Synovial fluid analysis in the diagnosis of joint disease

Paul Hermansen and Tony Freemont

Paul Hermansen is Consultant BMS in the Cytology Department, Central Manchester

University NHS Foundation Trust

Tony Freemont BSc, MD, FRCP, FRCPath is Professor of Osteoarticular Pathology, University

of Manchester, Director of the Manchester MRC/EPSRC Molecular Pathology Node,

Manchester NIHR Biomedical Research Centre. UK.

Conflicts of interest: none declared.

Abstract

Normal synovial fluid consists of a transudate of plasma from synovial blood

vessels supplemented with high molecular weight lipid and saccharide-rich

molecules. This produces a paucicellular, viscous fluid, which behaves like a

tissue. In primary (e.g. rheumatoid disease) and secondary (e.g. septic and

crystal-induced) inflammatory arthropathies, changes occur to the cell

numbers and cell type in the fluid forming the basis of a diagnostic test. The

diagnostic value is enhanced by the appearance of endogenous and exogenous

particles, particularly those associated with degenerative, crystal-induced, and

prosthesis-associated arthropathies (e.g. fibrin, cartilage, crystals, metal and

plastic). Cells and particles can be characterised under the microscope leading

to a simple, inexpensive, “cytological” test for every type of joint disease.

Keywords arthritis; crystals; microscopy; polarisation; synovial fluid

1

Why perform synovial fluid microscopy?

Histopathologists are frequently called upon to diagnose inflammatory or non-

inflammatory arthropathies based on the examination of a synovial biopsy.

Even experienced histopathologists can have difficulty in distinguishing

inflammatory from non-inflammatory arthropathies since the latter frequently

have a lymphocyte infiltrate in the synovium. Even if they can be distinguished

it is usually impossible to make specific diagnoses as there are few differences

that can be detected histologically between disorders in the same broad

group.

By contrast, synovial fluid (SF) microscopy is of greatest value in these

disorders, supporting clinicians in making early and accurate diagnoses of a

spectrum of inflammatory and non-inflammatory arthropathies often before

the full blown syndrome develops.

SF microscopy, on as little as a 0.5ml sample, permits the rapid diagnosis of

joint disease (the full test takes between 10 minutes for crystal arthropathies

and 2 hours for full cytological analysis), particularly disorders such as sepsis

and crystal related arthropathy where the prognosis is inversely related to

delay in diagnosis.1

The properties of synovial fluid

SF consists of a transudate of plasma from synovial blood vessels,

supplemented with high molecular weight lipid and saccharide-rich molecules

2, produced by one of the two main types of synovial cells (type B synoviocytes

derived from synovial fibroblasts). Type A synoviocytes are tissue macrophage-

derived phagocytes that remove debris from the synovial fluid.

SF differs from all other body fluids in that the synovium and cartilage, which

are the tissues it contacts, do not have an intact surface cellular layer seated

2

on a basement membrane. This means that the matrix of both tissues is in

direct contact with and through SF allowing an homogenous biological

environment to develop within the joint. Because of this it is probably better to

regard SF as a semi-liquid, avascular, hypocellular, connective tissue rather

than a true body fluid such as is seen in other situations (e.g. pericardial

effusion).

Cytological analysis of SF differs in three important ways from other body

fluids. Firstly synovial joints are only very rarely affected by primary or

secondary malignancies. Secondly the term “cytological analysis” of SF would

be better described as “microscopy” of SF, as many of the diagnostic features

found are not cellular but are particles such as cartilage, crystals and prosthetic

wear particles. Thirdly the greatest diagnostic information comes from the

recognition of individual cell types and their quantification.

Normal SF:

Is a supplemented ultra-filtrate of plasma

Contains no clotting factors

Is viscous

Is paucicellular (<500 cells/mm3)

Contains no particles

Any or all of these may change with joint disease.

High molecular weight clotting factors are introduced into the joint as a result

of vascular leakage following trauma or in inflammation. This results in the

need to anticoagulate specimens. The choice of anticoagulant can have

profound influence on the final diagnosis. Anticoagulants such as EDTA will,

because of their sequestering properties, dissolve complex calcium phosphates

such as seen in bone-derived hydroxyapatite and calcium pyrophosphate

3

making the diagnosis of osteoarthritis and pseudogout difficult, if not

impossible. Other anticoagulants, such as sodium heparin, can crystalise out

leading to the false diagnosis of crystal arthritis. Lithium heparin, by contrast,

has none of these faults and is the anticoagulant of choice.

As, at the time of aspiration it is impossible to predict which SF will clot or

contain crystals, all SF samples should be anticoagulated with lithium heparin.

The viscosity of the SF is dependent on the concentration and size of the

proteoglycans it contains. Normal SF and that from non-inflammatory

arthropathies is very viscous. By contrast, in inflammatory arthropathies,

inflammatory mediators and/or products cause abnormal synthesis or

breakdown of existing proteoglycans, reducing viscosity.

Viscosity of SF can be assessed by mixing SF and a 2% solution of acetic acid

leading to the formation of a white precipitate, produced by aggregation of

proteins and hyaluronans (the mucin clot test). The nature and amount of

precipitate varies from good to poor and reflects the quality and quantity of

the protein/hyaluronan complex, such that in inflammatory joint diseases and

haemarthroses there is poor clot formation. Non-inflammatory arthropathies

exhibit a good mucin clot.

The mucin clot test is not normally performed on routine specimens, but it is

very useful in identifying surgical joint washouts or dilution with local

anaesthetic, where, in the presence of a low cell count no clot will form. This is

an important observation as dilution of SF with water will affect cell counts and

remove monosodium urate crystals, adversely affecting diagnostics.

Why examine synovial fluid?

Basic examination of SF will result in identifying combinations of cellular and

non-cellular biomarkers, which can have diagnostic, prognostic and theranostic

4

(following responses to treatment) value, in primary, secondary and tertiary

care, where SF analysis can be used to influence the choice of care pathway,

and assess treatment options.

This valuable role for SF analysis is made possible because the absence of a cell

or basement membrane barrier between synovium, cartilage and SF means

that pathological changes in the tissues surrounding the joint are reflected in

the volume, cellular composition and particulate load within the SF.

SF is examined “fresh” requiring careful handling in the laboratory and a need

to minimise cell and crystal loss, which can dramatically adversely affect

diagnosis, during transport to the laboratory. Storage at 4oC is well tolerated by

the SF and prolongs transport times, but even so, adequate diagnosis that even

cooled samples must reach the laboratory within 48 hours of aspiration3.

Basic approach to synovial fluid analysis

In order that no part of the analysis is omitted there is a sequential

examination of SF specimens arriving in the laboratory. It follows four steps:

Gross analysis

Nucleated cell count

The “wet prep”

Preparation and examination of a stained cell monolayer

Gross analysis relies on the visual inspection of the SF to determine colour,

clarity and viscosity.

Colour will change, particularly following haemorrhage such as that seen in

trauma, anticoagulant use and in primary disorders of joints such as pigmented

villonodular synovitis.

5

Clarity of SF is variable and a cloudy or opaque appearance generally indicates

an increase in cellular concentration, crystal content or the presence of lipid,

microscopic clarification is then necessary.

Total nucleated cell count

A total nucleated cell count is performed in order that the degree of

inflammation can be assessed and so that an optimally diluted cell suspension

can be achieved for the subsequent cytocentrifuge preparation.

The traditional method of counting cells is to use a graduated pipette to dilute

the SF by a known amount with normal saline, to which crystal violet has been

added as a supravital stain. This cell suspension is then introduced into a

haemocytometer counting chamber where the cells are manually counted on

the microscope.

With the advent of newer technologies for cell counting, assessing cell number

has now changed. The nature of the proteoglycans makes microbore based cell

counting techniques tricky but computerised chamber counting methodologies

are highly practical and accurate. These automated methods, although in

themselves more costly in terms of initial investment and consumables, can

make SF analysis quicker and more accurate than the manual method, and, by

reducing the amount of biomedical scientist time required, reduce the overall

cost of the test.

The number of nucleated cells contained in the SF is the primary indicator of

inflammation. Where the cell count is less than 500/mm3 the patient can be

assumed to have a non-inflammatory arthropathy, whereas a cell count

greater than 1000/mm3 indicates an inflammatory arthropathy. If between

these figures, the percentage of neutrophils in a differential count

distinguishes inflammatory from noninflammatory arthropathies, where more

6

than 50% of the nucleated cells in the sample being polymorphs indicates

inflammation.

Wet preparation

The “wet prep” is an important part of the examination of SF and can often be

relied upon to give a definitive diagnosis. Two preparations are made each

serving a different purpose. Firstly using a glass Pasteur pipette for clarity, an

aliquot of SF is removed from the container and by slowly returning the SF to

the container visible particles of fibrin, cartilage or crystal aggregates can be

seen in the thin part of the pipette and placed onto a microscope slide, this we

call the “thick preparation”. It can be of great help if the viewing background

provides contrast, for example fragments of cartilage are most easily seen

against a dark background whereas prosthetic debris is most easily seen

against a pale background, white laboratory coats and dark laboratory bench

surfaces usually provide this contrast. It is of great importance that any pieces

of fibrin clot are found, as frequently identification of crystal arthropathies

with a low crystal burden relies on finding crystals trapped in the fibrin clot.

A second, much thinner preparation, is made avoiding particulate matter and

using only a few microlitres of SF. Coverslipping this preparation flattens the

cells and allow intracytoplasmic inclusions to be identified. Large inclusions

with specific refractile properties characterise a functional group of cells called

ragocytes. Ragocytes have diagnostic significance (see below).

“Wet preps” can be preserved for 24 hours by painting around the edge of the

coverslip with nail varnish to minimize evaporation. This also helps provide

additional safety features, as the SF is a potential infection hazard.

Thick preparation examination

Crystals

7

When screening for crystals it is of the greatest importance that aggregates of

fibrin and other particles are included in the slide preparation, as it is these

micro clots that will often contain the crystals even though the surrounding

fluid and cells may not. A recent review 4 shows that there is poor consistency

in crystal identification and characterisation between individuals and

laboratories unless adequate training of personnel is first carried out, but we

find the simple expediency of including clots in the “wet prep” reduces this

problem.

A list of the characteristics of the main crystal types can be seen below in Box

1.

Mono sodium urate crystals (MSU) are seen in microscopic preparations as

thin rod/needle shaped crystals or even collections of crystals radiating from a

central point, which are described as “beach balls”. When examined by

polarising microscopy they are highly birefringence, meaning they are bright

white against the black background of the fully crossed polarizing filters.

Additional information on the nature of the crystal can be gained by adding a

quarter wave retardation plate to the polarising system. This plate retards the

polarised light by one quarter of a wavelength which produces a magenta

coloured background.

The slow wave of the plate is often marked by a double headed arrow

frequently accompanied by the Greek symbol “ɤ”. In most cases the slow wave

of the plate is perpendicular to the long axis of the plate holder. If the

longitudinal axis of the crystal is aligned parallel to the long axis of the

retardation plate and the crystal appears blue the crystal is said to have a

positive sign. If the same crystal were rotated through 90o aligning the

longitudinal axis along the slow wave of the plate the crystal would be yellow.

8

This colour is characteristic of mono sodium urate crystals. With the light path

setup in this way crystals of calcium pyrophosphate (the crystals seen in

pseudogout) are blue (ie the opposite of urate crystals). Unfortunately some

microscopes are configured such that the image is reversed, so that urate

crystals would be blue and pyrophosphate yellow at the image seen through

the eyepieces has been reversed. It is therefore critically important that each

individual microscope is “calibrated “with a known urate specimen.

Often urate crystals have been phagocytosed becoming intracellular. When

found in an inflammatory setting as judged by a WBC count of >1000/mm3, the

presence of urate crystals is diagnostic of acute gout. If found on a background

of a non-inflammatory arthropathy, the presence of urate crystals still signifies

gout but in a quiescent or latent form at the time of aspiration.

Calcium pyrophosphate dihydrate crystals (CPPD) are of a much “weaker”

birefringence than the mono sodium urate crystals and appear as short rods

rhomboids or cubes. As described above they have an opposite sign to urate,

thus, when aligned along the fast wave of the plate they are yellow. In

common with monosodium urate crystals they are often seen within cells.

Pseudogout is diagnosed when the nucleated cell count exceeds

1000cells/mm3. CPPD in a non-inflammatory setting (nucleated cell count

<1000 cells/mm3) could indicate latent pseudogout but more commonly

osteoarthritis or chondrocalcinosis.

Pseudogout is a disease of old age. If seen in a relatively young patient,

particularly a male below 50 years of age, a secondary underlying diagnosis of

a systemic metabolic disease such as hypothyroidism, haemochromatosis or

hypomagnesaemia should be considered.

9

Both mono sodium urate and calcium pyrophosphate can occasionally be

found in the same fluid indicating a mixed crystal arthropathy [5] Figure 1.

Apatite as a true crystalline form of hydroxyapatite is somewhat of a rarity in

SF, only occurring as spherulites where it is then diagnostic of the peri-articular

calcification known as Milwaukie shoulder.6 Although originally described in

the shoulder this condition has been noted in many other joints. When apatite

is seen in SF it is normally in the form of bone dust or bone debris as a result of

the mechanical removal of bone or calcified cartilage, following cartilage loss,

typically as a result of osteoarthritis Figure 2. If seen in an inflammatory setting

destructive, erosive arthropathies such as rheumatoid or psoriatic arthritis, as

well as secondary osteoarthritis, should be considered. Apatite is also

commonly seen in joints in which there is loosening of joint prostheses.

Apatite crystaloids have low or zero birefringence. Although not difficult to

visualise microscopically a simple chemical test can be used to improve

detection. A drop of SF, preferably including a piece of fibrin clot, is placed on

the slide and a solution of 2% alizarin red S dye in 4% acetic acid added and a

coverslip applied. This simple chemical reaction produces an easily visualized,

bright red, highly birefringent compound co-localising with the apatite.7 This

test is specific and sensitive to any complex calcium phosphate, and therefore

CPPD crystals will also react.

Lipids can often be seen in SF as droplets of fat either within or (more

commonly) outside cells. Some lipids occur in crystalline forms such as the

plates of cholesterol or the droplets of cholesterol ester exhibiting the classic

“Maltese cross” birefringence. One of the pitfalls that can cause confusion in

the analysis of SF is the failure to recognize the difference between urate

crystals and the presence of fatty acid crystals. Urate crystals either singly or as

10

“beach balls” are never within a lipid droplet, whereas fatty acid crystals, which

are also highly positively birefringent, and which can also occur as single

needle shaped crystal or “beach balls”, can only occur within lipid droplets

(Figures 3 and 4).

Contamination of the SF by depot steroid crystals can be problematic and

causes difficulties during crystal examination where they are seen as highly

birefringent crystals that have no discernible optical sign.

There are many other rarer crystals seen in SF which can be discounted in day

to day practice, but one that can give excellent, reliable, clinical information is

hematoidin. This self-coloured orange/brown crystal that occurs in either

rhomboidal or fern like form, is highly birefringent, the rhomboidal form being

bright orange and the fern form bright green in polarized light. In a non-

inflammatory SF the significance is that of a chronic haemarthrosis, whereas,

much more importantly, when associated with an inflammatory SF they are

diagnostic of septic arthritis. Haematoidin is most often present in untreated

sepsis of at least 1 week’s duration Figure 5. Although formed from the

breakdown of blood, they have no iron content and are similar to bilirubin in

their chemical composition.

Debris

The inside of the synovial joint is surrounded by synovial tissue and cartilage, in

addition, the knee joint contains the fibrocartilagenous menisci and the

ligamentous tissue of the cruciate ligaments. In disease, particularly in acute or

low grade recurrent trauma, any of these tissues may fragment and the

fragments be seen by SF microscopy. However the absence of these tissue

fragments in SF does not mean they are not present in the joint as the particles

in the joint may be too large to be aspirated using a conventional needle. In

11

general the larger the bore of the needle, bearing in mind patient compliance,

the more likely is the detection of diagnostic particles.

Diagnostically, one of the most useful types of debris is fragmented cartilage

containing clusters of chondrocytes. Only in osteoarthritis does articular

cartilage undergo chondrocyte proliferation, so this feature is diagnostic of

osteoarthritis. Chondrocyte proliferation is associated with cartilage fibrillation

and so another feature of osteoarthritis is the presence of fragments of

fibrillated cartilage, fragments of which in SF are “striped” in polarising light an

appearance we describe as resembling a “tigers tail”. Figure 6.

Frequently tigers tails or cartilage showing chondrocyte proliferation are found

in combination with CPPD or apatite, and when seen together are absolutely

diagnostic of osteoarthritis.

Occasionally, fragments of cartilage are introduced into the joint where they

start to proliferate, the SF acting like a tissue culture medium. This

phenomenon is most often seen in the younger patient with chondromalacia

patellae, particularly in a patient with a sporting background. The fragments of

cartilage proliferate, forming cartilaginous loose bodies; any microscopic

cartilaginous loose bodies identified should indicate an arthroscopy as larger

cartilage nodules are often present.

Particles of ligament are found most often in cases of rotational trauma to the

knee joint where they are associated with a low cell count and a

haemarthrosis. In an inflammatory disease such as rheumatoid disease, they

are an indicator of non-traumatic cruciate ligament damage, which will result

in joint instability.

Synovial villi are most often introduced into the joint during the aspiration of

the joint fluid and can be considered an “artefactual biopsy”. Only when the

12

synovial villus changes its external morphology can its presence indicate a

specific disease. In osteoarthritis a characteristic fern or leaf like formation of

the synovial villus is found Figure 7.

SF analysis is increasingly being requested by orthopaedic surgeons in the

assessment of patients with a failing prosthetic joint replacement where

careful microscopic analysis can identify individual prosthetic components (eg.

Plastic 8, metal 9, cement etc.)

Ragocytes

The thin “wet preparation” is used to assess the presence of ragocytes which

were originally described as refractive intracytoplasmic spheres, which vary

from black to green, depending on the focal plane of the microscope.10

Pseudo-phase, where the condenser iris of the microscope is at its minimal

aperture, gives the best microscope setting for the detection of ragocytes

Figure 9. Ragocytes were originally thought of as a marker of rheumatoid

disease, but due to improvements in therapy they are now not often seen in

this condition. Ragocytes are counted and their number expressed as a

percentage of nucleated cells seen in the “wet preparation”. If this percentage

is between 70% and 90%, the diagnosis is most probably rheumatoid disease,

but when greater than 90%, septic arthritis is the most likely diagnosis.

Monolayer preparation

In order to identify and perform a differential cell count it is necessary to stain

the cells. Traditionally this has been done by making a smear, as would be

done routinely in haematology laboratories. Unfortunately SF contains much

more protein and hyluronans than blood, and if a smear is prepared, the

background staining obscures cellular detail. To overcome this, a monolayer of

cells is produced by cytocentrifugation. The recommended method is to dilute

13

the SF to a cell count of 400 cells/mm3 with sterile saline. 100µl of the

suspension is then loaded into the cytocentrifuge chamber with the filter

paper and slide, and spun at 800 rpm for 30 min. This produces a monolayer of

cells on the slide. Following air drying the specimen is methanol fixed for a

minimum of 5 min. The dilution of the fluid serves two purposes, firstly a

standard number of cells are available for microscopic examination and

secondly the hyluronans that would stain and obscure the cells are removed

giving a much clearer picture with little or no background staining. Following

fixation the cell monolayer is stained by the Jenner Giemsa technique, but any

cytological stain can be used, depending on personal preference or when a

specific objective such as the diagnosis of sepsis is required, where a Gram

stain would be necessary.

Septic arthritis

Sepsis in a joint can be a life threatening situation, either because the bacteria

are disseminating into the circulation from an infected joint following joint

penetration (accidental or surgical), or the joint itself is infected by the

haematological route. Either way both can be associated with potentially fatal

septicaemia.

Careful microscopic examination of SF cytocentrifuge preparations allows

micro-organisms to be identified in approximately 87% of instances of clinical

infective arthritis. Most cases of infective arthritis are caused by Gram positive

bacteria and organisms making them detectable in Gram stained preparations.

Unfortunately, significant numbers of cases are partially treated with

antibiotics before aspiration and rendering previously Gram positive organisms

Gram negative. It is therefore essential in this clinical setting to examine the

normally stained slide for evidence of organisms. Figure 10.

14

Sometimes no organisms can be identified, but the presence of a very high cell

count (>30,000 cells/mm3), particularly if crystals and reactive arthritis (see

below) have been ruled out, or haematoidin crystals are present should raise

the real possibility of a septic arthritis.

Another feature we have found useful for raising a high suspicion of septic

arthritis is the presence of “galaxy cartilage”, an appearance of cartilage

following partial breakdown by bacterial proteolytic enzymes. Figure 11. Most

bacterial infections result in a raised neutrophil response. However

mycobacterial infections may be associated with lymphocyte-rich fluids.

Periprosthetic infections are a special type of infective arthritis associated with

a joint replacement. Here, a florid cell response is never seen. However, there

is now a body of work that says if the SF cell count exceeds 1700/mm3 and

more than 60 % of the nucleated cells are polymorphs that patient has a

periprosthetic infection 11. When first made this observation changed joint

revision surgery as it meant the diagnosis of periprosthetic infection can be

made before, rather than during, the operation.

Care must also be taken in interpreting SF samples from patients with pre-

existing inflammatory arthropathies, particularly rheumatoid disease. These

patients are both at a greater risk of developing a superimposed infective

arthritis and of having a disease that may mask the presence of infection.

Cells

For reasons that are not clearly understood, primary and metastatic neoplastic

disease is exceptionally rare in joints. Occasionally, leukaemic cells may be

found in SF, but there are only a handful of cases of other neoplastic processes

involving joints. Malignant cells are therefore so rare as to be disregarded in

everyday practice.

15

The cells most commonly encountered in synovial flid are either cells derived

from connective tissues (chondrocytes, fibroblasts) or cells one would usually

associate with inflammation (neutrophils, macrophages, lymphocytes).

Patterns of “inflammatory” cell types and/or the presence of certain types of

cytoplasmic inclusion within the cells can be used to identify individual

diseases or more commonly a disease group.

The cells that are found in SF are therefore a reflection of the two major

groups of joint diseases, namely the inflammatory arthropathies such as septic

arthritis, gout, seronegative spondylarthropathies and rheumatoid disease,

and the non-inflammatory arthropathies resulting from trauma or due to

osteoarthritis.

For most of the non-inflammatory arthropathies a differential count of the

cells within the fluid is unhelpful and unnecessary and is therefore not

routinely performed. In inflammatory arthropathies, by contrast, identifying

certain types of inflammatory cell or the proportion of the three most common

cells, polymorphs, macrophages and lymphocytes, is the mainstay of

distinguishing different types of inflammatory arthritis.

In very general terms, in inflammatory arthropathies polymorphs dominate the

cytological picture, with other cells such a lymphocytes and macrophages

occurring in varying proportions in various arthropathies. This is often

completely different to the proportions of these cell types within the

synovium. Although making up the overwhelming majority of the cells within

diseased joints, these 3 cell types (polymorphs, lymphocytes and

macrophages) represent only a small proportion of the cell types that can be

identified regularly within diseased joints. Most of these cell types are

identifiable on morphological grounds in conventional Jenner Giemsa stained

16

cytocentrifuge preparations. The following are the most commonly identified

cells in SF.

Neutrophil polymorphs: these cells are recognized by their characteristic

nuclear morphology and eosinophilic cytoplasm in Jenner Giemsa stained

cytocentrifuge preparations. They are the predominant cells in inflammatory

arthropathies and in intra-articular haemorrhage, the former as a consequence

of specific traffic into the SF and the latter because they are the most abundant

nucleated cell in blood. In septic arthritis and acute crystal arthritis they

frequently amount to more than 95% of the total. In the absence of crystals,

finding >95% of the nucleated cells as polymorphs and a total nucleated cell

count >30,000/mm3 is practically diagnostic of septic arthritis, even when

organisms cannot be identified.

Lymphocytes:

This cell can be of the typical small type, although biological factors in SF can

frequently transform the lymphocyte into a larger activated forms or even cells

resembling lymphoblasts. In the circulation these changes may be interpreted

as leukaemic in origin but within SF they are “normal” in certain types of

inflammatory arthropathy and do not represent a malignancy.

Typically small lymphocytes are described as up to 12µm in diameter with a

nuclear/cytoplasmic ratio greater than 9:1. They predominate in approximately

10% of all cases in inflammatory arthritis, and in rheumatoid disease they

indicate a better long-term prognosis for the joint than when neutrophils

predominate. When seen in the company of LE cells (see below) they strongly

suggest the diagnosis of systemic lupus erythematosus. Lymphocytes up to

30µm with a nuclear/cytoplasmic ratio of about 1:1 indicate lymphocyte

transformation and activation. All lymphocytes, other than small lymphocytes ,

17

have a peripheral cytoplasmic blue colouration often containing vacuoles when

Jenner Giemsa stained.

Macrophages

These form one of the three morphologically distinct categories of large

mononuclear cells encountered in SF, the others being transformed

lymphocytes and synoviocytes. They are common in all types of arthritis and

are frequently the most common cell found in non-inflammatory

arthropathies.

Where the total nucleated cell count is high and the cellular pattern is

predominantly cells of the macrophage lineage, then a true viral arthritis, i.e.

one in which the virus is present in the joint should be suspected.

In non-inflammatory arthropathies, macrophages in combination with

eosinophils indicate a resolving haemarthrosis.

Synoviocytes:

It is perhaps unexpected, that the cells universally in contact with SF are one of

the least common cells found in SF. This is perhaps a reflection of the way

these cells adhere to the underlying synovial matrix.

Both types of synoviocyte are seen in SF. Type A synoviocytes, which are

functionally activated macrophages and type B synoviocytes which are

functionally connective tissue cells involved in the production of specialised

matrix molecules have distinct and distinctive morphologies.

Type A synoviocytes are large cells (>20µm) which generally have a vacuolated

cytoplasm and a small nucleus that occupies less that 20% of the cell. It is most

commonly seen in osteoarthritis.

18

The type B synoviocyte is smaller (approx. 20µm) and has a somewhat

basophilic stippled cytoplasm, with a faintly cytoplasmic frill. The nucleus of

the type B synoviocyte occupies 25%-50% of the cell. They are most commonly

seen in seronegative arthropathies where they generally form a small

percentage of the total number of cells.

Cytophagocytic mononuclear cells (CPM):

This is a functional cell type. It refers to a macrophage that has phagocytosed

apoptotic polymorphs Figure 12. The cells are normally seen whenever

apoptosis is occurring as phagocytosis by macrophages is the usual way

neutrophils are removed from the joint. They are, however, most abundant in

the seronegative spondylarthropathies and as such are seen on an

inflammatory background, either with polymorph or lymphocyte

predominance. The seronegative spondylarthropathies are a group of diseases

which includes the peripheral arthritis associated with psoriasis, inflammatory

bowel disease, Behcet’s disease, ankylosing spondylitis, many types of juvenile

idiopathic arthritis and reactive arthritis (an oligoarthropathy occurring in

association with extra- articular infection, notably of the gastrointestinal and

genitourinary tracts). If more than 1% of all large mononuclear cells are CPM, a

confident diagnosis of a seronegative spondylarthropathy can be made.12

CPMs are not found in SF from rheumatoid disease patients except when the

patient has an extra-articular infection, particularly of the lung or bowel.

Neither are they seen in septic arthritis, unless the patient has AIDS.

Mast cells:

Although mast cells can be found in most arthropathies they are seen most

commonly in inflammatory arthritis in patients with a seronegative

19

spondylarthropathy and in noninflammatory arthropathies associated with

trauma.

Cells in mitosis:

Neoplastic infiltration of joints is very rare. Mitotic figures are relatively

common by comparison and, no matter how bizarre they appear, are usually of

little diagnostic or prognostic significance.

Eosinophils

Eosinophils are most often seen following intra-articular haemorrhage or

arthrography, as well as in the allergic reaction to injected medications such as

artificial SF.

LE cells:

Phagocytes containing a cytoplasmic inclusion of nuclear material are not

uncommon and do not have the same significance in SF as they do in blood.

However, they should always raise the possibility of systemic lupus

erythematosus when found in a fluid rich in lymphocytes Figure 13.

Concluding remarks

By examining the type and number of cells, crystals and debris it is possible to

give either a specific single diagnosis (e.g. gout, osteoarthritis, rheumatoid

arthritis,) or where this is not possible, to give a more generalised conclusion

varying from relatively specific diagnoses such as “seronegative

spondylarthropathy” to the most general of diagnoses such as “inflammatory”

arthritis. Figures 14 and 15 below are simple algorithms that form the basis of

much of diagnostic SF microscopy.

SF microscopy is of greatest value in distinguishing inflammatory from non-

inflammatory arthropathies and in defining specific disorders within these two

20

groups. It is also important in the diagnosis of early inflammatory disease

where it might be possible, on the basis of cytology, to identify a specific

arthropathy before the clinical syndrome develops. In these cases accurate

early diagnosis often allows the implementation of specific therapies before

irreversible joint damage has occurred. The identification of non-cellular

particles particularly crystals, endogenous joint debris and prosthetic joint

debris has great significance, all of which will greatly influence clinical therapy

and surgical intervention. Finally, it permits the very rapid diagnosis of acute

joint disease, particularly in the difficult clinical differentiation of septic arthritis

and acute crystal arthropathies where a misdiagnosed septic arthritis has an

increased mortality and morbidity for the patient.

The simple observations described above are based on conventionally

illuminated and sometimes stained preparations examined microscopically.

These can simply represent an increase in the cytopathologists workload but

one of considerable importance to Rheumatologists, Orthopaedic Surgeons,

Accident and Emergency Physicians and General Practitioners. Of course the

greatest benefit is to the patient who receives an accurate and speedy

diagnosis of their joint problem.

21

REFERENCES1. Swan A, Amer H, Dieppe P. The value of synovial fluid assays in the

diagnosis of joint disease: a literature survey. Ann Rheum Dis 2002 Jun; 61: 493-8. Review.

2. Barton K, Ludwig T, Achari Y, Shrive N, Frank C, Schmidt T. Characterization of proteoglycan 4 and hyaluronan composition and lubrication function of ovine synovial fluid following knee surgery. J Orthop Res. 2013 Oct;31(10):1549-54.

3. Jones ST, Denton J, Holt PJ, Freemont AJ. Refrigeration preserves synovial fluid cytology. Ann Rheum Dis 1993 May; 52: 384.

4. Lumbreras B, Pascual E, Frasquet J, Gonz_alez-Salinas J, Rodrı´guez E, Hern_andez-Aguado I. Analysis for crystals in synovial fluid: training of the analysts results in high consistency. Ann Rheum Dis 2005 Apr; 64:612-5.

5. Heselden EL, Freemont AJ. Synovial Fluid Findings and Demographic Analysis of Patients With Coexistent Intra-articular Monosodium Urate and Calcium Pyrophosphate Crystals. J Clin Rheumatol. 2016;22:68-70.

6. Paul H, Reginato AJ, Schumacher HR. Alizarin red S staining as a screening test to detect calcium compounds in synovial fluid. Arthritis Rheum 1983 Feb; 26: 191-200.

7. Peterson C, Benjamin JB, Szivek JA, Anderson PL, Shriki J, Wong M. Polyethylene particle morphology in synovial fluid of failed knee arthroplasty. Clin Orthop Relat Res 1999 Feb; 359: 167-75.

8. Khan WS, Agarwal M, Malik AA, Cox AG, Denton J, Holt EM. Chromium, cobalt and titanium metallosis involving a Nottingham shoulder replacement. J Bone Jt Surg Br 2008 Apr; 90: 502-5.

9. Hollander JL, McCarty Jr DJ, Rawson AJ. The “R.A. cell”, “ragocyte”, or “inclusion body cell”. Bull Rheum Dis 1965 Sep; 16: 382-3. No abstract available.

10.Trampuz A, Hanssen AD, Osmon DR, Mandrekar J, Steckelberg JM, Patel R. Synovial fluid leukocyte count and differential for the diagnosis of prosthetic knee infection. Am J Med 2004 Oct 15; 117: 556-62.

22

11.Jones ST, Denton J, Holt PJ, Freemont AJ. Possible clearance of effete polymorphonuclear leucocytes from synovial fluid by cytophagocytic mononuclear cells: implications for pathogenesis and chronicity in inflammatory arthritis. Ann Rheum Dis 1993 Feb; 52: 121-6.

Box 1. Characteristics of the main crystal types

Mono sodium urate - highly birefringent, needle shaped, negative sign, water soluble

Calcium Pyrophosphate - low birefringence, rhomboid/rectangular shape, positive sign, water insoluble, alizarin positive

Apatite/bone - non-birefringent, amorphous granules or spherulites, alizarin positive

Lipid/fatty acid - intra-lipid rosettes or sheaves, birefringent, hydrocarbon soluble

Cholesterol - low birefringence, plates or crescents, hydrocarbon soluble

Haematoidin - orange, highly birefringent, rhomboid form intense orange, fern form green

CAPTIONS

Figure 1 Mixed crystal inflammatory arthropathy containing both calciumpyrophosphate and mono sodium urate crystals. In the centre of theimage are two parallel crystals, one showing a negative sign of birefringence,(yellow) and the other (blue) having the opposite, positive sign.The yellow urate crystal is needle shaped which contrasts with the stumpyrod shape of the calcium pyrophosphate.

Figure 2 Two large pieces of bone viewed in “pseudo phase “characteristicof bone fragments seen in osteoarthritis. Unstained.

Figure 3 Urate “beach ball” composed of a cluster of mono sodium uratecrystals that seem to radiate from a central point viewed in compensating,polarizing microscopy.

23

Figure 4 Intra-lipid droplet fatty acid crystal cluster mimicking a urate“beach ball” viewing conditions the same as Figure 3.

Figure 5 A cluster of brown haematoidin crystals. The insert in the bottomright if the image shows intra-cytoplasmic Gram-negative bacteria, thesewere only found after a prolonged search initiated by the finding of thehaematoidin crystal cluster. Gram stain.

Figure 6 A “tiger’s tail” of fibrillated cartilage showing the collagenfibrillation in full crossed polarizing filters. Unstained.

Figure 7 Synovial villus from a case of osteoarthritis with the leaf or fernlike configuration. Unstained.

Figure 8 Prosthetic derived UHMW polyethylene seen as finely divided,refractile shards with a piece of bone (far right) this combination isdiagnostic of a loosening and wear of the prosthesis. Unstained.

Figure 9 Intra-cytoplasmic ragocytes granules seen in “pseudo-phaseillumination. When the focus of the microscope is varied slightly theinclusions appear to change from black to green, whereas lipid dropletsjust go in and out of focus. Many inclusions, particularly if over 90% of thenucleated cells strongly suggest sepsis. Unstained.

Figure 10 Intra-cytoplasmic, blue pairs of cocci. Although the bodies arereadily identified as bacteria, a Gram stain would be necessary to gainfurther information as to their grouping. Jenner Giemsa stained.

Figure 11 Galaxy cartilage containing four chondrocytes at its centre. Thetypical effect of enzymatic degradation of the cartilage is seen wherethere is a separation of the cartilage into individual collagen fibrils givingthe “hairy” appearance of the surface. Unstained.

Figure 12 Cytophagocytic macrophages. The centre cell containinga freshly phagocytosed apoptotic neutrophil, other large macrophagesare in various stages of digesting apoptotic neutrophils. Jenner Giemsastained.

Figure 13 The lower of the central group of three cells containing a classical“LE” cell and to the upper left of the group another neutrophil butcontaining “small LE” inclusions. Jenner Giemsa stained.

24

25

26

Fig 1

Fig 7 Fig 8

Fig 6

Fig 4

Fig 2

Fig 5

Fig 3

27

Fig 9 Fig 10

Fig 11 Fig 12

Fig 13

28

29