J Shoulder Elbow Surg (2011) -, 1-10
www.elsevier.com/locate/ymse
Disorders of the long head of biceps tendon
Michael Khazzam, MDa,*, Michael S. George, MDb, R. Sean Churchill, MDc,
John E. Kuhn, MDd
a
UT Southwestern Medical Center, Sports Medicine & Shoulder Service, Dallas, TX, USA
KSF Orthopaedic Center, Houston, TX, USA
c
Aurora Advanced Healthcare, Milwaukee, WI, USA
d
Vanderbilt Sports Medicine & Shoulder Center, Nashville, TN, USA
b
Disorders of the long head of the biceps (LHB) tendon can
exist in conjunction with several other shoulder pathologies.
Currently, the function of the LHB tendon remains unresolved. It is clear, however, that this tendon can be a signiﬁcant
source of shoulder pain and dysfunction. We have reviewed
the anatomy, pathophysiology, classiﬁcation, diagnosis, and
treatment of disorders involving the LHB tendon. We also
have reviewed the literature to help make treatment decisions.
Anatomy
The LHB tendon originates from the superior labrum and
the supraglenoid tubercle. The exact location of labral
attachment varies, but is usually in the posterior portion of
the superior labrum.18,60
The intra-articular portion of the LHB tendon is partially
stabilized by the biceps reﬂection pulley, which consists of
the superior glenohumeral ligament (SGHL), the coracohumeral ligament, and deep ﬁbers of the subscapularis
and supraspinatus tendons64 (Fig. 1). The superior glenohumeral ligament runs spirally along the LHB tendon adjacent to the coracohumeral ligament3 (Fig. 2). The pulley
stabilizes the LHB tendon as the tendon makes a 30 to 40
turn while exiting the glenohumeral joint.32 The LHB tendon
travels deep to the coracohumeral ligament and through the
rotator interval before exiting the glenohumeral joint.
The extra-articular portion of the LHB tendon travels in
the bicipital (intertubercular) groove before attaching to the
biceps muscle belly. The bicipital groove is approximately 4
mm in depth with a medial wall angle of approximately 56 .17
The biceps vinculum is a loose membranous tissue
consisting of fat, arteries, and veins that loosely attaches the
biceps tendon to the periosteum of the bicipital groove.30
The transverse humeral ligament is a broad band of tissue
from the greater to the lesser tuberosity overlying the LHB
tendon. The ligament is formed by a continuation of
superﬁcial ﬁbers of the subscapularis that cross the bicipital
groove and attach to the greater tuberosity. The deeper
ﬁbers of the subscapularis insert on the lesser tuberosity.3,29
Longitudinal ﬁbers of the supraspinatus tendon and the
coracohumeral ligament also contribute to the ligament.29
The LHB tendon is innervated by a large neuronal
network of thinly myelinated sensory neurons. The innervation is most dense at the tendon origin.2 The vascular
supply to the LHB tendon is supplied by the thoracoacromial
artery via osteotendinous branches and the brachial artery via
musculotendinous branches. A zone of hypovascularity is
found 1.2 to 3 cm from the tendon origin.16 This zone of
hypovascularity correlates with the portion of the tendon that
starts intra-articularly and courses through the biceps pulley,
ending near the proximal bicipital groove.16
Function
Investigational Review Board approval was not required for this article.
*Reprint requests: Michael Khazzam, MD, UT Southwestern Medical
Center, Sports Medicine & Shoulder Service, 1801 Inwood Rd, Dallas, TX
75390-8883, USA.
E-mail address: [email protected] (M. Khazzam).
The function of the LHB tendon is controversial. Electromyography analysis has shown that the LHB tendon serves
as a glenohumeral joint stabilizer in the unstable
1058-2746/$ - see front matter Ó 2011 Journal of Shoulder and Elbow Surgery Board of Trustees.
doi:10.1016/j.jse.2011.07.016
2
M. Khazzam et al.
Figure 1 Biceps sling ()) is shown of the long head of biceps
tendon (LHBT) as it passes into bicipital groove. SSC, subscapularis tendon.
Figure 2 Superior glenohumeral ligament (SGHL) and its
relationship to the long head biceps tendon (LHBT).
glenohumeral joint but not in the stable shoulder.38
Biomechanical studies showed the LHB tendon was
a dynamic depressor of the humeral head, but this has not
been demonstrated clinically.43,63 Kuhn et al42 demonstrated that the biceps is an important restraint to external
rotation of the abducted arm.
adjacent rotator cuff tendons. Direction of tendon instability was seen as anterior, posterior, or anteroposterior.
Extent of instability was graded as none, subluxation
(deﬁned by translation of less than one-third of the LHB
without passing over the pulley insertion sites of the greater
and or lesser tuberosities), and dislocation. LHB tendon
lesions were graded as grade 0 (normal tendon), grade I
(minor lesion with fraying or erosion involving <50% of
the diameter of the tendon (Fig. 3), grade II (major lesion
fraying or erosion involving >50% of the tendon diameter;
Fig. 4). Status of the rotator cuff was evaluated and graded
as A (intact), B (partial-thickness tear), or C (full-thickness
tear). This was the ﬁrst classiﬁcation system to recognize
the importance of posterior LHB instability and its relationship to tears involving the supraspinatus tendon.
Classiﬁcation
Disorders involving the LHB tendon can be classiﬁed in 3
broad subgroups: (1) inﬂammation, (2) instability, or (3)
traumatic.65 A number of authors12,41,44,53,54,56,58,62 have
further subclassiﬁed biceps tendon lesions according to
anatomic location,12,62 pathologic process,58 and the status
of the biceps tendon.12
Burkhead et al12 and Walch et al62 classiﬁed lesions of
the LHB by the anatomic location of the pathology. Lesions
at the origin affect the biceps anchor at the attachment to
the supraglenoid tubercle and superior labrum. Rotator
interval lesions consist of biceps tendinitis, isolated tendon
ruptures, and subluxation.62 Habermeyer et al32 also classiﬁed biceps pathology by the status of the biceps sling,
which provides information about where the LHB tendon
will sublux. An isolated lesion of the SGHL is group 1,
lesions of the SGHL and partial articular sided tears of the
supraspinatus tendon (PASTA) are group 2 (results in
lateral LHB tendon subluxation), deep surface tears of the
subscapularis and SGHL are group 3 (results in medial
LHB tendon subluxation), and group 4 is when there is
tearing of the deep surface of the subscapularis tendon,
PASTA, and SGHL (LHB tendon can sublux or dislocate
medial or lateral).32
Lafosse et al44 classiﬁed LHB disorders by arthroscopic
ﬁndings related to LHB instability (both direction and
extent), macroscopic lesions of the LHB, and status of the
Pathology
The LHB tendon may be affected by tendinopathy, dislocation, and partial or complete tears. Disorders of the LHB
tendon are associated with rotator cuff tear in up to 90% of
cases.5,51 Glenohumeral arthritis is also commonly associated with LHB tendon pathology.51
LHB tendinopathy is characterized by chronic inﬂammation, ﬁbrotic degeneration, a decrease in the number of
axons in the distal portion of the tendon, and an increase in
the calcitonin gene-related peptide and substance P within
associated nerve roots and blood vessels.51,57 Hypertrophic
tendinopathy may result in entrapment of the tendon within
the glenohumeral joint. The resultant ‘‘hourglass biceps’’
may be unable to slide through the bicipital groove, causing
locking of the shoulder, which has been likened to the
trigger ﬁnger in the hand.8 In addition, inﬂammation and
hypertrophy or stenosis of the bicipital groove, or both, are
thought to incarcerate the LHB tendon, which may produce
Long head biceps tendon review
Figure 3 Fraying of long head of the biceps tendon (LHBT)
involving less than 50% of the tendon diameter. Also note abrasion of the humeral head ()) from the LHBT.
atraumatic superior labrum anteroposterior (SLAP) lesions
and abrasions on the humeral head.13
Rupture of the LHB tendon may be partial thickness or
complete. LHB tendon rupture typically occurs in the
hypovascular zone 1.2 to 3 cm from the tendon origin.16
Partial-thickness rupture is often seen in association with
tendinopathy and contributes to shoulder pain and
dysfunction, with or without concomitant rotator cuff tear.
Complete rupture of the diseased LHB tendon may actually
relieve the symptoms. Complete ruptures may cause biceps
muscle retraction, leading to the ‘‘Popeye’’ deformity.
Hypertrophic tendinopathy may prevent the muscle
retraction in this scenario.
The LHB tendon may be painfully unstable medially or
posterolaterally. Medial instability is associated with
a subscapularis tear, whereas posterolateral instability is
associated with a supraspinatus tear.44 Complete LHB
tendon dislocation is only seen medially.44,49 Medial
dislocation occurs underneath or into the substance of the
subscapularis tendon49 (Fig. 5). The deep ﬁbers of the
subscapularis tendon must be disrupted for medial LHB
tendon dislocation, although the anterior ﬁbers of the subscapularis may remain intact.29
Injury to the biceps reﬂection pulley has been shown to
be necessary for LHB tendon dislocation in either direction.3,11,29,58 The biceps reﬂection pulley is injured by
increased shear load at forward ﬂexion with internal or
neutral arm rotation.11 Although the transverse humeral
ligament was historically thought to contribute to biceps
tendon stability, this function has come into question. The
transverse humeral ligament may actually remain intact,
even in the setting of complete LHB tendon dislocation.49
Diagnosis
Physical examination is unreliable in the diagnosis of LHB
tendon pathology. Anterior shoulder tenderness may
3
Figure 4 Fraying of the long head of the biceps tendon (LHBT)
exceeding 50% of the tendon diameter. HH, humeral head.
indicate LHB tendon pathology. The belly press, Speed’s,
and Yergason’s tests have only moderate speciﬁcity.34,37
The bear hug and upper cut tests are highly sensitive but
have low speciﬁcity.37 In patients with rotator cuff tears, the
physical examination is even less reliable in the detection
of LHB tendon disorders.27
Holtby and Razmjou34 performed a prospective diagnostic study (Level of Evidence [LOE] I) to determine the
accuracy of the Speed’s and Yergason’s tests to detect
biceps pathology by using arthroscopy as the gold standard.
The authors reported 32% sensitivity, 75% speciﬁcity, 56%
accuracy, 50% positive predictive value, 58% negative
predictive value, 1.28 positive likelihood ratio, and 0.91
negative likelihood ratio for the Speed’s test. For the Yergason’s test, they reported 43% sensitivity, 79% speciﬁcity,
63% accuracy, 60% positive predictive value, 65% negative
predictive value, 2.05 positive likelihood ratio, and 0.72
negative likelihood ratio. These results indicated that
neither clinical examination maneuver is very good at
identifying biceps pathology, but when the result one of
these tests is positive, then it is quite reliable that some
shoulder pathology is present. The weakness of this study
was that the authors combined arthroscopic ﬁndings of
SLAP tears and biceps pathology, leaving no information
about which pathology (SLAP tears or LHB tendinitis)
these tests correlate more accurately with clinically.
In a similar study, Kibler et al37 (LOE II) examined the
accuracy of Yergason’s, Speed’s, uppercut, bear hug, belly
press, O’Brien’s, anterior slide, and dynamic labral shear
tests for identiﬁcation of LHB tendon pathology. The
authors reported that the bear hug (sensitivity, 0.79) and
upper cut (sensitivity, 0.73) tests were most sensitive, and
the belly press (speciﬁcity, 0.85) and Speed’s (speciﬁcity,
0.81) tests were most speciﬁc for biceps tendon pathology.
In addition, the upper cut test was test was most accurate
(0.77) and had the highest positive likelihood ratio (3.38).
Further regression analysis found that the combination of
Speed’s and the upper cut tests were signiﬁcantly better at
4
M. Khazzam et al.
Figure 5 Medial subluxation of the long head of the biceps
tendon (LHBT) from bicipital sling disruption. HH, humeral head.
Figure 6 Long head of the biceps tendon (LHBT) inﬂammation
()‘‘lipstick’’ biceps).
detecting biceps pathology (P ¼ .021, R2 ¼ 0.400, 80%
predictability). What makes these data stronger than the
results of the study by Holtby and Razmjou34 is that they
separated out the ﬁndings of biceps lesions and SLAP tears.
The available evidence from LOE I and II studies shows
that the combined use of the Speed’s test and the upper cut
test is recommended for the clinical detection of LHB
tendon pathology.
Magnetic resonance imaging (MRI), with or without
contrast, can be used in the detection of LHB tendon
disorders. Interobserver reliability is low in the detection of
LHB tendon tears.59 Noncontrast MRI may detect LHB
tendon tears with only 52% sensitivity.5 MRI arthrography
may raise the sensitivity to 90%.66 Computed tomography
arthrogram can also be a useful diagnostic tool to evaluate
the status of the LHB tendon. Although sensitivity of this
modality has been poor (31% to 46% sensitive), it was very
speciﬁc for biceps pathology (95% to 99% speciﬁc), with
a positive predictive value of 96% and negative predictive
value 88%.15,20
Ultrasound imaging may reliably detect complete
rupture and dislocation of the LHB tendon but poorly
detects partial-thickness tears of the LHB tendon.4
Armstrong et al4 correlated ultrasound ﬁndings to arthroscopy and found it was 100% speciﬁc and 96% sensitive for
subluxation or dislocation and 50% sensitive and 100%
speciﬁc for identifying pathology with 100% positive
predictive value and 71% negative predictive value. These
results indicate that normal ultrasound ﬁndings do not
guarantee normal biceps anatomy.
Arthroscopy is the gold standard for detection of LHB
tendon disorders. A thorough arthroscopic examination of
the LHB tendon includes using an arthroscopic probe to
pull the extra-articular portion of the biceps into the glenohumeral joint to allow complete visualization of the
tendon. A segment of the LHB tendon that is extra-articular
may be affected by tendinopathy or partial-thickness
tearing and can only be examined by probing the tendon
into the joint (Fig. 6). Surgeons should look for an abrasion
of the cartilage of the humeral head near the bicipital
groove, which may be a sign of biceps pathology.13
Treatment
Isolated LHB lesions are relatively uncommon; therefore, it
is important to properly recognize and treat associated
shoulder pathology such as rotator cuff dysfunction, scapular dyskinesis, adhesive capsulitis, or glenohumeral joint
arthritis. We review the nonoperative and operative treatment options for the LHB. We also have reviewed the
current literature to aid in clinical decision making.
Nonsurgical
Management of patients with symptoms related to inﬂammation or abnormalities involving the LHB tendon should
begin with nonoperative treatment, including activity
modiﬁcation, nonsteroidal anti-inﬂammatory medication,
and physical therapy directed at associated underlying
shoulder disorders as well as scapulothoracic mechanics.
An injection of steroid with local anesthetic may be
beneﬁcial, either intra-articularly into the glenohumeral
joint or directly into the bicipital tendon sheath within the
bicipital groove. Subacromial injection may also be used
for concomitant impingement symptoms. No published
studies have speciﬁcally evaluated the nonoperative treatment of LHB disorders.
Several studies12,14,47,65 report the results after nonsurgical treatment of complete LHB ruptures. Traumatic or
spontaneous ruptures of the LHB are commonly associated
with cosmetic deformity of the biceps muscle, Popeye
deformity, and biceps muscle spasm, particularly in thin
patients. Patients typically report resolution of biceps muscle
Long head biceps tendon review
pain and spasm after 6 to 8 weeks, with little residual
dysfunction related to the LHB. Carroll and Hamilton14
reviewed the nonoperative treatment results of 100 patients
after rupture of the LHB tendon in a case series (LOE IV) and
found no residual disability at ﬁnal follow-up.
Mariani et al,47 in a retrospective case-control study (LOE
III), compared the results of patients who had rupture of the
LHB and were treated with or without biceps tenodesis.
There was no difference in outcome for pain and shoulder or
elbow range of motion. They did ﬁnd a statistically signiﬁcant difference (P < .0001) between groups for patient
satisfaction, with 78% of patients in the surgery group
reporting being much better at an average follow-up of 13
years compared with no patients reporting being much better
(1% stated they were better, 37% the same, and 53% worse)
in the nonsurgery group at an average follow-up of 4.6 years.
In addition, patients in the nonoperative group had an 8%
decrease in elbow ﬂexion strength and a 21% decrease in
forearm supination strength at ﬁnal follow-up.
Deutch et al21 monitored 11 patients (mean age, 59
years) for an average of 1.7 years (range, 0.5-3.5 years) to
determine residual pain and disability after rupture of the
LHB treated nonoperatively (LOE IV). The authors reported continued complaints of pain or disability, or both, at
the ﬁnal follow-up. They found a 23% decrease in supination, 29% decrease in pronation, and 29% decrease in
elbow ﬂexion strength as measured by strain gauge dynamometers and compared with the contralateral unaffected
side. The authors concluded that LHB rupture may not be
completely benign with nonoperative treatment, resulting in
possible continued painful symptoms and dysfunction,
especially in younger patients.
Surgical
Surgical treatment options for LHB tendinitis that remains
symptomatic despite nonoperative treatment consists of
tenotomy or tenodesis in conjunction with addressing
any additional concomitant shoulder pathology, including
impingement syndrome, rotator cuff tears, and labral
pathology. The ideal surgical treatment for LHB
pathology continues to remain an area of signiﬁcant
controversy.25,35,52
Many surgical techniques have been reported for
arthroscopic and open biceps tenotomy and tenodesis.
Tenotomy or tenodesis of the LHB tendon is indicated
when there is partial tearing (>25% to 50% of the tendon
diameter), longitudinal tears that result in poor tendon
gliding in the bicipital groove (symptomatic catching),
medial subluxation of the tendon, disruption of the biceps
sling, or in the setting of a subscapularis tear.12,52,56,65
Several authors25,35,40 have attempted to answer the
question of which surgical treatment is betterdtenotomy or
tenodesisdand provide evidence to support one rather than
the other.
5
Frost et al,25 in a systematic review (LOE IV), hypothesized that there would be no difference in outcome
between tenotomy and tenodesis for treatment of pathology
involving the LHB tendon. This review analyzed the results
of 5 studies that evaluated the results of tenotomy, 8 that
examined tenodesis, and 6 that reported results of both.
Overall, tenodesis resulted in good or excellent results in
40% to 100% of patients, with a failure rate of 5% to 48%.
Tenotomy resulted in 65% to 100% good or excellent
results, with a failure rate of 13% to 35%. The comparative
studies did not show any signiﬁcant differences between the
groups, other than the Popeye sign being present in 3% to
70% of the tenotomy patients. The authors concluded that
the available evidence shows there is little difference in
outcomes using tenodesis or tenotomy; however, welldesigned, prospective, randomized controlled trials
comparing these procedures should be performed to
provide stronger evidence for treatment decisions.
Hsu et al35 conducted a systematic review (LOE IV) to
determine if there were any signiﬁcant differences in the
incident of cosmetic deformity or load-to-tendon failure, or
both, between tenotomy and tenodesis. They included 8
studies in the ﬁnal analysis. Postoperative bicipital pain was
present in 17% of patients in the tenotomy group and in
24% in the tenodesis group. The corresponding odds ratio
(1.5) and relative risk (1.4) reﬂected that bicipital pain was
more common in the tenodesis group. The average
Constant score was 66.9 for the tenotomy group and 76.1
for the tenodesis group, and the average University of
California, Los Angeles (UCLA) score was 33 for the
tenotomy group and 28 for the tenodesis group. In the
tenotomy group, 41% of patients had biceps muscle belly
deformity at ﬁnal follow-up compared with only 25% in the
tenodesis group. The corresponding odds ratio was 2.15 and
the relative risk was 1.7 for the tenotomy group to be more
likely to have a Popeye muscle deformity at ﬁnal follow-up.
Again, evidence-based recommendations are limited due
to the lack of high-quality studies in the current literature.
On the basis of what is available, tenotomy is recommended for older patients with sedentary lifestyle, obese
arms, or those not concerned with cosmesis, and tenodesis
is recommended for young (<40 years) active patients with
high physical demands, thin arms, or concern for cosmesis.
These authors similarly recommended more high-quality,
well-designed, prospective, randomized controlled studies
to help support the decision-making process.
Recently, Koh et al40 compared tenotomy and tenodesis
results in the setting of rotator cuff tears in a prospective
cohort study (LOE II). At a ﬁnal follow-up that averaged
27-months, 5% of patients in the tenodesis group and 10%
in the tenotomy group complained of arm pain during
resisted elbow ﬂexion (P ¼ .43), and 9.3% in the tenodesis
group had a Popeye deformity compared with 26.8% in the
tenotomy group (P ¼ .036, statistically signiﬁcant difference). The American Shoulder and Elbow Surgeons
(ASES) scores improved from 52.1 to 84.7 in the tenodesis
6
group and from 48.1 to 79.6 in the tenotomy group, and
these differences were clinically and statistically signiﬁcant
(P < .0001). Constant scores for tenodesis (38.9 to 82.9)
and tenotomy (35.2 to 78.3) were clinically and statistically
signiﬁcantly improved (P < .0001) in both groups. There
was no difference in outcome scores between the 2 groups.
The authors concluded from these results that besides
biceps deformity, there was no signiﬁcant difference
between these 2 treatment modalities.
Arthroscopic biceps tenotomy
Walch et al61 described arthroscopic biceps tenotomy
for the treatment of biceps tendinopathy in the setting
of massive irreparable rotator cuff tears in 307 shoulders (LOE IV). The tendon is arthroscopically transected at the superior labrum or supraglenoid tubercle
and allowed to retract out of the glenohumeral joint into
the bicipital groove. If this does not occur, the intraarticular portion is removed until the tendon can then
retract out of the joint. At the follow-up evaluation
(average, 57 months), the authors reported statistically
signiﬁcant (P < .0001) improvement in Constant score
(preoperatively, 48.4; postoperatively, 67.6) and that
87% of patients were satisﬁed or very satisﬁed with
their treatment results.
Gill et al28 reported the results after arthroscopic biceps
tenotomy in a retrospective case series (LOE IV) in 30
shoulders at an average follow-up of 19 months (range, 1269 months). They demonstrated improvement in symptoms
at the ﬁnal evaluation, with an ASES score of 81.8, average
time to return to work of 1.9 weeks, and 90% of patients
returned to their previous level of sporting activity with no
complaints of pain. The complication rate was 13.3% (4 of
30 shoulders), 2 related to impingement, 1 with persistent
pain, and 1 as a result of the cosmetic biceps deformity,
with no pain, who was subsequently revised to a tenodesis.
Kelly et al36 reviewed the results after arthroscopic
biceps tenotomy in 40 patients with an average follow-up of
2.7 years (range, 24-42 months) in a retrospective case
series (LOE IV). The authors reported the following
outcomes scores: L’Insalata, 75.6 (range, 29.1-100),
UCLA, 27.6 (range, 10-35), and ASES, 77.6 (range 13.3100). They also found that 70% of patients had a Popeye
deformity at rest or during active elbow ﬂexion, and 37.5%
of patients complained of fatigue, discomfort, and soreness
after resisted elbow ﬂexion. The authors concluded that this
technique might be useful for a select patient population,
speciﬁcally an older less active patient population with no
concern for cosmesis.
Bradbury et al10 described an arthroscopic biceps
tenotomy technique in which the tendon is released with
a portion of the superior labrum to prevent the Popeye
deformity. Tenotomy performed in this fashion creates
a bulbous proximal biceps stump that becomes entrapped in
the intra-articular portion of the bicipital groove, thereby
limiting tendon excursion and retraction down the arm.
M. Khazzam et al.
Biomechanical testing of this technique showed signiﬁcantly increased force was required to pull the tendon
through the bicipital groove (73.2 N, with-labrum group;
25.0 N, without-labrum group; P ¼ .001).
Open biceps tenodesis
Open biceps tenodesis historically has been the surgical
treatment of choice for young active patients, such as
athletes and heavy laborers, and those who wish to avoid
cosmetic deformity. The goal of tenodesis of the LHB
tendon is to maintain the length-tension relationship of the
biceps muscle.52 A variety of reattachment sites have been
described, including ﬁxation to the lesser tuberosity,12
coracoid,12,31 intertubercular (bicipital) groove33; suture
ﬁxation to the transverse humeral ligament, short head of
the biceps,6 or pectoralis major tendon1,6,47; or by a bone
tunnel in a subpectoralis major location.33,48 Ultimately, the
biggest controversy exists about optimal location of
tenodesis and if ﬁxation should occur above, within, or
below the bicipital groove. To our knowledge, no LOE I or
II studies have compared the outcomes after tenodesis
above, within, or below (subpectoral) the bicipital groove to
guide surgical decision making using an evidenced- based
approach.
Becker et al6 evaluated the results of 51 patients in
a retrospective case series (LOE IV) after open biceps
tenodesis at an average follow-up of 7 years (range, 1-21
years). Three different surgical techniques were used in
these patients: (1) suture of the tendon stump in or adjacent to the bicipital groove, as described by Hitchcock
et al,33 (2) a keyhole technique,12 or (3) side-to-side
suturing to the short head of the biceps.12 All patients
underwent tenotomy with resection of the intra-articular
potion of the LHB tendon in conjunction with tenodesis.
At ﬁnal follow-up, only 22 of the 51 shoulders (43%)
were still pain-free, whereas 26 (50%) continued to have
moderate to severe pain. The authors concluded that
biceps tenodesis should not be recommended as a primary
procedure but should be performed in conjunction with
a larger surgical procedure.
Crenshaw and Kilgore19 retrospectively reviewed (LOE
IV) the results of 3 tenodesis techniques in 92 shoulders at
1 year. The techniques included suturing LHB tendon to the
lesser tuberosity without tenotomy, transfer of the tendon to
the coracoid, or the Hitchcock procedure,12,33 in which the
tendon is reattached into the bicipital groove beneath
a periosteal ﬂap. Pain relief was good to excellent in 90%
of patients, but only 85% had return of their normal
shoulder motion at ﬁnal follow-up.
Froimson and Oh,24 Dines et al,22 and Berlemann and
Bayley,7 in small retrospective case series (LOE IV),
reviewed the results of keyhole tenodesis and found good to
excellent results at ﬁnal follow-up. These authors attributed
clinical failures of this technique to unrecognized additional shoulder pathologies such as subacromial impingement with rotator cuff tearing.
Long head biceps tendon review
Subpectoral biceps tenodesis
Subpectoral biceps tenodesis removes the tendon
completely from the bicipital groove, thus avoiding leaving
behind inﬂamed tendon in this area, which could cause
residual painful symptoms from persistent stenosis or
tenosynovitis. Mazzocca et al,48 in a prospective case series
(LOE IV), reviewed the results after subpectoral biceps
tenodesis with interference screw ﬁxation in 41 patients at
an average follow-up of 29 months (range, 12-49 months).
Mean scores were Rowe, 86 (range, 67-100); ASES, 81
(range, 32-100); Simple Shoulder Test, 9 (range, 3-12);
Constant score, 84 (range, 67-100); and Single Assessment
Numeric Evaluation score, 84 (range 50-100). There was
only 1 failure, with loss of ﬁxation at the bone tunnel
resulting in Popeye deformity on physical examination. No
patients reported any residual pain in the area of the
bicipital groove at ﬁnal follow-up. The authors concluded
that this procedure reliably and successfully eliminates
symptoms associated with biceps tendinitis and maintains
the anatomic length-tension relationship of the muscle.
In a similar, retrospective case-control study (LOE III),
Millett et al50 compared subpectoral suture ﬁxation vs
interference screw ﬁxation for LHB tenodesis. The authors
reviewed 88 patients (34 interference screw, 54 suture
anchor ﬁxation) at an average 13-month follow-up and
found no ﬁxation failures. All patients demonstrated clinical and statistically signiﬁcant improvement in scores
preoperatively to postoperatively, respectively, of 9 to 2 on
the visual analog scale (VAS) score (P < .0001), 28 to 76
on the ASES (P < .0001), and 29 to 59 on the Constant
(P < .0001). In addition, there were no statistically or
clinically signiﬁcant differences between the groups on the
VAS (P ¼ .04), ASES (P ¼ .2), and Constant (P ¼ .09)
scores. The authors concluded that both ﬁxation methods
provide a reliable means for treating symptomatic biceps
tendinitis by alleviating pain and no Popeye deformity at
short-term follow-up.
Arthroscopic biceps tenodesis
Arthroscopic biceps tenodesis may be performed using
interference screws, suture anchors, or soft tissue ﬁxation.
Arthroscopic tenodesis of the LHB tendon was described
by Boileau et al,9 in which the tenotomy was performed
and a bioabsorbable interference screw was used to anchor
the tendon stump to the proximal aspect of the bicipital
groove. The authors also reported the results after this
procedure in 43 patients at an average of 17 months (range,
12-34 months) in a prospective case series (LOE IV). The
average Constant score was signiﬁcantly improved clinically and statistically from 43 preoperatively (range, 13-60)
to 79 postoperatively (range, 59-87; P < .005). Strength
averaged 90% of the contralateral arm, and biceps shape
and contour was maintained in 41 of 43 patients (95%).
Drakos et al,23 in a retrospective case series (LOE IV),
reported the results of arthroscopic transfer of the LHB
tendon to the conjoint tendon ‘‘subdeltoid’’ in 40 shoulders
7
at an average follow-up of 28 months (range, 24-53
months). The average L’Insalata score was 78.9 (range,
35.7-100), UCLA score was 27.8 (range, 12-35), and ASES
score was 79.6 (range, 30-100) at the ﬁnal follow-up. Of the
40 shoulders, only 2 patients (5%) had a Popeye deformity
at rest or with active elbow ﬂexion, 5 (12.5%) had fatigue
soreness after resisted elbow ﬂexion, 38 (95%) reported
relief of tenderness on palpation of the bicipital groove, and
32 (80%) rated their shoulders as good, very good, or
excellent.
Several authors26,39,45,55 have reported arthroscopic
techniques for performing this procedure (LOE V), with no
real outcomes data to support or refute its use.
Once the decision has been made to perform tenodesis
of the LHB tendon, the next controversial question that
arises is: what is the optimal location of tenodesis? Lutton
et al,46 in a retrospective case-control study (LOE III),
compared results after tenodesis into the upper or lower
half of the bicipital groove to determine if location was
related to residual postoperative pain. A more distal location resulted in a decreased incidence of persistent postoperative pain in the bicipital groove. This study was
limited due to its retrospective nature and limited sample
size, but to our knowledge, this is the only study in the
literature that attempts to answer this important question.
Treatment algorithm
The diagnosis and management of disorders involving the
LHB tendon can be challenging. We present an algorithm
for the management of these patients (Fig. 7). It is clear
from the current literature that the approach to patients who
present with biceps pathology begins with an accurate
diagnosis, which can be the most challenging aspect in the
management of biceps pathology. Physical examination
ﬁndings with use of a biceps-speciﬁc test, such as the upper
cut test and Speed’s test, may provide useful clues about the
status of the LHB tendon. Imaging of the shoulder with
MRI, CT arthrogram, or ultrasound provides detailed
visualization of the anatomy of the shoulder and should be
considered a critical part of the workup. In our practice, we
typically use MRI as our diagnostic study of choice.
Treatment should initially begin with a trial of conservative measures such as nonsteroidal anti-inﬂammatory
medications and physical therapy. Physical therapy should
be tailored to other concomitant shoulder pathology with
a focus on regaining range of motion and eliminating
scapular dyskinesis. Corticosteroid injections placed in the
glenohumeral joint, subacromial space, or biceps tendon
sheath may be considered initially or after a 6-week trial of
physical therapy.
Surgical intervention is appropriate a minimum of 12
weeks of conservative management has failed. Surgical
treatment is generally dictated by the major underlying
shoulder pathology that coexists with LHB tendon
8
M. Khazzam et al.
Figure 7 Treatment algorithm for the long head of the biceps (LHB). MRI, magnetic resonance imaging; NSAIDS, nonsteroidal antiinﬂammatory drugs; OA, osteoarthritis; PT, physical therapy; RC, rotator cuff; RCT, rotator cuff tear.
abnormalities, including rotator cuff tear, osteoarthritis,
adhesive capsulitis, and SLAP tears. Age and activity level
may also help determine which biceps procedure would be
appropriate (tenotomy vs tenodesis; open vs arthroscopic).
In our hands, how the biceps is managed surgically is
dictated by patient activity level and functional demands and
the other shoulder pathology being treated. We believe an
open subpectoral biceps tenodesis is appropriate for an active
patient with high functional demands. In older patients (>60
years) with low functional demands, we recommend a biceps
tenotomy with a cuff of labrum using the technique described
by Bradbury et al.10 Finally, in the setting of other shoulder
pathology, such as a full-thickness rotator cuff tear, in our
opinion, it is appropriate to incorporate the tenodesis into the
anterior suture anchors for the cuff repair.
Conclusion
Patients presenting with symptomatic LHB tendon
lesions must be thoroughly evaluated for other
concomitant shoulder pathology, which may make
a deﬁnitive diagnosis challenging. There are few highlevel evidence studies in the current literature to support
decision-making strategies. Thorough clinical evaluation
Long head biceps tendon review
should include the Speed’s and upper cut tests. Treatment should begin with nonoperative treatment,
including activity modiﬁcation, physical therapy, antiinﬂammatories, and injections. Surgery should be
considered as a ﬁrst-choice treatment option in a select
subset of patients who have biceps pathology associated
with full-thickness subscapularis tears or painful
subluxation or dislocation of the tendon as a result of an
acute injury. Arthroscopy is the gold standard to identify
LHB tendon pathology.
There appears to be no difference in surgical treatment outcomes after LHB tenotomy or tenodesis. Clinical decision making should be tailored to individual
patients and their expectations and physical needs after
surgical intervention. Well-designed prospective cohort
or randomized controlled trails need to be performed to
provide evidence to support treatment protocols.
9
9.
10.
11.
12.
13.
14.
15.
Disclaimer
16.
The authors, their immediate families, and any research
foundations with which they are afﬁliated have not
received any ﬁnancial payments or other beneﬁts from any
commercial entity related to the subject of this article.
17.
18.
References
19.
20.
1. Ahrens P, Boileau P. The long head of biceps and associated tendinopathy. J Bone Joint Surg Br 2007;89:1001-9. doi:10.1302/0301620X.89B8
2. Alpantaki K, McLaughlin D, Karagogeos D, Hadjipavlou A,
Kontakis G. Sympathetic and sensory neural elements in the tendon of
the long head of the biceps. J Bone Joint Surg Am 2005;87:1580-3.
doi:10.2106/JBJS.D.02840
3. Arai R, Mochizuki T, Yamaguchi K, Sugaya H, Kobayashi M,
Nakamura T, et al. Functional anatomy of the superior glenohumeral
and coracohumeral ligaments and the subscapularis tendon in view of
stabilization of the long head of the biceps tendon. J Shoulder Elbow
Surg 2010;19:58-64. doi:10.1016/j.jse.2009.04.001
4. Armstrong A, Teffey S, Wu T, Clark A, Middleton W, Yamaguchi K,
et al. The efﬁcacy of ultrasound in the diagnosis of long head of the
biceps tendon pathology. J Shoulder Elbow Surg 2006;15:7-11. doi:10.
1016/j.jse.2005.04.008
5. Beall D, Williamson E, Ly J, Adkins M, Emery R, Jones T, et al.
Association of biceps tendon tears with rotator cuff abnormalities:
degree of correlation with tears of the anterior and superior portions of
the rotator cuff. AJR Am J Roentgenol 2003;180:633-9.
6. Becker D, Coﬁeld R. Tenodesis of the long head of the biceps brachii
for chronic bicipital tendinitis: long-term results. J Bone Joint Surg
Am 1989;71:376-81.
7. Berlemann U, Bayley I. Tenodesis of the long head of biceps brachii in
the painful shoulder: improving results in the long term. J Shoulder
Elbow Surg 1995;4:429-35.
8. Boileau P, Ahrens P, Hatzidakis A. Entrapment of the long head of the
biceps tendon: the hourglass bicepsda cause of pain and locking of
21.
22.
23.
24.
25.
26.
27.
28.
the shoulder. J Shoulder Elbow Surg 2004;13:249-57. doi:10.1016/j.
jse.2004.01.001
Boileau P, Krishnan S, Coste J, Walch G. Arthroscopic biceps
tenodesis: a new technique using bioabsorbable interference screw
ﬁxation. Arthroscopy 2002;18:1002-12. doi:10.1053/jars.2002.36488
Bradbury T, Dunn W, Kuhn J. Preventing the Popeye deformity after
release of the long head of the biceps tendon: an alternative technique
and biomechanical evaluation. Arthroscopy 2008;24:1099-102. doi:10.
1016/j.arthro.2008.06.002
Braun S, Millett P, Yongpravat C, Pault J, Anstett T, Torry M, et al.
Biomechanical evaluation of shear force vectors leading to injury of
the biceps reﬂection pulley: a biplane ﬂuoroscopy study on cadaveric
shoulders. Am J Sports Med 2010;38:1015-24. doi:10.1177/
0363546509355142
Burkhead W, Habermeyer P, Walch G, Lin K. The biceps tendon. In:
Rockwood C, Matsen F, Wirth M, Lippitt S, editors. The shoulder. 4th
ed, Vol 2. Philadelphia: Saunders Elsevier; 2009. p. 1309-60. ISBN:
978-1-4160-3427-8.
Byram I, Dunn W, Kuhn J. Humeral head abrasion: an association with
failed superior labrum anterior posterior repairs. J Shoulder Elbow
Surg 2011;20:92-7. doi:10.1016/j.jse.2010.05.013
Carroll R, Hamilton L. Rupture of biceps brachii. a conservative
method of treatment. J Bone Joint Surg Am 1967;49:1016.
Charousset C, Bellaiche L, Duranthon L, Grimberg J. Accuracy of CT
Arthrography in the Assessment of Tears of the Rotator Cuff. J Bone
Joint Surg Br 2005;87:824-8. doi:10.1302/0301-620X.87B6
Cheng N, Pan W, Vally F, Le Roux C, Richardson M. The arterial
supply of the long head of biceps tendon: anatomical study with
implications for tendon rupture. Clin Anat 2010;23:683-92. doi:10.
1002/ca.20992
Cone R, Danzig L, Resnick D, Goldman A. The bicipital groove:
radiographic, anatomic, and pathologic study. AJR Am J Roentgenol
1983;41:781-8.
Cooper D, Arnoczky S, O’Brien S, Warren R, DiCarlo E, Allen A.
Anatomy, histology, and vascularity of the glenoid labrum: an anantomical study. J Bone Joint Surg Am 1992;74:46-52.
Crenshaw A, Kilgore W. Surgical treatment of bicipital tenosynovitis.
J Bone Joint Surg Am 1966;48:1496-502.
De Maeseneer M, Boulet C, Pouliart N, Kichouh M, Buls N,
Verhelle F, et al. Assessment of the long head of the biceps tendon of
the shoulder with 3T magnetic resonance arthrography and CT
arthrography. Eur J Radiol 2011;. doi:10.1016/j.ejrad.2011.01.121
Deutch S, Gelineck J, Johannsen H, Sneppen O. Permanent disabilities
in the displaced muscle from rupture of the long head tendon of the
biceps. Scand J Sci Sports 2005;15:159-62. doi:10.1111j.1600-0838.
2004.00421.x
Dines D, Warren R, Inglis A. Surgical treatment of lesions of the long
head of the biceps. Clin Orthop Relat Res 1982;164:165-71.
Drakos M, Verma N, Gulotta L, Potucek F, Taylor S, Fealy S, et al.
Arthroscopic transfer of the long head of the biceps tendon: functional
outcome and clinical results. Arthroscopy 2008;24:217-23. doi:10.
1016/j.arthro.2007.07.030
Froimson A, Oh I. Kethole tenodesis of the biceps origin at the
shoulder. Clin Orthop Relat Res 1975;112:245-9.
Frost A, Zafar M, Maffulli N. Tenotomy versus tenodesis in the
management of pathologic lesions of the tendon of the long head of
the biceps brachii. Am J Sports Med 2009;37:828-33. doi:10.1177/
0363546508322179
Gartsman G, Hammerman S. Arthroscopic biceps tenodesis: operative
technique. Arthroscopy 2000;16:550-2. doi:10.1053/jars.2000.438
Gill H, El Rassi G, Bahk M, Castillo R, McFardland E. Physical
examination for partial tears of the biceps tendon. Am J Sports Med
2007;35:1334-40. doi:10.1177/0363546507300058
Gill T, McIrvin E, Mair S, Hawkins R. Results of biceps tenotomy
for treatment of pathology of the long head of the biceps brachi.
J Shoulder Elbow Surg 2001;10:247-9. doi:10.1067/mse.2001.
114259
10
29. Gleason P, Beall D, Sanders T, Bond J, Ly J, Holland L, et al. The
transverse humeral ligament: a separate anatomical structure or
a continuation of the osseous attachment of the rotator cuff? Am J
Sports Med 2006;34:72-7. doi:10.1177/0363546505278698
30. Gothelf T, Bell D, Goldberg J, Harper W, Pelletier M, Yu Y, et al.
Anatomic and biomechanical study of the biceps vinculum, a structure
within the biceps sheath. Arthroscopy 2009;25:515-21. doi:10.1016/j.
arthro.2008.10.026
31. Gumina S, Carbone S, Perugia D, Perugia L, Postacchini F. Rupture of
the long head biceps tendon treated with tenodesis to the coracoid
process: results at more than 30 years. Int Orthop 2011;35:713-6. doi:
10.1007/s00264-010-1099-0
32. Habermeyer P, Magosch P, Pritsch M, Scheibel M, Lichtenberg S.
Anterosuperior impingement of the shoulder as a result of pulley
lesions: a prospective arthroscopic study. J Shoulder Elbow Surg 2004;
13:5-12. doi:10.1016/j.jse.2003.09.013
33. Hitchcock H, Bechtol C. Painful shoulder: observations on the role of
the tendon of the long head of the biceps brachii in its causation. J
Bone Joint Surg Am 1948;30:263-73.
34. Holtby R, Razmjou H. Accuracy of the Speed’s and Yergason’s tests in
detecting biceps pathology and SLAP lesions: comparison with
arthroscopic ﬁndings. Arthroscopy 2004;20:231-6. doi:10.1016/j.
arthro.2004.01.008
35. Hsu A, Ghodadra N, Provencher M, Lewis P, Bach B. Biceps tenotomy versus tenodesis: a review of clinical outcomes and biomechanical results. J Shoulder Elbow Surg 2011;20:326-32. doi:10.1016/j.jse.
2010.08.019
36. Kelly A, Drakos M, Fealy S, Taylor S, O’Brien S. Arthroscopic release
of the long head of the biceps tendon: functional outcome and
clinical results. Am J Sports Med 2005;33:208-13. doi:10.1177/
0363546504269555
37. Kibler W, Sciascia A, Hester P, Dome D, Jacobs C. Clinical utility of
traditional and new tests in the diagnosis of biceps tendon injuries and
superior labrum anterior and posterior lesions in the shoulder. Am J
Sports Med 2009;37:1840-7. doi:10.1177/0363546509332505
38. Kim S, Ha K, Kim H, Kim S. Electromyographic activity of the biceps
brachii muscle in shoulders with anterior instability. Arthroscopy
2001;17:864-8. doi:10.1053/jars.2001.19980
39. Klepps S, Hazrati Y, Flatow E. Arthroscopic biceps tenodesis.
Arthroscopy 2002;18:1040-5. doi:0.1053/jars.2002.36467
40. Koh K, Ahn J, Kim S, Yoo J. Treatment of biceps tendon lesions in the
setting of rotator cuff tears: prospective cohort study of tenotomy
versus tenodesis. Am J Sports Med 2010;38:1584-90. doi:10.1177/
0363546510364053
41. Krupp R, Kevern M, Gaines M, Kotara S, Singleton SB. Long head of
the biceps tendon pain: differential diagnosis and treatment. J Orthop
Sports Phys Ther 2009;39:55-70. doi:10.2519/jospt.2009.2802
42. Kuhn J, Huston L, Blasier R, Soslowsky L. Ligamentous restraints and
muscle effects limiting external rotation of the glenohumeral joint in
the neutral and abducted positions. J Shoulder Elbow Surg 2005;14(15
Suppl):39-48S. doi:10.1016/j.jse.2004.09.016
43. Kumar V, Satku K, Balasubramaniam P. The role of the long head of
biceps brachii in the stabilization of the head of the humerus. Clin
Orthop Relat Res 1989;244:172-5.
44. Lafosse L, Reiland Y, Baier G, Toussaint B, Jost B. Anterior and
posterior instability of the long head of the biceps tendon in rotator
cuff tears: a new classiﬁcation based on arthroscopic observations.
Arthroscopy 2007;23:73-80. doi:10.1016/j.arthro.2006.08.025
45. Lo I, Burkhart S. Arthroscopic biceps tenodesis using a bioabsorbable
interference screw. Arthroscopy 2004;20:85-95. doi:10.1016/j.arthro.
2003.11.017
46. Lutton D, Gruson K, Harrison A, Gladstone J, Flatow E. Where to
tenodese the biceps: proximal or distal? Clin Orthop Relat Res 2011;
469:1050-5. doi:10.1007/s11999-010-1691-z
M. Khazzam et al.
47. Mariani E, Coﬁeld R, Askew L, Li G, Chao E. Rupture of the tendon
of the long head of the biceps brachii: surgical versus nonsurgical
treatment. Clin Orthop Relat Res 1988;228:233-9.
48. Mazzocca A, Cote M, Arciero C, Romeo A, Arciero R. Clinical
outcomes after subpectoral biceps tenodesis with an interference
screw. Am J Sports Med 2008;36:1922-9. doi:10.1177/
0363546508318192
49. Meyer A. Spontaneous dislocation and destruction of tendon of long
head of biceps brachii: ﬁfty-nine instances. Arch Surg 1928;17:493506.
50. Millett P, Sanders B, Gobezie R, Braun S, Warner J. Interference screw
vs. suture anchor ﬁxation for open subpectoral biceps tenodesis: does
it matter? BMC Musculoskelet Disord 2008;9:121. doi:10.1186/14712474-9-121
51. Murthi A, Vosburgh C, Neviaser T. The incidence of pathologic
changes of the long head of the biceps tendon. J Shoulder Elbow Surg
2000;9:382-5. doi:10.1067/mse.2000.108386
52. Nho S, Strauss E, Lenart B, Provencher M, Mazzocca A, Verma N,
et al. Long head of the biceps tendinopathy: diagnosis and management. J Am Acad Orthop Surg 2010;18:645-56.
53. Patton W, McCluskey G. Biceps tendinitis and subluxation. Clin
Sports Med 2001;20:505-29.
54. Post M, Benca P. Primary tendinitis of the long head of the biceps.
Clin Orthop Relat Res 1989;246:117-25.
55. Sekiya J, Elkousy H, Rodosky M. Arthroscopic biceps tenodesis using
the percutaneous intra-articular transtendon technique. Arthroscopy
2003;19:1137-41. doi:10.1016/j.arthro.2003.10.022
56. Sethi N, Wright R, Yamaguchi K. Disorders of the long head of the
biceps tendon. J Shoulder Elbow Surg 1999;8:644-54.
57. Singaraju V, Kang R, Yanke A, McNickle A, Lewis P, Wang V, et al.
Biceps tendinitis in chronic rotator cuff tears: a histologic perspective.
J Shoulder Elbow Surg 2008;17:898-904. doi:10.1016/j.jse.2008.05.
044
58. Slatis P, Aalto K. Medial disloction of the tendon of the long head of
the biceps brachii. Acta orthop Scand 1979;50:73-7.
59. Spencer E, Dunn W, Wright R, Wolf B, Spindler K, McCarty E, et al.
Shoulder Multicenter Orthopaedic Outcomes Network. Interobserver
agreement in the classiﬁcation of rotator cuff tears using magnetic
resonance imaging. Am J Sports Med 2008;36:99-103. doi:10.1177/
0363546507307504
60. Tuoheti Y, Itoi E, Minagawa H, Yamamoto N, Saito H, Seki N, et al.
Attachment types of the long head of the biceps tendon to the glenoid
labrum and their relationships with the glenohumeral ligaments.
Arthroscopy 2005;21:1242-9. doi:10.1016/j.arthro.2005.07.006
61. Walch G, Edwards T, Boulahia A, Nove-Josserand L, Neyton L, Szabo I.
Arthroscopic tenotomy of the long head of the biceps in the treatment of
rotator cuff tears: clinical and radiographic results of 307 cases. J
Shoulder Elbow Surg 2005;14:238-46. doi:10.1016/j.jse.2004.07.008
62. Walch G, Nove-Josserand L, Boileau P, Levigne C. Subluxations and
dislocations of the tendon of the long head of the biceps. J Shoulder
Elbow Surg 1998;7:100-8.
63. Warner J, McMahon P. The role of the long head of the biceps brachii
in superior stability of the glenohumeral joint. J Bone Joint Surg Am
1995;77:366-72.
64. Werner A, Mueller T, Boehm D, Gohlke F. The stabilizing sling for the
long head of the biceps tendon in the rotator cuff interval: a histoanatomic study. Am J Sports Med 2000;28:28-31.
65. Yamaguchi K, Keener J, Galatz L. Disorders of the biceps tendon. In:
Iannotti J, Williams G, editors. Disorders of the shoulder: diagnosis
and management. 2nd ed, Vol 1. Philadelphia: Lippincott, Williams, &
Wilkins; 2007. p. 218-60. ISBN: 0-7817-5678-2.
66. Zanetti M, Weishaupt D, Gerber C, Hodler J. Tendinopathy and rupture
of the tendon of the long head of the biceps brachii muscle: evaluation
with MR Arthrography. AJR Am J Roentgenol 1998;170:1557-61.