We evaluated forty-one hips in thirty-two patients who had had total hip replacement because of osteonecrosis of the femoral head in association with a radiographically normal acetabulum. Radiographs were evaluated independently by the Department of Radiology at our institution and by us. The articular cartilage was also evaluated grossly in all forty-one hips. In addition, a section of articular cartilage was removed, before reaming, from the superior, weight-bearing area of thirty-three of the acetabula and was examined histologically. The cartilage was graded grossly as grade 0 (normal) if no abnormalities were noted; grade I (mild degeneration) if there was superficial fibrillation and slight irregularity of the surface; grade II (moderate degeneration) if there was moderate fibrillation, alteration in color and consistency, and thinning of cartilage without complete erosion to bone; or grade III (severe degeneration) if there was marked fibrillation, alteration in color and consistency, and marked thinning of cartilage with areas of complete erosion to bone. The cartilage was graded histologically as grade 0 (normal) if no abnormalities were noted; grade I (mild degeneration) if there was mild loss of proteoglycan basophilia and there was superficial fibrillation; grade II (moderate degeneration) if there was extensive loss of proteoglycan basophilia, extension of clefts into the transitional or radial zone (or both), and mild thinning of articular cartilage; or grade III (severe degeneration) if there was extension of clefts into the calcific zone and marked thinning of cartilage with areas of complete erosion.
The radiographic stage of involvement was graded with use of the staging system of the University of Pennsylvania (Table I)22, and the patients were evaluated clinically with the Harris hip-rating system11. Certain clinical parameters were documented, including age, gender, duration of symptoms before the procedure, and etiological factors. We then looked for a possible relationship between the severity of the damage of the articular cartilage and the radiographic stage of involvement, the Harris hip score, and the clinical parameters.
Twenty-three patients had had a unilateral total hip replacement, and nine had had a bilateral replacement, with the interval between the two procedures ranging from seven days to eighteen months. The ages of the twenty-five men and seven women ranged from twenty-five to seventy-three years, with a mean of forty-one years. The duration of symptoms before the procedure ranged from three to forty-eight months, with a mean of seventeen months. The use of steroids was implicated as an etiological factor in the osteonecrosis in twenty-eight hips; the use of alcohol, in six (of which four were also associated with a history of steroid use); trauma without fracture, in three; and scuba diving, in two. The disease was considered idiopathic in six hips. Steroids were considered to be an etiological factor when the mean daily dose had been at least ten milligrams of prednisone or a prednisone-equivalent for several weeks. Alcohol was believed to be implicated when the patient stated that he or she had had at least three drinks per day for several months. Although the reasons for administration of steroids were recorded as part of the patient's history, they are not reported here as it was thought that the steroids themselves, and not the underlying disease, were responsible for the osteonecrosis.
The Harris hip scores ranged from 25 to 76 points before the arthroplasty. Although the mean score was 47 points, a total hip replacement was performed on three hips that had scores of 76, 75, and 70 points. This reflects one of the known weaknesses of the Harris hip-scoring system, which perhaps places undue emphasis on pain rather than on the results of physical examination or on function. A few patients in this study had relatively little pain but had a major loss of motion and function, which was not adequately reflected by the Harris hip score. This was the case for the three hips that had scores of at least 70 points. The scores of 70 and 75 points were given to a young, stoic individual with bilateral involvement of the hip who tolerated a poor gait and marked limitation of motion for some time before requesting a bilateral arthroplasty.
Only hips that had had a normal-appearing acetabulum on preoperative radiographic evaluation were included in the study. Radiographic evaluation of the femur revealed a pronounced crescent sign in one hip, but there was no clear evidence of collapse of the articular surface (stage III). In all other hips, there was collapse of the femoral head, which was usually moderate to severe (stage IV). Overall, one hip was stage III-C, four hips were stage IV-A, seventeen were stage IV-B, and nineteen were stage IV-C.
Only one acetabulum was found to have no cartilaginous abnormalities on gross examination at the time of the operation. However, mild degenerative changes were seen on histological examination of a specimen of the cartilage from this acetabulum. All of the other acetabula had gross irregularity of the articular surface that was confined to the superior, weight-bearing region and involved less than 20 percent of the articular surface. On gross examination, the cartilaginous surface of the acetabulum was graded as 0 (normal) in one hip, I (mild degeneration) in sixteen hips, II (moderate degeneration) in twenty hips, and III (severe degeneration) in four hips.
A section of cartilage was removed from the superior, weight-bearing region of thirty-three acetabula by sharp dissection from the underlying bone with use of a scalpel and an osteotome. The specimen did not include subchondral bone. Some acetabula had complete erosion of the articular cartilage and satisfactory specimens could not be obtained for histological examination. The cartilage was fixed in 10 percent neutral buffered formalin, embedded in paraffin, and cut into five-micrometer sections. Specimens were stained with hematoxylin and eosin and examined under light microscopy by an experienced bone pathologist.
All specimens were found to have some degree of degeneration on histological examination. These degenerative changes included loss of proteoglycan basophilia; superficial fibrillation; extension of clefts into the transitional, radial, or calcific zone; and mild-to-severe thinning of the articular cartilage. Sixteen specimens had mild degeneration; fourteen, moderate degeneration; three, moderate-to-severe degeneration; and none, severe degeneration (Figs. 1, 2-A, 2-B, 3 through 4).
There was a general similarity between the gross and histological grades of the articular cartilage. The two grades were the same for sixteen hips, they differed by one grade for fourteen, and they differed by two grades for three. With the numbers available, we were unable to detect a relationship between the status of the articular cartilage and the Harris hip score, the radiographic stage, or any of the clinical factors.
The observations in the present study were not unexpected. It is well known that many types of arthritis begin with inflammation of the joint and early degeneration of the cartilage that cannot be detected radiographically until later in the course of the disease. It could be anticipated that, in osteonecrosis, a certain degree of degeneration would occur in the articular cartilage of the acetabulum after it had been subjected to abnormal mechanical stresses and perhaps biomechanical degradation secondary to gross collapse and irregularity of the femoral head. As mentioned, these changes were limited to a relatively small area in the roof of the acetabulum where major weight-bearing occurs. It is somewhat surprising that these observations have received little attention in the literature on osteonecrosis to date1-3,7,8,15,17-19.
Marcus et al. made brief mention of a depression in the acetabular cartilage opposite a step-off in the femoral head in stage-IV hips15. Habermann examined specimens of acetabular cartilage that had been retrieved, during total hip replacement, from hips with osteonecrosis of the femoral head but with a radiographically normal acetabulum10. He found variable depletion of the proteoglycan on safranin-O staining. We previously presented a preliminary report of our observations21. However, most reports, whether they focused on the pathological evaluation or the staging of osteonecrosis, have concentrated on the initial changes in the femoral head. The acetabular abnormalities have been described at a later stage, after they had appeared on radiographs1-3,7,8,17-19.
Although we found a general similarity between the grades determined on the basis of gross and histological evaluation, there was an apparent inconsistency with regard to cases of more advanced degeneration. Four acetabula were graded as III (severe degeneration) on the basis of the gross examination, but although three were graded as II/III (moderate-to-severe degeneration) on the basis of the histological findings none were graded as III (severe degeneration). In part, this may be explained by the fact that the gross and histological evaluations were done separately by two different individuals, an orthopaedic surgeon and a pathologist, working independently and using somewhat different criteria. Perhaps more important is the method that was used to obtain the specimens for histological examination. The specimens did not include subchondral bone. Thus, the pathologist did not see the areas of most severe degeneration—that is, those in which cartilage had been completely eroded and no specimen was available for histological examination. However, although the inclusion of subchondral bone would have been better for the purpose of histological evaluation, it might have weakened the fixation and the support of the acetabular component to some extent.
The information obtained from the present study may have important clinical implications. Most would agree that, in patients who have advanced osteonecrosis with changes in the acetabulum already apparent on the radiographs, total hip replacement is the procedure of choice when an arthroplasty is clinically indicated. However, when an arthroplasty is indicated for the treatment of an earlier stage of osteonecrosis, in which the acetabulum still appears radiographically normal (stages III and IV), many surgeons might prefer techniques that replace or resurface only the femoral head and not the acetabulum13,16,20,23,24. Osteonecrosis is primarily a disorder of the femoral head, and it is often presumed that when radiographs of the acetabulum appear normal the acetabulum is normal. The present study showed that this may be a false assumption and that radiographs cannot demonstrate early or limited changes in articular cartilage.
It is generally recognized that the results of the use of older, nonmodular press-fit endoprosthetic replacements were poor. These poor results were due to several factors, including poor fixation of the femoral stem within the medullary cavity and a frequent mismatch between the size of the femoral head and that of the acetabulum. The results of endoprosthetic replacements improved considerably when components were solidly fixed either with cement or by biological ingrowth and when the size of the head was closely matched to that of the acetabulum. It was anticipated that even more improvement would be seen with the advent of the bipolar prosthesis. However, there is no evidence that the results with bipolar components are better than those with unipolar components if factors such as fixation and fit are equal. In addition, specific problems with bipolar components have been identified5.
A number of reports have indicated that the results of total hip replacement are better than those of endoprosthetic replacement, even with the use of current techniques and components4,5,14,23. Since the design and fixation of the stems of endoprostheses and total hip prostheses are now similar, it is possible that the difference in the results is attributable to the acetabulum. A metallic femoral head places abnormal stresses on the acetabular cartilage and creates a situation that is far from physiological6,9,12. These stresses are not always well tolerated even by normal articular cartilage, and, as the present study demonstrated, the cartilage in an acetabulum affected by osteonecrosis is no longer normal by the time that an arthroplasty is needed.
Recently, there has been renewed interest in modified cup arthroplasties and hemi-surface replacements for the treatment of osteonecrosis that primarily affects the femoral head. There are certain advantages to this approach. Studies of greater numbers of patients and with longer follow-up are required before the outcomes of procedures with these components can be adequately assessed.
There are several alternatives to consider before managing a patient who has an advanced stage of osteonecrosis. A more detailed discussion of the advantages and disadvantages of each is beyond the scope of this study. However, the surgeon should keep our observations in mind when deciding which component to use and when to do the operation in a patient who has osteonecrosis involving collapse of the femoral head in association with an acetabulum that appears normal radiographically.