Article Text

Download PDFPDF

Temporal sequence of greyscale ultrasound changes and their relationship with neovascularity and pain in the patellar tendon
  1. P Malliaras1,
  2. C Purdam2,
  3. N Maffulli3,
  4. J Cook4
  1. 1Brunel University, Uxbridge, Middlesex, UK
  2. 2Australian Institute of Sport, Bruce, ACT, Australia
  3. 3Keele University, Keele, UK
  4. 4Deakin University, Burwood, Victoria, Australia
  1. Correspondence to P Malliaras, Brunel University, Uxbridge, Middlesex, UK, UB8 3PH; p.malliaras{at}


Objective Ultrasound (US) imaging has revealed distinct types of greyscale abnormalities in the patellar and Achilles tendons, including diffusely thickened tendons and tendons containing a hypoechoic region. The relationship between these greyscale abnormalities and their clinical relevance is unknown. This study investigated the temporal sequence in greyscale abnormalities as well as the relationship between greyscale abnormalities, Doppler flow and pain.

Methods Patellar tendon pain (single leg decline squat test) and ultrasound imaging (greyscale (normal, diffuse thickening, hypoechoic), presence of Doppler flow) were assessed bilaterally among 58 volleyball players at monthly intervals during a 5-month season. The probability of transition between the greyscale ultrasound groups was calculated for each month and totalled over the season (prospective) and the relationship between these groups and the presence and intensity of pain and the presence of Doppler flow were investigated (cross-sectional investigation).

Results Tendons with normal US were more likely to transition to diffuse thickening than to a hypoechoic region. Tendons containing a hypoechoic region were more likely to transition to diffuse thickening rather than to a normal US appearance. Hypoechoic regions were more likely to be painful (59%) and contain Doppler flow (42%) than tendons with diffuse thickening (pain in 43% and Doppler flow in 6%).

Conclusions The transitions identified between normal, diffusely thickened tendons and those containing a hypoechoic region suggests that these greyscale US changes may represent different phases of tendon pathology. Tendons containing a hypoechoic region are more likely to be painful and contain Doppler flow than diffusely thickened tendons.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Patellar tendon pain is common among jumping athletes, including volleyball and basketball players. Some affected athletes experience a reduction in function, and some are unable to compete.1 Management of painful tendinopathy is often difficult,2 and may require several months of rehabilitation.3 Painful tendinopathy has typical clinical features, including pain localised to the inferior patellar pole and pain with patellar tendon loading.4

Ultrasound (US) imaging is used diagnostically in tendinopathy. Greyscale US abnormalities and changes in MRI signals tend to precede tendon pain for a variable time,5 and therefore have been reported to have limited diagnostic utility.6,,9 In contrast, neovascularity demonstrated on Doppler US is a common imaging finding that has been associated with tendon pain10,,12 more so than athletic cohorts,13,,15 and thereby has some diagnostic utility.

There are, however, distinct types of abnormalities seen on greyscale US, and their importance in the diagnosis and treatment of tendinopathy is unclear. In the patellar and Achilles tendons, diffuse thickening (associated with reduced echogenicity) as well as focal hypoechoic regions have been reported.5 8 15 In the patellar tendon, longstanding pain tends to be associated with a hypoechoic region,16 17 whereas hypoechoic regions and diffuse thickening have been reported among active athletes with or without pain.8 15 It is not known, however, whether greyscale US abnormalities change in appearance over time and, if so, if there is a customary sequence of change.

The relationship between greyscale US, neovascularity and patellar tendon pain also requires further clarification. This relationship may indicate the clinical relevance of each type of greyscale US abnormality. Neovascularity has been reported among Achilles tendons with diffuse thickening and hypoechoic regions.10 15 18 In the Achilles tendon, neovascularity is more likely to be associated with pain in the presence of an underlying hypoechoic region10 compared with diffuse thickening,15 suggesting that the likelihood of pain may differ between the two greyscale US abnormalities.

This study sought to clarify whether greyscale US abnormalities change over time. To achieve this, tendons must be imaged longitudinally to record changes in greyscale appearance over time. Active volleyball players were investigated as they have a high prevalence of patellar tendinopathy and are likely to demonstrate a range of imaging appearances. The relationship between greyscale US abnormality, neovascularity and pain was also investigated.



A convenience sample of men and women competing in a State League volleyball competition (elite to recreational) were recruited. Players under 18 years of age were excluded as they may have juvenile disorders, such as Osgood–Schlatter disease, that may be difficult to differentiate from patellar tendinopathy. The La Trobe University (Melbourne, Australia) human ethics committee approved the study and participants provided informed consent before commencement of the study.

Study design

Pain and US (greyscale and Doppler) were assessed at monthly intervals over a 5-month indoor volleyball season (six measurement occasions).

Pain assessment

The single leg decline squat (SLDS) test assessed pain at each month. This test was designed to preferentially load the patellar tendon and has been shown to increase patellar tendon load when compared to squatting on flat ground.19 20 Participants performed a single leg squat to of 60° knee flexion on a 25° decline board and pain at the proximal insertion of the patellar tendon was rated on a 100 mm visual analogue scale (0=no pain, 100=worse pain ever). Both sides were assessed for each participant.

US examination

Bilateral patellar tendon scans were performed with a high-resolution 13.5-MHz US machine (Siemens Accuson CV 70, Siemens AG, Berlin, Germany) by a single experienced musculoskeletal ultrasonographer. Colour Doppler settings were standardised with a gain of 68 dB, sensitivity of 8 cm/s and pulse repetition frequency of 1250 Hz.

The ultrasonographer, who was blind to the participants' pain status, performed greyscale and Doppler imaging. Tendons were placed into one of three greyscale imaging categories (fig 1); normal, diffuse thickening (thickened tendon with diffuse change in echogenicity) or hypoechoic (thickened tendon with focal change in echogenicity). Greyscale imaging was confirmed in the sagittal and axial planes. Tendons were categorised as vascular and non-vascular on Doppler imaging. Vascular tendons contained at least one vessel in the sagittal plane that was estimated to be greater than 1 mm in length. Doppler imaging was only investigated in the sagittal plane in this study, as normal vessels run in the sagittal plane, parallel to collagen fibres.

Figure 1

Ultrasound imaging appearance of a normal patellar tendon (A), diffuse thickening (B) and a hypoechoic region (C) (D, diffuse thickening; H, hypoechoic region; P, patella; T, tendon).

Data analysis

Data were analysed with SPSS (V.13.0; SPSS, Chicago, Illinois, USA). The frequency and probability of transition in each greyscale US group was calculated between each month. Frequency and probability of transitions were totalled over the 6 months (five transitions; one to two, two to three, three to four, four to five, five to six). Analysis using χ2 was used to investigate the relationship between greyscale US (normal, diffuse thickening, hypoechoic), neovascularity (present, absent) and pain (present, absent) as a previous study among the same cohort found no overall increase in the prevalence of pain and US abnormality over the season.5 The mean pain intensity among tendons in each greyscale US group was compared with one-way ANOVA. As neovascularity accompanied with a hypoechoic region has been associated with the presence of pain,10,,12 the intensity of pain was also compared between hypoechoic tendons with and without neovascularity (t test). The α level was set at 0.05 for all analyses.


A total of 58 athletes (116 tendons), including 36 men and 22 women (mean (SD) age;26.1 (5.3) years, mean height 181.6 (10.0) cm and mean body weight; 80.0 (12.3) kg), were investigated over the 5-month volleyball season.

Temporal sequence of greyscale US

The frequency and probability of monthly change in greyscale US were totalled for the 5-month season (table 1). Tendons in each greyscale US group were likely to remain unchanged, particularly tendons with normal greyscale US (probability=0.78) and hypoechoic regions (probability=0.81). When they did change, tendons with normal greyscale imaging were more likely to develop diffuse thickening (probability=0.20) rather than a hypoechoic region (probability=0.02). Likewise, tendons with a hypoechoic region were more likely to develop diffuse thickening (probability=0.16) rather than become normal (probability=0.03). Only 9 out of 580 tendon transitions (2%) involved a change from normal to hypoechoic or vice versa over a 1-month period.

Table 1

Combined 5-month transition frequencies (probability) for greyscale ultrasound (US)

Relationship between greyscale and neovascularity and pain

A significant relationship was identified between greyscale US and pain (table 2). Pain was least likely among tendons with normal greyscale US (24%), more common among tendons with diffuse thickening (43%) and most common among tendons containing a hypoechoic region (59%). There was also a significant relationship between greyscale US and neovascularity (table 3). Although neovascularity was absent among tendons with normal greyscale US and uncommon among those with diffuse thickening (6%), almost half (42%) of the tendons containing a hypoechoic region also contained neovascularity. Although pain intensity was greater among tendons that were abnormal on ultrasound (diffuse thickening, hypoechoic) compared with normal tendons, this difference was not significant (table 4). Pain intensity did not differ significantly (t test=1.7, p=0.17) among hypoechoic tendons with neovascularity (n=9, mean=49.4, SD=18.3) and without neovascularity (n=17, mean=31.7, SD=28.6).

Table 2

The relationship between greyscale ultrasound (US) and pain

Table 3

The relationship between greyscale ultrasound (US) and neovascularity

Table 4

Mean (SD) pain intensity among painful tendons in each greyscale ultrasound (US) group at month 1


Distinct tendon abnormalities on greyscale US have been previously described, including diffuse thickening and hypoechoic regions,5 8 15 but their relationship and relative clinical significance has not been clear. The current study identified a temporal sequence in greyscale ultrasound abnormality. The current data suggests that normal tendon is likely to develop diffuse thickening prior to a hypoechoic region. Likewise, tendons containing a hypoechoic region are likely to develop diffuse thickening before they have a normal structure on ultrasound imaging. This suggests that diffuse thickening is a transitional greyscale US abnormality and this is shown by its greater probability to change than normal tendon and tendon containing a hypoechoic region (table 1).

The two greyscale US abnormalities are likely to reflect different tendon structure, and may reflect different stages of underlying tendon pathology. It has been suggested that initial tendon pathology involves a proliferative response21 with increased numbers of activated tenocytes22 that preferentially produce ground substance as opposed to collagen,22 23 leading to disorganised and separated collagen.23 24 An animal model of overuse tendinopathy has shown that tenocyte activity and increased ground substance is seen in the short term (8 weeks), with no signs of apoptosis (programmed cell death) or tissue degeneration.25 It is possible that diffuse thickening on US reflects this proliferative tendon response, as ground substance is less echoic than collagen.26 The high probability for transition among tendons with diffuse thickening (table 1) is consistent with a transient proliferative phase. Human biopsy studies are required to confirm the association between a proliferative response and diffuse thickening on US.

Tendon pathology may progress beyond a proliferative response to what is described as degenerative tendinopathy.21 This phase is characterised by a hypocellular environment as a result of cell death or apoptosis, discontinuous and degenerated collagen and neurovascular ingrowth. Biopsy studies have confirmed that these pathological changes are associated with hypoechoic regions on US.27 28 The current data provide evidence for this final degenerative tendinopathy phase, as neovascularity was seldom present among normal tendons and those with diffuse thickening (table 4). Although the label degenerative tendinopathy suggests a reduced tendon function, the current study and previous studies8 9 have demonstrated a limited potential for these tendons to revert towards more normal structure on US.

The relationship between the various greyscale US categories, Doppler flow and pain was also investigated in the current study. There was a greater probability of tendon pain in hypoechoic tendons (59%) and diffusely thickened tendons (43%) compared with normal tendons (24%). Doppler flow was common among tendons containing a hypoechoic region (42%), uncommon among diffusely thickened tendons (6%) and absent among normal tendons. As previously demonstrated, abnormal tendon US is not necessarily painful,7 29 although associated Doppler flow may increase the risk of pain,10 18 and pain may arise from tendons with normal imaging.30 Further, the presence of neovascularity is not a prerequisite for tendon pain,18 as 43% of diffusely thickened tendons were painful even though only 6% contained neovascularity. Recent studies suggest that biochemical substances (including glutamate, substance P, calcitonin gene-related peptide) associated with neurovascular ingrowth may have a role in tendon pain.31 32 Importantly, these substances have also been demonstrated in pain-free tendons with normal structure,31,,33 providing a potential pain mechanism in tendons that do not contain greyscale and/or Doppler US abnormality. There are two other possible explanations for the apparent lack of neovascularity in many painful tendons. A previous study showed that some patellar tendons contain neovascularity but this may not always be identifiable with Doppler imaging.34 Further, some Doppler settings may be insensitive to mild neovascularity.

The current findings suggest that greyscale US imaging demonstrates stages of tendinopathy. There is a greater risk of pain and neovascularity among tendons containing a hypoechoic region. However, diffusely thickened patellar tendons may also be painful, usually in the absence of Doppler flow. This information may assist in interpreting ultrasound scans among people with pain in the region of the patellar tendon. Further studies are needed to determine whether there are characteristics that may assist in identifying people who are prone to developing diffusely thickened or hypoechoic tendons (eg, age, activity level) and whether these greyscale tendon abnormalities respond differently to clinical interventions, such as an eccentric training programme.35


This study demonstrates that there is transition between distinct types of tendon abnormalities on ultrasound imaging (diffuse thickening and hypoechoic regions) that may relate to different phases of tendon pathology. Tendons containing a hypoechoic region demonstrated the highest probability (59%) of pain and neovascularity (42%). Diffusely thickened tendons are also likely to be painful (43%) but rarely display neovascularity (6%).



  • Competing interests None.

  • Ethics approval Ethics approval was obtained from La Trobe University, Melbourne, Australia.