summries two articles

THE LANCET

1886 Vol 349 • June 28, 1997

(seven partial or complete responses, four stable disease for more than 6 months), while 12 (52%) did not. The median time to progression for responding patients was 11 months (range 5–29) and for non-responding patients 3 months (range 0–6). The rate of response did not depend on the type of chemotherapy given. There was no difference between responders and non-responders in age, lymph node status, or recurrence-free period. Two of eight (25%) postmenopausal patients responded, compared with nine of 15 (60%) premenopausal patients. The incidence of MRP-positive tumours was not different for patients with soft tissue (one of three), bone (two of eight), or visceral metastases (five of twelve) as predominant site of relapse. In patients receiving first-line chemotherapy, MRP was more often positive in non-responding tumours (50%) than in responding tumours (18%). Only one of eight (13%) MRP-positive tumours had an objective response (5 months), compared with six of 15 (40%) MRP-negative tumours. Analysing for overall response, including stable disease, two of eight (25%) MRP- positive tumours responded, compared with nine of 15 (60%) MRP-negative tumours (odds ratio: 0·22; 95% CI 0·03–1·49). Patients with MRP-positive tumours showed a shorter time to progression on first-line chemotherapy than those with MRP-negative tumours (Cox proportional hazard model, p=0·006, figure). The relative hazard rate for time to progression in patients with MRP-positive tumours, compared with MRP-negative tumours, was 4·08 (95% CI 1·50–11·12). At 9 months, all eight patients with MRP- positive tumours showed disease progression, while seven of 15 of those with MRP-negative tumours did not (four objective responses, three stable disease). In Cox multivariate analysis for time to progression, MRP was the only significant variable in the model.

Of the 41 patients who received chemotherapy after one or more lines of hormonal therapy, 19 (46%) responded (seven partial responses, 12 stable disease), whereas 22 (54%) did not. In these patients, there was no significant difference in the rate or duration of response, or in the time to progression between patients with MRP-positive and MRP-negative tumours, suggesting differences in tumour cell biology. Metastatic breast cancer patients who receive chemotherapy as the first choice of treatment usually are premenopausal, are oestrogen-receptor and progesterone- receptor negative, and may have visceral metastases. These are all unfavourable prognostic factors. Women first treated with hormonal therapy are usually postmenopausal, have receptor positive tumours, and have bone rather than visceral metastases. We conclude that MRP expression is an

important predictor of poor prognosis in patients with breast cancer who were treated with chemotherapy as first-line systemic therapy for recurrence.

We thank Maxime Look for statistical analysis. This work was supported by the Dutch Cancer Society (Grants DDHK95-1051, DDHK96-1236).

1 Cole SPC, Bhardwaj G, Gerlach JH, et al. Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 1992; 258: 1650–54.

2 Flens MJ, Zaman GJR, van der Valk P, et al. Tissue distribution of the multidrug resistance protein. Am J Pathol 1996; 148: 1237–47.

3 Nooter K, Westerman AM, Flens MJ, et al. Expression of the multidrug resistance-associated protein (MRP) gene in human cancers. Clin Cancer Res 1995; 1: 1301–10.

Departments of Medical Oncology (K Nooter) and Clinical Pathology and Division of Endocrine Oncology, University Hospital Rotterdam and Rotterdam Cancer Institute, PO Box 2040, Rotterdam, Netherlands

100

80

60

40

20

0P ro

g re

ss io

n -f re

e s

u rv

iv a l (%

)

0 3 6 9 12 15 18 21 24 27 30

15 12 7 6 5 2 1 1 1 1 0

Time (months)

p=0·006

RHR (95% CI) 4·08 (1·50–11·12)

MRP- positive (n=8)

MRP-negative (n=15)

MRP-negative 8 3 1 0 0 0 0 0 0 0 0MRP-positive

Time to progression for patients treated with first-line chemotherapy for recurrence as a function of MRP status. Patients at risk at start and at every 3 months are indicated RHE=relative hazard rate.

Complement C3 and factor B cerebrospinal fluid concentrations in bacterial and aseptic meningitis Philip F Stahel, David Nadal, Hans-Walter Pfister, P Maria Paradisis, Scott R Barnum

Establishing a diagnosis is difficult in most cases of acute meningitis, since its clinical signs are non-specific, and laboratory examination of cerebrospinal fluid (CSF) often does not accurately differentiate between bacterial and aseptic meningitis.1,2 Therefore the identification of a discriminating parameter, which might lead to a rapid and accurate clinical test, would be of value in the differential diagnosis of acute meningitis.

Several studies have suggested that the complement system contributes to intrathecal inflammation in bacterial meningitis.3 In a retrospective analysis, we measured the concentrations of the alternative pathway complement components C3 and factor B in CSF samples obtained by lumbar puncture from 39 patients with clinically suspected acute infectious meningitis, and from 64 controls without central nervous system infection, with an ELISA developed in our laboratory. 18 patients (median age 40 years; range 14–67 years; 9 female) were diagnosed as having bacterial meningitis, on the basis of positive bacterial culture or on detection of bacterial antigen in CSF. The pathogens were Streptococcus pneumoniae (n=10), Haemophilus influenzae (3), Neisseria meningitidis (3), Listeria monocytogenes (1), and Streptococcus bovis (1). 21 patients were diagnosed as having aseptic meningitis (median age 8 years; range 2 months to 13 years; 6 female) on the basis of CSF pleocytosis with a predominance of mononuclear cells, negative bacterial and fungal CSF and blood cultures, negative results on CSF antigen detection tests for S pneumoniae, H influenzae, and N meningitidis, and full recovery without antibiotic treatment. No patient had received antibiotics or steroids before diagnostic lumbar puncture, and all CSF samples were obtained on admission. The mean C3 concentration in the CSF of patients with bacterial meningitis (48·32 [SD 50·74] µg/mL) was significantly higher than in aseptic meningitis (2·16 [1·82] µg/mL; p<0·001, Wilcoxon rank sum test) or in controls (2·49 [2·18] µg/mL; p<0·001). Similarly, factor B CSF concentrations were significantly raised in patients with bacterial meningitis (15·89 [17·36] µg/mL) compared with those with aseptic meningitis (0·25 [0·20] µg/mL; p<0·001) or controls (0·29 [0·26] µg/mL; p<0·001). C3 and factor B CSF concentrations in bacterial meningitis did not correlate

THE LANCET

Vol 349 • June 28, 1997 1887

Nonsense mutation of prostacyclin synthase gene in a family Tomohiro Nakayama, Masayoshi Soma, Dolkun Rahmutula, Yoichi Izumi, Katsuo Kanmatsuse

We found a nonsense mutation in exon 2 of the human prostacyclin-synthase gene in a family with essential hypertension and cerebral infarction. Prostacyclin (PGI2) is an inhibitor of platelet aggregation, smooth muscle cell proliferation, and vasoconstriction. Prostacyclin synthase (PGIS), which catalyses the formation of PGI2 from prostaglandin H2, is widely distributed, predominantly in vascular endothelial and smooth muscle cells. We have reported the organisation of this gene.1

We searched for possible point mutations in the exons using peripheral blood from 100 patients with essential hypertension by PCR and single strand conformation polymorphism (PCR-SSCP) analysis. One patient had an abnormally migrating band on exon 2. Sequencing of this exon showed a nonsense mutation in codon 26 (CGA/TGA). This nucleotide change makes Bst EII the restriction site. 300 people (150 with essential hypertension and 150 healthy controls) were screened by PCR and Bst EII digestion. The mutation was found in one patient with essential hypertension and in none of the controls. The patient was shown to be heterozygous for this mutation. This mutation of the stop codon is 76 bp downstream from ATG, the start codon in cDNA, thus a large part of mRNA,

Male Female Dead

HT CI

Hypertension Cerebral infarction Patient first detected

66y HT

69y HT

66y HT

61y HT

68y

Apoplexy HT

63y HT CI

57y HT CI

–

The 247-bp fragment digested with Bst EII will give fragments of 189 bps, 58bps

– + +– + +

247bp 189bp

58bp

Mutation

Family tree and PCR

with CSF total white blood cell counts or CSF protein concentrations (r<0·6, Spearman’s rank correlation coefficient).

We found that complement concentrations in the CSF may be of clinical value in distinguishing bacterial from aseptic meningitis. With cut-off levels of the mean value +2 SD for the aseptic meningitis population (5·8 µg/mL for C3 and 0·65 µg/mL for factor B; figure), C3 and factor B CSF concentrations were highly sensitive (both 100%) and highly specific (95·2% and 100%, respectively) tests for the diagnosis of bacterial meningitis, and associated with a negative predictive value of 100%, and a positive predictive value of 94·7% (C3) and 100% (factor B). Quantification of C3 and factor B CSF concentrations can be completed within 3–4 hours. We plan to test these preliminary results in a multicentre prospective study.

1 Lindquist L, Linné T, Hansson L-O, Kalin M, Axelsson G. Value of cerebrospinal fluid analysis in the differential diagnosis of meningitis:

1000

100

10

1

0·1

0·01

1000

100

10

1

0·1

0·01

C 3 (

µg / m

L) F a c to

r B

( µg

/ m

L)

Bacterial meningitis

Aseptic meningitis

Controls

Mean value

Cut-off

Complement C3 and factor B concentrations in the CSF of patients with infectious meningitis and controls Each point represents the mean of duplicate sample analysis. Cut-off level=mean of aseptic meningitis group+2 SD for differentiation between bacterial and aseptic meningitis.

a study in 710 patients with suspected central nervous system infection. Eur J Clin Microbiol Infect Dis 1988; 7: 374–80.

2 Rodewald LE, Woodin KA, Szilagyi PG, Arvan DA, Raubertas RF, Powell KR. Relevance of common tests of cerebrospinal fluid in screening for bacterial meningitis. J Pediatr 1991; 119: 363–69.

3 Stahel PF, Barnum SR. Bacterial meningitis: complement gene expression in the central nervous system. Immunopharmacology. (In press)

Department of Microbiology, University of Alabama at Birmingham, Birmingham, AL 35294, USA (S Barnum); Unit of Infectious Diseases, University Children’s Hospital, Zürich, Switzerland; and Department of Neurology, Klinikum Grosshadern,University of Munich, Munich, Germany

THE LANCET

1888 Vol 349 • June 28, 1997

Colonic perforation and serosal tears associated with colonoscopy Yoshiharu Uno, Takayuki Morita

Serosal tears with no mucosal damage are a complication of colonoscopy.1–4 These tears have been attributed to the pressure of the air introduced through the colonoscope or to stretching the wall of the colon.

A colonoscope (CF230I, Olympus Inc, Japan) was pressed against the mucosa of a piece of sigmoid colon removed during surgery for rectal cancer at a continuous pressure of 2–3 kg/cm2 (figure, a). First, the muscularis propria ruptured (figure, b), and then the serosa tore (figure, c), before the mucosa ruptured and the colonoscope penetrated the wall (figure, d). We repeated this experiment at 40 different points (ten points in each of four segments: sigmoid, descending, transverse, and ascending colon) of resected colons from eight patients with mean age 62 (SD 11) years. Perforation occurred in the above order, regardless of age, colonic segment, or sex.

We asked 60 doctors in our hospital how injury occurred when the tip of a colonoscope was pressed against the colonic wall. They all replied that perforation started from the mucosa, proceeded to the muscularis propria and then the serosa. Our results suggest that, when the tip or a bend of a colonoscope is pressed hard against the colonic mucosa, a seromuscular tear will probably occur, even when there is no mucosal injury. If the mucosa does not rupture, peritonitis is unlikely.

1 Livstone EM, Cohen GM, Troncale FJ, Touloukian RJ. Diastatic serosal lacerations. Gastroenterology 1974; 67: 1245–47.

2 Livstone EM, Kerstein MD. Serosal tears following colonoscopy. Arch Surg 1976; 111: 88.

3 Kozarek RA, Earnest DL, Silverstein ME, Smith RG. Air-pressure- induced colon injury during diagnostic colonoscopy. Gastroenterology 1980; 78: 7–14.

4 Ehrlich CP, Hall FM, Joff N. Postendoscopic perforation of normal colon in an area remote from instrumentation. Gastrointest Endosc 1984; 30: 190–91.

First Department of Internal Medicine and Second Department of Surgery, Hirosaki University School of Medicine, Aomori 036, Japan (Y Uno)

Perforation of colon by colonoscope a, b, c, d depict how serosal tears occur as complication of colonoscopy.

6 kbp,2 can not be translated. Consequently the activity of PGIS may be decreased.

The patient is a 57-year-old woman with essential hypertension who presented with a blood pressure of 177/113 mm Hg. Her electrocardiogram showed left ventricular hypertrophy. Although she had never smoked and rarely consumed alcohol, at the age of 50 she had a transient ischaemic attack. We looked for this mutation in her family. Her father died of stroke aged 70. Her mother was healthy until her death at age of 92. She had eight siblings including two elder brothers one of whom died in the war and the other of chronic renal failure. Three of the five living siblings had the mutation (figure); all were hypertensive. One with the mutation had had a cerebral infarction. As essential hypertension is thought to be a multifactorial disorder, PGIS may be one of the genes involved.

1 Nakayama T, Soma M, Izumi Y, Kanmatsuse K. Organization of the human prostacyclin synthase gene. Biochem Biophys Res Commun 1996; 221: 803–06.

2 Miyata A, Hara S,Yokoyama C, Inoue H, Ullrich V, Tanabe T. Molecular cloning and expression of human prostacyclin synthase. Biochem Biophys Res Commun 1994; 200: 1728–34.

Second Department of internal Medicine, Nihon University School of Medicine, Ooyaguchikamimachi 30–1, Tokyo 173, Japan (T Nakayama)