Thalidomide neuropathy was the first serious side-effect of the drug to be detected and this side-effect alone would probably have led to the demise of the drug' (1), had not the teratogenic disaster followed. Yet in reintroducing the drug in 1965 for the management of the lepra reaction, Sheskin ignored this complication (2).

Despite initial concern by prominent leprologists that there may be a recurrence of the teratogenic disaster which had occurred only four years previously, with the drug being given inadvertently to a woman in the reproductive age-group, it soon became used widely, especially in developing countries, until it became the drug of choice for this complication of the disease. Initially it was regarded as a research drug. By restricting the use of the drug to males and post-menopausal women it would be possible to avoid the teratogenic effects. On the other hand it would be impossible to avoid the neuropathic side-effects which were observed in adults of both sexes when thalidomide was used as a sedative. It should be emphasized that everyone who takes the drug is at risk as there is no evidence that the neuropathy is due to hypersensitivity and is related to the dose and total amount of the drug taken(3). Leprosy patients have received 400 mg daily which is well over twice the daily dose when used as a sedative. Also, there is no minimum dosage below which it is safe to give the drug.

This primarily affects the axon rather than the myelin sheath and is initially sensory (3). Burning pain in the feet was a common feature together with cramping pain in the calves. Although all modalities of sensation could be affected the sensory loss was sometimes confined to pain and temperature sensation, leaving joint and position sense and vibration intaet, an important point when considering the differentiation of thalidomide neuropathy from leprous neuropathy.

Although primarily a peripheral dying-back neuropathy with the maximal involvement of the longest nerve fibres and with loss of ankle and knee reflexes, other patients showed evidence of pyramidal tract damage with increased reflexes and extensor-plantar responses (4). If thalidomide was continued then motor involvement occurred affecting the proximal muscles of the lower limbs.

Sural nerve biopsies in patients with neuropathy showed a loss of mainly larger myelinated fibres and teased fibres confirming that the changes were those of axonal degeneration, without segmental demyelination (5). Unmyelinated fibres did not appear to be affected. Nerve conduction studies, both motor and sensory, were normal but sensory action potentials were reduced.

In patients who stopped taking the drug the motor and pyramidal signs recovered but sensory loss was permanent, even years after stopping the drug in over half the patients (5,·6).

The frequency of this neuropathy was originally uncertain with estimates ranging from 0,5% to 5-20 %. Recently, more reliable assessments are available because the drug has been used in a variety of non-leprous disorders and the frequency has been found to be very high; at least 21% if electrophysiological studies, especially the recording of sural nerve action potentials, are carried out (7).

Leprologists continue to maintain that although the frequency of neuropathy in non-leprous disorders is over 20% not a single case of neuropathy has been detected in a leprosy patient given thalidomide, even though the drug has been in use for 25 years.

Unfortunately, the current classification (8) is based on the cutaneous manifestations and does not include any information about the presence or absence of nerve damage in the peripheral nervous system. A spectrum of disease is defined with at one end the presence of well defined skin lesions, very few bacteria and a cell-mediated response limiting the spread of the disease. At the other end of the spectrum is the lepromatous form where there are masses of bacteria, an inadequate cell-mediated response and widespread disease. Paradoxically, although there are masses of bacteria in this multibacillary form, nerve damage is slight and late in appearance. This point is not emphasized in the current classification and it is necessary to go back to Hansen's description of the disease(9) where he described this form as progressive due to uncontrolled bacillary proliferation, but with sensory loss only occurring in a few cases and late in the course of the disease. If it did occur, it was associated with loss of baeteria. Thus Hansen concluded: "The patients were cured that is to say they became anaesthetic" (9). The other form of the disease with few bacteria was called maculo-anaesthetic where the first signs were skin lesions, followed early by severe sensory loss in the extremities leading to mutilation, but again sensory loss was associated with disappearance of bacteria. As thalidomide is used to treat a complication which only occurs in the multibacillary or lepromatous form it is necessary to emphasize that sensory loss due to the disease may or may not be present. In my own series this symmetrical loss of sensation was present in four out of 14 patients (10)o who had the complication for which thalidomide is being used. It is therefore possible that the use of thalidomide may inflict intractable sensory loss on a leprosy patient who would not have developed it as a result of the disease. In any event, it is imperative that a detailed sensory examination of the peripheral nervous system must be carried out and recorded before starting treatment with thalidomide. This must include testing for all modalities of sensation and can be achieved by simple and purely clinical examination without any special equipment.

Any deterioration in function while thalidomide is being administered must be assumed to be due to the drug, the more so as we know from Hansen's observation that sensory loss in this form of the disease is insidious and if the treatment is successful, deterioration due to the disease should not occur.

The other problem is that the nature of the sensory loss in the disease and due to the drug may be remarkably similar in that both pain and temperature sensation may be selectively involved with preservation of deep sensation, as described in leprosy patients over 70 years ago (11). Of course if evidence occurs that deep modalities of sensation become impaired during Thalidomide administration then this is due to the drug and not the disease. There are other symptoms which are distinctive of thalidomide neuropathy but do not occur in leprosy such as burning pains in the legs and cramps in the calf. The fact that a drug causing a peripheral neuropathy was being used in a disorder where damage to the peripheral nerves was occurring might lead to difficulties in detection of toxic effects did not appear to occur to leprologists advocating its use.

Although many studies on the use of thalidomide in leprosy have been published, details of the findings in the peripheral nervous system either before or during drug treatment have not been recorded so that neuropathy cannot be excluded. Much has been made of electrophysiological studies which it is claimed have shown that neuropathy has been excluded. In particular it is claimed(12) that the comments by Sagher satisfactorily have shown no neuropathy (13). It has been pointed out that as thalidomide is an axonal neuropathy, a decrease in nerve conduction would not be expected (14). This point has been reiterated by Jones in response to Jakeman and Smith's editorial (15). Furthermore, as thalidomide neuropathy is an axonopathy, or a dying-back neuropathy affecting the longest axons, the earliest and most severe damage will occur in the legs so there is little point in carrying out measurements in the ulnar nerve at the elbow. Finally, motor nerve conduction studies are unlikely to be helpful since thalidomide neuropathy is primarily a sensory neuropathy. The most useful electrophysiological test is the recording of sensory nerve action potentials, as demonstrated by Wulff et al. where they found abnormalities in all nine patients given the drug for prurigo nodularis (14). It has been claimed that the presence of ankle reflexes means that thalidomide neuropathy has not occurred (16,17). It is true that absent reflexes has been listed as a sign but this only occurred in under half the cases (5). In one recent case of an autoimmune deficiency syndrome (AIDS) patient who developed a neuropathy after taking thalidomide the reflexes were preserved(18) and in cases where the pyramidal tracts are affected the reflexes were present and even increased. The absence of reflexes would not be regarded as a useful diagnostic sign according to the comprehensive analysis of thalidomide neuropathy in a review article (19). The most common symptom in thalidomide neuropathy, as emphasized above, was cramp-like pains in the calf present in 16 of 22 patients (5).

Another disadvantage of the current classification is that it does not provide any information about the clinical findings in the peripheral nervous system. As thalidomide is not a cure for leprosy, having no effect on Mycobacterium leprae, and is only used to treat a complication of lepromatous leprosy, it is necessary to consider the management of the uncomplicated disease. Once again, Hansen's approach has an advantage over the current classification in describing the complications, morbidity and causes of death in the untreated multibacillary form of the disease. Thus the life expectancy was only eight years and the cause of death was an affliction of the kidneys, almost certainly amyloidosis, due to uncontrolled bacillary proliferation and therefore similar to tuberculosis. The other complications were laryngeal obstruction and blindness with susceptibility to tuberculosis being a possible factor. ENL was hardly mentioned in Hansen's account and it was not until 1913 that this complication was described in detail and concluded that it was "generally benign"(20). It is interesting and significant that as late as 1958 "lepra reaction" or, ENL, was not regarded as a serious complication and had been regarded as a good sign. Deliberate induction of ENL by giving drugs such as potassium iodide has been practised (21).

From a study of the natural history of the disease it is clear that the object of treatment in lepromatous or multibacillary leprosy is to kill the bacteria, prevent amyloid occurring and so to increase life expectancy to that of the normal population. The prevention of the sensory loss which leads to mutilation is not so important as this is pot an invariable feature and tends to occur later in the course of the disease. In contrast, sensory loss is the major complication of the other, maculo-anaesthetic form of the disease as the term anaesthetic implies. Here killing of the bacteria is important at an early stage before nerve damage has occurred, otherwise it is not necessary because the body is able to do that on its own, even at the cost of the nerves being destroyed.

Successful treatment with dapsone monotherapy or, more recently, multidrug therapy has lead to an increase in life expectancy which is now the same as in the normal population(22) suggesting that chemotherapy has prevented amyloidosis, a finding similar tt that in the reatment of tuberculosis. The other complications of lepromatous leprosy such as blindness and laryngal obstruction have been prevented with chemotherapy.

Although ENL was originally regarded as a benign condition it has generally been regarded much more seriously since the introduction of chemotherapy. Undoubtedly, the frequency of ENL has increased markedly since the introduction of chemotherapy. Thus Waters considers that it is life-threatening (16), and hence the use of thalidomide is justifiable. Before the introduction of thalidomide the main treatment for ENL was corticosteroids. If ENL was considered to be life threatening, then it was necessary to suppress it completely necessitating the use of very high doses of corticosteroids. In fact Waters has used 160 mg of prednisone/day(23). At this stage serious side-effects would have occurred and thalidomide was a welcome alternative. What then is the evidence for ENL being a serious complication?

ENL only occurs in multibacillary or lepromatous leprosy and characteristically starts about six months after the commencement of chemotherapy when the percentage of viable M. leprae, the morphological index (M.I.) has fallen to zero, but where there are still masses of dead bacteria. Although the cause of this syndrome is still uncertain, it may be the failure to clear these dead bacteria which is an important contributory factor in the chronic nature of the syndrome. At this stage the nodules of lepromatous leprosy have disappeared and the skin thickening has regressed to such an extent that the patient will not be identified by the public as having leprosy, because the neurological complications such as mutilation and trophic ulcers only appear late in this form of the disease, as we have already seen from Hansen's observations on the natural history of the disease. Thus the patient can lead a normal life without fear of being ostracized. The complications which occurred before the introduction of chemotherapy have also been prevented. If the serious complications of lepromatous leprosy such as amyloidosis are due to the continued presence of viable bacteria then the primary aim of treatment is to keep the M.I. at zero with continuous chemotherapy for a sufficient period to prevent relapse.

ENL is readily recognized clinically and is exactly the same in appearance as erythema nodosum in other diseases except that it is generalized and is not confined to the legs. It is also chronic and recurrent and lasts for as long as chemotherapy is being given. Self-limiting episodes of a few months have not been personally observed, in contrast to the experience of other leprologists. The syndrome is sufficiently well-defined to justify the term ENL. Thalidomide has a specific action in suppressing ENL (24,25) and in reducing the associated temperature but is of questionable value in treating other complications arising during chemotherapy. In view of its teratogenic and neuropathic side-effects its use should have been strictly confined to patients with ENL. Unfortunately, in several studies describing its use (26 - 29), it is not clearly stated whether attacks of ENL did actually occur. Also, many patients have been designated merely as having "lepra reaction", where complications other than ENL have been included (2,30,31) and the term has even been used to describe reactions in non-lepromatous leprosy where ENL does not occur (23,32). Because ENL has not been strictly defined its reported frequency is found to vary widely between 3% (10) and 50 % of lepromatous cases (23,·33,·34). Another reason for the wide disparity is that the higher figure is based only on a concentrated group of institutional cases, whereas the low figure includes lepromatous cases treated as out-patients.

The lowest figure of 3 % lepromatous cases was calculated in an area where the prevalence of all types of leprosy, determined by population survey, was 0.2 % and this agreed very closely with the figure found by out-patient attendance and population as determined from the census (35). A similar frequency of ENL of 5% lepromatous cases was found in the coastal region of Tanzania (personal observation). A figure of 5.9 % lepromatous cases from Nepal supports these findings (36). Despite this evidence, ENL has been assumed to be much more common than it really is and hence a large number of patients have received thalidomide unnecessarily.

It has been claimed that recurrent bouts of ENL have led to amyloidosis (37,·38) justifying the need for thalidomide and also the need to suppress the attacks completely using high doses of corticosteroids. Lepromatous leprosy patients given dapsone monotherapy have been shown to have a normal life expectancy (22,39) which is the main argument against ENL being a life-threatening condition. This is because the mortality of treated lepromatous patients would not have fallen since ENL has greatly increased in frequency since the introduction of chemotherapy. The other complications of untreated lepromatous leprosy such as blindness and laryngeal obstruction have also been prevented by chemotherapy. The most intractable problem in management is a recurrent ulnar mononeuritis with pain and tenderness which can lead to permanent motor loss (10). However, this mononeuritis also occurs in the absence of ENL and thalidomide appears to have no effect in preventing it (33,40).

The other complications such as arthritis, iritis or orchitis were not seen in my patients and although proteinuria can occur as a result of amyloidosis this is a complication of untreated disease. Besides, if these complications were common and serious there should be sufficient patients to conduct a therapeutic trial, comparing for example, the effect of thalidomide with corticosteroids but no such trial has been carried out. Finally, support for the view that ENL

is not serious, at least in African patients, comes from the Medical Research Council (Bunje, personal communication). lt can thus be argued that a patient with ENL is well on the way to being cured and this complication represents a "nuisance factor" (41) which interferes with continuous chemotherapy.

Although neuropathy as a side-effect of thalidomide was recognized by German neurologists soon after the drug was introduced in 1958, the drug firm involved withheld information about this side-effect (42), and the first publication did not occur until December 1960 (43). A year later, in 1961, the drug was withdrawn following reports of its teratogenic effects which, however, did not occur in the USA where Kelsey (42) of the Food and Drug Administration refused to licence thalidomide because she was concerned that a drug with neuropathic side effects might damage the fetus. In 1962 Mellen and Katzenstein described how the tragedy had occurred with the recognition of the neuropathy preceding the teratogenic effects (19). Nevertheless, despite the wide publicity given to the fact that thalidomide could cause a severe neuropathy, this side-effect was not mentioned in Sheskin's initial publications (2,30), or indeed subsequently by Pearson (24), a leprologist working for the Medical Research Council (MRC) in Malaysia, or by Hastings et al. (25) in the USA. In 1969, Sagher and Sheskin were criticized for not mentioning the neuropathic side effects in three publications (44) and in the accompanying editorial, reference was made to the Lancet editorial on thalidomide neuropathy (3).

The reply by Sagher in response to this criticism has been used by Jakeman and Smith as evidence that neuropathy has not occurred in leprosy patients, but as emphasized earlier, these electrophysiological tests are of no use in detecting thalidomide neuropathy (13). Furthermore, Sagher's comments were written only four and not 25 years after its use in leprosy and before any of the evidence for the very high frequency of neuropathy following the use of thalidomide for non-leprous disorders in the 1980s. The 1969 editorial in the Lancet commented on Fullerton and O'Sullivan's clinical, pathological, and electrophysiological observations and the very poor prognosis, yet Waters, another MRC leprologist, claimed that in using thalidomide there was only a slight theoretical risk of neurotoxicity (45). As a senior member of the MRC, Waters's statement would carry extra weight, especially so, when leprosy workers live in very isolated areas without access to medical literature and were virtually solely dependent on information published in leprosy journals including Leprosy Review where this statement appeared. They would therefore naturally assume that the drug was safe to use provided it was not given to a woman in the reproductive age group. Indeed, Waters stated as much when he wrote in Leprosy Review, that thalidomide was: "apart from its teratogenic properties", a "generally safe and effective drug" (23).

In 1973 another warning was issued about the dangers of neuropathy in leprosy patients given the drug (10), with the recommendation that thalidomide be withdrawn as alternative treatment was available. It had always been assumed that because of the teratogenic dangers, the drug should only be used for research purposes, a view endorsed by the WHO expert committee on leprosy in 1970 (46) who stated that thalidomide should only be used for: "strictly investigative purposes", yet by 1973 it was being used in at least 62 centres many of which were in developing countries. By 1974 it had become the drug of choice in males and post-menopausal females for the management of ENL. The drug was obtainable directly from Chemie-Grünenthal but WHO set up no advisory body to monitor the side-effects, or even to record how many patients were being treated. The widespread use in developing countries created a new danger because of the absence of any statutory body to monitor side-effects and because of the frequency with which drugs are stolen and misused.

There were other grounds for concern. A paper appeared jointly by Sheskin, Sagher, Dorfmann and Von Schrader-Beielstein (47) but the last named was an employee of Chemie-Grünenthal yet this fact was not recorded. In this study, dapsone was withheld and thalidomide was given on its own to test if it had any chemotherapeutic effect. Not surprisingly, several patients deteriorated. It was not necessary to conduct this trial when thalidomide could have been tested for its chernotherapeutic action againsl M. leprae in the footpads of mice.

Thalidomide was even given to women of reproductive age without contraception (48,49). New derivatives obtained from Chemie-Grünenthal were tested in countries such as Tanzania and Ethiopia, presumably without the detailed and expensive procedure which is mandatory in developed countries. In 1976, according to information from David Owen, then the Minister of Health to Jack Ashley MP, some 12 derivatives of thalidomide were in the process of being investigated but it is not clear how many of these drugs were given to leprosy patients or what safety tests were carried out. In this letter, Owen also stated: "you will also wish to know that sir Ludwig Guttmann, among others, is satisfied that the allegation of neurotoxicity was unfounded and has given evidence to this effect".

All this time the neuropathic side-effects were being ignored and in the WHO expert committee on leprosy in 1977, there is no mention of neuropathy as a side-effect (51). Severe neurological disturbances did occur after administration of thalidomide but were considered to be of "uncertain causation", or ascribed to Guillain-Barré polyneuritis, a rare disease and not nearly as common as thalidomide neuropathy. One particular patient developed muscular paralysis involving muscles around the hip, consistent with the pattern of thalidomide nerve damage of considerable severity as muscle paralysis only follows sensory loss, but this was diagnosed as a rare form of dimorphous leprosy (47).

Another warning about the dangers of thalidomide neuropathy was issued in 1978, but was ignored (51). In the 1980s the drug was now being used for a variety of non-leprous disorders such as prurigo nodularis (14) chronic discoid lupus erythematosis (52) and latterly "graft -versus-host" disease (53). Now the true frequency of neuropathy was realized and found to be as high as 25 %, especially when clinical neurological findings were supplemented by electrophysiological testing. Yet it was still maintained that despite this high frequency neuropathy was not occurring in leprosy patient(54), even though the electrophysiological tests used by Sheskin and Yaar were flawed (55) and this fact was pointed out (56).

In 1987, thalidomide ernbryopathy was detected by ultrasound and an abortion carried out at 17 weeks to a Brazilian woman who had leprosy and had been given thalidomide (57).

The WHO expert committec on leprosy in 1988 again did not mention that thalidomide could cause a neuropathy (58) but while emphasizing that thalidomide should only be given to men and post-menopausal women stated that in this group the: "treatment of choice for ENI. is thalidomide and the importance of making this drug available cannot be overstressed". Inevitably, this recommendation must have increased the risk of the drug being given to a pregnant woman by mistake. The 19th edition of Martindae, The Extra Pharmacopeia published in 1989 stated that the nerve damage produced by thalidornide could be: "severe and irreversible" (59), but the increasing use of thalidomide prompted articles with complacent headings like: "thalidomide's back in the news but under more favourable circumstances" (60). In the same issue, claims were made that the nerve damage resulting from thalidomide was a rare cornplication and could be: "'reversed by stopping treatment or reducing dosage" (61), a view repeated in 1992 (62).

Even condescending statements from a member of medical profession were printed: "Mention thalidomide in non-medical company and people recoil in horror" (63).

Two publications (Bryceson(64) and Nordeen(65)), still do not mention the neuropathy as a side-effect and Waters, although now recognizing the serious nature of the neuropathy, claims its frequency was only 0.5% in 1991. He considered that there is likely to be a higher incidence in prurigo nodularis because of histological evidence of cutaneous nerve involvernent in this disease. Yet even if clinical evidence of sensory loss is not invariable in leprornatous leprosy, most patients have sorne thickening of the peripheral nerves which would predispose to nerve damage from thalidomide. In 1992 he accepted a figure of 25 %, but claimed that: "the nerves of lepromatous patients are relatively insusceptible to thalidomide-induced nerve damage" (17).

In view of the evidence presented here, this conclusion cannot be entertained and in all probability a large number of leprosy patients must have sustained severe and irreversible nerve damage.

The publication of an article on the Brazilian woman with an embryopathy (57) might suggest that the control of thalidomide was extremely lax, as confirmed by the film: "Thalidomide: the drug which came back" (66,67), which revealed the horrifying events which had been taking place over many years and which were not disclosed by the leprologists using the drug.

One is entitled to ask whether this teratogenic disaster could have been avoided if misleading information about the neuropathic side-effects of thalidomide had not been published and if more serious efforts had been taken to detect it in leprosy patients. The realization that the drug was too toxic to use could then have averted this tragedy.

It has been the view of several leprologists that lowering the dose of the chemotherapeutic agent will lessen the severity of ENL. Indeed, when thalidomide was first introduced, both Jopling and Davey were surprised that Sheskin gave 700 mg of dapsone weekly and Jopling recommended lowering the dose to only 20 mg weekly (68). Management with low doses of 10-25 mg weekly was also the rule in northern Nigeria in the 1960s where ENL could be successfully managed, although not completely suppressed, with the addition of 5-10 mg prednisone daily (10). This association between low dose dapsone and severity of ENL was challenged (69); but the lowest dose used was 100 mg weekly. Since the 1960s, dapsone resistance has increased and considered to be in part due to low doses of the drug. Hence it has been mandatory to give full doses of dapsone for lepromatous leprosy and also for those patients with ENL. There is, however conflicting evidence about the relation between low doses and dapsone resistance. It has been found that 8 % of patients given low dose sulphones as sulphetrone (solapsone) developed resistance compared to 2 % given sulphones in larger doses as dapsone (70). In Karamui, New Guinea, long-acting sulphones were given by injection of acedapsone at 225 mg every 75 days which is equivalent to giving only 17 mg weekly of dapsone (71), whereas the recommended dose to prevent resistance is 700 mg weekly. Moreover, sulphone blood levels in patients given acedapsone were only half those given sulphetrone (72). Twenty eight multibacillary patients were followed for up to six years using mouse foot pad tests to detect resistance, but none was found (73). In countries like the USA or UK where there is no spread of the disease, drug resistance to dapsone is not a public health risk and even in countries where the disease is endemic, spread is unlikely in patients with ENL because of the low M.I. Indeed it must be difficult to prove resistance in patients with ENL because of the low numbers of viable bacteria and hence their failure to multiply in the foot pads of mice.

Clofazimine has a chemotherapeutic effect as well as suppressing ENL so only one drug is necessary. It causes hyperpigmentation, which can be a problem in light-skinned patients, but it has been used in the multi-drug therapy regimen with good effect. Moreover, some recent phenazine analogues have chemotherapeutic activity in vivo, do not cause hyperpigmentation and deserve a trial (74, 75). Nonsteroidal antiinflammatory drugs have also shown some effect against ENL (76). Fucidic acid has been shown to have a chemotherapeutic activity against M. Leprae and also it appears to suppress the "lepra reaction"(78). Once again it is not clear if this reaction is ENL, but full reports of this drug's action are awaited with interest. It can therefore be seen that alternative forms of management are feasible and it should be possible to avoid the use of thalidomide. These regimens have drawbacks but are surely preferable to the teratogenic effects of thalidomide, recently demonstrated in Brazil.

Although thalidomide has been demonstrated to be effective in suppressing ENL, trials have not been carried out comparing the benefit of thalidomide against clofazimine or low doses of corticosteroids. One obvious trial would be to assess morbidity and mortality between males, and females in the reproductive age group. If thalidomide is so beneficial, then this group of females should be at a serious disadvantage. Some countries such as South Africa have continued to ban thalidomide so that leprosy patients here should be at greater risk. In Karamui, New Guinea(27) lepromatous leprosy has only been treated with sulphones. Reactions occurred in 13 of 38 lepromatous patients but did not seem severe enough to warrant additional treatment. Unfortunately, the type of reaction was not defined but merely called "lepra reacton" so it is unclear whether this was ENL. Nevertheless, neither corticosteroids nor thalidomide were given because it was not possible to provide medical supervision. The follow up of these patients from the bacteriological view was closely monitored with good progress for at least six years. It is therefore possible to treat lepromatous leprosy, in a highly endemic area to good effect without using thalidomide.

A very small study in Tanzania demonstrated that it was possible to prevent attacks of ENL if chemotherapy was stopped (79). While ENL persists indefinitely, it is probably not a sign of bacterial activity as most if not all of the bacteria are dead. With the introduction of multidrug therapy the duration of chemotherapy for uncomplicated lepromatous leprosy has been reduced to two years. It would be rational then to restrict the duration of chemotherapy in lepromatous leprosy complicated with ENL. By ceasing chemotherapy the ENL will also stop or be greatly reduced. This policy could be tried out in well-supervised leprosy centres, initially in Europe and the USA. Any relapse is easily detected by taking skin smears and determining if there has been a rise in the M.I. Without a relapse the need for thalidomide both in the short term and for prolonged use would not be necessary.

While the teratogenic tragedy in Brazil has caught public attention the failure to recognize thalidomide neuropathy in leprosy patients is still of serious concern. It must be unprecedented that a drug can cause toxic effects in one group of patients and not at all in others. Thalidomide has been used in the UK on a "named patient" basis to treat lepromatous patients with ENL since 1977. Drugs given on this basis are meant to be used only as a last resort when all other treatment has failed, yet thalidomide is recommended as the: "drug of choice for ENL reactions in male patients and in females without child-bearing potential" (December 1991, Mrs Virginia Bottomley, Minister for Health in a letter to Jack Ashley MP). Once again, it is claimed that no cases of neuropathy due to the drug have occurred and this conclusion has been supported by the Committee on the Safety of Medicines (Asscher, personal communication) and also by Mrs Bottomley. Yet none of the precautions to exclude a neuropathy detailed in this article have been met. In particular, the recording of sural nerve action potentials has not been carried out. Belatedly, the Committee on the Safety of Medicines is now issuing guidelines on the use of the drug in the UK.so These guidelines include a statement that the patient should be given an information sheet on which the side-effects of the drug are listed including nerve damage. Until 1979, leprosy patients in the UK were not warned that thalidomide could cause a neuropathy (Waters, personal communication). The present position is unclear. My request to the Committee on the Safety of Medicines to find out whether the patients are still not being warned was passed on to the Panel of Leprosy Opinion in October I992 (81,82). As the patients are now to be given an information sheet on which the side-effects are listed this will ensure that they

become aware that the drug can cause a neuropathy which: "can be both severe and irreversible"(80). At present any sensory symptoms will have been attributed to the disease and the chance of reversing nerve damage will have been lost. The guidelines state to the patients: "Should you develop pins and needles YOU. MUST STOP THALIDOMIDE IMMEDIATELY." Yet a leprosy patient, even if they do not have a symmetrical sensory loss, is likely to have some sensory disturbances because the peripheral nerves are usually thickened. Therefore, if the guidelines are followed this will preclude the use of thalidomide in leprosy in most cases. Even if leprosy patients are given thalidomide, the guidelines recommend that sensory nerve action potentials should be performed before starting and at frequent intervals during the course of treatment. Yet these potentials may well be altered because of the underlying disease, making detectionof thalidomide ncuropathy even more difficult. Furthermore, these tests cannot be carried out in developing countries where leprosy exists because of the lack of facilities and of experienced neurophysiologists.

Because of the delay in issuing guidelines for the use of thalidomide in the UK, and in particular, the failure to emphasize that the drug causes neuropathies which: "can be severe and irreversible", leprosy workers in developing countries where the disease is endemic are entitled to assume that the drug is safe provided it is not given to a woman in the reproductive age group, knowing that the Committee on the Safety of Medicines has sanctioned its use and knowing that this Committee sets very high standards on drug safety. Judging from the latest contribution on the subject this is still the case, as workers from Mexico do not mention that neuropathy is a side-effect of the drug (83).

Since the film disclosing the tragedy in Brazil was shown, little has been done to rectify the situation. The Brazilian authorities have issued a statement claiming that the guidelines for the use of thalidomide are stricter than those of WHO (84), but as the drug is used for domicilary treatment to patients who cannot be fully aware of the risks, further tragedies are likely to arise. This has already happened and it is not certain whether the mother even had leprosy (85). Brazil is unique in manufacturing the drug on a large scale, the estimates varying between two and eight million tablets per year(66,86). Despite the assurance of WHO, given in the film, that the use of thalidomide would be reviewed nothing has been done to prevent a recurrence of this tragedy. Although the tragedy in Brazil was revealed over a year ago none of the leprosy journals has even mentioned this disaster. In view of the arguments advanced here the right solution should be to withdraw thalidomide for use in leprosy and substitute alternative treatment regimens, while vigorously establishing the efficacy of newer forms of treatment especially fucidic acid and phenazine analogues of the clofazimine type, which do not cause skin pigmentation.

Thalidomide neuropathy is a very serious side-effect and has not been excluded in leprosy patients given the drug. By failing to take this side-effect seriously leprologists have recommended the widespread use of the drug especially in developing countries resulting in the teratogenic tragedy recently revealed in Brazil.

Thalidomide suppresses ENL, but this complication of lepromatous leprosy is not serious and its frequency has been exaggerated. Treatment with thalidomide can be avoided as alternative methods of management are available. These include low dose dapsone together with low doses of corticosteroids or clofazimine. The use of phenazines which do not cause pigmentation and fucidic acid should be encouraged. The view that low dose dapsone leads to resistance is challenged. Chemotherapy in lepromatous leprosy complicated by ENL should not be prolonged indefinitely. By ceasing chemotherapy, the attacks of ENL may be greatly diminished.

Recently introduced guidelines to prevent thalidomide neuropathy in the UK are not appropriate for leprosy patients and are too sophisticated to be applicable for use in developing countries. Thalidomide should not be used any longer for the management of ENL.

2. Sheskin J. Thalidomide in the treatment of lepra reactions. Clin Pharmacol Ther 1965; 6: 303-11.

3. Editorial. Thalidomide neuropathy. Lancet 1969; I: 713-14.

4. Fullerton PM, Kremer N. Neuropathy after intake of thalidomide (Distaval). BMJ 1961; 2: 855-8.

5. Fullerton PM, O'Sullivan DJ. Thalidomide neuropathy: a clinical, electrophysiological and histological follow-up study. J Neurol Neurosurg Psychiat 1968; 31: 543-51.

6. Hafström T. Polyneuropathy after Neurosedyn. Acta Neuro Scand 1967; 43 (suppl. 32): 1-41.

7. Ochonisky S, Verroust J, Bastuji-Garin S, Gheradi R, Rewz J. Thalidomide neuropathy incidence and clinicoelectrophysiologic findings in 42 patients. Arch Dermatol 1994; 130: 66-9.

8. Ridley DS, Jopling WH. Classification of leprosy according to immunity. Int J Lepr 1966; 34: 255-73.

9. Hansen GA, Looft C. Leprosy in its clinical and pathological aspects. Bristol: John Wright 1973.

10. Crawford CL. Treatment of erythema nodosum Ieprosum with thalidomide. Lancet 1973; 2: 567-8.

11. Monrad-Krohn GH. The neurological aspect of leprosy. Christiania: Jacob Dybwad 1923.

12. Lakeman P, Smith WCS. Thalidomide in leprosy reaction. Lancel 1994; 343: 432-3.

13. Sagher F. Thalidomide neuropathy. Lepr Rev 1969; 40: 126-8.

14. Nulff CH, Hoyer H, Asboel-Hansen G, Brodthagen H. Development of poly-neuropathy during thalidomide therapy. Br J Dermatol 1985; 112: 475-80.

15. Jones GRN. Thalidomide: 35 years on and still deforming. Lancet 1994; 343: 1041.

16. Waters MFR. Use of thalidomide in leprosy. BMJ 1991; 303: 470.

17. Waters.MFR. Thalidomide peripheral neuropathy. Trop Doct 1992;22: 139-140.

18.Fuller GN, Jacobs JM, Guiloff RJ. Thalidomide, peripheral neuropathy and AIDS. Int J STD & AIDS 1991; 2: 369-70,

19. Mellin GW, Katzenstein M. The sage of thalidomide neuropathy to embryopathy N Engl J Med 1962; 267: 1238-44.

20. Jopling WH, Summary of Murata's paper on erythema nodosum leprosum. Lepr Rev 1958; 29: 116- 18

21. Editorial. Induced leprotic reaction. Lepr Rev 1957; 28: 136-8.

22. Oleinick A. Survival among leprosy patients with special consideration of cancer as a cause of death. Int J Lepr 1968; 36: 318-27.

23. Waters MFR. Treatment of reactions in leprosy. Lepr Rev 1974; 45: 337-41.

24. Pearson JMH, Vedagiri M. Treatment of moderately severe erythema nodosum leprosum with thalidomide. A double blind controlled trial. Lepr Rev 1969; 40: 111-16.

25. Hastings RC, Trautman JR, Enna CD, Jacobson RR. Thalidomide in the treatment of erythema nodosum leprosum. Clin Pharmacol Ther 1970; 11: 481-7.

26. Convit J, Browne SG, Languillon J, et al. Therapy of leprosy. Bull WHO 1970; 42: 667-72.

27. Boughton CR, Scott GC, Russell DA, Vincin DR. A preliminary report on the use of the depot sulphone preparation acedapsone ("Hansolar") in the control of leprosy.·Med J Ausl 1971; 1: 1258-62.

28. Sheskin J, Sagher F. Survey of thalidomide treatment in lepra reaction. Int J Lepr 1973; 41: 693.

29. Awofeso A. Thalidomide peripheral neuropathy. Trop Docl 1992; 22: 139-40.

30. Sheskin J. Further observations with thalidomide in lepra reactions. Lepr Rev 1965; 36: 183-7.

31. Convit J, Soto JM, Sheskin J. Thalidomide therapy in lepra reaction. Int J Lepr 1967; 35: 446-51.

32. Manson-Barr PEC, Bell DR. Manson',s Tropicol Diseases,19th ed. London: Bailliere Tindall,1987; 757-85.

33. Pfaltzgraff RE, Bryceson A. Clinical Ieprosy. In: Hastings RC (ed.) Leprosy. Edinburgh: Churchill Livingstone 1985; 135-76.

34. Cohn ZA, Kaplan G. Leprosy-Hansen's Disease. In: Wyngaarden JB, Smith LH, Bennett JC (eds) Cecil Textbook of Medicine. 19th ed. Philadelphia: WB Saunders 1992; 1745-51.

35. Crawford CL. The effect of orrt-patient dapsone in an area of endemic leprosy. Lepr Rev 1969; 40: 159-63.

36. Van Brakel WH, Khawas IB, Lucas Sßr Reactions in leprosy: an epidemiological study of 386 patients in West Nepal. Lepr Rev 1994; 65: 190-203.

37. McAdam KPWJ, Andcrs RF, Smith SR, Russell DA, Price MA. Association of amyloidosis with erythema nodosum leprosum and recurrent leucocytosis in leprosy. Lancel 1975; ii: 572-5.

38. Editorial. Amyloiaosis and leprosy. Lancel 1975; ii: 589-90.

39. Crawford CL. Amyloidosis and erythema nodosum leprosum. Lancet 1975; 2: 703-4.

40. Levy L, Fasal P, Levan NE, Freedman RI. Treatment of erythema nodosum leprosum with thalidomide. Lanret I973; ii: 324-5.

41. Cochrane RG. Cornplicating conditions due to leprosy. In: Cochrane RG, Davey TF (cds) Leprosy in Tlreory and Practice. 2nd ed. Bristol: John Wright 1964; 331-43.

42. Sjöström H, Nilsson R. Thalidomide and thre power of thre drug companies. Harmondsworth, Middlesex: Penguin Books 1972.

43. Florence AL. Is thalidomide to blame? BMJ I960; 2: 1954.

44. Crawford CL. Thalidomide neuropathy. Lepr Rev 1969; 40: 126-8.

45. Waters MFR. An internally controlled double-blind trial of thalidomide in severe crythema nodosum leprosum. Lepr Rev 1971; 42: 26-42.

46. WHO Expert Committee on Leprosy, third report. WHO Tech Rep Ser No 459,1970.

47. Shakin J, Sagher F, Dorfman M, von Schrader W, Van-Beielstein H. Unsatisfactory results with thalidomide as a specific treatment for leprosy. Israel J Med Sci 1968; 4: 901 -4.

48. Graver V, Use of drugs and their effects on leprosy patients. Int J Lepr 1974; 42: 72.

49. Sheskin J. Response to Sr Graver. Inl J Lepr 1974; 42: 209.

50. WHO Expert Committee on Leprosy, fourth report. WHO Tech Rep Ser No 607, 1977.

51. Crawford C. Fresh concern over thalidomide. New Sci 1978; Feb 2: 277-9.

52. Knop L, Bonsmann G, Happle R, et al. Thalidomide in the treatment of 60 cases of chronic discoid lupus erythematosis. Br J Dermotol 1983; 108: 461-6.

53. Heney D, Norfolk DR, Wheeldon J, Bailey CC, Lewis IJ, Barnard DL. Thalidomide treatment for chronic graft-versus-host disease. Br J Haematol 1991; 78: 23-7.

54. Leading article. Thalidomide in dermatology and leprosy. Lancet 1985; ii: 80-1.

55. Sheskin J, Yaar I. Motorische leitungsgeschwindigkeit der kubitalnerven bei patienten mit leprareaktion. Zusammenenfassung einer dreizehnjahrigen beobachtungsreiche wahrend der thalidomidebehandlung. Hautarzt 1979; 30: 376-9.

56. Crawford CL. Thalidomide and leprosy. Lancet 1985; 2: 331.

57. Gollop TR, Eigier A, Neto JG. Prenatal diagnosis of thalidomide syndrome. Prenat Diagn 1987; 7: 295-8.

58. WHO Expert Committee on Leprosy, sixth report. WHO Tech Rep Ser No 768,1988.

59. Anonymous. Thalidomide. In: Reynolds JEF (ed.) Martindale: the Extra Pharmacopeia. 29th ed. London: The Pharmaceutical Press 1989; 1621-2.

60. Randell T. Thalidomide's back under more favorable circumstances. JAMA 1990; 263: 1467-8.

61. Randell T. Research focuses on immunosuppressive effect, unknown teratogenic mechanism of thalidomide. JAMA 1990; 263: 1473.

62. Crawford CL, Vogelsang GB. Thalidomide neuropathy. N Engl J Med 1993; 327: 735.

63. Minerva. BMJ 1992; 305: 322.

64. Bryceson ADM. Leprosy. In: Champion RH, Burton JL, Ebling FJG (eds) Rook Wilkinson Ebling Textbook oJ Dermatology. 5th ed. Oxford: Blackwell Scientific Publications 1992; 1065-83.

65. Nordeen SK. Leprosy. In: Rakel RE (ed.) Conn's Current Therapy. Philadelphia: WB Saunders 1994; 89-94.

66. Thalidomide: the drug that came back. First Tuesday. 1993; June 1. Yorkshire Television.

67. Cutler J. Thalidomide revisited. Lancet 1994; 343: 795-6.

68. Davey TF, Jopling WH. Comments on Sheskin J. Further observations with thalidomide in lepra reaction. Lepr Rev 1965; 36: 183-7.

69. Waters MFR. The management of erythema nodosum leprosum with particular reference to continued dapsone therapy. Int J Lepr 1968; 36: 638.

70. Meade TW, Pearson JMH, Rees RJW, North WRS. The epidemiology of sulphone-resistant leprosy. Int J Lepr 1973; 41: 684.

71. Shepard CC, Tolentino JG, McRae DH. The therapeutic effect of 4,4,1-diactyl-diamino-diphenyl sulphone (DADDS) in leprosy. Amer J Trop Med Hyg 1968;17: 192-201.

72. Gelber RH, Gooi HC, Waters MFR, Rees RJW. The pharmacology of sulphetrone and its implications for sulphone resistance. Lepr Rev 1974; 45: 308-12.

73. Russell DA, Shepard CC, McRae DH, Scott GC, Vincin DR. Acedapsone (DADDS) treatment of leprosy patients in the Karamui District of New Guinea: status at 6 years. Amer J Trop Med Hyg 1975; 24: 485-95.

74. Hooper M. Use of thalidomide in leprosy. Br Med J 1991; 303: 1404-5.

75. Van Landingham RM, Walker LL, O'Sullivan JF, Shinnick TM. Activity of phenazinc analogs against Mycobacterium leprae infections in mice. Int J Lepr 1993; 61: 406-16.

76. Theirs H, Rousset J, Coudert J, Battesti MR, Than LH. Indomethacine dans les états réactionnels lépreux. Bull Soc Franc Dermatol Syph 1966; 73: 347-9.

77. Franzblau SG, Chan GP, Garcia-Ignacio BG, et al. Clinical trial of fucidic acid for lepromatous leprosy. Antimicrob Agents and Chemother 1994; 38: 1651-4.

78. Rowe PM. Drug therapy for tropical diseases. Lancet 1994; 343: 1352.

79. Crawford CL. Erythema nodosum leprosum. Lancet 1973; 1: 568.

80. Powell RJ, Gardner-Medwin JMM. Guideline for the clinical use and dispensing of thalidomide. Postgrad Med J 1994 (in press).

81. Crawford CL. Safety of thalidomide. BMJ 1994; 308: 1437-8.

82. Asscher W. Safety of thalidomide. BMJ 1994; 309: 193-4.

83. Teixeira F, Hojyo MT, Arenas R, et al. Thalidomide: can it continue to be used? Lancet 1994; 344: 196-7.

84. The use of thalidomide in the treatment of Hansen's disease in Brazil. Ministry of Health, Brazil 1993.

85. Gorman C. Thalidomide's return. Time 1994; 24: 50.

86. Rocha J. Thalidomide in Brazil. BMJ 1994; 308: 1061.