Ultrasound for suspected Baker’s cyst: A test of limited clinical value?

Abstract

Introduction:

Patients with posterior knee swellings are commonly encountered in clinical practice with the vast majority referred for an ultrasound scan to assess for a clinically suspected Baker’s cyst.

Aims:

Our aim was to determine the incidence of different pathologies in patients investigated for a posterior knee swelling of a clinically suspected Baker’s cyst using ultrasound. We also wanted to investigate the incidence of significant pathologies in the popliteal fossa to assess whether the ultrasound scan findings influenced patient management.

Methodology:

Retrospective analysis was performed on all patients who underwent an ultrasound scan for a suspected Baker’s cyst or posterior knee swelling at our institution, between January 2017 and December 2018. Patient demographics, ultrasound findings, further imaging investigations and histopathology were recorded. Patient records were followed up for a minimum period of 3 years after ultrasound.

Results:

A total of 680 patients had a posterior knee ultrasound scan. Of that, 51% (347/680) had a Baker’s cyst and 40% (273/680) had a normal ultrasound scan. Sixty (9%) patients had other abnormalities identified within the popliteal fossa. Four patients had significant findings on ultrasound that required a change of management. No malignancies identified.

Conclusion:

This study confirms the high incidence of Baker’s cysts in patients presenting with a posterior knee swelling. In the absence of red flag symptoms, ultrasound of the posterior knee for a clinically suspected Baker’s cyst is of limited clinical value.

Introduction

Requests for posterior knee swellings are a common indication for a musculoskeletal (MSK) ultrasound scan (USS), with around 350 scans performed each year at our institution. The most common pathology encountered is a Baker’s cyst, a benign fluid filled synovial cyst arising from the posterior knee joint and characteristically extends between the medial head of the gastrocnemius and semimembranosus muscles. ¹ Baker’s cysts develop as a consequence of underlying degenerative change within the knee joint. ² The prevalence of Baker’s cysts ranges from 5% to 38% with a peak incidence between the ages of 35–70 years. ³ Often patients with Baker’s cysts have no symptoms, although when symptomatic present as a smooth, fluctuant swelling within the popliteal fossa and may be associated with pain. ⁴ On occasion, Baker’s cysts may present with more generalised lower limb swelling, redness and pain secondary to rupture, mimicking deep vein thrombosis (DVT). ⁵ In a large proportion of cases, the diagnosis can be made clinically assessing for the Foucher’s sign without the need for further investigation. Magnetic resonance imaging (MRI) is the gold standard imaging modality for assessing the knee for abnormalities including differentiating Baker’s cysts from other conditions. ⁶ However, MRI is expensive, not widely available and has some contraindications, whereas ultrasound is widely available, inexpensive and a non-ionising imaging tool that can be used to assess specifically for Baker’s cysts. ⁷ Previous studies have reported ultrasound as highly accurate in diagnosing Baker’s cysts when compared to MRI.^7,8 Furthermore, a recent meta-analysis comparing the diagnostic accuracy of ultrasound in detecting Baker’s cysts to both surgical and MRI findings recommended ultrasound for the assessment of Baker’s cysts. ⁹ Where there is clinical uncertainty, or a more sinister pathology is suspected, guidelines dictate ultrasound is recommended to confirm a Baker’s cyst and exclude other significant abnormalities. ¹⁰

Baker’s cysts are commonly managed conservatively, not requiring radiological or surgical intervention (unless causing neurovascular compromise) as they would recur because the underlying pathology is unchanged. ⁵ Therefore, it appears the justification to perform these scans, where the likely diagnosis is a Baker’s cyst, which itself will not change patient management, is only to exclude sinister pathology of the popliteal fossa.

Nationally, there has been a 1% increase in ultrasound examinations between 2019 and 2020 with a shortfall of 1939 Radiologists (33%) and a Sonographer vacancy rate of 12.6%.^11–13 This evidence demonstrates a national problem in Radiology with increasing demands for imaging, compounded by staff shortages. While ultrasound is a quick and relatively inexpensive imaging modality used to assess clinically suspected Baker’s cysts, in the absence of red flag symptoms and given the ever-increasing strains on Radiology, it is fundamental we image the right patients and avoid performing imaging tests of limited clinical value. This not only reduces waiting times, but additionally is a better use of resources.

We retrospectively reviewed a large cohort of patients referred for a posterior knee swelling to establish the incidence of Baker’s cysts and other pathologies and how the scan influenced patient management. Our specific aim was to determine whether knee ultrasound for posterior knee swellings is a test of limited clinical value.

Methods

A retrospective analysis of all USS knee requests referred from both primary and secondary care over a 2-year period, between January 2017 and December 2018, was carried out. The clinical details for all these patients were reviewed and only those patients with a suspected posterior knee abnormality were included in the study.

Demographic details, referral source and USS findings were recorded. Patients were followed up over a minimum period of 3 years and any subsequent radiological knee examinations were recorded using the Radiology Information System at our institution. Similarly, any patients found to have a solid soft tissue mass at USS were followed up using the hospital histology database.

The ultrasound examinations were performed on a Philips IU22 ultrasound machine (Philips; Bothell, WA, USA) using a 17–5 MHz linear transducer and an MSK superficial pre-set. The examinations were undertaken by six MSK Consultant Radiologists with between 5 and 20 years of experience and three MSK-trained Sonographers with between 5 and 10 years of experience. There was no blinding from the clinical indications for the scan operator. Ethical approval was obtained from the Clinical Effectiveness Unit at our institution.

Results

A total of 1144 patients were referred for a USS of the knee between January 2017 and December 2018. A total of 464 patients were excluded from the study as the clinical details specifically indicated an abnormality of the anterior knee (291), medial knee (94) and lateral knee (79). A total of 680 patients were referred for a posterior knee abnormality and were included in the study. Of the 680 patients referred for a posterior knee USS, 95% (646) patients were referred from primary care. The remaining patients were referred from Orthopaedics (11), Physiotherapy (10), in-patients (8), Accident and Emergency (3), out-patients (1) and Rheumatology (1).

The age range of the study population was 20–92 years with a mean age of 59 years. The study group comprised 32% (221/680) males and 68% (459/680) females. The cases were grouped into posterior knee USS reported as normal (273), Baker’s cyst including ruptured Baker’s cyst (347) or other abnormalities (60) as shown in Figure 1. A list of all the different findings reported in the patients where an alternative abnormality was reported, is summarised in Table 1.

Figure 1.

Ultrasound knee for posterior knee swellings.

Table 1.

Other abnormalities in the popliteal fossa.

Abnormality	N
Lipoma	11
Prominent subcutaneous adipose tissue	9
Ganglion	8
Varicosity	7
Indeterminate	4
Subcutaneous oedema	4
Epidermoid cyst	2
Haematoma	2
Popliteal aneurysm	2
Semi-membranosis bursa	2
Calf tear	1
Deep vein thrombosis	1
Fat necrosis	1
Intramuscular lipoma	1
Peripheral nerve sheath tumour	1
Scarred common peroneal nerve	1
Semi-tendinosis tendinopathy	1
Skin appendage lesion	1
Tenosynovitis	1

Normal popliteal fossa

A total of 273 patients (40%) had a normal USS, with 13 patients having an MRI scan confirming the ultrasound findings of no Baker’s cyst. Of these MRI scans, two were normal, six reported osteoarthritis (OA), four reported either a meniscal tear or degeneration and one reported mucoid degeneration of an anterior cruciate ligament.

Baker’s cyst

A total of 347 patients (51%) had a Baker’s cyst (342) or ruptured Baker’s cyst (5) diagnosed on ultrasound (see Figure 2). Of these patients, the majority of patients (338) were referred from primary care. The remaining patients were referred from Physiotherapy (three), Orthopaedics (three), in-patients requests (two) and Accident and Emergency (one). The size of the Baker’s cyst ranged from 0.6 to 10 cm. The mean size was 3.9 cm and the median was 3.7 cm. A total of 20 patients had an MRI scan confirming the presence of a Baker’s cyst. Of these patients, six reported a Baker’s cyst and a meniscal tear, one reported medial OA and one had the Baker’s cyst closely related to the tibial nerve, although did not require any intervention.

Figure 2.

Transverse USS image of a Baker’s cyst located between the semi-membranosis tendon (Semi M T) and medial head of gastrocnemius muscle (MHG).

Other abnormalities in the popliteal fossa

A total of 60 (9%) patients had other abnormalities identified within the popliteal fossa (see Table 1). The most common abnormalities were lipomas (11), prominent subcutaneous adipose tissue (9), ganglions (8) and varicosities (7). One of the patients with a lipoma had an excision biopsy confirming the USS findings of a benign lipoma and needed no further follow-up. A total of five patients required a follow-up MRI to confirm the USS findings, and these corroborated with the USS reporting a ganglion (two), haematoma (one), intramuscular lipoma (one) and benign peripheral nerve sheath tumour (one). None of these patients had any further investigations. Two patients had a popliteal aneurysm and were referred to vascular surgery. One patient had a DVT on the USS and was subsequently referred to Accident and Emergency for further assessment.

A total of four patients had an indeterminate USS as the soft tissue lesions seen within the popliteal fossa had complex imaging features including cysts with solid components and areas of intrinsic vascularity and therefore an MRI scan was recommended for further evaluation. Of these patients, two had an MRI scan reporting a Baker’s cyst (one) and an intramuscular myxoma (one). The intramuscular myxoma was subsequently confirmed on histology and had no further follow-up. The remaining two patients with indeterminate findings had no further follow-up in a 3-year period following the initial USS.

All the remaining abnormalities had no further imaging or follow-up. No malignancies were identified in the cohort.

Discussion

A posterior knee swelling is a commonly encountered symptom in both primary and secondary care with a significant number of patients referred for a USS. Baker’s cysts are benign, synovial-lined cysts arising from the posterior knee joint and are a consequence of underlying knee joint OA. ¹⁴ On imaging, they characteristically extend into the popliteal fossa from the knee joint, between the medial head of the gastrocnemius and semimembranosus muscles. ¹ The most common clinical finding is a smooth swelling behind the knee that may or may not be painful. Baker’s cysts can also be associated with difficulty in walking or incomplete flexion and extension of the knee. ² The cyst can disappear or decrease with 45 degrees of knee flexion (Foucher’s sign), which is a useful discriminator to distinguish Baker’s cysts from fixed, solid masses that do not change position. ⁶ Less commonly, if a Baker’s cyst ruptures (see Figure 3), a patient can present more acutely with redness, pain and swelling mimicking a DVT. In cases where there is diagnostic uncertainty or red flag symptoms (see Table 2), the National Institute for Health and Care Excellence (NICE) recommends imaging for further characterisation. ¹⁰ Following either clinical or radiological diagnosis, the vast majority of cases are managed conservatively. Surgical management is only reserved for very large cysts that cause neurovascular compromise or for patients receiving knee replacement for the underlying degenerative change.

Figure 3.

Longitudinal USS image of a ruptured Baker’s cyst extending superficially along the posterior aspect of the calf. Callipers outline the cyst.

Table 2.

NICE recommendations for management of suspected Baker’s cysts in adults (2020). ³

□ If any red flags then arrange appropriate imaging/referral: □ Sudden ‘pop’ with increased pain, redness and warmth □ Suspected alternative diagnoses, for example, DVT or mass □ Sudden increase in the size of a lump, change in consistency, increased pain or neurovascular symptoms.

NICE: National Institute for Health and Care Excellence; DVT: deep vein thrombosis.

The aim of the study was to determine the incidence of Baker’s cysts and other pathologies in patients investigated for a posterior knee swelling using USS. This study retrospectively reviewed all patients (680) referred with a posterior knee swelling over a 2-year period and patients had a 3-year follow-up period from their initial USS.

In the literature, the prevalence of Baker’s cysts depends on the imaging modality used and the study population. The vast majority of USS in this study corroborated with the clinical suspicion and identified a Baker’s cyst (51%). A total of five patients in this group (0.7%) had a ruptured Baker’s cyst on USS. The prevalence of Baker’s cysts is higher in this study compared to other studies, with the prevalence of Baker’s cysts reported between 7% and 30%^14–23 (see Table 3). Patients in these studies were predominantly Rheumatology patients or patients with knee pain assessing the diagnosis and management of knee OA. Baker’s cysts in these studies were identified incidentally rather than symptomatically as in this study. Therefore, the prevalence of Baker’s cysts in their populations were likely to be lower than this study.

Table 3.

Prevalence of Baker’s cysts in other studies.

Authors	Year	Patient cohort	Total patients	Baker’s cyst
Iagnocco et al. 15	2010	Rheumatology patients with osteoarthritis (OA) and knee pain	164	10 (7%)
Ragab et al. 16	2012	Rheumatology patients with OA and knee pain	30	8 (27%)
Picerno et al. 17	2014	Rheumatology patients with OA and knee pain	399	102 (26%)
Artul et al. 18	2014	Patients with knee pain	276	47 (16%)
Akgul et al. 14	2014	Patients with knee pain or OA	220	23 (10%)
Abogamal et al. 19	2017	Rheumatology patients with knee OA	1568	203 (13%)
He et al. 20	2017	Patients with OA and knee pain	77	19 (25%)
Phatak et al. 21	2019	Patients with knee pain	100	10 (10%)
Shaaban et al. 22	2019	Orthopaedic/Rheumatology patients with knee OA	20	6 (30%)
Kandemirli et al. 23	2020	Patients with knee OA and knee pain	198	23 (12%)

A large proportion (40%) of patients had no positive findings and were reported as normal. Similarly, Artul et al. ¹⁸ reported 100 (36.2%) normal examinations in their ultrasound study. Another study researched into the prevalence of Baker’s cysts in patients with known OA and reported 87.1% (1365) patients had no Baker’s cyst. ¹⁹ However, this study did not compare the clinical examination to the ultrasound findings. In addition, patients were only included because they had OA rather than a clinical suspicion of a Baker’s cyst; therefore, the number of normal examinations is expected to be higher than in this study.

In the remaining 60 (9%) patients, an alternative pathology was identified. These findings are summarised in Table 1 and the vast majority were not significant. A total of seven (1%) patients had significant abnormalities in the popliteal fossa. Two patients had a popliteal aneurysm and were referred to vascular surgery for further assessment. Of these, one patient did not require any further intervention due to the aneurysm size and was placed on annual aneurysm surveillance. The other patient had a femoral-popliteal bypass graft, although the ultrasound referral was made following a soft tissue abnormality identified on a previous radiograph, which suggested a possible Baker’s cyst or aneurysm. One patient who had a DVT on the USS was referred to Accident and Emergency and was treated with rivaroxaban. However, the DVT was identified as an incidental finding and was not the cause of the patients’ posterior knee swelling. A total of four patients had an indeterminate USS and recommended an MRI scan for further clarification. Of these patients, two had an MRI scan reporting a Baker’s cyst and an intramuscular myxoma, which was subsequently confirmed on histology and required no further management. On ultrasound, the imaging features of the patient with the Baker’s cyst was solid with some peripheral vascularity and was not cystic, hence the recommendation for further imaging. On ultrasound, the imaging features of the myxoma (see Figure 4) were a predominantly complex solid and cystic mass with a small amount of vascularity. The abnormality could not be definitively diagnosed on the subsequent MRI scan and histological diagnosis was recommended. The remaining two patients with indeterminate findings had no further follow-up in a 3-year period following the initial USS. Therefore, only four (0.6%) patients had significant findings on their USS that required a change in management.

Figure 4.

Transverse USS image of an indeterminate complex solid and cystic abnormality on the posterior knee with a small amount of internal vascularity. On histology, this was confirmed as a myxoma.

The retrospective nature of the study meant multiple operators, both MSK Radiologists and MSK-trained Sonographers performed the examinations in the study. Previous studies have typically used experienced Consultant Radiologists and Rheumatologists; however, the use of both MSK Consultant Radiologists and MSK-trained Sonographers reflects practice in most imaging departments in the United Kingdom. A high proportion of patients did not have further investigations following the USS. However, patients were followed up for at least 3 years after the initial scan and these patients did not have any further imaging, biopsy or surgical excision at our institution to disprove the ultrasound diagnosis.

Conclusion

To conclude, the dual purpose of this study was to identify the incidence of Baker’s cysts and other clinically significant pathologies and to determine if the ultrasound findings led to a change in patient management. To the authors’ knowledge, this is the first study to address the clinical value of USS for posterior knee swellings for clinically suspected Baker’s cysts and with a large study group. Our findings confirm the high incidence of Baker’s cysts in patients presenting with a posterior knee swelling. They also show that although there are a variety of other rarer pathologies that can cause posterior knee swellings, only a tiny proportion needed further active management. While ultrasound is a quick and relatively inexpensive imaging modality to assess for clinically suspected Baker’s cysts, the demand for imaging is high and it is fundamental imaging departments avoid performing imaging tests of limited clinical value. It is, therefore, clear that routine ultrasound for posterior knee swelling of clinically suspected Baker’s cysts, in the absence of red flag symptoms, is a test of limited clinical value.

As a result of this study and following consultation with our local orthopaedic department and clinical commissioning group (CCG), we no longer routinely offer USS for posterior knee swellings in the absence of red flag symptoms as described in Table 2. The Integrated Clinical Environment (ICE) referral system has also been adapted to reflect this, so that clinicians are aware of the change in referral guidelines at the point of requesting and additionally all USS requests are now vetted by either MSK Consultant Radiologists or MSK Sonographers. This has helped to significantly reduce the burden on our busy ultrasound department by saving the equivalent of 69 MSK ultrasound sessions per year. In addition, this has helped to improve waiting times for other ultrasound investigations. Given the ever-increasing demand on Radiology, it is fundamental to utilise this finite resource by trying to ensure that we image the right patients with the right tests and avoid performing tests of limited clinical value.