609

SERIALT-CELL SUBSETS

(cells)

IN 3 PATIENTS

Spickett GP, Webster ADB, Farrant J, Chapel HM. Analysis of lymphocyte populations in common variable immunodeficiency by flow cytometry. In: Chapel HM, Levinsky RJ, Webster ADB, eds Progress in immune deficiency III. London: Royal Society of Medicine (Int Cong Symp Ser), 1991 132-33. 3. Lebranchu Y, Thibault G, Degenne D, Bardos P. Deficiency of CD4- CD45R +T cells in common variable immunodeficiency. N Engl J Med 1990; 323: 276. 4. Esolen LM, Fasano MB, Flynn J, et al. Pneumocystis carinii osteomyelitis in a patient with common variable immunodeficiency. N Engl J Med 1992; 326: 999-1001.

2.

**This letter has been shown follows.-ED. L.

’Month/year Monoclonals for CD4, CD8, and CD19 (B cells, not shown) were Leu3, Leu2, and Leul2, respectively, except for 1986 data for case 2 for which OKT4 and OKT8 were used

negative for diseases or syndromes associated with congenital immunodeficiencies. Case 1 illustrates a progressive panlymphocytopenia with a significantly inverted CD4/CD8 ratio. Case 2 has a panlymphocytopenia with a profound diminution in B cells, that has not appreciably changed over 6 years. Case 3 has a CD4 + T-lymphocytopenia that has remained stable over 2

are

years. Cases 1 and 2 are consistent with Centers for Disease Control criteria for idiopathic CD4 + T-lymphocytopenia4 and seem to be the first such to be reported among injecting

drug users. Supported in part by the New Jersey State Department of Health (grants 90-195, 91-105, and 92-808 AIDS), the Foundation of UMDNJ (grant 15-89), and NIH biomedical research support grant 2-507-RR05393. STANLEY H. WEISS Division of Infectious Diseases Epidemiology, CAROL WESTON KLEIN Department of Preventive Medicine, RAGHAVAN K. MAYUR UMDNJ-New Jersey JAMES BESRA Medical School, THOMAS N. DENNY Newark, New Jersey 07107, USA 1

Caussy D, Weiss SH, Blattner WA, et al. Exposure factors for HIV-1 infection among heterosexual drug abusers in New Jersey treatment programs. AIDS Res Hum

Retroviruses 1990; 6: 1459-67. 2. Weiss SH, French J, Klein CW, Mayur RK, Altman R. Mortality predictors in a 7 year IVDA cohort study: roles of RIV, HTLV-2, age, gender and entry symptoms (abstr 1555). VIII International Conference on AIDS; 1992 July; Amsterdam. 3 Palumbo PE, Weiss SH, McCreedy BJ, et al. Evaluation of HTLV infecting in a cohort of injection drug users. J Infect Dis 1992; 166: 896-99. 4 CDC Update CD4 + T-lymphopenia in persons without evident HIV infection— United States. MMWR 1992; 41: 578-79.

in Dr the basis of previously good health and a negative family history. The clinical importance of their report necessitates that primary immunodeficiency be excluded with the same vigour with which HIV infection was ruled out. Depletion of circulating CD4 + T lymphocytes may occur in several primary immunodeficiencies, mcluding purine nucleoside phosphorylase and adenosine deaminase deficiency and CVI,1-3 and CVI may present for the first time in middle age. Pulmonary and extrapulmonary Pneumocystis cannii infections have been reported in some of these patients,4 resulting in diagnostic confusion with AIDS if serum unmunoglobulin levels are not measured.

SiR,—The possibility of

a

primary immunodeficiency

Laurence and colleagues’ patients is dismissed

on

John Radcliffe

Department of Immunology, Hospital,

S. A. MISBAH

Oxford 0X3 9DU, UK

H. M. CHAPEL

1

Zegers BMJ, Stoop JW treatment

Metabolic causes of immune deficiency: mechanisms and ADB, ed. Immunodeficiency and disease. Lancaster:

In: Webster

MTP, 1989 113-31.

to

Dr

Laurence, whose reply

SIR,-We have measured serum immunoglobulins in the first four patients. Patients 1, 2, and 3 had normal concentrations of IgG, IgM, IgA, and IgE. IgG subclasses were studied in case 1; they were normal. Patient 4 had an oligoclonal IgM spike without reciprocal depression of any other isotype. Thus, while adult onset of a congenital CVImay certainly occur, and should be investigated as part of any evaluation of the putative new syndrome of idiopathic CD4 + T-lymphopenia, it is unlikely to be involved in our patients. We also believe it is important to document progressive decrements in absolute CD4 counts in such individuals. The rate of decline and extraordinarily low values noted in some of these patients would be an unlikely consequence of a congenital disorder. of Medicine, Division of Hematology-Oncology, New York Hospital-Cornell Medical Center, New York, NY 10021, USA

Department

JEFFREY LAURENCE

Hair loss with minoxidil withdrawal SiR,—The vasodilator minoxidil is used occasionally for severe hypertension. It has a well-recognised effect of stimulating hair growth in about 80% of patients. This effect is said to be reversible withdrawal. Hair loss below baseline after minoxidil withdrawal has been reported only in men and has been assumed to be due to a return to natural balding. We report a woman in whom discontinuation of oral minoxidil led to substantial hair loss below baseline. A 53-year-old woman with long-standing hypertension had been an outpatient at our hospital since March, 1983. She had been taking captopril 100 mg three times daily (the maximum recommended doses were higher then), frusemide 80 mg daily, and minoxidil 5 mg four times daily. She had received minoxidil since 1980 and had developed excessive hair growth on the face to the extent of having to shave every two to three days. She reported no abnormality of hair growth before starting minoxidil. Over the next three years blood pressure control was erratic. At her annual visit in April, 1986, she reported increasingly troublesome gastrointestinal upset that was attributed to minoxidil which was stopped. She then started losing her hair over the next two months. Hair loss started from the temples, and then fell from the top of her head to reveal scalp; she lost most of her eyebrows. The excessive facial hair cleared. The hair loss was enough to be socially unacceptable and she had to wear a wig. Minoxidil was, therefore, reintroduced about three months after withdrawal. Hair growth was apparent to the patient within a few days and appreciable growth had occurred by one month. She stopped wearing her wig after about two months. Facial hair requiring shaving reappeared as did the gastrointestinal symptoms, although they were less troublesome. The data sheet on minoxidil states "spontaneous reversal to the pretreatment state can be expected one to three months after cessation of therapy". Olsen et aF found that 4 of 10 men had non-vellus hair counts while off topical minoxidil that fell below baseline, inferring, perhaps indicating, a return to the natural balding that was interrupted by minoxidil. Bamford3reported a case of telogen effluvium after discontinuation of oral minoxidil in a 57-year-old man. Again the speculation was that this was a return to normal progression of male pattern baldness that had been slowed down by minoxidil. However, our female had no previous abnormality of hair growth and no systemic or dermatological illness or endocrine dysfunction to explain the hair loss, which on

610

suggests that this effect is not sex dependent and cannot be attributed to a simple return to natural balding. Our observation suggests a direct effect of minoxidil on the hair follicle, perhaps in some way sensitising it and making it dependent on the drug for future growth. B. J. KIDWAI North Devon District Hospital, M. GEORGE Barnstaple EX31 4JB, UK 1. Kosman ME. Evaluation of a

new antihypertensive agent, minoxidil. JAMA 1980; 244: 73-75. 2. Olsen EA, Weiner MS. Topical minoxodil in male pattern baldness: effects of discontinuation of treatment. Am Acad Dermatol 1987; 17: 97-101. J 3. Bamford JT. A falling out following minoxidil: telogen effluvium. J Am Acad Dermatol 1987; 16: 144-45.

Britton and

Gresty suggest that the sharp low frequency peak

may result from an increase in the output of a mechanism that contributes to physiological tremor. In an analysis of the tremor of 127 healthy subjects the averaged spectrum resembled that of a resonant system with broad-band forcing.2 There was no evidence

of a specific input at a low frequency. However, as we noted in our report, a few apparently healthy subjects do have a substantial low-frequency peak in their tremor spectrum. We do not know whether this is a symptom of inchoate disease or a physiological (but in our experience rare) oscillation. Applied Physiology Research Unit, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

MARTIN LAKIE

CD, Meadows JC, Lange GW, Watson RS. The relation between physiological tremor of the two hands in healthy people. Electroenceph Clin Neurophysiol 1969; 27: 179-85. 2. Arblaster LA, Lakie M, Walsh EG. Human physiological tremor: a bilateral study. J Physiol 1990; 429: 123P. 1. Marsden

Mechanisms for essential tremor SIR,-Dr Lakie and colleagues (July 25, p 206) report postural hand tremor before and after left-sided thalamotomy in a patient with bilateral essential tremor. Postoperatively the spectrum of tremor in the right hand had a broad range of frequencies (2-14 Hz) characteristic of physiological tremor, together with a distinctive peak at 5-7 Hz that they claim to be a residuum of the pathological tremor. However, the remaining symptomatic tremor of the left hand, recorded simultaneously, was also at 57 Hz, suggesting that the peak in the right hand spectrum was due to passive mechanical transmission of tremor from the left. This possibility could have been assessed statistically by calculating the coherence between the right and left hand spectra and tested by restraining the left arm

during recording. Lakie et al interpret the distinct peak amidst a broad spread of frequencies as evidence that essential and physiological tremor have separate mechanisms. In fact many mechanisms are known to contribute to physiological tremor, including mechanical properties of the limb, the ballistocardiogram, and spinal and supraspinal neuromuscular mechanisms.1 These mechanisms combine to give low and high frequency spectral components that vary in relative magnitudes according to circumstances, such as constitutional factors, anxiety, and drugs. Physiological tremor may well receive a contribution from a mechanism whose oscillations become magnified in patients with essential tremor. MRC Human Movement and Balance Unit, Section of Neuro-Otology, Institute of Neurology, National Hospital for Neurology and Neurosurgery, London WC1 N 3BG, UK

T. C. BRITTON M. A. GRESTY

1. Marsden CD.

Origins of normal and pathological tremor. In: Findley LJ, Capildeo R, eds. Movement disorders: tremor. New York. Oxford University Press, 1984.

**This letter has been shown -ED.L.

to

Dr

Lakie, whose reply follows.

SiR,—Dr Britton and Dr Gresty raise two interesting points. We considered the possibility that the persistent low frequency rhythmic peak in the tremor spectrum of the right hand could have been due to transmission of energy from the tremulous side. In a series of measurements, both before and after operation, the low frequency peak ranged from 5-1Hz to 6Hz. The peak frequency was similar on both sides, but was not in general identical, both before and after surgery-for example, on one occasion after surgery there was a difference of 0-5 Hz between left and right sides. It seems unlikely that the oscillations have a common source and are phasically related. Furthermore, low frequency tremor of the right hand persisted when the left hand was at rest. Marsden et aP have noted that the size and frequency of physiological tremor mostly correspond quite well in both hands, and these variables usually alter in the same way on both sides with repeated measurements, although the tremor does not have a common origin. The situation is probably the same in essential tremor; in this patient, and in others with essential tremor, we have seen similar bilateral covariation of these indices.

Intrapartum fetal monitoring SiR,—Dr Westgate and colleagues’ findings (July 25,

p

194)

that fetal electrocardiographic waveform (FECG) monitoring reduces the proportion of deliveries for fetal distress. We are doing similar research but so far our results differ from those of Westgate et al: the number of fetal blood samples needed in labour was reduced ten-fold in our study if ST analysis was used to guide management. There was no reduction in the number of operative interventions for fetal distress, and there was an equal number of babies with metabolic acidosis at delivery in each group (pH < 7-2, base deficit> 10). There may be several reasons for this difference: firstly, although our management guidelines for labour in the trial are broadly similar to those of Westgate et al, monitoring in the FECG group is based wholly on the ST waveform. Conventional cardiotocography (CTG) is not included since it has not been shown to be useful in successive randomised trials1 and the search for other indices of fetal wellbeing should not be hampered by the use of the CTG. Second, the senior obstetrician in our study withdrew 8% of labours from the FECG group. Third, we found that interference with the ECG signal in the second stage of labour from maternal muscles during pushing led to variation in the signal; this reduced the reliability and interpretability of the ST analysis. On average, only 64% of the second-stage traces printed had ST analysis, and a smaller percentage of the second-stage trace was accompanied by a normal check ECG assuring an acceptable signal for analysis. Prospective assessment is vital in such clinical decisions. Lastly, technical difficulties with monitors or signals were dealt with by an expert, dedicated team from the labour ward and the medical physics department at Plymouth. Without this immediately available expertise, such difficulties could hamper the use of FECG monitoring by clinical obstetricians. We, too, look forward to improvements in on-line computer analysis techniques, and are continuing our study. suggest

Department of Obstetrics and Gynaecology, St George’s Hospital Medical School,

J. E. COCKBURN J. M. PEARCE

London SW17 ORE, UK

G. V. P. C. CHAMBERLAIN

M, Kierse JNC, Chalmers I. A guide to effective care childbirth. Oxford: Oxford University Press, 1989 192-93

1. Enkin

m

pregnancy and

Chemoresistance of Plasmodium falciparum in central Africa appearance of drug-resistant strains of Plasmodium subsaharan Africa has made malaria control more difficult. Our organisation (OCEAC), with the government health services of the six member states, maintains epidemiological surveillance of chemoresistance. In-vivo studies of autochtonous

SiR,—The falciparuna in