of the latex. The
specimens were then fixed in formalin for 48 hours.
We investigated the extraosseous blood supply to the lunate.
In three wrists we took transversal sections and clarified using
 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20 15

the Spalteholz technique. In the other 24 cases, the radiocarpal

and midcarpal joints were dissected. The radiocarpal and mid-
carpal joints were entered through a dorsal approach. We eval-
uated the incidence and distribution of anatomic features,
arthrosis, condromalacia and soft tissue lesions. In all cases
we showed the lunate morphology, arthrosis distribution and
its relationship with ligament tears. The number and location
of the entry vessels was recorded. After observations on the
external vascularity had been made the lunates were removed.
The 24 lunates were separated and cleared by the Spalteholz
technique. The lunates were immersed in ethanol and dehy-
drated. The process of dehydratation consisted of placing the Fig. 1. Drawing of the three different morphological types of lunate as
lunates in 50% ethanol for 1 week, and then changing them classified by Antuña-Zapico. The lunate was classified into three different
shapes on the basis of the angle between the lateral scaphoid side of the lunate
each week for higher ethanol concentrations up to 100% etha-
and the proximal radial side of the lunate: a) type I lunate had an angle greater
nol. After that, we immersed the lunates in methylbenzene for than 130°, b) type II lunate had an angle of approximately 100°, and c) type III
2 weeks. Finally, the lunates were placed in a liquid mixture lunate had two distinct facets on its proximal surface, one that articulated with
made up of one part methyl salicylate and two parts benzyl the radius and the other that articulated with the triangular fibrocartilage.
benzoate for another 2 weeks. The specimens were immersed
in this liquid mixture and the bones were completely transpar-
ent. Photographs of the internal vascularity were taken with
specimens immersed in methyl salicylate and benzyl benzoate.
The information from anatomic dissections was organized
by means of the SPSS computer software system and this soft-
ware was also used for the statistical analysis. We correlated
the lunate morphology of Antuña-Zapico classification and the
lunate morphology of Viegas classification with the arthritic
changes. We correlated the ligaments and TFCC disruptions
with the arthritic changes. We calculated the proximal surface
of the two most frequent lunate variants (types I and II as
described by Antuña-Zapico) in mm2. Categorical variables
were analyzed in a univariate model by means of X2 contin-
gency table analysis. Statistical analysis of data were per-
formed with the Chi-square test and the level of significance
was P < 0.05. Comparison between means of the proximal sur-
face of the lunates type I or II of Antuña-Zapico was made
with the Student’s t-test. The presence of lunate morphology
type I or II as described by Antuña-Zapico [12,13] was related
with the presence or not of a higher dorsal or volar foramina Fig. 2. Dorsal aspect of the lunate type II (angle of 110°) with one foramina for
number. Statistical analysis of this data were performed with vascular nutrition.
the Mann–Whitney test.
and lunotriquetral interosseous ligament (LTIL). We showed
3. Results in all the specimens that the ulnar facet of the lunate had
more surface than the radial facet and presented a well differ-
3.1. Lunate anatomy entiated area in the middle of the ulnar facet with two textures
of different characteristics: a) the dorsal middle had a square-
The lunate morphology results as classified by Antuña- shaped surface articulating with triquetrum; and b) the volar
Zapico [12,13] (Fig. 1) were five type I (20.8%), 18 type II middle had ligamentous insertions: radio-lunate-triquetrum
(75%) and one type III (4.2%). The most frequent type II mor- (RLT) ligament and ulnar-lunate-triquetrum (ULT) ligament.
phology was 100° (29.2%) and 110° (25%) (Fig. 2). The lunate With black latex we showed the ligament insertion and its ves-
was found to have a separate facet for the hamate (type II as sels (Fig. 3).
described by Viegas) [22] in 11 cases (45.8%). The most com- We calculated the proximal surface of the lunate in mor-
mon size of the facet was found to be 3 mm (range, 3–6 mm). phology type I as described by Antuña-Zapico [12,13] (mor-
The proximal and distal surfaces of the lunate are completely phology associated with Kienböck’s disease with an
covered with articular cartilage and possess no vascular fora- angle > 130°) and type II (angle >110) [12,13]. In the five
mina or sites for ligament attachment. The radial and ulnar type I cases of Antuña-Zapico the average was 237.2 mm2
aspects are covered by articular cartilage except for insertions (S.D. 49.8). In the 18 type II cases of Antuña-Zapico the aver-
of the ligaments: scapholunate interosseous ligament (SLIL) age was 251.4 mm2 (S.D. 123.7). Both surfaces are similar and
16 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20
Fig. 3. Surface between the lunate and the triquetrum. Showed the two well
differentiated areas in the middle: a) the dorsal middle has a square-shaped
surface articulating with triquetrum; and b) the palmar middle has a
ligamentous insertions and vessels.
the results are not statistically significant (Student’s t-test 0.56).
We calculated the surface of the radial and ulnar aspect of the
lunate and its relationship with the lunate morphology. The
difference was not significant (Mann–Whitney test 0.15 and
0.86, respectively).
3.2. Arthrosis, ligament tears and associated findings
3.2.1. Correlation lunate morphology of Antuña-Zapico with
arthritic changes
Arthritic changes were identified in radius in 18 cases
(75%), in scaphoid in 14 cases (58.3%), in lunate in 22 cases
(94.1%), in triquetrum in three cases (12.5%), in capitate in 14
cases (58.3%) and in hamate in 16 cases (66.6%). The loss of
cartilage with exposure of subchondral bone was measured
from 3 to 8 mm. The lunate morphology was type I in five
cases, type II in 18 cases and type III in one case.
We have correlated the lunate morphology type I and type II
of Antuña-Zapico with the arthritic changes in radius, sca-
phoid, lunate, triquetrum, capitate and hamate. Statistical ana-
lysis with X2-test showed this dates, P = 0.608, P = 0.605,
P = 0.283, P = 0.539, P = 0.367, P = 0.508, respectively. Sta-
tistical analysis did not find a correlation between the lunate
morphology of Antuña-Zapico [12,13] and arthrosis of the
wrist.
3.2.2. Correlation lunate morphology of Viegas with arthritic
changes
The lunate was a facet for the capitate (type I as described
by Viegas) [22] in 13 cases (54.2%) and a separate facet for the
capitate and the hamate (type II as described by Viegas) [22] in
11 cases (45.8%). We have correlated the lunate morphology
type I and type II of Viegas [22] with the arthritic changes in
radius, scaphoid, lunate, triquetrum, capitate and hamate. Sta-
tistical analysis with X2-test found this dates, P = 0.048,
Fig. 4. Viegas’ type II lunate with two facets, one facet articulating with the
hamate (number 1) and another with the capitate (number 2). Photograph of
midcarpal joint demonstrating arthrosis at the proximal pole of the hamate and
capitate associated with this facets.
P = 0.223, P = 0.717, P = 0.435, P = 0.185, P = 0.027, respec-
tively. Statistical analysis found a significant correlation
between arthrosis of the hamate and the presence of a lunate
facet. In lunates, without a facet for hamate, arthrosis in the
capitate was found in 46.2% of cases and in the hamate in
46.2% of cases. In lunates, with a facet for hamate, arthrosis
in the capitate was found in 72.7% of cases and in the hamate
in 90.9% of cases. The correlation lunate morphology type II
of Viegas with arthrosis in hamate was statistically significant
(X2-test P = 0.027) (Fig. 4).
3.2.3. Correlation SLIL, LTIL and TFCC tears with arthritic
changes
The TFCC was found torn in 14 cases (58.3%), the LTIL
was torn in five cases (20.8%) and the SLIL was torn in 13
cases (54.2%) of the wrists.
We have correlated the SLIL tears with the arthritic changes
in radius, scaphoid, lunate, triquetrum, capitate and hamate.
Statistical analysis with X2-test found this results P = 0.239,
P = 0.002, P = 0.717, P = 0.435, P = 0.527, P = 0.444, respec-
tively. There was a relationship between the presence of a tear
in the SLIL and the presence of arthrosis in the scaphoid
(P = 0.002) (Fig. 5). The presence of erosion of the scaphoid
take place in the radiocarpal joint, near the scapholunate liga-
ment. We did not observe scaphoid arthrosis in the midcarpal
joint.
We have correlated the LTIL tears with the arthritic changes
in radius, scaphoid, lunate, triquetrum, capitate and hamate.
Statistical analysis with X2-test found this results P = 0.634,
P = 0.668, P = 0.620, P = 0.479, P = 0.668, P = 0.445, respec-
tively. Statistical analysis did not find a statistically significant
correlation between LTIL disruptions and arthrosis of the wrist.
We have correlated the TFCC tears with the arthritic
changes in radius, scaphoid, lunate, triquetrum, capitate and
hamate. Statistical analysis with X2-test found this results
P = 0.506, P = 0.290, P = 0.670, P = 0.629, P = 0.611,
 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20
Fig. 5. Photograph of radiocarpal joint demonstrating arthrosis at the scaphoid
and lunate associated with tears of the SLIL.
P = 0.561, respectively. Statistical analysis did not find a sta-
tistically significant correlation between TFCC tears and
arthritic changes of the wrist.
3.3. Vascularity of the lunate
The vascularity of the lunate was studied (Fig. 6). The nutri-
ent vessels entered the lunate through the dorsal and volar
poles in all the specimens. Dorsal lunate vascularity is supplied
by branches originating from the dorsal plexus as well as the
radiocarpal, intercarpal and transverse metacarpal arch. The
dorsal intercarpal and radiocarpal arches supply blood to the
lunate from plexus of vessels located directly over the lunate’s
dorsal pole (Fig. 7). Vessels entered the dorsal bone through
one to three foramina located in the proximal, central or ulnar,
non-articular aspect of the bone (Fig. 8). Palmar lunate vascu-
larity is supplied by branches originating from a palmar plexus
as well as direct ulnar, radial and anterior interosseous artery
Fig. 6. Dorsoradial view of the external wrist vascularity: dorsal radiocarpal and
intercarpal arches with a plexus of vessels over the lunate (S = scaphoid,
T = trapezium and Tz = trapezoid).
17
Fig. 7. Photograph of radiocarpal joint demonstrating the proximal and dorsal
aspect to the lunate. Showed vessels entering dorsal surface from the dorsal
radiocarpal arch and branching within the bone to provide the pole blood
supply.
branches. Palmar radiocarpal arches were present in 100% of
all specimens, while the intercarpal arch was present in 70%.
One to five nutrient vessels were observed entering the volar
pole through various ligament insertions, including the liga-
ment of Testut-Kuentz (radio-scapho-lunate (RSL) ligament)
and the radiolunate triquetrum ligament (or dorsoradial carpal
ligament) and ulnar lunate triquetral ligament. By means of
latex injections and the Spalteholz technique we were able
show the ligament insertion and its vessels (Figs. 9 and 10A–
C).
3.3.1. Correlation dorsal and volar foramina number with
lunate morphology of Antuña-Zapico
We have correlated the number of dorsal and volar vascular
foramina in the lunate with the lunate morphology of Antuña-
Zapico. Vessels entered the dorsal bone through one foramina
in seven cases (29.2%), two foramina in 15 cases (62.5%) and
Fig. 8. Dorsal aspect of the lunate which showed three dorsal foramina and
blood supply vessels with anastomoses entering the bone.
18 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20
Fig. 9. Transverse section of the wrist and Spalteholz technique. Arterial supply
of the dorsal and palmar aspect of the wrist. Dorsal nutrient vessels entering the
lunate through the dorsal radiocarpal arch. Palmar nutrient vessels entering the
lunate through the RSL ligament of Testut-Kuentz, RLT ligament and ULT
ligament.
three foramina in two cases (8.3%). Vessels entered the volar
bone through one foramina in one case (4.2%), two foramina in
11 cases (45.8%), three foramina in eight cases (33.3%), four
foramina in three cases (12.5%) and five foramina in one case
(4.2%). Lunate morphology of Antuña-Zapico was type I in
five cases and type II in 18 cases. Statistical analysis did not
find a statistically significant correlation between the lunate
morphology and foramina number in the lunate (Mann–Whit-
ney test 0.478 and 0.690, respectively).
4. Discussion
The central position of the lunate results in it being exposed
to compressive forces from different directions [23]. In Hara et
al.’s [24] biomechanical study the lunate receives 35% of the
load. The pressure of the distal radius, overload of the lunate
and ligament disruptions may cause condromalacia and arthro-
sis [23]. Osteonecrosis may also be caused by ligament rup-
tures or compression fractures. Kienböck [1,2] described the
progressive nature of this disease and its diagnosis on the
basis of radiographic appearance. He attributed it to trauma
that produces disturbance in the blood supply by affecting the
wrist ligaments. The pattern of lunate arthrosis in the speci-
mens is uniform and present in the margins of all the ligaments
and at risk in the event of ligament ruptures. However, the
most typical feature of avascular necrosis of the lunate was a
radial and proximal lesion [4,8,14]. The Testut-Kuentz RSL
ligament is inserted at this point. This ligament may be consid-
ered a mesocapsule extending from the palmar radiocarpal joint
capsule [25]. Studies of wrists in adults and fetuses have
shown that they contain terminal branches of the anterior inter-
osseous nerve and vessels emanating from the distal radial arch
[26]. Histologically, the RSL ligament does not have a fascicle
pattern to the collagen fascicles and carries very high concen-
trations of small caliber nerves and vessels [26]. In our study
we observed the existence of these vascular pedicles and their
entrance from this ligament to the lunate. It would be logical to
assume a lesion of the RSL ligament could lead to ischemia in
the radial and proximal side of the lunate with secondary osteo-
necrosis. However, if the blood vessels that enter the dorsal
aspect of the lunate anastomose with the volar blood vessels,
then it may be possible that ischemia would not develop.
Distal surface of the lunate has one or two facets [22,27]. In
Viegas classification [22], Viegas’ type I lunate has only a
facet articulating with the capitate, and a Viegas’ type II lunate
has two facets, one facet articulating with the capitate and
another with the hamate. Viegas [22] reported incidences of
type I and type II lunates about 27% and 73% of the time,
respectively, in 393 cadaver wrists studied. The findings of
Nakamura et al. [28] are similar to those of previous reports
of the incidence of type I versus type II lunates (29% versus
71%). In our study, the lunate was found to have a separate
facet for the hamate in 11 cases (45.8%). These differences
could be attributed to features associated with population,
race or number of specimens.
In cadaver arterial injections, Stahl [8] found in 31 speci-
mens only one or two narrow arteries entering the lunate
from the volar side, and he was able to demonstrate only one
dorsal artery. The suggestion that only a volar blood supply
exists directly contrasts with our study and other authors,
which have noted a dorsal supply in a majority of cases [9,
10,29,30]. Travaglini [31] described the extraosseous vascular
anatomy of the carpus based on injections of four specimens.
He noted volar and dorsal arterial arches. Lee [9] examined 53
lunates injected with 50% micropaque suspension and clarified
by the Spalteholz method. He noted that 66% of his specimens
had a volar and dorsal blood supply with anastomosis of the
volar and dorsal arterial systems. He found that 7.5% of the
specimens had both a palmar and dorsal blood supply but no
anastomosis between the two arterial systems. In his series,
26% of the specimens were supplied by volar (15%) or dorsal
(11%) vessels alone.
Gelberman et al. [32] and Panagis et al. [10] examined a
total of 60 lunates in two studies examining the intraosseous
blood supply. The authors noted a consistent volar supply to
the lunate and a frequent, but inconsistent, dorsal supply [11,
32]. They noted anastomoses between the volar and dorsal ves-
sels and found that the anastomoses occurred distal to the mid-
line of the lunate. The authors noted three patterns of intraoss-
eous anastomosis (Y, I and X). Panagis et al. [10]
demonstrated only a volar blood supply in the 20% of the
lunates. The remaining 80% of the specimens had a volar and
dorsal supply and X, Y, I pattern of anastomosis. Lee [9] and
Gelberman et al.’s studies [10,32] showed that in 7–26% of
specimens there is only a single volar or dorsal blood supply
to the lunate. In our study there was a consistent vascularity of
the lunate. Dorsal and volar arteries entering the bone in all the
specimens. The foramina number is greater in the volar pole of
the lunate than the dorsal pole. The dorsal intercarpal and
radiocarpal arches supply blood to the lunate from plexus of
vessels located directly over the lunate’s dorsal pole. Vessels
entered the dorsal bone through one to three foramina. The
 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20 19

Fig. 10. Transverse section of the wrist and Spalteholz technique. A) Intraosseous vascular pattern formed by two dorsal and one palmar vessels: Y-pattern of
Gelberman. B) Microscopic image of the lunate where the vascular entrances are appreciated in the RLT and ULT ligaments. C) Microscopic image of the lunate
where the vascular entrances are appreciated in the RSL ligament.

palmar lunate vascularity is supplied by the palmar intercarpal
and radiocarpal arches. One to five nutrient vessels were
observed entering the volar pole through various ligament
insertions, including the ligament of Testut-Kuentz (RSL liga-
ment) and the radiolunate triquetrum ligament (or dorsoradial
carpal ligament) and ulnar lunate triquetral ligament.
The etiology of Kienböck’s disease is unknown [4,5],
although there are different hypotheses which can basically
be divided into traumatic and non-traumatic theories [5,12].
Traumatic theory subdivides, into three areas: a) the trauma
acts on the vascular system [1–5], b) the trauma acts on the
osseous system: fracture theory [5,8,12], c) the trauma acts
on the vascular nervous system [33]. The trauma can be an
avulsion of the ligaments and a vascular break [1,2], or act
on the nervous system of the vessels [33]. In the last case, it
is important to observe the vasomotor reactions which follow
all serious traumas, vasodilatation and vasoconstriction alter-
nate. Vasodilatation leads to stasis and decalcification with
the resulting reduction in osseous resistance and vasoconstric-
tion can lead to a partial osseous infarct. The trauma can also
act on the osseous system: fracture theory. The fracture shows
up as a transverse line caused by compression along which
decalcification sets in with a sinking of the upper or proximal
half of the lunate [8]. Lee [9] defines the possibility of a trans-
verse fracture interfering with the entrance of the arterials leav-
ing the proximal half of the lunate ischemic which is some-
thing very often observed in the disease. Watson and Guidera
[34] postulates the “fault plate hypothesis”, in which as a result
of various factors trabeculae ruptures occur and multiple oss-
eous lamina form which have the effect of a wall and interfere
with the vascularity. Non-traumatic vascular theory [5,12],
explains among other causes, necrosis of the lunate caused by
embolus of the vessels [7], vasculitis and treatment with corti-
coids [5].
In the cross-sections which appear in the figures it is
demonstrated that the vessels injected with black latex enter
by ligament insertions (Figs. 9 and 10A–C). According to our
study there is a double vascularity of the lunate in all the speci-
20 C. Lamas et al. / Chirurgie de la main 26 (2007) 13–20
mens. If the vascular anatomy of the lunate found in this study
is representative of the lunates which develop avascular necro-
sis, then lunate necrosis would not be due to primary lack of
osseous circulation. If anatomically the vessels enter in the
lunate via different capsulo-ligamentous structures, as we
have demonstrated in this study, it is logical to think that an
injury to these ligaments could damage their blood supply. In
the etiopathogeny of Kienböck’s disease, it is possible that an
acute or chronic, traumatic or non-traumatic lesion of the ves-
sel bearing RSL ligament (and particularly because of their
location and structure) could have a decisive influence in the
appearance of lunate necrosis. In Kienböck’s disease vascular-
ized bone grafts is an alternative to load altering procedures
[21]. Because the special vascularization of the lunate we
think that seems better to revascularize the lunate by volar
approach than the dorsal approach.
Acknowledgements
The authors thank Dr. Ignasi Gich of the Preventive Medi-
cine and Statistics Department for his assistance in the statisti-
cal analysis of the data obtained in this study.
References
[1] Kienböck R. Über traumatische malazie des mondbeins und ihre folge-
zustände: Entartungsformen und kompressionsfrakturen. Fortschr Geburt-
shilfe Roentgenol 1910;16:77–103.
[2] Kienböck R. Concerning traumatic malacia of the lunate and its conse-
quences: degeneration and compression fractures. Clin Orthop Relat Res
1980;149:4–8.
[3] Preiser G. Zur Frage der Typischen Traumatischen Ernabrungsstorungen
der Kaurzen Hand. Und. Fusswurzel Knochen. Fortsthr a. d. Geb. D.
Röntgenstr 1911;17:360.
[4] Irisarri C. Aetiology of Kienböck’s disease. J Hand Surg 2004;29B:279–
85.
[5] Lamas C. In: Kienböck’s disease. Barcelona: JM Bosch editor; 2005. p.
1–181.
[6] Lamas C. Osteotomía de sustracción lateral del radio en casos Zero Var-
iant en la enfermedad de Kienböck. Doctoral thesis. Autonomous Uni-
versity of Barcelona, 1998.
[7] Axhausen G. Nicht Malacie sondern Nekrose des os lunatum. Capri.
Arch Klin Chir 1924;129:26.
[8] Stahl F. On lunatomalacia (Kienbock’s disease): a clinical and roentgen-
ological study, especially on its pathogenesis and the late results of
immobilization treatment. Acta Chir Scand 1947;126(Suppl):3.
[9] Lee MLH. The intraosseous arterial pattern of the lunate bone and its
relation to avascular necrosis. Acta Orthop Scand 1963;33:43.
[10] Panagis JS, Gelberman RH, Taleisnik J, Baumgaertner M. The arterial
anatomy of the human carpus. Part II: the intraosseous vascularity. J
Hand Surg 1983;8A:375.
[11] Gelberman RH, Panagis JS, Taleisnik J, Baumgaertner M. The arterial
anatomy of the human carpus. Part I: the extraosseous vascularity. J
Hand Surg 1983;8A:367.
[12] Antuña-Zapico JM. Malacia del semilunar. Doctoral thesis. University of
Valladolid, 1966.
[13] Antuña-Zapico JM. Enfermedad de Kienböck. Rev Ortop Traumatol
1993;37 IB(Suppl. I):100–13.
[14] Hultén O. Über anatomische der handgelenknochen. Acta Radiol 1928;9:
155–68.
[15] Persson M. Causal treatment of lunatomalacia. Further experiences of
operative ulnar lengthening. Acta Chir Scand 1950;100:531–44.
[16] Tsuge S, Nakamura R. Anatomical risk factors for Kienböck’s disease. J
Hand Surg 1993;18:70–5.
[17] Werner FW, Palmer AK. Biomechanical evaluation of operative proce-
dures to treat Kienböck’s disease. Hand Clin 1993;9:431–43.
[18] Nakamura R, Tsuge S, Watanabe K, Tsunoda K. Radial wedge osteot-
omy for Kienböck’s disease. J Bone Joint Surg 1991;73A:1391–6.
[19] Garcia-Elias M. Radial wedge osteotomy for Kienböck’s disease. J Bone
Joint Surg 1992;74A:1431.
[20] Lamas C, Mir X, Llusà M, Navarro A. Dorsolateral biplane closing radial
osteotomy in zero variant cases of Kienböck’s disease. J Hand Surg
2000;25A:700–9.
[21] Sheetz KK, Bishop AT, Berger RA. The arterial blood supply of the dis-
tal radius and ulna and its potential use in vascularized pedicled bone
grafts. J Hand Surg 1995;20A:902–14.
[22] Viegas SF, Patterson R, Hokanson JA, Davis J. Wrist anatomy: inci-
dence, distribution, and correlation of anatomic variations, tears, and
arthrosis. J Hand Surg 1993;18A:463–75.
[23] Golimbu CN, Firooznia H, Rafii M. Avascular necrosis of carpal bones.
MRI Clin North Am 1995;3(2):281–303.
[24] Hara T, Horii E, Ann KN, Cooney WP, Linscheid RL, Chao EY. Force
distribution across wrist joint. Application of pressure-sensitive conduc-
tive rubber. J Hand Surg 1992;17A:339–47.
[25] Berger RA. The anatomy of the ligaments of the wrist and distal radio-
ulnar joints. Clin Orthop Relat Res 2001;383:32–40.
[26] Berger RA, Kauer JMG, Landsmeer JMF. The radioscapholunate liga-
ment: a gross and histologic study of adult and fetal wrist. J Hand Surg
1991;16A:350–5.
[27] Testut L, Latarjet A. Tratado de Anatomía Humana. Salvat Editores SA
Barcelona 1977;339:612.
[28] Nakamura K, Patterson R, Moritomo H, Viegas SF. Type I versus type II
lunates: ligament anatomy and presence of arthrosis. J Hand Surg 2001;
26A:428–36.
[29] Grettve S. Arterial anatomy of the carpal bones. Acta Anat (Basel) 1955;
25:331.
[30] Barber H. The extraosseous arterial anatomy of the adult human carpus.
Orthopedics 1972;5:1.
[31] Travaglini F. Arterial circulation of the carpal bones. Bull Hosp Joint Dis
1959;20:19.
[32] Gelberman RH, Bauman TD, Menon J, Akeson WH. The vascularity of
the lunate bone and Kienböck’s disease. J Hand Surg 1980;5A:272–8.
[33] Leriche R, Fontaine R. Contribution a l´etude de la maladie de Kienböck;
son traitment par la sympathectomie périhumerale. Strasbourg Med 1929;
89:581.
[34] Watson HK, Guidera PM. Aetiology of Kienböck’s disease. J Hand Surg
1997;22B:5–7.