Background/Purpose: It is widely thought that patellofemoral joint (PFJ) pathology, a source of symptoms in knee osteoarthritis (OA), can be identified using patient history and pain location but this has not been evaluated in an OA population where tibiofemoral joint (TFJ) OA is also frequent. Treatment for PFJ and TFJ disease may differ making the identification of PFJ disease clinically important. Using MRI to identify knees with isolated PFJ disease, we sought to determine the diagnostic utility of anterior knee pain (AKP) and pain with specific activities to identify knees with isolated PFJ structural damage.

Methods: The Multicenter Osteoarthritis Study (MOST) is a NIH-funded prospective cohort study of older adults with or at risk of knee OA. We used data from the 60-month visit, the first visit at which a knee pain map was obtained, limiting our sample to knees with pain, aching, or stiffness in the past year. Subjects identified painful areas around their knee on the knee pain map, which was used to define AKP. Maximal WOMAC pain score was used to classify pain with stairs and walking on level ground (its absence might rule in PFJ pathology) to test their utility in identifying isolated PFJ damage. On MRIs from the same study visit, cartilage damage and bone marrow lesions (BMLs) were semi-quantitatively scored using the Whole Organ Magnetic Resonance Imaging Score. Knees with isolated PFJ damage (full-thickness cartilage loss or a BML in the PFJ and without either full-thickness cartilage loss or BML in the TFJ), isolated TFJ damage, and no structural damage were included (i.e., those with damage in both compartments were excluded). We determined the sensitivity (Sn), specificity (Sp), positive and negative predictive values (PPV and NPV) for AKP, pain with stairs, absence of pain with walking, and the combination of AKP and pain with stairs in identifying isolated PFJ damage. In sensitivity analyses we assessed pain with stair climbing vs. descending separately.

Results: 407 knees met our inclusion criteria (of 1185 knees with MRI read at 60 months). Of these, 193 (47%) had isolated PFJ damage, while 214 (53%) had either no damage (102; 25%) or isolated TFJ damage (112; 28%). AKP or activity pain showed only moderate or low Sn or Sp and correspondingly low PPV and NPV for identifying isolated PFJ (see table). Absence of moderate pain with walking had the greatest Sn (93%) but poor Sp. The combination of AKP and pain with stairs had the greatest Sp (82%) and but had low Sn (29%). We found similar results with other definitions of structural damage, WOMAC pain, and pain with stair climbing vs. descending.

Conclusion: Location of and specific activity-related pain do not appear to highly discriminate underlying isolated PFJ pathology from isolated TFJ or lack of damage. Clinical exam parameters (e.g., localized tenderness on exam) may perform better in identifying isolated PFJ structural damage than patient-reported measures.