54 J Clin Pathol 1992;45:54-57 Plasma viscosity in inflammatory bowel disease

A J Lobo, S C Jones, L D Juby, A T R Axon J Clin Pathol: first published as 10.1136/jcp.45.1.54 on 1 January 1992. Downloaded from

Abstract The advantages of plasma viscosity are that, Aims: To assess the relation of plasma unlike the ESR, its value is independent of viscosity to disease activity in patients packed cell volume, sex, and for practical with inflammatory bowel disease. purposes, age.6 Viscosity is lower in neonates7 Methods: Crohn's disease (n = 60) and and rises slightly in the elderly due to changes ulcerative colitis (n = 71) were diag- in the plasma fibrinogen concentration.8 The nosed on the basis of typical histological use of automated systems reduces the bio- or radiological features. Active Crohn's hazard risks and produces reproducible, disease was defined as a Crohn's disease precise results and these can subjected to activity index of 150 or over. Active quality assurance. Plasma must be separated ulcerative colitis was defined as a liquid as soon as possible and never more than six stool passed three times a day or more hours after venesection, but once separated, with blood. Blood samples were assessed its viscosity may remain stable for several days for haemoglobin concentration, total if kept at 15-200C.2 white cell count, platelets, plasma vis- The initial capital cost of an automated cosity, erythrocyte sedimentation rate, viscometer is currently about £7500 with very serum albumin, and C-reactive protein. low running costs. Measurement of ESR Results: Plasma viscosity was higher in requires a very low initial outlay, but the those with active Crohn's disease com- subsequent running costs are higher. If the pared with those with inactive Crohn's annual running costs are said to include disease or active ulcerative colitis. laboratory disposables, appropriate venesec- Plasma viscosity correlated significantly tion containers, a maintenance contract for the with erythrocyte sedimentation rate, C- viscometer cost, interest repayments on the reactive protein, and platelet count in initial outlay for the viscometer and a sum to patients with Crohn's disease. In ulcer- contribute to its eventual replacement, the ative colitis plasma viscosity correlated approximate respective costs for performing only with serum C-reactive protein. 32 000 (the annual number performed at this

Plasma viscosity showed a low sensitivity institution in 1990) measurements of ESR and http://jcp.bmj.com/ for detecting active Crohn's disease, with plasma viscosity would be £10 848 and 48% of those with active disease having a £6464, respectively (P Day, personal plasma viscosity within the laboratory communication). The initial capital outlay reference range. would therefore be recouped within two years Conclusions: Plasma viscosity is related and more rapidly if plasma viscosity was to disease activity in Crohn's disease, but measured from the same blood container as

is insufficiently sensitive for it to replace the blood count. on October 1, 2021 by guest. Protected copyright. erythrocyte sedimentation rate as a There are few data on the use of plasma measure of the acute phase response in viscosity in clinical practice. In unselected Crohn's disease. patients attending a general medical and rheumatological clinic plasma viscosity was a more reliable measure of changes in acute phase reactants.4 In rheumatoid arthritis Plasma viscosity can be used as a non-specific plasma viscosity was said to be at least as measure of the acute phase response" and reliable as ESR and C-reactive protein (CRP) may have several advantages over the more in terms of diagnosis, and as an index of widely used erythrocyte sedimentation rate improvement.9 More recently however, (ESR).24 plasma viscosity was found to compare less The first clinical use of plasma viscosity well than ESR with a battery of other measurement was described in 1942, when its measures of inflammation in rheumatoid use in pulmonary tuberculosis was described,5 arthritis.'° and an association between red cell sedi- There are no previous reports of the use mentation and was of plasma viscosity in inflammatory bowel Gastroenterology velocity plasma viscosity Unit, The General noted. The widespread use of plasma viscosity disease. This paper reports the prospective Infirmary at Leeds, measurement, however, was delayed until an evaluation of plasma viscosity measurement in Leeds, LS1 3EX automated system for its measurement a large series of patients with inflammatory A J Lobo S C Jones became available. One such commercially bowel disease. L D Juby available system has been shown to be reliable A T R Axon and precise3 and detailed guidelines for the Correspondence to: assessment of the acute phase response, Methods Dr A J Lobo including the role of plasma viscosity, have One hundred and thirty one patients with Accepted for publication 4 July 1991 been published.2 inflammatory bowel disease attending the gas- Plasma viscosity in inflammatory bowel disease 55

trointestinal service of the General Infirmary, Results Leeds, were studied. Crohn's disease and By the criteria defined above, 21 patients had ulcerative colitis were diagnosed on the basis of active Crohn's disease, 39 had inactive disease; typical histological or radiological features. 24 had active ulcerative colitis (11 mild, 10 Sixty had Crohn's disease. This affected the moderate, three severe) and 47 had inactive J Clin Pathol: first published as 10.1136/jcp.45.1.54 on 1 January 1992. Downloaded from small bowel in 19 (32%), the large bowel in 22 disease. (37%), and both small and large bowel in 19 The plasma viscosity was significantly higher (32%). Seventy one had ulcerative colitis and in those with active Crohn's disease (median the extent was proctitis or proctosigmoiditis in 1-73 mPa.s; range 1 50-2 04) than in those with 39 (55%). Left sided disease (extending not inactive disease (median 1 62 mPa.s; range further than the splenic flexure) in 12 (17%), 1-5-1-92 (figure)). There was a large overlap, and subtotal or total colitis in 17 (24%). In however, between the two groups, and 10 of 21 three patients the precise extent was not known (48%) patients with active disease had values but had been documented to extend to at least within the reference range. In eight of these the splenic flexure. this normal plasma viscosity was associated with an increased CRP (median 77.5 mg/l; CLINICAL ASSESSMENT OF DISEASE ACTIVITY range 23-145), and in two the CRP was not Active Crohn's disease was defined as a Crohn's increased. disease activity index" (CDAI) of 150 or over. Plasma viscosity was also higher in patients The CDAI is a widely used clinical index in with active Crohn's disease compared with which the number of liquid stools per day, those with active ulcerative colitis (median 1 65 general well-being, and abdominal pain over mPa.s; range 15-1-92). There was a trend the previous week are scored and weighted. towards increasing plasma viscosity with The presence of the following clinical features increasing disease severity in ulcerative colitis also contribute to the score: iritis, aphthous but this was not significant (r = 0-22; p = 0 07). ulceration of the mouth, arthritis or arthralgia, In patients with ulcerative colitis plasma erythema nodosum, pyoderma gangrenosum, viscosity correlated with C-reactive protein fever, abdominal mass, perianal disease or (r = 0-25; p = 0-03), but not with any other fistula elsewhere. Use of lomotil or opiates, laboratory measurement. percentage below standard bodyweight, and In Crohn's disease plasma viscosity showed a haematocrit also contribute to the final score. significant correlation with the ESR (Spearman Active ulcerative colitis was defined as a r = 0 77, p = 0 00001), CDAI (r = 0.34, liquid stool passed three times a day or more p = 0009), CRP (r = 0 44, p = 0 001), platelet with blood. The severity of active ulcerative count (r = 0-43, p = 0-001), and correlated colitis was assessed by the criteria of Truelove negatively with albumin (r = -0 44, and Witts.'2 By these criteria patients with a p = 0.002). The inverse correlation with bowel frequency of four motions a day or less, haemoglobin did not reach significance = = and no more than small amounts of macro- (r -0-23,p 0-07). http://jcp.bmj.com/ scopic blood in the stool are considered to have For comparison, the ESR values according mild disease; six or more motions a day with to disease activity are shown alongside the macroscopic blood and one or more of: fever on plasma viscosity values in fig 1A and the CRP two days out of four, tachycardia (mean pulse values are shown in fig 1B. rate greater than 90 beats/min), anaemia The sensitivity and specificity of ESR and (haemoglobin concentration below 10-5 g/l) or plasma viscosity in distinguishing active from

ESR greater than 30 mm in one hour, are inactive Crohn's disease depends on the value on October 1, 2021 by guest. Protected copyright. considered to have severe disease. Moderate taken as a cut-off point. Examples are given in severity is attributed to attacks intermediate the table. between mild and severe.

LABORATORY ASSESSMENT OF DISEASE ACTIVITY Discussion All patients had blood drawn for haemoglobin Despite the claimed advantages of plasma vis- concentration, total white cell count, platelet cosity over ESR, our data highlight limitations count (Coulter counter; Coulter electronics, in the use of this test in inflammatory bowel Luton), plasma viscosity (Coulter viscometer), disease. The assessment of disease activity in ESR (Westergren method), C-reactive protein Crohn's disease is difficult, and a major clinical (CRP) (Boehring nephelometer; Hoechst UK, problem is to differentiate those with symptoms Hounslow), and serum albumin (sequential due to active disease from those with symptoms multiple autoanalyser with computer; Tech- due to other causes. Laboratory indices may be nicon, Basingstoke). The laboratory reference particularly helpful here, but no laboratory test range for plasma viscosity was 1-5-1-72 or clinical assessment has been shown to be millipascal seconds (mPa.s), and for C-reactive without limitations. No gold standard exists for protein, the lower limit ofdetection ofthe assay the assessment of disease activity in Crohn's is 5 mg/l and the reference range is 5-10 mg/l. disease. The most commonly used evaluation Differences between plasma viscosity in for clinical studies, including therapeutic trials, active and inactive disease and between is the CDAI, which is closely related to Crohn's disease and ulcerative colitis were physicians' overall assessment of disease assessed by the Mann-Whitney U-test. The severity."I correlation between plasma viscosity and the Forty eight percent of those with active CDAI, ulcerative colitis disease severity and Crohn's disease had plasma viscosity values laboratory variables was performed using the within the reference range. The presence of Spearman rank order correlation coefficient. symptoms not due to active disease may 56 Lobo, Jones, Juby, Axon

2.110 overlap between active Crohn's disease and 0 both inactive disease and the reference range makes a single normal value a poor predictor 2-0- .

for differentiating active from inactive disease. J Clin Pathol: first published as 10.1136/jcp.45.1.54 on 1 January 1992. Downloaded from 0 There is also an overlap of values for ESR 0 0 0 -;, 1.9 0 between active and inactive Crohn's disease. In 5, 0- cP 0 0 contrast, however, this is due to several E 0 0 increased values in patients without symptoms, > 1.8 0 CLCI * 0 "I 0 and in the clinical context does not pose a n lb 0 0 o 0 diagnostic problem. > 17 o° This can also be expressed in terms of the E 0 * 0 0 . ctp 0 sensitivity and specificity ofthe tests. These are, la oA OQOP * in turn, critically dependent on where the cut- 0 off lines are drawn. The upper limit of the we* 0 £00% reference range in our laboratory is 1-72 mPa.s. 1*5 0 0 O As This gives a specificity of 90% but a sensitivity 0 o o ofjust 52%. In comparison, using an ESR of20 as an upper limit, the sensitivity is much 1.4 SC nactive Active Inactive Active improved at 85%, but with a lower specificity of Crohn's disease Ulcerative colitis 71%. To obtain similar figures for the plasma viscosity would require the upper limit of the 2501 reference range to be dropped to 1 -65 mPa.s. At @ the levels currently used therefore, the ESR is more sensitive at detecting active Crohn's disease, but the plasma viscosity is more 200- specific. It has been pointed out that plasma viscosity 0 may be less useful in conditions associated with 0c anaemia and hypoalbuminaemia as both tend to c 150 0 00 increase the sedimentation rate.2 Crohn's .0 disease is associated with both, although the 4.1 hypoalbuminaemia may be partly due to the *4 100 0 acute phase response. The anaemia of Crohn's 0 disease may be multifactorial, but is partly (U 00 . related to disease activity, and the effect of an CL associated anaemia in increasing the sedimen- 50 tation rate may make the ESR more sensitive. It

is therefore of interest that in patients with http://jcp.bmj.com/ 8 Crohn's disease there was only a trend towards o0o e a correlation between haemoglobin concentra- I nactive Active I nactive Active tion and plasma viscosity. This presumably Crohn's disease Ulcerative colitis reflects a balance between the lack of effect of packed cell volume on viscosity and the com- Figure (A) Plasma viscosity (0) and erythrocyte sedimentation rate (ESR) (0) in mon effect of disease activity by haemoglobin active and inactive Crohn's disease and ulcerative colitis. (B) Serum C-reactive protein in active and inactive Crohn's disease and ulcerative colitis. via the plasma protein concentration). The on October 1, 2021 by guest. Protected copyright. ESR has also been noted to perform better than plasma viscosity in rheumatoid arthritis, where similar considerations obtain.'0 The higher values of plasma viscosity found Sensitivity and specificity ofplasma viscosity (PV) and in those with active Crohn's disease compared ESR in distinguishing activefrom inactive Crohn's with active ulcerative colitis again reflect the diseasefor different levels ofboth PV and ESR acknowledged difference in the extent of the acute phase response between the two condi- Sensitivity (00) Specificity (00) tions.314 Although there was a trend towards Plasma viscosity increasing plasma viscosity with increasing 1-65 mPa.s 90 67 1-72 mPa.s 52 90 severity in ulcerative colitis, this was not ESR significant, but may reflect the low number of 14 mm/h 90 54 which are now 20 mm/h 85 71 cases in the severe group, quite 42 mm/h 35 90 uncommon. Our data cannot support the replacement of ESR by plasma viscosity for the assessment of inflammatory bowel disease, especially Crohn's spuriously increase the CDAI, but eight of 10 disease. We have documented a considerable patients with a normal plasma viscosity and overlap in plasma viscosity between those with high CDAI, also had increased concentrations active and those with inactive disease. of circulating CRP suggesting that the symp- Although plasma viscosity may provide a toms causing a high CDAI were genuinely due greater specificity in the differentiation ofactive to active inflammation rather than the fibrotic from inactive Crohn's disease, ESR is more strictures that complicate Crohn's disease or sensitive, and in the clinical context this is more coexisting irritable bowel symptoms. This useful. Plasma viscosity in inflammatory bowel disease 57

1 Stuart J, Lewis SM. Monitoring the acute phase response 8 Meade TW, Chakrabati R, Haines AP, North WRS, Stirling (editorial). Br Med J 1988;297:1 143-4. Y. Characteristics infectingfibrinolytic activity and plasma 2 Intemational Committee for Standardization on fibrinogen concentrations. Br Med J 1979;1:153-6. Haematology (Expert Panel on Blood Rheology). 9 Pickup ME, Dixon JS, Hallett C, Bird HA, Wright V. Guidelines on the selection oflaboratory tests for monitor- Plasma viscosity-a new appraisal of its use as an index of ing the acute phase response. J Clin Pathol 1988;41: disease activity in rheumatoid arthritis. Ann Rheum Dis 1203-12. 1981;40:272-5. J Clin Pathol: first published as 10.1136/jcp.45.1.54 on 1 January 1992. Downloaded from 3 Cooke BM, Stuart J. Automated measurement of plasma 10 Bull BS, Westengard JC, Farr M, Bacon PA, Meyer PJ, viscosity by capillary viscometer. J Clin Pathol 1988; Stuart J. Efficacy of tests used to monitor rheumatoid 41:1213-16. arthritis. Lancet 1989;ii:965-7. 4 Hutchinson RM, Eastham RD. Comparison of the 11 Best WR, Becktel JM, Singleton JW, Kern F. Development erythrocyte sedimentation rate and plasma viscosity in of a Crohn's disease activity index. National co-operative detecting changes in plasma proteins. J Clin Pathol Crohn's disease study. Gastroenterology 1976;70:439-44. 1977;30:345-9. 12 Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final 5 Miller AK, Whittington RB. Plasma viscosity in pulmonary report on a therapeutic trial. Br Med J 1955;2:1041-8. tuberculosis. Lancet 1942;ii:5 10-11. 13 Fagan EA, Dyck RF, Maton PN, Hodgson HJF, Chadwick 6 Jung F, Roggenkamp HG, Ringelstein EB, et al. Effect of sex VS, Pepys MB. Serum levels of C-reactive protein in age, body weight, and smoking on plasma viscosity. Klin Crohn's disease and ulcerative colitis. Eur J Clin Invest Wochenschr 1986;64:1076-81. 1982;12:351-60. 7 Linderkamp 0. Blood rheology in the newborn infant. Clin 14 Saverymuttu SH, Hodgson HJF, Chadwick VS, Pepys MB. Haematol 1987;1:801-25. Differing acute phase responses in Crohn's disease and ulcerative colitis. Gut 1986;27:809-13. http://jcp.bmj.com/ on October 1, 2021 by guest. Protected copyright.