Evaluation of activated protein C resistance in stored plasma

THE LANGET System the aPTT is further prolonged (eg, 78s) but the ratio

remains less than 2 (eg, 78/40= l-9), which is abnormal. If one instead takes into account a basal value of 32 s, this becomes 78/32=2-4, which is normal. Therefore, tests on stored plasma may be reliable only if the aPTT is identical to that for fresh plasrna, which is quite unlikely.

This Interpretation does not rule out the possibility of an acquired APC resistance that might have disappeared with time in certain patients. Therefore the proposed prevalence of APC resistance plasma Legnani and co-workers show (21%) in frozen plasma samples from patients with thrombosis seems imjustified. The actual frequency in their patients is probably only about 10%. This figure is arrived at by extrapolating the results for patients in whom a confirmatory test on fresh plasma could be done (12 positive of 24 examined).

Anton/o Girolami, Paolo Simioni, Luigi Scarano

University of Padua Medtcal School, Institute of Medical Semeiotics, IV Chair of Internal Medicme. 35100 Padua, Italy

l Dahlback B, Hildebrand B. Inhented resistance to activated protein C is corrected by anticoagulant cofactor activity found to be a property of factorV. ProcNatlAcadSci USA 1994; 91: 1396-400.

Authors' reply

SIR—Although Koster and colleagues' reported that APC response is not significantly affected by subsequent cycles of freezing and thawing and all published results on APC response have been obtained on frozen samples, aPTT results (which are at the basis of APC response measurement) are undoubtedly affected by freezing, äs Girolami and colleagues point out. Their comments give us an opportunity to clarify some aspects that, for the sake of brevity, were not discussed in our previous report on APC in patients with juvenile venous thrombosis (JVT).

First, all tests in our study—in patients (at first and second control), normal pool, and healthy subjects—were done on frozen samples. We considered the fact that the duration of storage might have been a possible cause of variability in results. The mean storage time of samples used for first screening was significantly longer than that of samples used for the second control (10-5 vs 3-1 months,

ρ<0·001). However, the storage time was not different in' samples of patients with confirmed or unconfirmed altered APC resistance (12-8 vs 8-2 months). Moreover, aPTT ratios (frozen samples vs frozen normal pool) did not differ in the first and second controls (mean 1-04 vs 1Ό5), nor did they in patients with confirmed or unconfirmed altered APC response. Thus, the observations of Girolami and co-workers though important and germane, do not seem to explain our inconsistent results.

We have no defmite explanation for the high variability recorded. The possibility of acquired and transient APC resistance cannot be ruled out. However, on the basis of some of our more recent work, it now seems that assay reproducibility is less consistent than we reported. By using more kits (APC Resistance, Chromogenix, Sweden) we have recorded an inter-batch co-efficient of Variation (CV) of 17%, which is much higher than the 8% inter-assay CV previously measured. Since handling of samples and Instrumentation used (ACL 300 Research) were carefully controlled, we surmise that differences in activity of APC. reagent or instability oi" aPTT reagents in the kits, or both, might substantially affect the reproducibility of results. We believe that the expression of results äs a ratio of the values

obtained in samples and in normal pools for the same testing Session (äs originally proposed by the manufacturer), should

be adopted to improve assay reproducibility, at least until the reagents are better standardised.

We have now increased the number of patients in our JVT series (n=261) and we can confirm our previous results. We agree with Girolami that the prevalence of APC resistance in these patients, on the basis of only the cases confirmed at re-sampling, seems to be around 10%. Since in other studies1"1 the prevalence of APC resistance varied from about 20% to 65%, we are surprised at the low prevalence in our patients and wonder whether methodological problems may be at least partly responsible for the wide Variation between studies.

Cnstina Legnani, Gualtiero Pa/areti, Roberte Biagi, Sergio Coccheri

Department of Angiology and Blood Coagulation, University Hospital S Orsole, 40138 Bologna, Italy

Koster T, Rosendaal FR, de Ronde H, Briet E, Vandenbroucke JP, Bertma RM. Venous thrombosis due to poor anticoagulant response to activated protem C' Leiden Thrombophilia Study. Lancel 1993; 342: 1503-06.

Gnffin JH, Evat B, Wideman C, Fernandez JA. Anticoagulant protem C pathway defecüve m majonty of thrombophilic patients. Blood 1993; 82: 1989-93.

Svensson PJ, Dahlback B. Resistance to activated protein C äs a basis for venous thrombosis. NEnglJMed 1994; 330: 517-22.

Halbmayer WM, Haushofer A, Schön R, Fischer M. Prevalence of poor anticoagulant response to activated protein C (APC resistance) among patients suffenng from stroke or venous thrombosis and among healthy subjects Blood Coagul FibnnolysK 1994; 5: 51-57

SIR—Our group reported (Koster and colleagues, Dec 18/25, p 1503) that a poor anticoagulant response to ACP is a common and strong risk factor for venous thrombosis. This ACP resistance becomes apparent by insufficient Prolongation of the clotting time (APTT) when ACP is added to patient plasma.1 We set a reference value for the ACP-sensitivity ratio (ACP-SR, ratio of APTT after addition of ACP to APTT in the absence of ACP) from results in 301 healthy individuals (mean-2 SD=2-17).

In 301 consecutive patients with a first, objectively confirmed episode of deep venous thrombosis, we found ACP resistance in 21% (64 cases). In 301 age-matched and sex-matched healthy controls, ACP resistance was present in 5% (14 cases): this yielded an odds ratio for thrombosis associated with ACP resistance of 66. Our cut-off points were strictly statistical, which leaves some uncertainty about the diagnosis in individuals. Moreover, Legnani and colleagues descnbe poor reproducibility of an abnormal ACP-SR. Our finding2 that ACP resistance is associated with a mutation in the factor V gene (FV-Leiden) at the site where ACP cleaves and inactivates clotting factor V, made it possible to re-examine the previously diagnosed cases directly by DNA analysis. Mutant and normal factor V alleles were discriminated with Mnl I digestion of an amplified factor V DNA fragment and visualisation of the cleavage products on ethidium bromide-stained agarose gels.' In controls with normal ACP-SRs—ie, greater than 2-17—we invariably find normal results with this technique.

, We present the results of the DNA analysis for all individuals (patients and controls) who had an abnormal ACP-SR (APTT ratio <2·17) in our initial analysis (this figure is the lower part of our original figure). Of 64 patients with ACP-SR less than 2-17, 53 carried the mutation. 6 of these were homozygous for the mutation. Among the 14 controls with an ACP-SR less than 2-17, 9 carried the mutation. The distribution of individuals with normal and abnormal results in DNA analysis, shows that the cut-off point should be set at an ACP-SR of about 1-90 rather than

THE LANGET l APC-SR 2.25 2 - 1.751.5 1.25 -1. OO OOOO OOO

OO

000

00

00

o

0

000

000

0

0

patients controls

Rgure: DNA analysis of APC resistance

Analysis of 64 patients and 14 healthy controls who were APC resistant (APC-SR <2·17). Open Symbols mdicate mdividuals not carrymg factor V Leiden mutation, semi-closed Symbols mdividuals who were heterozygous, and closed Symbols individuals who proved homozygous for this defect.

at 2-17, äs we suggested in our report because of the bimodal distribution of ACP-SRs, with a gap between Γ-8

and 2-0. It is noteworthy that there is no overlap of ACP-SR values of individuals with and without the factor V mutation. These results show that in all patients with ACP-SR less than 1-9, the diagnosis of ACP-resistance can be confirmed by DNA analysis. It also shows that, by contrast with Legnani's experience, in our laboratory the clotting test with the ratio of two APTTs performs perfectly.

F R Rosendaal, R M Bertina, P H Reitsma

Departements of Climcal Epidemiology and the Haemostasis and Thrombosis Research Centre, Umversity Hospital Leiden, PO Box 9600, 2300 Leiden, Netherlands

impairment of pain perception and proprioception in all four limbs, and he could hardly walk because of striking sensory ataxia.

Routine blood tests were normal. Cerebral spinal fluid analysis showed a slightly raised protein concentration of 0-57 g/L but no cells. GQlb-ganglioside antibodies were present with a titre of 200. Nerve conduction studies revealed evidence typical of a widespread acute inflammatory polyradiculoneuropathy with * unrecordable sensory nerve action potentials in the upper and lower limbs, delayed F waves, and evidence of motor conduction block.

We managed his bulbar palsy with intravenous fluids and soft diet and monitored his pulmonary function with spirometry and peak flow; he had physiotherapy for his mobility difficulties. He recovered rapidly and was discharged three weeks after the onset of Symptoms.

This case adds further evidence for the strong association between the Miller-Fisher syndrome and GQlb-ganglioside antibodies (Roberts and co-workers and refs l and 2). The wide spectrum of clinical features in the Miller-Fisher syndrome has been emphasised in a recent review,3 and

bulbar presentation is uncommon. Our patient did not complain of diplopia, and the diagnosis may easily be overlooked. Furthermore, most published cases have striking neurological impairments, and mean recovery time is over ten weeks.3 We suspect that milder cases with atypical

presentation and rapid recovery such äs in this man are

underreported.

Alexander Hammers, Richard J Hardie

Department of Neurology, Royal Devon and Exeter Hospital (Wonford), Exeter EX2 50W, UK

1 Chiba A, Kusonoki S, Shimuzu T, Kanazawa I. Serum IgG antibody to ganglioside GQlb is a possible marker of Miller Fisher syndrome. Ann Neurol 1992; 31: 677-79.

2 Wilhson HJ, Veitch J, Paterson G, Kennedy PGE. Miller Fisher syndrome is associated with serum antibodies to GQlb ganglioside. J Neurol Neurosurg Psychiatry 1993; 56: 204-06.

3 Berht P, Rakicky J. The Miller Fisher syndrome: review of the literature. JClm Neuro-Ophthalmol \<)<)2, 12: 57-63

Dahlback B, Carlsson M, Svensson PJ. Familial thrombophilia due to a previously unrecognised mechamsm charactenzed by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Pmc NatIAcadSa USA 1993; 90: 1004-08. Bertina RM, Koeleman B, Koster T, et al. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994; 369: 64-67.

Miller-Fisher syndrome with rapid recovery

SIR—Roberts and colleagues (Feb 19, p 454) report a strong association between GQlb-ganglioside antibodies and the Miller-Fisher syndrome in three patients. We describe another case.

A 65-year-old retired farmer presented with a one-week history of an altered nasal voice, increasing dysphagia, unsteadiness, and tingling in his hands and feet. He had been feeling generally unwell since an upper respiratory tract infection two weeks before admission. His medical, family, and social history were unremarkable. On examination, he was alert and fully oriented. The main findings were a mild ophthalmoparesis with diplopia on left lateral and vertical gaze, bilateral Horner's syndrome, absent voluntary palatal elevation with reduced gag refiex, impaired cough, and severely impaired swallowing. He had slight upper limb weakness, normal lower limb power, and complete areflexia with flexor plantar responses. There was symmetrical distal

Proliferating cell nuclear antigen äs a marker

of hepatocyte proliferation

SIR—Fang et al (April 2, p 820) report on the use of the monoclonal antibody PC 10 against proliferating cell nuclear antigen (PCNA) for the assessment of hepatocyte proliferation in acute alcoholic hepatitis. They imply a good correlation of PCNA staining with another proliferation marker, human histone 3. In 11 of 25 cases, PCNA expression was seen in 31-80% of hepatocytes. However, PCNA has limitations äs a marker for hepatocyte proliferation.

We have compared the expression of PCNA with the proliferation-associated Ki-67 antigen in liver biopsies containing metastatic tumour or showing hepatitis, including alcoholic hepatitis.1 Ki-67 is a reliable immunohistochemical marker of proliferating cells.2 We found that PCNA can be expressed by non-proliferating hepatocytes—indeed, it may be expressed in large numbers of hepatocytes in histologically normal liver. In alcoholic hepatitis, we found that there was generally high expression of PCNA in hepatocyte nuclei compared with low expression of Ki-67 antigen. Furthermore, PCNA is overexpressed by hepatocytes in post-transplant liver biopsies showing acute rejection (unpublished). We 'postulate that cytokines or other growth factors released from nearby inflammatory cells