- Email: [email protected]
Ultrasonography as a prognostic and objective parameter in Achilles tendinopathy: A prospective observational study
European Journal of Radiology 84 (2015) 458–462
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Ultrasonography as a prognostic and objective parameter in Achilles tendinopathy: A prospective observational study Mads Bakkegaard a,1 , Finn E. Johannsen b,∗ , Betina Højgaard c,2 , Henning Langberg d,3 a
Department of Rheumatology, Holbæk Hospital, Smedelundsgade 60, 4300 Holbæk, Denmark Private Department of Rheumatology, Furesø-reumatologerne, Farum and ISMC, Institute of Sports Medicine Copenhagen, Bispebjerg Hospital, Bispebjerg Bakke 23, 2. Tværvej, Indgang 8, 1. sal, 2400 København NV, Denmark c Danish Institute for Local and Regional Government Research, Købmagergade 22, 1150 København K, Denmark d Institute of Social Medicine, Department of Public Health and Centre for Healthy Ageing, Faculty of Heath Sciences, University of Copenhagen, Øster Farimagsgade 5, bygn. 24, postboks 2099, 1014 København, Denmark b
a r t i c l e
i n f o
Article history: Received 22 March 2014 Received in revised form 30 July 2014 Accepted 23 November 2014 Keywords: Ultrasonography Achilles tendinopathy Prognostic markers Cause of pain
a b s t r a c t Objectives: To study prospectively whether structural changes determined by ultrasound scanning (US) can be used as prognostic markers for outcome in patients with symptomatic Achilles tendinopathy (AT) and to investigate whether there exists an association between US findings and pain measured by visual analog scale (VAS) and a general assessment score (GA). Methods: 92 consecutive patients with AT symptoms were recruited from two outpatient clinics in rheumatology. The patients underwent a conservative treatment protocol consisting of reduced activities, controlled rehabilitation including eccentric exercises of the calf muscles and if needed supplemented with corticosteroid injections. The patients were examined clinically and by US (tendon thickness, hyperand hypoechogenicity, calcification, bursitis, calcaneusspure, tenosynovitis, gray scale and color Doppler focusing on increased flow intra- or peritendinous). The clinical and US examination were performed at entry, 1, 2, 3 and at 6 month. Results: 42 women and 50 men were included (mean age of 47 years). They had symptoms for more than 13 months and a symptomatic Achilles tendon mean thickness of 7.4 ± 2.3 mm. Heterogeneity at the initial examination was found to be a prognostic marker for the clinical outcome. Tendon thickness, hypoechogenicity and increased flow at any time point were significantly correlated to pain at function, palpatory pain and morning pain at the same time points. A reduction in tendon thickness was statistically associated with a decrease in palpatory pain. Conclusion: Heterogeneity is a prognostic marker in AT. Tendon thickness, hypoechogenicity and increased Doppler activity can be used as objective outcome parameters for the treatment effect of AT. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Achilles tendinopathy (AT) is a very common injury in sports and a difficult condition to treat [1,2]. The exact mechanisms behind AT and the source of pain has not yet been clarified [3]. At the moment there is no consensus on the treatment regime. Eccentric exercises have proven effect on AT [4,5] as well as stretching
∗ Corresponding author. Tel.: +45 35 31 21 54; fax: +45 35 31 27 33. E-mail addresses: [email protected] (M. Bakkegaard), [email protected], fi[email protected] (F.E. Johannsen), [email protected] (B. Højgaard), [email protected] (H. Langberg). 1 Tel.: +45 59 48 40 00. 2 Tel.: +45 44 45 55 00. 3 Tel.: +45 35 32 79 62. http://dx.doi.org/10.1016/j.ejrad.2014.11.028 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
exercises [6,7]. In these efficacy training studies on motivated patients 60–90% are cured, and we cannot expect as high cure rate in the daily clinic, in fact one clinical effectiveness study found only 10% cure rate from a home exercise program consisting of eccentric exercises [8]. Glucocorticosteroid (GCS) injection is also found to have a good short term effect on chronic AT [9], but we have found no studies that combine exercises and GCS injections. Ultrasonography (US) is increasingly used in the diagnosis of soft tissue diseases such as tendinopathy. However there is no consensus on what is defined as normal, abnormal or a pathological US. In a study by Fredberg and Bolvig on high level soccer players, asymptomatic Achilles tendons were investigated [9]. They found US changes (tendon thickness >7 mm) in 11% of the asymptomatic Achilles tendons and demonstrated that the risk of developing symptoms within 1 year was very high in the group of players with increased
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
thickness on US (45%) when continuing playing soccer. This means that increased thickness probably is a precursor for developing AT. It has been suggested, that increased tendon thickness and the presence of Doppler signals are correlated with the level of discomfort and dysfunction and that reduction in symptoms is accompanied with a decrease in US changes [10]. On the other hand neovascularisation is also found in asymptomatic tendons [11]. In the present study we evaluated whether different US changes can be prognostic markers for patients with symptomatic Achilles tendinopathy undergoing a conservative treatment protocol. The hypothesis was that increased tendon thickness, increased peri- and intratendinous flow and heterogeneity (hypo-, hyperechogenicity and calcifications) at the initial examination were associated with a poor outcome at follow-up 6 months later measured by visual analog scale (VAS) and a general assessment score (GA). Furthermore we investigated whether a correlation between symptoms (VAS and GA) and US findings can be determined pointing towards objective signs of AT severity and AT improvement. 2. Materials and methods Over an 8 month period 113 consecutive patients (52 women and 61 men) with AT from two outpatient clinics in rheumatology in Denmark were recruited for this prospective study. AT was defined as tenderness at palpation of the tendon, tenosynovium or tendon insertion impairing the daily activities of the patient. A total of 21 (18.6%) patients were excluded; 2 patients were diagnosed with a rupture of the Achilles tendon and in 5 patients erosions were found. 14 patients were excluded due to missing follow up data. There were no significant differences between the patients that completed the study and the patients that dropped out with regard to age, sex, tendon thickness, morning pain, intra or peritendinous flow, pain at function or palpatory pain. All patients were treated with the same treatment regime: Reduction in impact activities (ex running and jumping), controlled training with concentric/eccentric exercises and stretching, and if these exercises were impossible due to pain or did not lead to improvement, the treatment was supplemented with 1, up to maximum 3, GCS injections. 26% succeeded with training alone, 58% had one GCS injection, 14% had 2 injections and 2% had 3 injections. The clinical results of this treatment regime are presented in another study (E. Wetke submitted). 2.1. Clinical examination The patients’ symptoms were followed over 6 months, with a total of 5 visits (entry, 1, 2, 3 and 6 months). The patients were evaluated clinically and with US at each visit. Pain was measured on a 100 mm VAS. Symptoms were evaluated as pain at function consisting of 20 times one foot heel raises (VAS function), palpatory pain by pinging the tendon gently between two fingers (VAS palpation) and morning pain (VAS morning). General assessment was evaluated as relative to symptoms at the first consultation on the following scale: 0 = no more symptoms, 1 = much better, 2 = slightly better, 3 = no changes, 4 = slightly worse, 5 = much worse. Ultrasound scanning (US) was performed at all visits before exercises by one of the investigators, who were experienced in US (Siemens G-50 with a linear transducer VF 13–5 (5–13 Hz)). US was performed with the patient prone with the feet hanging free over the edge of the bench and tendon relaxed. The tendon was scanned longitudinally and transversally. The tendon maximal thickness was measured. The two investigators had beforehand agreed upon how to evaluate tenosynovitis on gray scale (thickening of hyper/hypo echoic paratenon with poorly defined boarders),
459
tendinosis (hyper/hypoechoic areas within the tendon), bursitis (clearly visible bursa over 2 mm in size), insertional tendinopati (echo changes at the insertion and 5 mm above), calcifications within the tendon (hyperechogenicity with echo shadow underneath), calcaneal spur (an elevation of the calcaneal compacta as a spur at the distal insertion of the tendon). Color-Doppler was used to measure the presence of flow within the tendon or peritendon. We used the factory setting with pulse repetition frequency on 780 Hz. At an asymptomatic area of the gastrocnemius part of the tendon the Doppler was calibrated to zero-pixels. Peritendinous flow and intratendinous flow was registered as clearly visible color-Doppler signal just as visible arteries with pulsation within the paratenon and tendon were registered. Tenosynovitis could therefore be diagnosed as either on gray scale and/or with color Doppler. 2.2. Data analysis Statistical analyses were performed using STATA version 13.1. Statistical significance was set at p < 0.05. Completed patients and dropouts were compared according to central baseline measurements using Fischer exact tests for dichotomous variables, t-test for continuous variables, and Wilcoxon rank-sum test for ordinal data. Descriptive statistics (means and standard deviations as relative frequencies (%)) were obtained on completed patients. Association between tendon thickness, increased peri- and intratendinous flow and heterogeneity at the initial examination and outcome (measured by VAS) at 6 months follow up were tested using Wilcoxon rank-sum test for dichotomous variables and by correlation for continuous variables. Association between decrease in tendon thickness and decrease in VAS measures between baseline and 6 months follow-up were analysed by linear regression. Due to non normally distributed data, the VAS scores were divided into 4 clinically meaningful categories, VAS 0 = no more pain, VAS 1–25 = low pain, VAS 26–50 = acceptable pain during rehabilitation, VAS 51–100 = unacceptable pain. Univariate ordinal logistic regression was used to assess associations between US findings and pain measured by visual analog scale (VAS) and a general assessment score (GA). The analyses were run for the total number of observations (n = 460). To correct inference for the dependence among observations from a single subject, the analyses were performed as a cluster analysis. Hence, each patient was treated as a cluster. The Brandt test [12] was used to test the proportional odds constraint that the regression coefficients for the comparison of the 4 VAS categories were similar. If this test was statistically significant at the 5% level, a generalized ordered logistic regression [13] was estimated instead. This output is a series of binary logistic regressions. First, it is category 1 vs. categories 2, 3, and 4; then categories 1 and 2 vs. 3 and 4; and then categories 1, 2, and 3 vs. 4. 3. Results 92 patients completed the study, 42 women and 50 men with a mean age of 46.9 (SD 15.7) years, symptoms for 13.2 (SD 15.5) months and a symptomatic and asymptomatic Achilles tendon mean maximal thickness of 7.4 (SD 2.3)mm and 5.5 mm (SD 1.1), respectively. The US findings at entry are shown in Table 1. Eight (9%) symptomatic tendons were found with no US changes and 23 (30.7%) asymptomatic tendons were found with US abnormalities. Patients with or without heterogeneity at entry were compared with regard to results at 6 month follow up. We found significant differences in VAS palpation (p = 0.04) and borderline significance
Tendon thickness 5 10 0
0
1-25
26-50
51-100
26-50
51-100
26-50
51-100
0
Tendon thickness 5 10
15
VAS function
0
1-25
15
VAS Palpation
0
regarding VAS function (p = 0.05). A decrease in tendon thickness was statistically associated with a decrease in VAS palpation (p = 0.02) i.e. for each decrease of 1 mm in tendon thickness, there was a correlated decrease in VAS palpation of 5 on the 0–100 VAS. However a decrease in tendon thickness was not associated to a decrease in VAS morning or VAS function. The association between tendon thickness, hypoechogenicity, flow and pain measured by VAS is shown in Figs. 1–3. With regard to objective parameters of AT, 5 ultrasound examinations per patient were performed and we found a significant relation between tendon thickness and VAS function (p < 0.001), tendon thickness and VAS palpation (p = 0.002), i.e. thicker tendons were associated with increased odds of having more severe pain. For the comparison of tendon thickness and VAS morning, a significant increased odds ratio (p = 0.004) was only found for the comparison of those with no pain (VAS = 0) vs. those with pain (VAS > 0), see Fig. 1. Increased flow i.e. intratendinous and/or peritendinous flow was associated with VAS function (p < 0.001) and VAS palpation (p < 0.001) i.e. increased flow was associated with increased odds of having more severe pain. With regards to morning pain, a significant increased odds ratio was found for the comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001), and for those with no pain/low pain (VAS < 26) compared with those with acceptable pain during rehabilitation/unacceptable pain (VAS > 25) (p = 0.002). However, for comparison of those with no pain/low pain/acceptable pain during rehabilitation (VAS < 51) vs. those with unacceptable pain (VAS > 50) the increased odds ratio was borderline significant (p = 0.068), see Fig. 2a–c. Hypoechogenicity was associated with VAS palpation (p < 0.001). With regards to VAS morning a significant odds ratio was found for comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001). With regards to VAS function, a significant odds ratio was found for comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001) and of those with no pain/low pain (VAS < 26) vs. those with acceptable pain during rehabilitation/unacceptable pain (VAS > 25) (p = 0.01), see Fig. 3a–c. The conservative treatment protocol was very effective and of the 92 patients, 72 patients (78%) regained normal function (46 patients (50%) no more symptoms (GA = 0), 26 patients (28%) much better (GA = 1)), 16 patients (18%) slightly better (GA = 2), 2 patients (2%) unchanged (GA = 3), 2 patients (2%) slightly worse (GA = 4) and none (0%) much worse (GA = 5). We noticed that most of the patients with mid tendon affection were close to the insertion of soleus into the tendon. Fig. 4 illustrates the localization of maximal tendon thickness above calcaneus. We found no differences in clinical outcome with regard to insertional and midsubstance AT. Calcifications within the tendons were found in 4 cases, but although the pain diminished the calcifications did not change, furthermore calcifications were also found in 1 asymptomatic tendon. Calcaneal spur was seen in the 18 symptomatic tendons, but also found in 11 asymptomatic tendons.
15
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
Tendon thickness 5 10
460
Table 1 US findings at baseline.
0
1-25 VAS Morning
No. Tendon thickness >7 mm, no. (%) Hyperechogenicity, no. (%) Hypoechogenicity, no. (%) Calcification, no. (%) Bursitis, no. (%) Calcaneusspure, no. (%) Tenosynovitis on gray scale, no. (%) Intratendinous flow, no. (%) Peritendinous flow, no. (%) Normal US, no. (%)
Symptomatic tendons
Asymptomatic tendons
92 46 (50) 5 (5) 41 (45) 4 (4) 7 (8) 18 (20) 16 (17) 47 (51) 38 (41) 8 (9)
75 6 (8) 1 (1) 4 (5) 1 (1) 2 (3) 11 (15) 2 (2) 5 (7) 6 (8) 52 (69)
Fig. 1. Boxplot–tendon thickness, by VAS group. The ends of the whiskers represent the lowest/highest value still within 1.5 interquantile range (Irq) of the lower/upper quartile. Diamond symbols indicate means (n = 460).
4. Discussion The primary aim of this study was to find US changes that could be prognostic markers for the outcome. We found US heterogeneity (hyper- and/or hypoechogenicity) at entry to be a prognostic marker for poor outcome at 6 months. Our findings are in contrast to a 2 year prospective study where they found no US
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
a
a
b
b
c
c
461
Fig. 2. (a)–(c) Flow/no flow by VAS group. Percentage and absolute numbers (n = 460).
Fig. 3. (a)–(c) Homogeneity/hypoechogenicity by VAS group. Percentage and absolute numbers (n = 460).
changes that could predict the 12 months outcome [14]. This could be explained partly in the difference in follow up time. In another prospective study clinical outcome, after a conservative treatment regime, could not be predicted by neovascularization at baseline [15]. On the other hand it has been shown that patients with AT and normal tendon structure had a significant shorter period to full recovery compared to patients with enlarged tendons and tendons with hypoechoic areas [16]. Inhomogeneous tendons have also in another study been found to have an unfavorable outcome, in that study there was no association with either neovascularisation or tendon spindle form and clinical outcome measured on VAS [17]. Limitations in our study concerning prognostic markers are the short follow up time of 6 months and also we don’t know if the same result could be found if symptoms were registered with a validated score (i.e. VISA-A score) instead of pain measured on a VAS scale. The other aim of the study was to find US changes that correlated with pain or function and thereby could be an objective parameter. We demonstrated that maximal tendon thickness is associated with increased pain at function, morning pain and palpatory pain (Fig. 1). Furthermore an improvement in symptoms can be objectified as a reduction in tendon thickness. This is in accordance with a study describing a decrease in tendon thickness and a normalization of tendon structure at 3.8 year follow up [18]. In the same study they found that remaining US changes (hypoechoic areas) were associated with pain, which supports our findings of a correlation between hypoechoic changes and
palpatory pain, morning pain and pain at function. Neovascularisation and increased Doppler signal have previously been suggested to be partly the cause of pain in AT [10], which supports our and others findings, of flow being associated with morning pain, function pain and palpation pain [19]. In contrast calcification within the tendon and calcaneal spurs were found both in the symptomatic and asymptomatic tendons and did not change with improvement of symptoms, and is therefore not the cause of the pain. Somewhat to our surprise we found no differences between the effect of treatment on insertional or mid substance AT. This is in
Fig. 4. Location of pain at palpation/maximal tendon thickness.
462
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
contrast to other studies, where insertional AT is associated with a less favorable outcome in patients doing excentric exercises as described in Section 2 [5,7]. This could partly be due to the very slow and cautious progression in the eccentric exercises that was used in the present study. In our study we noticed that the midsubstance AT were close to the insertion of soleus into the tendon, which supports the theory that AT could be caused by shear forces [20]. Furthermore US is important to exclude differential diagnosis such as rupture, where we found 2, and erosions, where we found 5, all being one of the primary symptoms of a later diagnosed seronegative arthritis. In conclusion, we found heterogeneity to be a prognostic ultrasound parameter, which could predict the clinical outcome at 6-month follow up with regard to palpatory pain. Changes in tissue structure on US corresponded to the patients’ pain and symptoms at palpation, loading and morning pain, indicating, that the reason for the pain is likely to be mechanical. Along the same line maximal tendon thickness and hypoechogenicity were found to be associated with pain and hence an objective parameter for the severity of Achilles tendinopathy. Also we found that increased flow was suggestive for the severity of AT. A reduction in tendon thickness might be a useful objective outcome score in the future as it was associated with a decreased palpatory pain. Conflict of interest statement None. Acknowledgements Thanks to Arne Gam, MD, specialist in rheumatology, private rheumatology clinic, Brønshøj for including patients into this project. Thanks to Ditte Sæbye, Cand. Scient. in Statistics, Institute of Preventive Medicine, Frederiksberg Hospital for statistical advice. References [1] Kvist M. Achilles tendon injuries in athletes. Sports Med 1994;18(3):173–201. [2] Rees JD, Maffulli N, Cook J. Management of tendinopathy. Am J Sports Med 2009;37(Sep (9)):1855–67.
[3] Maffulli N, Sharma P, Luscombe KL. Achilles tendinopathy: aetiology and management. J R Soc Med 2004;97(Oct (10)):472–6. [4] Silbernagel KG, Brorsson A, Lundberg M. The majority of patients with Achilles tendinopathy recover fully when treated with exercise alone: a 5-year followup. Am J Sports Med 2011;39(Mar (3)):607–13. [5] Fahlström M, Jonsson P, Lorentzon R, Alfredson H. Chronic Achilles tendon pain treated with eccentric calf-muscle training. Knee Surg Sports Traumatol Arthrosc 2003;11(5):327–33. [6] Porter D, Barrill E, Oneacre K, May BD. The effects of duration and frequency of Achilles tendon stretching on dorsiflexion and outcome in painful heel syndrome: a randomized, blinded, control study. Foot Ankle Int 2002;23(Jul (7)):619–24. [7] Nørregaard J, Larsen C, Bieler T, Langberg H. Eccentric exercise in treatment of Achilles tendinopathy. Scand J Med Sci Sports 2007;17(2): 133–8. [8] Ram R, Meeuwisse W, Patel C, Wiseman DA, Wiley JP. The limited effectiveness of a home-based eccentric training for treatment of Achilles tendinopathy. Clin Invest Med 2013;36(4):E197–206. [9] Fredberg U, Bolvig L. Significance of ultrasonographically detected asymptomatic tendinosis in the patellar and achilles tendons of elite soccer players: a longitudinal study. Am J Sports Med 2002;30(Jul–Aug (4)):488–91. [10] Öhberg L, Alfredson H. Effects on neovascularisation behind the good results with eccentric training in chronic mid-portion Achilles tendinosis? Knee Surg Sports Traumatol Arthrosc 2004;12(5):465–70. [11] Van Snellenberg W, Wiley J, Brunet G. Achilles tendon pain intensity and level of neovascularization in athletes as determined by color Doppler ultrasound. Scand J Med Sci Sports 2007;17(5):530–4. [12] Brant R. Assessing proportionality in the proportional odds model for ordinal logistic regression. Biometrics 1990;46(4):1171–8. [13] Williams R. Generalized ordered logit/partial proportional odds models for ordinal dependent variables. Stata J 2006;6(1):58–82. [14] Khan KM, Forster BB, Robinson J, Cheong Y, Louis L, Maclean L, et al. Are ultrasound and magnetic resonance imaging of value in assessment of Achilles tendon disorders? A two year prospective study. Br J Sports Med 2003;37(Apr (2)):149–53. [15] de Vos RJ, Weir A, Cobben LP, Tol JL. The value of power Doppler ultrasonography in Achilles tendinopathy: a prospective study. Am J Sports Med 2007;35(Oct (10)):1696–701. [16] Archambault JM, Wiley JP, Bray RC, Verhoef M, Wiseman DA, Elliott PD. Can sonography predict the outcome in patients with achillodynia? J Clin Ultrasound 1998;26(7):335–9. [17] Zanetti M, Metzdorf A, Kundert H, Zollinger H, Vienne P, Seifert B, et al. Achilles tendons: clinical relevance of neovascularization diagnosed with power Doppler US. Radiology 2003;227(2):556–60. [18] Ohberg L, Lorentzon R, Alfredson H. Eccentric training in patients with chronic Achilles tendinosis: normalised tendon structure and decreased thickness at follow up. Br J Sports Med 2004;38(Feb (1)):8–11 (discussion 11). [19] Reiter M, Ulreich N, Dirisamer A, Tscholakoff D, Bucek R. Colour and power Doppler sonography in symptomatic Achilles tendon disease. Int J Sports Med 2004;25(04):301–5. [20] Finni T. Structural and functional features of human muscle–tendon unit. Scand J Med Sci Sports 2006;16(3):147–58.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Ultrasonography as a prognostic and objective parameter in Achilles tendinopathy: A prospective observational study Mads Bakkegaard a,1 , Finn E. Johannsen b,∗ , Betina Højgaard c,2 , Henning Langberg d,3 a
Department of Rheumatology, Holbæk Hospital, Smedelundsgade 60, 4300 Holbæk, Denmark Private Department of Rheumatology, Furesø-reumatologerne, Farum and ISMC, Institute of Sports Medicine Copenhagen, Bispebjerg Hospital, Bispebjerg Bakke 23, 2. Tværvej, Indgang 8, 1. sal, 2400 København NV, Denmark c Danish Institute for Local and Regional Government Research, Købmagergade 22, 1150 København K, Denmark d Institute of Social Medicine, Department of Public Health and Centre for Healthy Ageing, Faculty of Heath Sciences, University of Copenhagen, Øster Farimagsgade 5, bygn. 24, postboks 2099, 1014 København, Denmark b
a r t i c l e
i n f o
Article history: Received 22 March 2014 Received in revised form 30 July 2014 Accepted 23 November 2014 Keywords: Ultrasonography Achilles tendinopathy Prognostic markers Cause of pain
a b s t r a c t Objectives: To study prospectively whether structural changes determined by ultrasound scanning (US) can be used as prognostic markers for outcome in patients with symptomatic Achilles tendinopathy (AT) and to investigate whether there exists an association between US findings and pain measured by visual analog scale (VAS) and a general assessment score (GA). Methods: 92 consecutive patients with AT symptoms were recruited from two outpatient clinics in rheumatology. The patients underwent a conservative treatment protocol consisting of reduced activities, controlled rehabilitation including eccentric exercises of the calf muscles and if needed supplemented with corticosteroid injections. The patients were examined clinically and by US (tendon thickness, hyperand hypoechogenicity, calcification, bursitis, calcaneusspure, tenosynovitis, gray scale and color Doppler focusing on increased flow intra- or peritendinous). The clinical and US examination were performed at entry, 1, 2, 3 and at 6 month. Results: 42 women and 50 men were included (mean age of 47 years). They had symptoms for more than 13 months and a symptomatic Achilles tendon mean thickness of 7.4 ± 2.3 mm. Heterogeneity at the initial examination was found to be a prognostic marker for the clinical outcome. Tendon thickness, hypoechogenicity and increased flow at any time point were significantly correlated to pain at function, palpatory pain and morning pain at the same time points. A reduction in tendon thickness was statistically associated with a decrease in palpatory pain. Conclusion: Heterogeneity is a prognostic marker in AT. Tendon thickness, hypoechogenicity and increased Doppler activity can be used as objective outcome parameters for the treatment effect of AT. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Achilles tendinopathy (AT) is a very common injury in sports and a difficult condition to treat [1,2]. The exact mechanisms behind AT and the source of pain has not yet been clarified [3]. At the moment there is no consensus on the treatment regime. Eccentric exercises have proven effect on AT [4,5] as well as stretching
∗ Corresponding author. Tel.: +45 35 31 21 54; fax: +45 35 31 27 33. E-mail addresses: [email protected] (M. Bakkegaard), [email protected], fi[email protected] (F.E. Johannsen), [email protected] (B. Højgaard), [email protected] (H. Langberg). 1 Tel.: +45 59 48 40 00. 2 Tel.: +45 44 45 55 00. 3 Tel.: +45 35 32 79 62. http://dx.doi.org/10.1016/j.ejrad.2014.11.028 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
exercises [6,7]. In these efficacy training studies on motivated patients 60–90% are cured, and we cannot expect as high cure rate in the daily clinic, in fact one clinical effectiveness study found only 10% cure rate from a home exercise program consisting of eccentric exercises [8]. Glucocorticosteroid (GCS) injection is also found to have a good short term effect on chronic AT [9], but we have found no studies that combine exercises and GCS injections. Ultrasonography (US) is increasingly used in the diagnosis of soft tissue diseases such as tendinopathy. However there is no consensus on what is defined as normal, abnormal or a pathological US. In a study by Fredberg and Bolvig on high level soccer players, asymptomatic Achilles tendons were investigated [9]. They found US changes (tendon thickness >7 mm) in 11% of the asymptomatic Achilles tendons and demonstrated that the risk of developing symptoms within 1 year was very high in the group of players with increased
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
thickness on US (45%) when continuing playing soccer. This means that increased thickness probably is a precursor for developing AT. It has been suggested, that increased tendon thickness and the presence of Doppler signals are correlated with the level of discomfort and dysfunction and that reduction in symptoms is accompanied with a decrease in US changes [10]. On the other hand neovascularisation is also found in asymptomatic tendons [11]. In the present study we evaluated whether different US changes can be prognostic markers for patients with symptomatic Achilles tendinopathy undergoing a conservative treatment protocol. The hypothesis was that increased tendon thickness, increased peri- and intratendinous flow and heterogeneity (hypo-, hyperechogenicity and calcifications) at the initial examination were associated with a poor outcome at follow-up 6 months later measured by visual analog scale (VAS) and a general assessment score (GA). Furthermore we investigated whether a correlation between symptoms (VAS and GA) and US findings can be determined pointing towards objective signs of AT severity and AT improvement. 2. Materials and methods Over an 8 month period 113 consecutive patients (52 women and 61 men) with AT from two outpatient clinics in rheumatology in Denmark were recruited for this prospective study. AT was defined as tenderness at palpation of the tendon, tenosynovium or tendon insertion impairing the daily activities of the patient. A total of 21 (18.6%) patients were excluded; 2 patients were diagnosed with a rupture of the Achilles tendon and in 5 patients erosions were found. 14 patients were excluded due to missing follow up data. There were no significant differences between the patients that completed the study and the patients that dropped out with regard to age, sex, tendon thickness, morning pain, intra or peritendinous flow, pain at function or palpatory pain. All patients were treated with the same treatment regime: Reduction in impact activities (ex running and jumping), controlled training with concentric/eccentric exercises and stretching, and if these exercises were impossible due to pain or did not lead to improvement, the treatment was supplemented with 1, up to maximum 3, GCS injections. 26% succeeded with training alone, 58% had one GCS injection, 14% had 2 injections and 2% had 3 injections. The clinical results of this treatment regime are presented in another study (E. Wetke submitted). 2.1. Clinical examination The patients’ symptoms were followed over 6 months, with a total of 5 visits (entry, 1, 2, 3 and 6 months). The patients were evaluated clinically and with US at each visit. Pain was measured on a 100 mm VAS. Symptoms were evaluated as pain at function consisting of 20 times one foot heel raises (VAS function), palpatory pain by pinging the tendon gently between two fingers (VAS palpation) and morning pain (VAS morning). General assessment was evaluated as relative to symptoms at the first consultation on the following scale: 0 = no more symptoms, 1 = much better, 2 = slightly better, 3 = no changes, 4 = slightly worse, 5 = much worse. Ultrasound scanning (US) was performed at all visits before exercises by one of the investigators, who were experienced in US (Siemens G-50 with a linear transducer VF 13–5 (5–13 Hz)). US was performed with the patient prone with the feet hanging free over the edge of the bench and tendon relaxed. The tendon was scanned longitudinally and transversally. The tendon maximal thickness was measured. The two investigators had beforehand agreed upon how to evaluate tenosynovitis on gray scale (thickening of hyper/hypo echoic paratenon with poorly defined boarders),
459
tendinosis (hyper/hypoechoic areas within the tendon), bursitis (clearly visible bursa over 2 mm in size), insertional tendinopati (echo changes at the insertion and 5 mm above), calcifications within the tendon (hyperechogenicity with echo shadow underneath), calcaneal spur (an elevation of the calcaneal compacta as a spur at the distal insertion of the tendon). Color-Doppler was used to measure the presence of flow within the tendon or peritendon. We used the factory setting with pulse repetition frequency on 780 Hz. At an asymptomatic area of the gastrocnemius part of the tendon the Doppler was calibrated to zero-pixels. Peritendinous flow and intratendinous flow was registered as clearly visible color-Doppler signal just as visible arteries with pulsation within the paratenon and tendon were registered. Tenosynovitis could therefore be diagnosed as either on gray scale and/or with color Doppler. 2.2. Data analysis Statistical analyses were performed using STATA version 13.1. Statistical significance was set at p < 0.05. Completed patients and dropouts were compared according to central baseline measurements using Fischer exact tests for dichotomous variables, t-test for continuous variables, and Wilcoxon rank-sum test for ordinal data. Descriptive statistics (means and standard deviations as relative frequencies (%)) were obtained on completed patients. Association between tendon thickness, increased peri- and intratendinous flow and heterogeneity at the initial examination and outcome (measured by VAS) at 6 months follow up were tested using Wilcoxon rank-sum test for dichotomous variables and by correlation for continuous variables. Association between decrease in tendon thickness and decrease in VAS measures between baseline and 6 months follow-up were analysed by linear regression. Due to non normally distributed data, the VAS scores were divided into 4 clinically meaningful categories, VAS 0 = no more pain, VAS 1–25 = low pain, VAS 26–50 = acceptable pain during rehabilitation, VAS 51–100 = unacceptable pain. Univariate ordinal logistic regression was used to assess associations between US findings and pain measured by visual analog scale (VAS) and a general assessment score (GA). The analyses were run for the total number of observations (n = 460). To correct inference for the dependence among observations from a single subject, the analyses were performed as a cluster analysis. Hence, each patient was treated as a cluster. The Brandt test [12] was used to test the proportional odds constraint that the regression coefficients for the comparison of the 4 VAS categories were similar. If this test was statistically significant at the 5% level, a generalized ordered logistic regression [13] was estimated instead. This output is a series of binary logistic regressions. First, it is category 1 vs. categories 2, 3, and 4; then categories 1 and 2 vs. 3 and 4; and then categories 1, 2, and 3 vs. 4. 3. Results 92 patients completed the study, 42 women and 50 men with a mean age of 46.9 (SD 15.7) years, symptoms for 13.2 (SD 15.5) months and a symptomatic and asymptomatic Achilles tendon mean maximal thickness of 7.4 (SD 2.3)mm and 5.5 mm (SD 1.1), respectively. The US findings at entry are shown in Table 1. Eight (9%) symptomatic tendons were found with no US changes and 23 (30.7%) asymptomatic tendons were found with US abnormalities. Patients with or without heterogeneity at entry were compared with regard to results at 6 month follow up. We found significant differences in VAS palpation (p = 0.04) and borderline significance
Tendon thickness 5 10 0
0
1-25
26-50
51-100
26-50
51-100
26-50
51-100
0
Tendon thickness 5 10
15
VAS function
0
1-25
15
VAS Palpation
0
regarding VAS function (p = 0.05). A decrease in tendon thickness was statistically associated with a decrease in VAS palpation (p = 0.02) i.e. for each decrease of 1 mm in tendon thickness, there was a correlated decrease in VAS palpation of 5 on the 0–100 VAS. However a decrease in tendon thickness was not associated to a decrease in VAS morning or VAS function. The association between tendon thickness, hypoechogenicity, flow and pain measured by VAS is shown in Figs. 1–3. With regard to objective parameters of AT, 5 ultrasound examinations per patient were performed and we found a significant relation between tendon thickness and VAS function (p < 0.001), tendon thickness and VAS palpation (p = 0.002), i.e. thicker tendons were associated with increased odds of having more severe pain. For the comparison of tendon thickness and VAS morning, a significant increased odds ratio (p = 0.004) was only found for the comparison of those with no pain (VAS = 0) vs. those with pain (VAS > 0), see Fig. 1. Increased flow i.e. intratendinous and/or peritendinous flow was associated with VAS function (p < 0.001) and VAS palpation (p < 0.001) i.e. increased flow was associated with increased odds of having more severe pain. With regards to morning pain, a significant increased odds ratio was found for the comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001), and for those with no pain/low pain (VAS < 26) compared with those with acceptable pain during rehabilitation/unacceptable pain (VAS > 25) (p = 0.002). However, for comparison of those with no pain/low pain/acceptable pain during rehabilitation (VAS < 51) vs. those with unacceptable pain (VAS > 50) the increased odds ratio was borderline significant (p = 0.068), see Fig. 2a–c. Hypoechogenicity was associated with VAS palpation (p < 0.001). With regards to VAS morning a significant odds ratio was found for comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001). With regards to VAS function, a significant odds ratio was found for comparison of those with no pain (VAS = 0) and those with pain (VAS > 0) (p < 0.001) and of those with no pain/low pain (VAS < 26) vs. those with acceptable pain during rehabilitation/unacceptable pain (VAS > 25) (p = 0.01), see Fig. 3a–c. The conservative treatment protocol was very effective and of the 92 patients, 72 patients (78%) regained normal function (46 patients (50%) no more symptoms (GA = 0), 26 patients (28%) much better (GA = 1)), 16 patients (18%) slightly better (GA = 2), 2 patients (2%) unchanged (GA = 3), 2 patients (2%) slightly worse (GA = 4) and none (0%) much worse (GA = 5). We noticed that most of the patients with mid tendon affection were close to the insertion of soleus into the tendon. Fig. 4 illustrates the localization of maximal tendon thickness above calcaneus. We found no differences in clinical outcome with regard to insertional and midsubstance AT. Calcifications within the tendons were found in 4 cases, but although the pain diminished the calcifications did not change, furthermore calcifications were also found in 1 asymptomatic tendon. Calcaneal spur was seen in the 18 symptomatic tendons, but also found in 11 asymptomatic tendons.
15
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
Tendon thickness 5 10
460
Table 1 US findings at baseline.
0
1-25 VAS Morning
No. Tendon thickness >7 mm, no. (%) Hyperechogenicity, no. (%) Hypoechogenicity, no. (%) Calcification, no. (%) Bursitis, no. (%) Calcaneusspure, no. (%) Tenosynovitis on gray scale, no. (%) Intratendinous flow, no. (%) Peritendinous flow, no. (%) Normal US, no. (%)
Symptomatic tendons
Asymptomatic tendons
92 46 (50) 5 (5) 41 (45) 4 (4) 7 (8) 18 (20) 16 (17) 47 (51) 38 (41) 8 (9)
75 6 (8) 1 (1) 4 (5) 1 (1) 2 (3) 11 (15) 2 (2) 5 (7) 6 (8) 52 (69)
Fig. 1. Boxplot–tendon thickness, by VAS group. The ends of the whiskers represent the lowest/highest value still within 1.5 interquantile range (Irq) of the lower/upper quartile. Diamond symbols indicate means (n = 460).
4. Discussion The primary aim of this study was to find US changes that could be prognostic markers for the outcome. We found US heterogeneity (hyper- and/or hypoechogenicity) at entry to be a prognostic marker for poor outcome at 6 months. Our findings are in contrast to a 2 year prospective study where they found no US
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
a
a
b
b
c
c
461
Fig. 2. (a)–(c) Flow/no flow by VAS group. Percentage and absolute numbers (n = 460).
Fig. 3. (a)–(c) Homogeneity/hypoechogenicity by VAS group. Percentage and absolute numbers (n = 460).
changes that could predict the 12 months outcome [14]. This could be explained partly in the difference in follow up time. In another prospective study clinical outcome, after a conservative treatment regime, could not be predicted by neovascularization at baseline [15]. On the other hand it has been shown that patients with AT and normal tendon structure had a significant shorter period to full recovery compared to patients with enlarged tendons and tendons with hypoechoic areas [16]. Inhomogeneous tendons have also in another study been found to have an unfavorable outcome, in that study there was no association with either neovascularisation or tendon spindle form and clinical outcome measured on VAS [17]. Limitations in our study concerning prognostic markers are the short follow up time of 6 months and also we don’t know if the same result could be found if symptoms were registered with a validated score (i.e. VISA-A score) instead of pain measured on a VAS scale. The other aim of the study was to find US changes that correlated with pain or function and thereby could be an objective parameter. We demonstrated that maximal tendon thickness is associated with increased pain at function, morning pain and palpatory pain (Fig. 1). Furthermore an improvement in symptoms can be objectified as a reduction in tendon thickness. This is in accordance with a study describing a decrease in tendon thickness and a normalization of tendon structure at 3.8 year follow up [18]. In the same study they found that remaining US changes (hypoechoic areas) were associated with pain, which supports our findings of a correlation between hypoechoic changes and
palpatory pain, morning pain and pain at function. Neovascularisation and increased Doppler signal have previously been suggested to be partly the cause of pain in AT [10], which supports our and others findings, of flow being associated with morning pain, function pain and palpation pain [19]. In contrast calcification within the tendon and calcaneal spurs were found both in the symptomatic and asymptomatic tendons and did not change with improvement of symptoms, and is therefore not the cause of the pain. Somewhat to our surprise we found no differences between the effect of treatment on insertional or mid substance AT. This is in
Fig. 4. Location of pain at palpation/maximal tendon thickness.
462
M. Bakkegaard et al. / European Journal of Radiology 84 (2015) 458–462
contrast to other studies, where insertional AT is associated with a less favorable outcome in patients doing excentric exercises as described in Section 2 [5,7]. This could partly be due to the very slow and cautious progression in the eccentric exercises that was used in the present study. In our study we noticed that the midsubstance AT were close to the insertion of soleus into the tendon, which supports the theory that AT could be caused by shear forces [20]. Furthermore US is important to exclude differential diagnosis such as rupture, where we found 2, and erosions, where we found 5, all being one of the primary symptoms of a later diagnosed seronegative arthritis. In conclusion, we found heterogeneity to be a prognostic ultrasound parameter, which could predict the clinical outcome at 6-month follow up with regard to palpatory pain. Changes in tissue structure on US corresponded to the patients’ pain and symptoms at palpation, loading and morning pain, indicating, that the reason for the pain is likely to be mechanical. Along the same line maximal tendon thickness and hypoechogenicity were found to be associated with pain and hence an objective parameter for the severity of Achilles tendinopathy. Also we found that increased flow was suggestive for the severity of AT. A reduction in tendon thickness might be a useful objective outcome score in the future as it was associated with a decreased palpatory pain. Conflict of interest statement None. Acknowledgements Thanks to Arne Gam, MD, specialist in rheumatology, private rheumatology clinic, Brønshøj for including patients into this project. Thanks to Ditte Sæbye, Cand. Scient. in Statistics, Institute of Preventive Medicine, Frederiksberg Hospital for statistical advice. References [1] Kvist M. Achilles tendon injuries in athletes. Sports Med 1994;18(3):173–201. [2] Rees JD, Maffulli N, Cook J. Management of tendinopathy. Am J Sports Med 2009;37(Sep (9)):1855–67.
[3] Maffulli N, Sharma P, Luscombe KL. Achilles tendinopathy: aetiology and management. J R Soc Med 2004;97(Oct (10)):472–6. [4] Silbernagel KG, Brorsson A, Lundberg M. The majority of patients with Achilles tendinopathy recover fully when treated with exercise alone: a 5-year followup. Am J Sports Med 2011;39(Mar (3)):607–13. [5] Fahlström M, Jonsson P, Lorentzon R, Alfredson H. Chronic Achilles tendon pain treated with eccentric calf-muscle training. Knee Surg Sports Traumatol Arthrosc 2003;11(5):327–33. [6] Porter D, Barrill E, Oneacre K, May BD. The effects of duration and frequency of Achilles tendon stretching on dorsiflexion and outcome in painful heel syndrome: a randomized, blinded, control study. Foot Ankle Int 2002;23(Jul (7)):619–24. [7] Nørregaard J, Larsen C, Bieler T, Langberg H. Eccentric exercise in treatment of Achilles tendinopathy. Scand J Med Sci Sports 2007;17(2): 133–8. [8] Ram R, Meeuwisse W, Patel C, Wiseman DA, Wiley JP. The limited effectiveness of a home-based eccentric training for treatment of Achilles tendinopathy. Clin Invest Med 2013;36(4):E197–206. [9] Fredberg U, Bolvig L. Significance of ultrasonographically detected asymptomatic tendinosis in the patellar and achilles tendons of elite soccer players: a longitudinal study. Am J Sports Med 2002;30(Jul–Aug (4)):488–91. [10] Öhberg L, Alfredson H. Effects on neovascularisation behind the good results with eccentric training in chronic mid-portion Achilles tendinosis? Knee Surg Sports Traumatol Arthrosc 2004;12(5):465–70. [11] Van Snellenberg W, Wiley J, Brunet G. Achilles tendon pain intensity and level of neovascularization in athletes as determined by color Doppler ultrasound. Scand J Med Sci Sports 2007;17(5):530–4. [12] Brant R. Assessing proportionality in the proportional odds model for ordinal logistic regression. Biometrics 1990;46(4):1171–8. [13] Williams R. Generalized ordered logit/partial proportional odds models for ordinal dependent variables. Stata J 2006;6(1):58–82. [14] Khan KM, Forster BB, Robinson J, Cheong Y, Louis L, Maclean L, et al. Are ultrasound and magnetic resonance imaging of value in assessment of Achilles tendon disorders? A two year prospective study. Br J Sports Med 2003;37(Apr (2)):149–53. [15] de Vos RJ, Weir A, Cobben LP, Tol JL. The value of power Doppler ultrasonography in Achilles tendinopathy: a prospective study. Am J Sports Med 2007;35(Oct (10)):1696–701. [16] Archambault JM, Wiley JP, Bray RC, Verhoef M, Wiseman DA, Elliott PD. Can sonography predict the outcome in patients with achillodynia? J Clin Ultrasound 1998;26(7):335–9. [17] Zanetti M, Metzdorf A, Kundert H, Zollinger H, Vienne P, Seifert B, et al. Achilles tendons: clinical relevance of neovascularization diagnosed with power Doppler US. Radiology 2003;227(2):556–60. [18] Ohberg L, Lorentzon R, Alfredson H. Eccentric training in patients with chronic Achilles tendinosis: normalised tendon structure and decreased thickness at follow up. Br J Sports Med 2004;38(Feb (1)):8–11 (discussion 11). [19] Reiter M, Ulreich N, Dirisamer A, Tscholakoff D, Bucek R. Colour and power Doppler sonography in symptomatic Achilles tendon disease. Int J Sports Med 2004;25(04):301–5. [20] Finni T. Structural and functional features of human muscle–tendon unit. Scand J Med Sci Sports 2006;16(3):147–58.