A prospective study of the diagnostic potential of the knee tunnel view radiograph in assessing anterior knee pain☆

1. Introduction 

Routine investigation of anterior knee pain includes the use of radiological assessment. A recent survey of members of the British Orthopaedic Association has revealed a wide variation in the radiographs requested. 18.9% of orthopaedic surgeons requested a tunnel view in the assessment of anterior knee pain and 7.6% requested a tunnel view in the assessment of osteoarthritis (OA). There was no consensus as to the ‘gold standard’ of radiographic assessment [1].

The theoretical advantages of the tunnel view include assessment of intercondylar notch osteophytes, intercondylar notch shape, loose intra-articular osseous bodies and tibiofemoral joint space narrowing [2]. However no previous study has assessed the advantage of the tunnel view over a series of AP, lateral and skyline radiographs of the knee in the assessment of knee pain. There is also a lack of consensus on the technical aspects of how it should be taken.

Determination of the diagnostic potential of the tunnel view would facilitate standardisation of radiographic assessment of the knee joint and reduction of unnecessary radiation exposure for patients. In this prospective diagnostic study we compared a standardised tunnel view radiograph with a series of routine AP, lateral and skyline views in a group of consecutive patients referred with anterior knee pain. Our aim was to determine whether the tunnel view was better than a series of AP, lateral and skyline radiographs for diagnosing intra-articular osseous loose bodies and intercondylar notch osteophytes. Our secondary objective was to determine what other radiographic features a tunnel view radiograph was better at assessing compared to a series of AP, lateral and skyline radiographs.

3. Results 

Our study showed that for assessment of cysts, medial femoral defects, intercondylar notch osteophytes and loose bodies, there was a poor agreement (k<0.4) between the tunnel view and a series of AP, lateral and skyline (Table 1). For all these features there was a significant difference at P<0.05 for the Bowker value comparing the two sets of radiographs. There was a significantly higher score (P<0.05) using the tunnel view alone for determining the presence of cysts, medial femoral defects and intercondylar notch osteophytes. There was a significantly higher score (P<0.05) using the AP, lateral and skyline views for determining the presence of cyst number and loose body numbers.

Table 1. Bowker's test and kappa statistics for the tunnel radiograph compared to the series of AP, lateral and skyline radiographs

	Radiographic feature	Type of radiograph	Number of positive radiographs	Mean score as a % of maximum score	Kappa value (CI 95%)	Bowker value
	Cyst number	A	57	75	0.14 (0–0.34)	0.01
		T	193	30
	Presence of cysts	A	610	9	0.25 (0.14–0.36)	0.00
		T	600	14
	Medial femoral defects	A	609	6	0.18 (0.06–0.30)	0.04
		T	597	9
	Medial tibial defects	A	609	13	0.63 (0.53–0.73)	0.00
		T	597	10
	Intercondylar notch osteophytes	A	609	5	0.16 (0.11–0.20)	0.00
		T	598	25
	Loose body numbers	A	610	15	0.17 (0.07–0.28)	0.00
		T	600	7
	Chondrocalcinosis lateral tibiofemoral joint	A	610	15	0.47 (0.37–0.58)	0.00
		T	600	11
	Osteophytes medial tibiofemoral joint	A	609	21	0.65 (0.60–0.70)	0.00
		T	598	18
	Chondrocalcinosis medial tibiofemoral joint	A	610	9	0.49 (0.35–0.62)	0.00
		T	600	6
	Medial femoral sclerosis	A	610	16	0.47 (0.38–0.56)	0.00
		T	600	23
	Lateral tibial sclerosis	A	610	9	0.43 (0.29–0.56)	0.00
		T	600	5
	Medial tibial sclerosis	A	610	32	0.61 (0.54–0.68)	0.00
		T	600	39
	Tibial spine osteophytes	A	609	39	0.57 (0.51–0.63)	0.00
		T	598	54
	Width medial tibiofemoral joint	A	610	31	0.53 (0.48–0.58)	0.00
		T	600	24

Key: CI=confidence interval, A=series of AP in extension, lateral and skyline radiographs, T=tunnel radiograph.

In the assessment of medial tibial defects, chondrocalcinosis of the lateral tibiofemoral joint, medial tibiofemoral osteophytes, medial tibiofemoral joint chondrocalcinosis, medial femoral sclerosis, lateral tibial sclerosis, medial tibial sclerosis, tibial spine osteophytes and width of the medial tibiofemoral joint agreement between the two series of views varied from fair to good. For all these features there was a significant difference at P<0.05 for the Bowker value comparing the two sets of radiographs. There was a significantly higher score (P<0.05) using the tunnel view for assessing medial femoral sclerosis, medial tibial sclerosis, tibial spine osteophytes and width of the medial tibiofemoral joint. There was a significantly higher score (P<0.05) using the AP, lateral and skyline views to assess chondrocalcinosis of the lateral tibiofemoral joint, medial tibiofemoral joint osteophytes, chondrocalcinosis and lateral tibial sclerosis.

There was no significant difference between the tunnel view and the series of AP, lateral and skyline radiographs when assessing lateral femoral defects, lateral tibial defects, lateral tibiofemoral joint osteophytes, lateral femoral sclerosis and width of the lateral tibiofemoral joint.

4. Discussion 

Osteoarthritis is a degenerative disease of articular cartilage [2], [4]. The prevalence of osteoarthritis (OA) has been reported as 27% of subjects under the age of 70 years and 44% of subjects over the age of 80 years [5]. Degenerative change has been found in the medial compartment alone in 32.5% with 6.3% having changes to the lateral tibiofemoral compartment alone and 9.2% affecting the patellofemoral joint alone [6].

The radiographic definition of osteoarthritis has included the presence of joint space narrowing, sclerosis, cysts, osteophytes and bone defects [4], [7]. A recent survey of orthopaedic surgeons has shown that whilst 100% requested an AP and lateral radiograph for the assessment of OA, only 23.9% requested a skyline radiograph and 7.6% requested a tunnel view radiograph. For the assessment of anterior knee pain the number requesting a tunnel view radiograph rose to 18.9% [1]. There is as yet no consensus on the gold standard for radiographic assessment.

Advantages of the tunnel radiograph include assessment of intercondylar notch osteophytes, tibial spine osteophytes, intra-articular loose bodies and the dimensions of the intercondylar notch. However, prominent intercondylar osteophytes are also visible on lateral radiographs as a line parallel to the condyles and continuous with the trochlear groove [2]. Weight bearing postero-anterior radiographs in 30° of flexion have also been shown to demonstrate significantly more advanced erosion in both the medial and lateral tibiofemoral compartments compared to those in full extension [9] but are technically more difficult to obtain in the elderly population compared to non-weight bearing tunnel view.

Our study has shown that the tunnel radiograph does contribute further information to the radiographic assessment of osteoarthritis. Our results concur with previous studies that the tunnel view radiograph is better for assessing intercondylar notch osteophytes and tibial spine osteophytes than a series of AP, lateral and skyline radiographs [1], [2], [8]. However, we found that for the assessment of loose intra-articular osseous bodies, a series of AP, lateral and skyline radiographs showed more positive findings than a tunnel view radiograph alone. We conclude that when a loose body in the knee is suspected as a cause of anterior knee pain or locking, a tunnel view radiograph should form part of the radiographic assessment in conjunction with an AP, lateral and skyline view but should not be used alone as the only form of radiographic assessment.

Our study found more positive findings in the medial tibiofemoral joint than the lateral tibiofemoral joint. The tunnel view showed more positive findings when assessing medial femoral condylar defects, medial femoral sclerosis and medial tibial sclerosis. The series of AP, lateral and skyline radiographs showed more positive findings when assessing medial tibial defects, medial tibiofemoral joint osteophytes, width of the medial tibiofemoral joint, chondrocalcinosis of the lateral and medial tibiofemoral joints and lateral tibial sclerosis. This may reflect the higher incidence of degenerative changes in the medial as opposed to the lateral tibiofemoral joint in osteoarthritis [6] rather than the tunnel view being better at assessing the medial tibiofemoral joint.

We used a non-weight bearing tunnel view to compare the results with a series of standing AP, lateral and skyline views. Previous studies have shown the advantage of a weight bearing flexed PA view over a standing AP radiograph [9] which our study has not addressed. We have not compared the advantage of a tunnel view radiograph to other imaging modalities such as computed tomography or magnetic resonance imaging for the assessment of anterior knee pain. Our study included both bilateral and unilateral patients with a wide age range (19 to 93 years) since the aim of the study was to test the advantage of a tunnel view radiograph. We did not stratify our data according to age since only 60 of the 309 subjects were below 40 years of age and unequal study groups would have made analysis misleading. Further more, when comparing the unilateral to bilateral subject groups, there was a significant difference in age range and diagnosis which would have resulted in small sample sizes for stratification according to age and number of knees imaged. In order to investigate the diagnostic potential of the tunnel view radiograph according to age, diagnosis and number of knees involved a larger study group with different study design would be needed. This would be the subject of further work.

We concur with previous studies that the tunnel view is valuable in the routine assessment of anterior knee pain [6], [8]. In cases of osteoarthritis, where the clinician is specifically interested in the anatomy of the intercondylar notch and tibial spines a tunnel view radiograph provides valuable additional information in assessing intercondylar notch osteophytes and tibial spine osteophytes in osteoarthritis that cannot be provided by a series of AP, lateral and skyline views alone. A tunnel view radiograph should not be requested as the only means of radiographic investigation if intra-articular loose osseous bodies are suspected.