Joint hypermobility leading to osteoarthrosis and chondrocalcinosis

[Guest guest]

http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pmcentrez & artid=1000208

Abstract

We have reviewed 21 adults referred to rheumatology clinic and considered to

have generalised joint hypermobility by the criteria of and Wilkinson

(1964), modified by Beighton et al. (1973). They fell into two categories. 5

patients had a raised plasma viscosity (PV) and in each case a definite

pathology was found to account for this, superimposed on hypermobile joints.

The remaining 16 had a normal PV and this group was thought to represent the

late natural history of hypermobility. 5 of these (aged 32 to 54 years) had

no evidence of osteoarthrosis but the remaining 11 (aged 34 to 80 years) had

widespread radiological osteoarthrosis. Synovial histology was obtained at

arthroscopy in 6 of these patients and 4 (aged 60 to 75) had

chondrocalcinosis. This previously undescribed finding may be the end result

of hypermobile joints. Hypermobile patients with joint deformity (lax

connective tissue), widespread synovial thickening (traumatic), and hot

joint effusions (chondrocalcinosis) may mimic rheumatoid arthritis. They

must be distinguished from patients who develop rheumatoid arthritis in

hypermobile joints.

Full text:

ls ofthe Rheumatic Diseases, 1978, 37, 203-211

Joint hypermobility leading to

osteoarthrosis and chondrocalcinosis

H. A. BIRD, C. R. TRIBE, AND P. A. BACON

From the Royal National Hospitalfor Rheumatic Diseases, Bath, and the

Department ofPathology, Southmead

Hospital, Bristol

SUMMARY We have reviewed 21 adults referred to a rheumatology clinic and

considered to have

generalised joint hypermobility by the criteria of and Wilkinson (1964),

modified by Beighton

et al. (1973). They fell into two categories. 5 patients had a raised plasma

viscosity (PV) and in

each case a definite pathology was found to account for this, superimposed on

hypermobile joints.

The remaining 16 had a normal PV and this group was thought to represent the

late natural history

of hypermobility. 5 of these (aged 32 to 54 years) had no evidence of

osteoarthrosis but the remaining

11 (aged 34 to 80 years) had widespread radiological osteoarthrosis.

Synovial histology was obtained

at arthroscopy in 6 of these patients and 4 (aged 60 to 75) had

chondrocalcinosis. This previously

undescribed finding may be the end result of hypermobile joints.

Hypermobile patients with joint deformity (lax connective tissue),

widespread synovial thickening

(traumatic), and hot joint effusions (chondrocalcinosis) may mimic rheumatoid

arthritis. They must

be distinguished from patients who develop rheumatoid arthritis in hypermobile

joints.

The hypermobility syndrome describes hypermobile

individuals with musculoskeletal complaints who

lack the stigmata of other hereditary connective

tissue disorders (Kirk et al., 1967). Wood (1971)

suggested that these individuals simply represent one

extreme in a normal variation of joint mobility

throughout the population. Hypermobility can cause

knee effusions (Sutro, 1947) and an association with

many orthopaedic complaints has been described

including dislocation of the hips ( and

Wilkinson, 1964). Although the early rheumatic

manifestations of hypermobility are well documented

there is less information on the late natural

history of these individuals. An association with

premature osteoarthrosis has been described

(Rowatt-Brown and Rose, 1966; Rotes Querol,

1971) and Kirk et al. found degenerative joint

disease in 5 adult females out of their 24 cases. The

only published report of synovial histology is a

single needle biopsy of synovium from an adolescent

with knee effusions that showed no serious inflammation

(Ansell, 1972). We therefore decided

to review 21 adult patients with generalised

Accepted for publication September 8, 1977

Correspondence to Dr H. A. Bird, Rheumatism Research

Unit, University of Leeds School of Medicine, 36 Clarendon

Road, Leeds LS2 9PJ.

hypermobility who have presented to this hospital

with rheumatic complaints over the past 2 years and,

where possible to obtain synovial histology.

Patients and methods

Twenty-one patients were considered to have

generalised joint hypermobility on first attendance.

All were recalled for further interview by one of us

(H.A.B.) and hypermobility was assessed by the

method of Beighton et al. (1973) modified from

and Wilkinson (1964). Patients were thus given a

numerical score of 0 to 9, one point being allocated

for the ability to perform each of the following

tests. (1) Passive dorsiflexion of the little fingers

beyond 900. (2) Passive opposition of the thumbs to

the flexor aspects of the forearms. (3) Hyperextension

of the elbows beyond 100. (4) Hyperextension of the

knees beyond 100. (5) Forward flexion of the trunk

with knees extended so that the palms of the hands

rested on the floor.

A full clinical and family history was taken in

each case and the details of patients are given in

Table 1. Full blood count, plasma viscosity, and

Rose-Waaler tests were performed on all patients

and other investigations done as indicated. X-rays

of the knees and of any painful joints were taken

on each patient.

203

204 Bird, Tribe, Bacon

Eight patients had synovitis with effusion in the hypermobility score of more

than 1. 7 age- and sexknee

joint. 7 of these agreed to arthroscopy which matched nonhypermobile

patients attending for

was done under local anaesthetic with a Storz diagnostic arthroscopy

for persistent

knee effusion

instrument and routine biopsies and photographs completed the control group.

were taken. In 3 patients additional surgical biopsy

material was available (Table 2). Results

An age- and sex-matched group of 21 nonhypermobile

patients also referred to a rheumatology HISTORY

clinic with musculoskeletal complaints were chosen All patients presented

with musculoskeletal sympat

random and investigated as above. None had a toms. Most gave a past history

of such complaints,

Table 1 Details of 21 patients

Case Age Sex Hypermobility score Dislocations Other pathology

Radiological Viscosity

Rose-Waaler Arthroscopy

no. (years) (figures in brackets evidence of

represent score in osteoarthrosis

youth)

1 34 F 6 Cervical spine 1.57 -ve No

2 32 F 4 - 1.56 -ve

3 61 F 6 Cervical spine 1-59 -ve

4 42 F 8 - 1.57 -ve

5 80 F 8 Knees 1.62 -ve

6 64 F 2 (4) Patella Knees 1-70 -ve

7 36 F 9 Patella - 1.64 -ve

8 51 F 7 - 1.63 -ve

9 43 F 4 Vasculitis - 1.63 -ve

10 70 F 4 Spine 1.58 -ve

11 43 F 6 Myxoedema Knees 1.61 -ve Yes

12 60 F 2 (6) Both hips Wrists/knees 1 *50 -ve

13 66 F 7 Thyrotoxicosis Cervical spine/

thumb base/

knees 1.69 -ve

14 61 F 5 Knees/spine 1.62 -ve

15 50 F 2 (7) Both hips Hips/spine 1.59 -ve

16 75 F 4 (6) Hips/knees 1.58 -ve

17 70 F 6 Psoriasis with RA - 228 + ve Yes

18 65 F 6 Rheumatoid

arthritis - 1-99 +ve No

19 44 M 4 Psoriatic

arthropathy - 178 -ve

20 54 F 9 Raised

immunoglobulins

- 2-20 +ve

21 48 F 4 Rheumatoid

arthritis Cervical spine

Thumb base 1-82 +ve

Table 2 Arthroscopy and histology findings

Case Age Viscosity Arthroscopy findings Histology*

no. (years)

11 43 1 *61 Thickened synovium, mild osteoarthrosis, no No biopsy taken

crystals

15 50 1 .59 Synovial proliferation; minimal osteoarthrosis, No abnormality

seen

no crystals

13 66 1.69 Osteoarthrosis, chondrocalcinosis, bleeding tendency Severe

chondrocalcinosis;

wrist synovectomy;

proliferation of lining membrane with marked iron

deposition; no inflammatory response

14 61 1-62 Osteoarthrosis, chondrocalcinosis Osteoarthrosis and

chondrocalcinosis

16 75 1 .58 Osteoarthrosis, chondrocalcinosis Changes suggestive of

chondrocalcinosis

12 60 1.50 Calcified capsule, severe osteoarthrosis, Osteoarthrosis and

chondrocalcinosis

chondrocalcinosis

10 70 1 58 Not done Shoulder synovectomy; osteoarthrosis and changes

suggestive of chondrocalcinosis

17 70 2.28 Vascular synovium, inflammatory arthritis, fibrin, Changes

consistent

with psoriatic arthritis

no crystals

*Findings in biopsies taken at knee anhroscopies unless stated otherwise.

Joint hypermobility leading to osteoarthrosis and chondrocalcinosis 205

including muscle cramps aggravated by moving

joints to extremes of their range, 'growing pains',

spontaneous joint effusions, ease of dislocation, and

in 3 patients a symptom complex resembling

Raynaud's phenomenon. There was a higher

incidence of easy bruising, poor healing of the skin,

and varicose veins than in the control group.

Patients also described symptoms attributable to

secondary osteoarthrosis and several of the older

patients described intermittent symptoms typical of

chondrocalcinosis especially in their most mobile

joints. All patients except one were female. The

symptoms increased with age and there was an

impression of rapid deterioration, in the fourth or

fifth decade. 3 patients had a history of spontaneous

tendon rupture. Half the patients considered themselves

double-jointed though most agreed that joint

laxity had decreased with age. The 5 patients with a

raised PV described the above features in addition

to symptoms attributable to their concomitant

disease.

FAMILY HISTORY

This was only accepted where the history of joint

hypermobility and osteoarthrosis could be confirmed

by hospital notes, practitioners' records, or examination

of relatives. Although most patients claimed to

have relatives with 'loose joints', or 'arthritis',

careful checking validated only 9 histories. One of

these families had hypermobility without osteoarthrosis

but in the other 8, osteoarthrosis was seen

in hypermobile patients. Two families had more than

one member with seropositive rheumatoid arthritis.

EXAMINATION

Although patients with a hypermobility score of less

than 4 were excluded from the study, these scores

(Table 1) fall with age and 3 patients with a score of

less than 4 were admitted to the study when a

higher score in youth could be accurately confirmed.

Additional diagnoses confirmed by the appropriate

investigations are shown in Table 1 along with sites

of dislocation and radiological distribution of

osteoarthrosis.

Synovial thickening, often bilateral and mainly in

the more hypermobile joints, was observed in most

patients. Occasionally there was abnormal skin

laxity and a tendency to bruising. Tendon imbalance

across lax peripheral joints caused swan neck deformity

in 2 patients in the absence of demonstrable

rheumatoid arthritis. 3 patients later reattended

with transient acute monarthritis resembling

crystal deposition disease at the thumb base.

None of these findings, all considered typical of

hypermobility, was seen in the control group.

PLASMA VISCOSITY

This was normal in 16 patients and abnormal in 5.

This test formed the basis for distinguishing those

patients with hypermobility arthritis alone from

patients with a further pathology. Investigation

showed the following diagnoses in patients with a

raised viscosity: 2 with classical rheumatoid arthritis

(ARA criteria), 1 rheumatoid arthritis with psoriasis,

1 psoriatic arthropathy, 1 suspected myeloma. A

further patient (Case 9) had a suspected connective

tissue disorder with vasculitis in spite of a normal

PV.

RADIOLOGY

Osteoarthrosis was found at the sites indicated in

Table 1. In general the older the patient the worse

the radiological appearances. In several patients

there was a progression over 20 years from mild

radiological appearances with osteosclerosis, loss of

joint cartilage, and osteophytes to a bizarre appearance

with extensive calcification more typical of a

Charcot joint (Figs. 1 and 2). No patient with a

normal PV had erosions or any radiological features

of rheumatoid arthritis. However, in patients with a

raised PV radiological changes of rheumatoid

arthritis were superimposed on the osteoarthrotic

pattern of their hypermobile joints.

Examination of x-rays in the nonhypermobile

control group showed only 2 patients with radiological

evidence of polyarticular osteoarthrosis

comparable in severity to that seen in the hypermobile

group. There was a clinical impression that

the onset of osteoarthrosis was earlier in hypermobile

patients but it was not possible to make an exact age

comparison in this retrospective study.

Synovial thickening, presumed traumatic, was

extensive in many hypermobile patients and Fig. 3

shows a synoviogram of the wrist in a patient where

rheumatoid arthritis could not be shown.

ARTHROSCOPY

Seven patients (Table 1) agreed to arthroscopy of the

knee. 6 were from the normal PV group and 1 from

the raised PV group. Biopsies were taken in 6 of

these patients and arthroscopy findings are compared

with the histological findings in Table 2 in which

details of further available biopsy material is also

given. Arthroscopic appearances in the 2 patients

with the least severe radiological change (Cases 11, 15)

were identical. There was thickened synovium with

no increased vascularity, no crystals, and no bleeding

tendency. There were minimal osteoarthrotic changes

and bone surfaces looked almost normal. 3 patients

with more severe clinical and radiological change

(Cases 13, 14, 16) had all the above changes but more

severe osteoarthrosis and crystals ofchondrocalcinosis

206 Bird, Tribe, Bacon

Fig. 1 Case 14, aged 61.

Arthroscopy and histology

subsequently confirmed

chondrocalcinosis. X-rays had

been abnormal for 5 years.

Fig. 2 Case 12, aged 60.

X-rays had shown progressive

deterioration over 15 years.

Arthroscopy confirmed

chondrocalcinosis.

Joint hypermobility leading to osteoarthrosis and chondrocalcinosis 207

Fig. 3 Case 12. Wrist synoviogram showing extent ofsynovitis in the absence of

demonstrable rheumatoid arthritis.

in each case (Figs. 4, 5, 6). These crystals had not been

apparent either on x-ray or on examination of

aspirated joint fluid. A sixth patient (Case 12) had

the most severe change and arthroscopy was

technically unsatisfactory because of a calcified

capsule and gross joint destruction though chondrocalcinosis

was seen. However, biopsy from a previous

arthroscopy was available.

Fig. 4

Figs. 4, 5, 6 Case 13. Arthroscopic appearances

include a proliferative but relatively avascular synovium

andfine avascular villi, both with crystals. Examination

under a polarising microscope of crystals removed under

direct vision confirmed chondrocalcinosis. Fig. 5

208 Bird, Tribe, Bacon

The single patient arthroscoped from the raised crystals were in keeping

with the clinical diagnosis

PV group (Case 17) had completely different of psoriatic arthritis appearing

in a hypermobile

synovial appearances. The vascular synovium with joint and histology

confirmed this.

broad based villi, fibrin formation, and absence of

PATHOLOGY

Synovial biopsies from the knee joints of 6 patients

were taken at arthroscopy. A wrist synovectomy

specimen was available from one of these (Case 13)

and a further patient (Case 10) had had a shoulder

joint synovectomy for symptoms resembling polymyalgia

rheumatica. In retrospect these may have

been attributed to hypermobility.

Three biopsies showed changes characteristic of

chondrocalcinosis. In one (Case 13) the changes were

very florid with the villi virtually replaced by rounded

masses ofcrystalline material which in areas provoked

a low grade foreign body giant cell reaction (Fig. 7).

The other two biopsies (Cases 12, 14) showed smaller

masses of crystalline material lying within, or just

below, the synovial lining (Fig. 8). A further biopsy

(Case 16) and the shoulder synovectomy specimen

(Case 10) showed suggestive but not diagnostic

changes of chondrocalcinosis. Most of the biopsies

showed some degree of lining membrane hyperplasia

up to three cells wide and in the wrist synovectomy

Fig. 6 specimen from Case 13 this was prominent with

: :~.:...............

....-. 3i--||...

..:

B.: ..:: .:^..:::.: * .:

..:

A .......

..9.-t .a

Fig. 7 Case 13. Photomicrograph from synovial biopsy of the knee. Note

proliferating synovial membrane with

masses of crystalline material © in the underlying tissues. H & E x 535.

Joint hypermobility leading to osteoarthrosis and chondrocalcinosis 209

Fig. 8 Case 12. Photomicrograph from synovial biopsy of the knee. Note two

villi showing synovial lining

proliferation up to three cells thick with a small mass of crystalline

material lying within the thickened membrane.

©. H & E X 1100.

marked iron deposition consistent with traumatic

synovitis. This may be attributed to a bleeding tendency

more marked at arthroscopy in this patient than

in any other. The remaining two biopsies showed no

evidence of crystal deposition but one (Case 17) had

changes consistent with the diagnosis of psoriatic

arthritis (Table 2).

Inflammatory changes were virtually absent from

all these specimens and the histological changes

mirrored the arthroscopy findings confirming the

presence of chondrocalcinosis in the 4 cases where it

was visualised. In 7 age- and sex-matched nonhypermobile

patients with a joint score of less than 1,

referred for diagnostic arthroscopy, no examples of

chondrocalcinosis were seen or found on histology.

Discussion

Hypermobility represents one extreme of a normal

variation in joint mobility throughout the population

(Wood, 1971; Beighton et al., 1973) and may prove

advantageous in some professions (Grahame and

, 1972). There is also some evidence of a

distinct autosomal dominant inheritance (Beighton

and Horan, 1970) though these patients may be

spared premature osteoarthrosis. Ethnic differences

have been noted (Schweitzer, 1970) and hypermobility

also varies with age, sex, and athletic

training. All this leads to difficulty in defining hypermobility

and the criteria generally used take no

account of the reduction in mobility with age and

the even greater reduction that can occur when

osteoarthrosis develops. We defined generalised

hypermobility as a score of 4 or more, but this is an

arbitrary level. We preferred a simple scoring

system in conjunction with a careful clinical history

to some of the more complex scoring systems that

are said to cover all these variations.

Superimposed on this range of joint mobility are

specific inherited disorders of connective tissue. We

have attempted to exclude such patients from this

study but the division is not always distinct. Thus

although the benign hypermobile variant of Ehlers-

Danlos syndrome is said to be distinguished from

210 Bird, Tribe, Bacon

joint hypermobility by abnormal healing of the skin,

we have found that this distinction is hard to make in

practice. It is possible our series includes some mild

examples of this syndrome and we have been

impressed by the frequency of varicose veins and easy

bruising in many of our patients. Whatever the

genetic inheritance, we suspect joint hypermobility

is part of a more generalised connective tissue disorder

that involves all parts of the body. An elastic

joint capsule can only be used if there is play in the

skin, blood vessels, and ligaments along with

adequate muscle relaxation.

Hypermobile patients are as susceptible as any to

the whole spectrum of rheumatic disease. We have

used the plasma viscosity in conjunction with a

careful clinical history to distinguish the patients

who developed a new disease late in life in their

hypermobile joints from those with a gradual

progressive history who might be considered to

display the late results of possessing hypermobile

joints.

It has not been possible to assess the prevalence

of osteoarthrosis and hypermobility in a small

study of this kind. Patients are already preselected

by their attendance at hospital and this may account

for the impression sometimes claimed that hypermobility

is more frequent in the higher social classes.

It is also hard to quantify osteoarthrosis. We have

simply reinforced the strong clinical impression that

hypermobile joints are associated with premature

polyarticular osteoarthrosis.

The association of chondrocalcinosis and hypermobility

has not previously been recorded. No

examples were seen in the matched nonhypermobile

arthroscopy group and in the last 50 arthroscopies

performed at this hospital chondrocalcinosis has

been diagnosed seven times, four of these in hypermobile

patients. We have been impressed with the

arthroscope as a method of demonstrating condrocalcinosis

when it was not detectable radiologically

and when joint aspiration had failed to show crystals.

Synovial biopsy under direct vision at arthroscopy

with immediate polarising microscopy may be the

method of choice for diagnosing chondrocalcinosis.

The initial symptomatology of hypermobility has

been described (Kirk et al., 1967). It is suggested that

the subsequent natural history of hypermobility

leads to traumatic synovitis and later to osteoarthrosis,

normally in the fourth or fifth decades.

Chondrocalcinosis appears about 10 years later and

the final stage in the progression almost resembles a

Charcot joint with gross deforming osteoarthrosis,

chondrocalcinosis, and a tough calcified synovium.

We believe this progression deserves better recognition

as 'the arthritis of hypermobility'.

Many of our late x-rays from hypermobile patients

show changes similar to those seen in the premature

osteoarthrosis of ochronosis (Schumacher and

Holdsworth, 1977). If hypermobility is accepted as a

hereditary variant in the structure ofconnective tissue

there is a common link between these disorders since

in both the joints are subjected to an abnormal biochemical

environment throughout life. McCarty

(1977) has suggested that chondrocalcinosis may be

the final end point of all such hereditary disorders

and our findings support this view. It is possible

that chondrocalcinosis would be found in all late

cases of hypermobility in which it was looked for and

further studies are required using the arthroscope as

a means of detecting previously unsuspected chondrocalcinosis.

It is even possible that chrondrocalcinosis

is the end point of all osteoarthrosis but

only individuals with abnormal joint biochemistry

and therefore premature osteoarthrosis see it within

their lifetime.

Rheumatoid arthritis may develop in hypermobile

patients and Ansell (1972) described such a case. 3

patients in our series of 21 fit this diagnosis. In one

(Case 17) inflammatory arthritis with abundant

fibrin was confirmed at arthroscopy and in another

there was clear radiological evidence of rheumatoid

arthritis and osteoarthrosis coexisting. Such patients

must be distinguished from those with rheumatoid

arthritis who may have acquired hypermobility by

regular exercise. Thus patients with fixed deformities

at the ankles and knees find it advantageous to develop

abnormally good hip flexion in order to put on shoes.

It remains uncertain whether this is purely an

acquired hypermobility or whether they have hereditary

hypermobility before the onset of disease.

We also suggest that the later consequences of

hypermobility can easily mimic rheumatoid arthritis.

Dorwart and Schumacher (1974) drew attention to

hypermobile hands that resembled rheumatoid

arthritis and several of our patients were initially

referred with this erroneous diagnosis. They had

bilateral synovial thickening, presumed traumatic,

tending to involve the most mobile joints. The

fingers could sometimes be placed into a swan neck

position because of ligamentous laxity and spontaneous

tendon rupture can occur in these patients in

the presence of a thickened synovium. Osteoarthrosis

causes joint pain and when chondrocalcinosis

appears the joints become hot as well as swollen.

Although rheumatoid arthritis may be suspected, it

can usually be excluded by a normal PV, absence of

radiological erosions, and the standard serological

tests for rheumatoid remaining negative.

We thank Dr J. A. Cosh and Dr A. St. J. Dixon for

permission to report their patients, and Miss E. F.

Lupton for secretarial help.

Joint hypermobility leading to osteoarthrosis and chondrocalcinosis 211

References

Ansell, B. M. (1972). Hypermobility ofjoints. Modern Trends

in Orthopaedics, 6, 419-425.

Beighton, P., and Horan, F. T. (1970). Dominant inheritance

of familial generalised articular hypermobility. Journal of

Bone and Joint Surgery, 52B, 145-147.

Beighton, P., Soloman, L., and Soskolne, C. L. (1973).

Articular mobility in an African population. ls of the

Rheumatic Diseases, 32, 413-418.

, C., and Wilkinson, J. (1964). Persistent joint laxity

and congenital dislocation of the hip. Journal of Bone and

Joint Surgery, 46B, 40-45.

Dorwart, B. B., and Schumacher, H. R. (1974). Hand

deformities resembling rheumatoid arthritis. Seminars in

Arthritis and Rheutmatism, 4, 53-71.

Grahame, R., and , J. M. (1972). Joint hypermobility

-asset or liability? ls of the Rheumatic Diseases, 31,

109-111.

Kirk, J. A., Ansell, B. M., and Bywaters, E. G. L. (1967).

The hypermobility syndrome. ls of the Rheumatic

Diseases, 26, 419-425.

McCarty, D. J. (1977). Calcium pyrophosphate dihydrate

crystal deposition diseases (pseudogout syndrome).

Clinics in Rheumatic Diseases, 3, 61-89.

Rot6s-Querol, J. (1971). El sindrome de la laxitud articular.

Revista de Informaci6n Medico-Terapeutica, 3, 67-74.

Rowatt-Brown, A., and Rose, B. S. (1966). Familial precocious

polyarticular osteoarthrosis of chondrodysplastic

type. New Zealand Medical Journal, 65, 449-453.

Schumacher, H. R., and Holdsworth, D. E. (1977). Ochronotic

arthropathy: (i) cunico-pathologic studies. Seminars in

Arthritis and Rheumatism, 6, 207-246.

Schweitzer, G. (1970). Laxity of metacarpophalangeal joints

of finger and interphalangeal joint of thumb: a comparative

interracial study. South African Medical Journal, 44,

246-252.

Sutro, C. J. (1947). Hypermobility of knees due to overlengthened

capsular and ligamentous tissues. Surgeryv 21,

67-76

Wood, P. H. N. (1971). Is hypermobility a discrete entity?

Proceedings of the RoYal Society of Medicine, 64, 690-692.

--

" Life- Like the flutter of wings... feel your hollow wings rushing... " (AFI-

Silver and Cold).

my Flight in life is a metamorphosis of growth and this flutter of wings is

within me awaiting to find a space to find it's flow...