Definition of Early Symptomatic Knee Osteoarthritis (KOA)

Definition of early symptomatic knee osteoarthritis (KOA)

and referral criteria from primary care to specialists

Systematic Review 1 – Definition of early KOA

Initials / Name / e-mail / Skype / Phone
CS / Carlo Alberto Scirè / / caskire / 347 8578041
EB / EmanueleBizzi /
LC / Loreto Carmona / / loreto.carmona / +34-628578576
RC / Rayma Castalia
MPR / Maria Piedad Rosario

Objective

To systematically review the literature on the existing definitions of early KOA.

Methods

To search studies with definitions of early knee osteoarthritis, we decided to restrict the PICO(T/S) formula to only P/O. This is because a definition of early knee osteoarthritis can be found either in the description of the target population—eligibility criteria—or the outcome—healthy subjects that are followed-up until the development of OA.

The designs accepted in our search were many different ones: trials, observational longitudinal studies (prospective or retrospective), consensus and Delphi exercises, qualitative studies (interviews, focus groups, etc…). Therefore, we decided not to restrict by design in the search and make the search sensitive enough.

Medline, Embase, and Cochrane library databases, plus ACR/ EULAR abstracts were queried to retrieve relevant information. Searches were restricted to adults, humans, without language limitation. A librarian (MPRL) developed the search strategies, compiled the registries into a bibliographic library (EndNote), disregarded duplicates between databases, and provided the pdfs of the citations selected to be evaluated in full-text. Two reviewers (LC, RC) screened the titles and abstracts of the retrieved hits and read in detail the selected studies. Then they collected the information on the studies on ad hoc extraction forms. These forms included bibliographic data plus study design, setting, inclusion/exclusion criteria, definition of OA, and summary of the results. Risk of bias was evaluated parsimoniously by an experienced clinical epidemiologist based on checklists (SIGN) specific for the design of the individual study. The results of the review are presented according to the PRISMA guidelines[1], with flow-chart and summary tables.

RESULTS

Search strategy

The search strategy run in PubMed is shown in table 1. Similar searches were run in the rest of databases. Besides combining synonym terms for osteoarthritis AND knee, we combined the following potentially descriptors or early disease: “early onset “, “early-onset”, “pre-radiological”, “pre-radiographic”, “pre-clinical”, “recent onset”, “early osteoarthritis”.

Table 1. Search strategy in PubMed

Query (Filters:Humans)
early[MeSH Major Topic]
early[Text Word]
early[Title/Abstract]
onset[MeSH Major Topic]
onset[Text Word]
onset[Title/Abstract]
pre-clinical
pre-clinical[MeSH Major Topic]
pre-clinical[Text Word]
pre-clinical[Title/Abstract]
pre-radio*[Text Word]
pre-radio*[Title/Abstract]
early-onset[Text Word]
"early onset"[Text Word]
((((((early[Title/Abstract] AND Humans[Mesh])) OR (onset[Title/Abstract] AND Humans[Mesh])) OR (pre-clinical[Text Word] AND Humans[Mesh])) OR (pre-radio*[Text Word] AND Humans[Mesh])) OR (early-onset[Text Word] AND Humans[Mesh])) OR ("early onset"[Text Word] AND Humans[Mesh])Filters:Humans
Osteoarthritis[MeSH Major Topic]
Osteoarthritis, Knee[MeSH Major Topic]
osteoarthritis[Text Word]
osteoarthritis[Title/Abstract]
osteoarthritides[Text Word]
osteoarthrosis[Text Word]
osteoarthroses[Text Word]
Arthritis, Degenerative[Text Word]
Arthritides, Degenerative[Text Word]
knee[Title/Abstract]
((((((Osteoarthritis[MeSH Major Topic] AND Humans[Mesh])) OR osteoarthritis[Text Word]) OR osteoarthritides[Text Word]) OR osteoarthrosis[Text Word]) OR osteoarthroses[Text Word]) OR Arthritis, Degenerative[Text Word]
((((((((Osteoarthritis[MeSH Major Topic] AND Humans[Mesh])) OR osteoarthritis[Text Word]) OR osteoarthritides[Text Word]) OR osteoarthrosis[Text Word]) OR osteoarthroses[Text Word]) OR Arthritis, Degenerative[Text Word])) AND knee[Title/Abstract]
((((((((((Osteoarthritis[MeSH Major Topic] AND Humans[Mesh])) OR osteoarthritis[Text Word]) OR osteoarthritides[Text Word]) OR osteoarthrosis[Text Word]) OR osteoarthroses[Text Word]) OR Arthritis, Degenerative[Text Word])) AND knee[Title/Abstract])) OR (Osteoarthritis, Knee[MeSH Major Topic] AND Humans[Mesh])
((((((((((((Osteoarthritis[MeSH Major Topic] AND Humans[Mesh])) OR osteoarthritis[Text Word]) OR osteoarthritides[Text Word]) OR osteoarthrosis[Text Word]) OR osteoarthroses[Text Word]) OR Arthritis, Degenerative[Text Word])) AND knee[Title/Abstract])) OR (Osteoarthritis, Knee[MeSH Major Topic] AND Humans[Mesh]))) AND ((((((((early[Title/Abstract] AND Humans[Mesh])) OR (onset[Title/Abstract] AND Humans[Mesh])) OR (pre-clinical[Text Word] AND Humans[Mesh])) OR (pre-radio*[Text Word] AND Humans[Mesh])) OR (early-onset[Text Word] AND Humans[Mesh])) OR ("early onset"[Text Word] AND Humans[Mesh])) AND Humans[Mesh])

A total of 1522 records were captured in the three major databases. After removal of duplicates and exclusion by title/abstract, 26 articles were reviewed in full (See Figure 1). Of these, 17 were excluded, being the reasons explicit in table 2. The main reason was the absence of a definition of early OA.

Figure 1.

Table 2. Excluded studies and reasons for exclusion.

Author (year) / Exclusion criteria
Cai (2013)[1] / No definition of early knee osteoarthritis
Hart (1991)[2] / No definition of early knee osteoarthritis
Hart (1999)[3] / No definition of early knee osteoarthritis
Hunter (2011)[4] / No definition of early knee osteoarthritis
Keen (1997)[5] / No definition of early knee osteoarthritis
Kinds (2013)[6] / No definition of early knee osteoarthritis
Lane (2011)[7] / No definition of early knee osteoarthritis
Laxafoss (2010)[8] / No definition of early knee osteoarthritis
McGonagle (2010)[9] / No definition of early knee osteoarthritis
Menche (2003)[10] / Letter to the editor
Mezghani (2012)[11] / Congress abstract without clear definition
Mobasheri (2012)[12] / Congress abstract without clear definition
Morshedi (2011)[13] / Congress abstract without clear definition
Rousseau (2012)[14] / No definition of early knee osteoarthritis
Schiphof (2008)[15] / No definition of early knee osteoarthritis
Sowers (2012)[16] / No definition of early knee osteoarthritis
Zhang (2010)[17] / No definition of early knee osteoarthritis

Of the 9 studies included, 6 were observational studies, 1 a RCT, and 2 expert opinion. The risk of bias of the studies depended mainly on the objectives, but in relation to how the definition of early KOA was decided, none had low risk of bias.

Not all studies defined early OA unambiguously; they mainly defined other types of recently diagnosed KOA, such as incident OA, or MRI definitions of KOA.

Table 3 shows the designs, objectives and early OA definitions in the studies included.

Table 3. Summary table

Author (year) / Country/setting / Study design / N participants / Inclusion/exclusion criteria / Study objectives (recommendations, treatment efficacy, disease course….) / Early OA definition / Risk of bias*
Ling (2009) / USA / Population-based / Case-control nested within the Baltimore Longitudinal Study of Aging (BLSA) / 88 / Subjects with 2 sets of X-rays, the first of which was scored as no evidence of OA, and a set of banked serum samples obtained over the corresponding time interval. / To test the hypothesis that the early stage of knee and hand OA development is characterized by detectable changes in serum proteins relevant to inflammation, cell growth, activation, and metabolism several years before OA becomes radiographically evident. / Incident OA cases: Radiographic OA of one or both knees in the second X-ray. / Low - moderate
Schiphof (2014) / Netherlands / Population Study (Rotterdam study, a
population-based study that investigates prevalence,
incidence and risk factors for various chronic disabling
diseases among persons aged 55 years) / Cohort study / 888 women (1766 knees) / Women aged 45 to 60 years / To identify early signs of knee OA by:

1. Radiographic examination of both knees
2. MRI
3. Physical exam
4. Knee-specific questionnaire.

At baseline and after 2 years.
Objectives:

1. To examine whether the MRI definition for knee OA in an open-population cohort of middle-aged women is more sensitive in detecting structural knee OA than the radiographic K&L criteria.
2. To assess construct validity by assessing the association between the knee OA definitions and
3. knee pain at baseline
4. persistent knee pain during the 2-year follow-up period,
5. new onset of knee pain 2 years later,
6. body mass index (BMI).

/ MRI definition for tibiofemoral OA (TFOAMRI) was defined as
the presence of a definite osteophyte and full-thickness cartilage loss, or as a definite osteophyte or full-thickness cartilage loss plus two of the following features:

1. subchondral BML or cyst not associated with meniscal or ligamentous attachments,
2. meniscal subluxation, maceration or degeneration (including a horizontal tear),
3. partial-thickness cartilage loss,
4. or bone attrition.

‘Meniscal subluxation, maceration or degeneration (including a horizontal tear)’ was considered present if meniscal degeneration or a horizontal tear was present.
New onset knee pain was defined as knee pain at follow-up while not having current knee pain or knee pain in the last year at baseline measurement. / Low - moderate
Baert (2013) / Belgium / multicenter, private practice / Cross-sectional observational study / 65 / Women with symptomatic KOA (ACR classification criteria) and asymptomatic women.
Excluded:
-MSK disorders other than KOA in one or both lower limbs in the last 6 months,
-previous surgery of lower extremities or spine,
-neurological disorders,
-chronic intake of corticosteroids,
-contra-indications for MRI. / To compare knee proprioception between early knee OA subjects, established knee OA subjects, and healthy controls. / Recurrent knee pain with pre-radiological degenerative joint changes that are observed by either arthroscopy and/or MRI. / Moderate
Larsson (1998) / Sweden / population-based / Cross-sectional observational study / 204 / Subjects 35–54 years old with chronic knee pain (>3 months duration).
Exclusion:
-inflammatory joint diseases
-knee joint traumas related to knee pain. / To test the association between functional measurement variables and early radiographic signs of tibiofemoral OA of the knee in a population aged 35–54 years with long-standing knee pain. / Early radiographic KOA:
-35–54 years old
-chronic knee pain (>3 months duration)
-OA in X-Ray (Ahlbäck criteria). / Moderate
Nakamura (2006) / Japan / Hospital-based study / Cross-sectional study / 63 patients (21 males and 42 females) from 40 to 59 years / To have some knee complaints.
Exclusion:
-Ligament injuries caused by trauma
-Kellgren Radiographic Grades III and IV / To evaluate the relationship between the shape of the tibial spur on X-ray and the
meniscal changes on MRI in early knee OA.
All pts underwent X-ray and MRI. / K-L < II / High
Qazi (2007) / Denmark / Hospital-based study / Cross-sectional study / 76 persons (34 men and 42 women) (114 knees) / Healthy individuals with no or only minor OA symptoms with a broad age range from 22 to 73 years.
+
Subjects with known
OA symptoms of various degrees.
Exclusion:
-knee joint replacement,
-inflammatory arthritis in the knee
-any contraindication for MRI / To investigate whether intrinsic changes in the articular cartilage related to water distribution visualized by MRI (cartilage homogeneity on the medial compartment of the tibial cartilage) would enable separation of early OA vs healthy individuals.
Subjects underwent both clinical and radiological examinations and classified into groups of healthy subjects or OA patients according to the ACR (American College of
Rheumatology) definition of OA. / Unspecified: “none or only minor OA symptoms” but Kellgren-Lawrence>0 / High
Lotke (2000) / USA / practice-based / Case-series.
Description / expert opinion. / 41 / Subjects with acute or subacute onset of pain on the medial side of the knee associated with well-localized tenderness along the medial tibial plateau, just distal to the joint line.
Initial X-Rays could not show evidence of osteonecrosis and MRI must show abnormal changes in the tibial metaphysis. / To classify the MRI findings on the tibial side of the knee joint and to define the clinical course associated with these findings. / Radiographic staging:
Stage 2 was considered early degenerative changes with spurring but no narrowing. / Moderate-high
Luyten (2012) / International / Expert opinion / 5 experts / Participant’s expertise. / To define classification criteria of early knee OA. / Early knee OA:

1. Knee pain: at least two episodes of pain for >10 days in the last year
2. Standard radiographs: Kellgren–Lawrence grade 0 or I or II (osteophytes only)
3. At least one:
4. Arthroscopy ICRS grade I-IV in at least two compartments or grade II-IV in one compartment with surrounding softening and swelling
5. MRI, at least two:

b.i.Cartilage morphology WORMS 3–6
b.ii.Cartilage BLOKS grade 2 and 3
b.iii.Meniscus BLOKS grade 3 and 4
b.iv.Bone marrow lesions (BMLs)
b.v.WORMS 2 and 3 demonstrating articular cartilage degeneration and/or meniscal degeneration, and/or subchondral BMLs. / High
Farr (2010) / USA, multicenter / Randomized controlled trial / 293 / Subjects 35 to 68 years old and early-onset KOA / To determine the effect of a structured resistance training intervention on overall daily levels of activity by using accelerometry to measure moderate and vigorous intensity physical activity in individuals with early-onset KOA.
Subjects were randomly assigned to an RT program and a self-management program. / Early-onset KOA:
-pain on 4 or more days of the week
-in one or both knees
-for at least 4 months during the previous year;
-less than 5 years’ symptom duration;
-radiographic status of grade II OA (and no higher) in at least one knee, as defined by the Kellgren and Lawrence classification;
-and disability due to knee OA, as assessed with WOMAC Index. / Moderate

* Risk of bias for the definition of early OA.

Abbreviations: ACR, American College of Rheumatology; KOA, knee osteoarthritis; MRI, magnetic resonance imaging; MSK, musculoskeletal; WOMAC, Western Ontario and McMaster Universities Osteoarthritis;

DISCUSSION

A systematic review [15] found 25 classification criteria for OA, which could be summarized in three categories:

1. Radiological classification criteria, based on pathological changes such as JSN, OP, cysts, sclerosis, bone deformity, including the K&L radiological classification criteria, the Ahlback radiological classification criteria, and other radiological classification criteria;
2. Clinical criteria, such as ACR clinical, ACR clinical & laboratory criteria, and other clinical classification criteria; and
3. Clinical and radiological OA combined, such as ACR clinical & radiological classification criteria and other clinical & radiological classification criteria.

In this review by Schiphof, the majority of the radiological classification criteria showed good to very good reliability (k between 0.6 and 0.8 and above 0.8), while the reliability of the clinical OA and combined classification criteria was poorer. Moderate-to low sensitivity and somewhat higher specificity were reported for radiological classification criteria compared to clinical classification criteria. The associations for pain and obesity with the criteria were more or less comparable between the different classification criteria. None of the classification criteria was assessed with good content validity. Table 4 shows the results of this SR.

Knee OA is mostly defined by clinical criteria and radiography with the Kellgren and Lawrence (K&L) classification criteria [18]. Early OA as detected by the K&L score may be the result of long ongoing biochemical processes leading to both bone and cartilage alterations. Prior to cartilage fibrillation, local edema and swelling are well recognized features, and might be detected by MRI and X-ray based techniques.

Several MRI follow-up studies in high-risk populations for knee OA are ongoing, e.g., the incidence cohort of the Osteoarthritis Initiative (OAI)[19], the Prevention of knee Osteoarthritis in Overweight Females (PROOF) study[20], and a subpopulation of the Rotterdam Study [21]. These studies all address the definition of early predictive features of knee OA using MRI, but before this is possible, a definition for established knee OA on MRI is required first. In 2011 an MRI definition of structural knee OA was proposed by an expert group using a Delphi exercise [4]. Compared to radiography, more knee tissues are visible on MRI; consequently, this MRI definition of OA contains more features than the radiographic K&L definition. The MRI definition needs further evaluation, especially in persons with early signs of knee OA or at high risk for developing knee OA.

Table 4. Results of the systematic review by review by Schiphof [15].

The Schiphof study [21] shows that knee pain at baseline and a definition of tibial femoropatellar MRI OA has a larger associations than between knee pain at baseline and K&L >=2 alone. The construct validity of a definition can be determined by the strength of associations between the definition and factors that are known to be related to OA symptoms (such as knee pain), or risk factors (such as body weight).

Although many features of OA can be detected in totally asymptomatic persons, the combination of MRI features, biomarkers, and clinical symptoms, might detect knee OA at an earlier stage than the K&L criteria.

A review evaluated the diagnostic an prognostic value of biomarkers in clinical studies [14]. Despite promising results with synovial fluid levels of PIICP and CTX-II, serum COMP (elevated in hip but not in KOA), serum keratan sulfate, determined by HPLC after digestion with keratanase II, mutation Arg519- Cys in type II collagen, type II collagen markers uC2C, uC1,2C and uCTX-II, Fib3-1 and Fib3-2, and uCTX-II levels, (table 5), the conclusion of this review is that no biomarker can still be considered as a valid tool for the diagnosis and the prognosis of OA in routine clinical practice.

Table 5. List of biomarkers in the review by Rousseau [14].

REFERENCES

1. Cai S, Yang S, Zheng F, Lu M, Wu Y, Krishnan S: Knee joint vibration signal analysis with matching pursuit decomposition and dynamic weighted classifier fusion. Comput Math Methods Med 2013, 2013:904267.

2. Hart DJ, Spector TD, Brown P, Wilson P, Doyle DV, Silman AJ: Clinical signs of early osteoarthritis: Reproducibility and relation to x ray changes in 541 women in the general population. Annals of the rheumatic diseases 1991, 50(7):467-470.

3. Hart DJ, Doyle DV, Spector TD: Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women: The Chingford study. Arthritis and rheumatism 1999, 42(1):17-24.

4. Hunter DJ, Arden N, Conaghan PG, Eckstein F, Gold G, Grainger A, Guermazi A, Harvey W, Jones G, Hellio Le Graverand MP et al: Definition of osteoarthritis on MRI: Results of a Delphi exercise. Osteoarthritis and Cartilage 2011, 19(8):963-969.

5. Keen RW, Hart DJ, Lanchbury JS, Spector TD: Association of early osteoarthritis of the knee with a Taq I polymorphism of the vitamin D receptor gene. Arthritis and rheumatism 1997, 40(8):1444-1449.

6. Kinds MB, Marijnissen AC, Bijlsma JW, Boers M, Lafeber FP, Welsing PM: Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort. J Rheumatol 2013, 40(1):58-65.

7. Lane NE, Brandt K, Hawker G, Peeva E, Schreyer E, Tsuji W, Hochberg MC: OARSI-FDA initiative: Defining the disease state of osteoarthritis. Osteoarthritis and Cartilage 2011, 19(5):478-482.

8. Laxafoss E, Jacobsen S, Gosvig KK, Sonne-Holm S: Case definitions of knee osteoarthritis in 4,151 unselected subjects: Relevance for epidemiological studies: The copenhagen osteoarthritis study. Skeletal radiology 2010, 39(9):859-866.

9. McGonagle D, Tan AL, Carey J, Benjamin M: The anatomical basis for a novel classification of osteoarthritis and allied disorders. J Anat 2010, 216(3):279-291.