The surgical treatment of arthrofibrosis of the knee.

Cosgarea AJ, DeHaven KE and Lovelock JE. Am J Sports Med 22:184-191;1994.

This is the editor's interpretation of a paper published in the orthopaedic literature in 1994 - our attempt to make relevant medical articles accessible to lay readers. If you wish to read the original it is easy to ask your librarian to obtain a reprint for you from any medical library.

This paper discusses 61 procedures (in 55 patients) of lysis ('breaking up') of adhesions where the arthrofibrosis was secondary to -

  • prior knee ligament surgery (43 procedures)
  • fractures about the knee (10 procedures)
  • miscellaneous (8 procedures)
  • previous lysis of adhesions (6 procedures)


The authors felt that there were two major concerns -

  • Loss of range of motion (ROM). Loss of 10 degrees of flexion affects running speed. Loss of more than 5 degrees of extension creates a noticeable limp, and puts a strain on the patellofemoral joint. In general loss of extension is more poorly tolerated - and at 10-20 degree extension loss the disability can be far greater than the initial problem for which surgery was undertaken.
  • Progressive joint surface damage due to the alteration of the forces through the knee and also due to repeated attempts to force the knee into extension.

During the surgery it was noted that the extent of the arthrofibrosis varied from patient to patient -

  • The least involved knees showed only 'anterior intraarticular' involvement, that is to say that there was abnormal fibrous scar tissue only at the front of the notch where the cruciate ligaments lie. All these patients had had earlier ACL surgery, and the fibrous tissue was bunched up to form what is known as a 'cyclops' lesion, causing a loss primarily of extension.
  • Other patients showed more extensive involvement at the front of the joint - the fat pad was extensively scarred and the scarring involved the tissue around the patellar tendon.
  • In other patients there was an even more global pattern - besides the fat pad and tissues around the patellar tendon, dense scar tissue had obliterated the suprapatellar pouch and the medial and lateral parapatellar recesses. These structures are like the cheeks in your mouth - lubricating spaces that allow movement of the joint. These patients generally had deficits of both flexion and extension, and most had been treated originally with prolonged immobilisation.
  • In the worst cases there was this same global pattern, and the patella was 'entrapped' and unable to move and in addition it had pulled downwards, a condition known as 'patella infera'.


Three-stage surgical algorithm

These patients were all dealt with surgically, and the surgeons followed what they called their 'three-stage surgical algorithm'.

That is, starting at stage one of their algorithm all the surgery was begun using the arthroscope, and an initial evaluation was made. Then the procedure progressed according to the findings - via the arthroscope cutting away excessive fibrous tissue scarring ('arthroscopic debridement'), breaking or melting away adhesions ('percutaneous lysis of adhesions'), and where indicated performing a lateral retinacular release (freeing the tight anatomical structures on the outer aspect of the patella) and performing notchplasty (widening the bony notch in the femur where the cruciate ligaments lie). At this stage the ROM of the knee was carefully checked and a controlled manipulated under anaesthesia attempted. If it was considered adequate the surgery was stopped there - if not, the surgeon progressed to stage two - where a small cut was made at the side of the patellar tendon, and scar tissue was cut away in this region. Again the ROM was checked, and if this was adequate the surgery stopped there. If still deemed inadequate the surgeon progressed to stage three. This stage involved opening the knee joint towards the back on the medial (inner) aspect, and cutting away adhesions in that region, as well as releasing any tight capsule at the back of the knee. Again the ROM was evaluated.



Their post op and rehabilitation protocol was only briefly referred to. Indwelling epidural catheters were used for all patients unless there were reasons not to - this offered anaesthesia during surgery and also pain relief after surgery, allowing the physical therapist to commence active and passive ROM exercises on the first post-operative day. CPM machines were used between therapy sessions. Where extension was a problem bi-valved casts were used (casts split down the side to allow removal from time to time for ROM exercises).

During follow up visits, the ROM was measured, instability was assessed and X-rays were taken to look for arthritic changes, patella infera and calcification of the soft tissues (a complication which can occur in arthrofibrosis). The patients were scored on a scoring system to assess whether their function had improved.



The patients were followed up over for an average of 3.6 years. The authors at this stage reported that, although significant gains had been achieved in both flexion and extension, the ultimate functional outcome in many cases was disappointing, and X-rays frequently demonstrated degenerative changes in the joint surface, soft tissue calcification and patella infera.

In those patients required a second lysis of adhesions, both motion and function further improved.

There were two groups of patients, however, who did have good results -

  • the group of patients with the localised intraarticular variant - such as the cyclops lesion
  • the group of patients who had their lysis of adhesions earlier than six months after the trigger incident