1611
High frequency of concomitant pancreatitis in salmonella enteritis SIR,-Dr Baird-Parker (Nov 17, p 1231) underscores the increasing incidence of foodborne salmonellosis in many countries of the western world.l We report some observations that are relevant to the clinical picture and differential diagnosis of this diarrhoeal condition. Stimulated by a case of salmonellosis with severe concomitant pancreatitis we prospectively followed all culture-proven cases of salmonella enteritis for symptoms and laboratory signs of concomitant pancreatitis. Within 18 months we treated 47 patients with salmonella enteritis (17 men, mean age 45 [range 16-79] years). 16 patients had Salmonella typhimurium and 31 had S enteritidis infections, 1 patient having both. The frequency of concomitant pancreatitis as assessed by raised serum amylase and lipase was: Concomitant pancreatitis No
29(62%)
Men/women Mean age (range)
No pancreatitis 18(38%)
8/21 44.8
9/9
(18-79)
45-8
(16-71)
Concomitant pancreatitis was found in 7 patients with typhimurium and in 22 with S enteritidis (p < 005, Chi-square test). The course of pancreatitis was usually mild to moderate, with some pancreatic enlargement, as shown by abdominal sonography, S
in about half the patients. Pancreatitis was unrelated to treatment of salmonellosis either by fluid replacement alone or antibiotics
(mainly ciprofloxacin). The high frequency
of concomitant pancreatitis in salmonella series is in contrast with other publications. In gastroenterological textbooks salmonella pancreatitis is not mentioned2 or is merely listed in tables.3 Only in paediatric publications has some attention been paid to these forms of pancreatitis complicating viral or bacterial infections.4 The pathogenetic mechanisms are unknown as far as we are aware. By analogy to other infections, a haematogenous attack on parenchymal pancreatic cells seems probable. These observations indicate that patients with upper abdominal pain in salmonella enteritis should be investigated for concomitant pancreatitis, and in those with hyperamylasaemia and diarrhoea, salmonellosis should be considered as a possible explanation for this clinical condition. enteritis in
Fig 2-Localised proton
MR spectra and
paramedian
grey-
matter.
Upper: Patient with CJD exhibiting reduction of NAA. Lower. control. Volume of interest 18 ml.
The NAA data, a marker of vital neuronal tissue,’ indicate severe neuronal loss in CJD. Most notably, MRS in this patient revealed metabolic alterations in regions of the brain that had appeared normal on magnetic resonance imaging. Structural preservation in the presence of gross functional impairment in CJD is often seen with computerised tomography. Positron emission tomography5,1 reveals hypoperfusion and hypometabolism but proton MRS has the potential advantage of selectively indicating neuronal damage, the cell type most severely affected in the spongiform encephalopathies. This non-invasive investigation could be of use in the early diagnosis of CJD. Financial support by the Bundesminister fur Forschung und Technologie (grant 01 VF 8606/6) is gratefully acknowledged. Max Planck Institute for Biophysical D-3400 Gottingen, Germany,
Neurological Clinic, University of Gottingen; and Klinik Schildautal, Seesen
Chemistry,
H. BRUHN T. WEBER V. THORWIRTH
J. FRAHM
1. Pearl
GS, Anderson RE. Creutzfeldt-Jakob disease: high caudate signal on magnetic imaging. South Med J 1989; 82: 1177-80. 2. Levy SR, Chiappa KH, Burke CJ, Young RR. Early evolution and incidence of electrencephalographic abnormalities in Creutzfeldt-Jakob disease. J Clin Neurophysiol 1986; 3: 1-21. 3. Frahm J, Michaelis T, Merboldt KD, Bruhn H, Gyngell ML, Hanicke W. resonance
Improvements m localized proton NMR spectroscopy of human brain: water suppression, short echo times, and 1 ml resolution. J Magn Reson 1990; 90: 464-73. 4. Bruhn H, Frahm J, Gyngell ML, Merboldt KD, Hänicke W, Sauter R. Cerebral metabolism in man after acute stroke: new observations using localized proton NMR spectroscopy. Magn Reson Med 1989; 9: 126-31. 5. Hunter R, Gordon A, McLuskie R, et al. Gross regional cerebral hypofunction with normal CT scan in Creutzfeldt-Jakob disease. Lancet 1989; i: 214-15. 6. Holthoff VA, Sandmann J, Pawlik G, Schroder R, Heiss WD. Positron emission tomography in Creutzfeldt-Jakob disease. Arch Neurol 1990; 47: 1035-38.
our
Department of Medicine, Krankenhaus der Barmherzigen Schwestern, 4910 Ried/l, Austria
FRIEDRICH RENNER CHRISTIAN NIMETH NIKOLAUS DEMMELBAUER
1. World Health Organisation. Salmonellosis control: the role of animal and product hygiene. WHO Tech Rep Ser 1988, no 774. 2. Soergel KH. Acute pancreatitis. In: Sleismger MH, Fordtran JS, eds. Gastrointestinal disease, 3rd ed. Philadelphia Saunders, 1983: 1462-84. 3. Creutzfeldt W, Lankisch PG. Acute pancreatitis: etiology and pathogenesis. In: Berk JE, ed. Bockus gastroenterology, 4th ed. Philadelphia: Saunders, 1985. 3971-92. 4. Seifert G. Zur Pathologie der Pankreatitis im Kindesalter. Mschr Kinderheilk 1960; 108: 225-29.
Mega-dose methylprednisolone for chronic idiopathic thrombocytopenic purpura SIR,-Dr Balint and colleagues (May 4, p 1106) mention, among other approaches, conventional prednisolone (1-2 mg/kg daily) administration for the treatment of idiopathic thrombocytopenic purpura (ITP). With colleagues, I have shown that such treatment for 2 weeks is not effective in ITP; on the contrary it delays spontaneous recovery in childhood acute ITP.1 Mega-dose oraF or intravenous1.3 methylprednisolone (MDMP; 30 mg/kg daily for 3 days followed by 20 mg/kg for 4 days; each dose given before 0900 h) has, however, proved very effective. Platelet counts increased to over 150 000/µl within 3 days in about two-thirds of patients.1,2 Intravenous MDMP has also been used successfully in children4-6 and adults7,S with chronic ITP. We have used oral MDMP (30 mg/kg and 20 mg/kg daily, each for 1 week; every dose before 0900 h) in 10 children with chronic ITP (ie, duration longer than 6 months). 1 child could not be followed. Platelet count increased to over 150 000/µl in 6 patients within 3 days and in 8 by 2 weeks.
1612
Since oral MDMP administration is easier, cheaper, and platelet rise faster than in Balint and colleagues’ treatment method, I suggest that our approach should be tried before extracorporeal immunoadsorbtion. counts
Department of Paediatrics, Haematology Section, Hacettepe University Faculty of Medicine, Ankara, Turkey, and Hacettepe Children’s Hospital, Ankara
SINASI ÖZSOYLU
Özsoylu S, Irken G,
Karabent A. High-dose intravenous methylprednisolone for childhood idiopathic thrombocytopenic purpura. Eur J Haematol 1989; 42: 431-35. 2. Özsoylu S, Erturk G. Oral megadose methylprednisolone for childhood acute idiopathic thrombocytopenic purpura. Blood 1991; 77: 1856-57, 3. Van Hoff J, Ritchey AK. Pulse methlyprednisolone therapy for acute childhood idiopathic thrombocytopenic purpura. J Pediatr 1988; 113: 563-66. 4. Hara T, Miyazaki S, Yoshido N, Goya N. High dose of gammaglobulin and methylprednisolone therapy for idiopathic thrombocytopenic purpura in children. Eur JPediatr 1985; 144: 240-42. 5. del Principe D, Menichelli A, Mori PG, et al. Phase II trial of methylprednisolone pulse therapy in childhood chronic thrombocytopenia. Acta Haematol 1987; 11: 226-30. 6. Özsoylu S. High dose intravenous methylprednisolone for the treatment of childhood chronic idiopathic thrombocytopenic purpura. Pediatr Rev Commun 1987; 2: 29-40. 7. Von Dem Borne AEGKR, Vos JJE, Pegels JG, et al. High dose intravenous methylprednisolone or high dose intravenous gammaglobulin for autoimmune thrombocytopenia. Br Med J 1988; 296: 249-50. 8. Akoğlu T, Paydas S, Bayik M, et al. Megadose methylprednisolone pulse therapy in adult idiopathic thrombocytopenic purpura. Lancet 1991; 337: 56. 1.
acute
Triazolam SIR,-May we respond to Dr Girard’s letter (June 15, p 1483) on April 6 paper. Our study was not designed to assess sex differences in memory impairment. Five patients were women and all experienced memory impairment episodes; the one man did not. Nevertheless, many published case-reports on memory our
impairment/amnesia with triazolam refer to men as well as women. In the twelve studies we cited there were 29 reports of memory impairment, 69% in men and 31% in women. Furthermore, we only included insomniacs who did not have major psychiatric disorders. Thus, one would expect, as Girard points out, that the rate of such episodes in clinical practice would be even greater than that found in our study. The parametric Dunn statistical test was appropriate; when we analysed the data with a Poisson regression model,’ which does not assume a normal distribution, the results were the same. Furthermore, we presented our data in two ways-as average rate per patient (ie, probability of recurrence of amnesia in a patient), which was 40% for those on triazolam; and as the prevalence per drug or placebo condition (ie, the probability of amnesia occurring at least once within), which was 83% for the triazolam group. To be concise we reported in figure form raw data for the triazolam group only. However, both prevalence and average rate per subject of amnesia/memory impairment within each treatment group are in the text at p 829 (temazepam 0% and 0%; placebo group 16-7% and
in considering actions related to the current dose of triazolam (025 mg), should not accept the objections and positions presented by the manufacturer when triazolam 1 ’0 mg was removed from the market in the Netherlandss and, subsequently, when the 0-5 mg dose was withdrawn in France,6 Italy,’ and West Germany.8
community,
Sleep Research and Treatment Center and Department of Psychiatry, Pennsylvania State University College of Medicine, Hershey, Pennsylvania 17033, USA 1.
EDWARD O. BIXLER ANTHONY KALES Rocco L. MANFREDI ALEXANDROS N. VGONTZAS KATHY L. TYSON JOYCE D. KALES
McCullagh P, Nelder JA. Generalized linear models. New York: Chapman & Hall, 1989.
Meyboom RHB. Benzodiazepines and pilot error. Br Med J 1991; 302: 1274. Ayd FJ. Next-day memory impairment/amnesia associated with triazolam (editor’s comment). Int Drug Ther Newsl 1991; 26: 21-24. 4. Ayd FJ. Triazolam overdosage (editor’s comment). Int Drug Ther Newsl 1991; 26: 24 5. Offerhaus L. Trials and tribulations of triazolam. JClin Pharm 1980; 20: 700-02. 6. Halcion 0·5 mg suspended in France. Scrip 1987; 1189: 34. 7. Halcion 0·5 mg Italian suspension (F). Scrip 1987; 1207: 27. 8.0 5 Halcion withdrawn in FRG. Scrip 1988; 1296: 3. 2. 3.
Chondroprotection and NSAIDs SIR,-Your editorial (March 30, p 769) highlights the recent debate about the effects of non-steroidal anti-inflammatory drugs (NSAIDs) on articular cartilage.1,2 There are no convincing data that show an effect of NSAIDs on cartilage metabolism in man and extrapolation from laboratory studies may not be appropriate. The availability of more sensitive diagnostic methods such as quantitative microfocal radiography3 and magnetic resonance imaging (MRI),4 together with new immunoassaysS that can detect cartilage breakdown products should overcome these difficulties. Osteoarthritis (OA) leads to changes not only in cartilage, but also in subchondral bone, the joint capsule, and the synovium. Suitable laboratory animal models that simulate the human disorder within a time-scale acceptable for practical investigation are therefore difficult to find. However, OA is also a common complication after meniscectomy and anterior cruciate ligament rupture, and research has shown that when these injuries are reproduced experimentally in knee joints of large animals, such as the sheep6 and dog,7,8 biochemical and histological changes occur in the cartilage, subchondral bone, and synovium which are analogous to those seen in the early stages of human OA. We believe that these animal models qualify as suitable analogues of human OA, albeit of mechanical origin. Over the past 10 years, we have adopted the meniscectomy model to evaluate the effects of postoperative immobilisation,8 exerciseand therapeutic measures on the progression of cartilage and bone damage after induction of joint
3%, respectively). The clinical significance of the findings reported in our table 11 is obvious in that memory impairment observed with objective tests is consistent with the striking episodes of memory impairment/ amnesia for daytime activities reported by the patients. For further appreciation of these data we refer to the many clinical reports and controlled studies cited in
our
paper, to
a recent
letter
on
the
potential for pilot error due to amnesia associated with triazolam,2 and to Ayd’s review and editorial comments3,4 in which he indicates that triazolam may have considerably greater toxicity at lower doses than do other benzodiazepines. Girard feels that "it would be premature to withdraw triazolam from the market". Does this mean that he is unconcerned about the drug’s profile of side-effects and narrow margin of safety? Does he not agree with us that, at the very least, "the labelling for triazolam should be strengthened to warn patients that its adverse effects are common and severe", especially since serious cognitive impairments such as amnesia are frequent, yet the drug is promoted for situations where next-day "alertness" is desired? On the basis of the compelling evidence that has now accumulated, the scientific
Proteoglycan fragments D
=
Control group,
drug therapy. *p
High frequency of concomitant pancreatitis in salmonella enteritis SIR,-Dr Baird-Parker (Nov 17, p 1231) underscores the increasing incidence of foodborne salmonellosis in many countries of the western world.l We report some observations that are relevant to the clinical picture and differential diagnosis of this diarrhoeal condition. Stimulated by a case of salmonellosis with severe concomitant pancreatitis we prospectively followed all culture-proven cases of salmonella enteritis for symptoms and laboratory signs of concomitant pancreatitis. Within 18 months we treated 47 patients with salmonella enteritis (17 men, mean age 45 [range 16-79] years). 16 patients had Salmonella typhimurium and 31 had S enteritidis infections, 1 patient having both. The frequency of concomitant pancreatitis as assessed by raised serum amylase and lipase was: Concomitant pancreatitis No
29(62%)
Men/women Mean age (range)
No pancreatitis 18(38%)
8/21 44.8
9/9
(18-79)
45-8
(16-71)
Concomitant pancreatitis was found in 7 patients with typhimurium and in 22 with S enteritidis (p < 005, Chi-square test). The course of pancreatitis was usually mild to moderate, with some pancreatic enlargement, as shown by abdominal sonography, S
in about half the patients. Pancreatitis was unrelated to treatment of salmonellosis either by fluid replacement alone or antibiotics
(mainly ciprofloxacin). The high frequency
of concomitant pancreatitis in salmonella series is in contrast with other publications. In gastroenterological textbooks salmonella pancreatitis is not mentioned2 or is merely listed in tables.3 Only in paediatric publications has some attention been paid to these forms of pancreatitis complicating viral or bacterial infections.4 The pathogenetic mechanisms are unknown as far as we are aware. By analogy to other infections, a haematogenous attack on parenchymal pancreatic cells seems probable. These observations indicate that patients with upper abdominal pain in salmonella enteritis should be investigated for concomitant pancreatitis, and in those with hyperamylasaemia and diarrhoea, salmonellosis should be considered as a possible explanation for this clinical condition. enteritis in
Fig 2-Localised proton
MR spectra and
paramedian
grey-
matter.
Upper: Patient with CJD exhibiting reduction of NAA. Lower. control. Volume of interest 18 ml.
The NAA data, a marker of vital neuronal tissue,’ indicate severe neuronal loss in CJD. Most notably, MRS in this patient revealed metabolic alterations in regions of the brain that had appeared normal on magnetic resonance imaging. Structural preservation in the presence of gross functional impairment in CJD is often seen with computerised tomography. Positron emission tomography5,1 reveals hypoperfusion and hypometabolism but proton MRS has the potential advantage of selectively indicating neuronal damage, the cell type most severely affected in the spongiform encephalopathies. This non-invasive investigation could be of use in the early diagnosis of CJD. Financial support by the Bundesminister fur Forschung und Technologie (grant 01 VF 8606/6) is gratefully acknowledged. Max Planck Institute for Biophysical D-3400 Gottingen, Germany,
Neurological Clinic, University of Gottingen; and Klinik Schildautal, Seesen
Chemistry,
H. BRUHN T. WEBER V. THORWIRTH
J. FRAHM
1. Pearl
GS, Anderson RE. Creutzfeldt-Jakob disease: high caudate signal on magnetic imaging. South Med J 1989; 82: 1177-80. 2. Levy SR, Chiappa KH, Burke CJ, Young RR. Early evolution and incidence of electrencephalographic abnormalities in Creutzfeldt-Jakob disease. J Clin Neurophysiol 1986; 3: 1-21. 3. Frahm J, Michaelis T, Merboldt KD, Bruhn H, Gyngell ML, Hanicke W. resonance
Improvements m localized proton NMR spectroscopy of human brain: water suppression, short echo times, and 1 ml resolution. J Magn Reson 1990; 90: 464-73. 4. Bruhn H, Frahm J, Gyngell ML, Merboldt KD, Hänicke W, Sauter R. Cerebral metabolism in man after acute stroke: new observations using localized proton NMR spectroscopy. Magn Reson Med 1989; 9: 126-31. 5. Hunter R, Gordon A, McLuskie R, et al. Gross regional cerebral hypofunction with normal CT scan in Creutzfeldt-Jakob disease. Lancet 1989; i: 214-15. 6. Holthoff VA, Sandmann J, Pawlik G, Schroder R, Heiss WD. Positron emission tomography in Creutzfeldt-Jakob disease. Arch Neurol 1990; 47: 1035-38.
our
Department of Medicine, Krankenhaus der Barmherzigen Schwestern, 4910 Ried/l, Austria
FRIEDRICH RENNER CHRISTIAN NIMETH NIKOLAUS DEMMELBAUER
1. World Health Organisation. Salmonellosis control: the role of animal and product hygiene. WHO Tech Rep Ser 1988, no 774. 2. Soergel KH. Acute pancreatitis. In: Sleismger MH, Fordtran JS, eds. Gastrointestinal disease, 3rd ed. Philadelphia Saunders, 1983: 1462-84. 3. Creutzfeldt W, Lankisch PG. Acute pancreatitis: etiology and pathogenesis. In: Berk JE, ed. Bockus gastroenterology, 4th ed. Philadelphia: Saunders, 1985. 3971-92. 4. Seifert G. Zur Pathologie der Pankreatitis im Kindesalter. Mschr Kinderheilk 1960; 108: 225-29.
Mega-dose methylprednisolone for chronic idiopathic thrombocytopenic purpura SIR,-Dr Balint and colleagues (May 4, p 1106) mention, among other approaches, conventional prednisolone (1-2 mg/kg daily) administration for the treatment of idiopathic thrombocytopenic purpura (ITP). With colleagues, I have shown that such treatment for 2 weeks is not effective in ITP; on the contrary it delays spontaneous recovery in childhood acute ITP.1 Mega-dose oraF or intravenous1.3 methylprednisolone (MDMP; 30 mg/kg daily for 3 days followed by 20 mg/kg for 4 days; each dose given before 0900 h) has, however, proved very effective. Platelet counts increased to over 150 000/µl within 3 days in about two-thirds of patients.1,2 Intravenous MDMP has also been used successfully in children4-6 and adults7,S with chronic ITP. We have used oral MDMP (30 mg/kg and 20 mg/kg daily, each for 1 week; every dose before 0900 h) in 10 children with chronic ITP (ie, duration longer than 6 months). 1 child could not be followed. Platelet count increased to over 150 000/µl in 6 patients within 3 days and in 8 by 2 weeks.
1612
Since oral MDMP administration is easier, cheaper, and platelet rise faster than in Balint and colleagues’ treatment method, I suggest that our approach should be tried before extracorporeal immunoadsorbtion. counts
Department of Paediatrics, Haematology Section, Hacettepe University Faculty of Medicine, Ankara, Turkey, and Hacettepe Children’s Hospital, Ankara
SINASI ÖZSOYLU
Özsoylu S, Irken G,
Karabent A. High-dose intravenous methylprednisolone for childhood idiopathic thrombocytopenic purpura. Eur J Haematol 1989; 42: 431-35. 2. Özsoylu S, Erturk G. Oral megadose methylprednisolone for childhood acute idiopathic thrombocytopenic purpura. Blood 1991; 77: 1856-57, 3. Van Hoff J, Ritchey AK. Pulse methlyprednisolone therapy for acute childhood idiopathic thrombocytopenic purpura. J Pediatr 1988; 113: 563-66. 4. Hara T, Miyazaki S, Yoshido N, Goya N. High dose of gammaglobulin and methylprednisolone therapy for idiopathic thrombocytopenic purpura in children. Eur JPediatr 1985; 144: 240-42. 5. del Principe D, Menichelli A, Mori PG, et al. Phase II trial of methylprednisolone pulse therapy in childhood chronic thrombocytopenia. Acta Haematol 1987; 11: 226-30. 6. Özsoylu S. High dose intravenous methylprednisolone for the treatment of childhood chronic idiopathic thrombocytopenic purpura. Pediatr Rev Commun 1987; 2: 29-40. 7. Von Dem Borne AEGKR, Vos JJE, Pegels JG, et al. High dose intravenous methylprednisolone or high dose intravenous gammaglobulin for autoimmune thrombocytopenia. Br Med J 1988; 296: 249-50. 8. Akoğlu T, Paydas S, Bayik M, et al. Megadose methylprednisolone pulse therapy in adult idiopathic thrombocytopenic purpura. Lancet 1991; 337: 56. 1.
acute
Triazolam SIR,-May we respond to Dr Girard’s letter (June 15, p 1483) on April 6 paper. Our study was not designed to assess sex differences in memory impairment. Five patients were women and all experienced memory impairment episodes; the one man did not. Nevertheless, many published case-reports on memory our
impairment/amnesia with triazolam refer to men as well as women. In the twelve studies we cited there were 29 reports of memory impairment, 69% in men and 31% in women. Furthermore, we only included insomniacs who did not have major psychiatric disorders. Thus, one would expect, as Girard points out, that the rate of such episodes in clinical practice would be even greater than that found in our study. The parametric Dunn statistical test was appropriate; when we analysed the data with a Poisson regression model,’ which does not assume a normal distribution, the results were the same. Furthermore, we presented our data in two ways-as average rate per patient (ie, probability of recurrence of amnesia in a patient), which was 40% for those on triazolam; and as the prevalence per drug or placebo condition (ie, the probability of amnesia occurring at least once within), which was 83% for the triazolam group. To be concise we reported in figure form raw data for the triazolam group only. However, both prevalence and average rate per subject of amnesia/memory impairment within each treatment group are in the text at p 829 (temazepam 0% and 0%; placebo group 16-7% and
in considering actions related to the current dose of triazolam (025 mg), should not accept the objections and positions presented by the manufacturer when triazolam 1 ’0 mg was removed from the market in the Netherlandss and, subsequently, when the 0-5 mg dose was withdrawn in France,6 Italy,’ and West Germany.8
community,
Sleep Research and Treatment Center and Department of Psychiatry, Pennsylvania State University College of Medicine, Hershey, Pennsylvania 17033, USA 1.
EDWARD O. BIXLER ANTHONY KALES Rocco L. MANFREDI ALEXANDROS N. VGONTZAS KATHY L. TYSON JOYCE D. KALES
McCullagh P, Nelder JA. Generalized linear models. New York: Chapman & Hall, 1989.
Meyboom RHB. Benzodiazepines and pilot error. Br Med J 1991; 302: 1274. Ayd FJ. Next-day memory impairment/amnesia associated with triazolam (editor’s comment). Int Drug Ther Newsl 1991; 26: 21-24. 4. Ayd FJ. Triazolam overdosage (editor’s comment). Int Drug Ther Newsl 1991; 26: 24 5. Offerhaus L. Trials and tribulations of triazolam. JClin Pharm 1980; 20: 700-02. 6. Halcion 0·5 mg suspended in France. Scrip 1987; 1189: 34. 7. Halcion 0·5 mg Italian suspension (F). Scrip 1987; 1207: 27. 8.0 5 Halcion withdrawn in FRG. Scrip 1988; 1296: 3. 2. 3.
Chondroprotection and NSAIDs SIR,-Your editorial (March 30, p 769) highlights the recent debate about the effects of non-steroidal anti-inflammatory drugs (NSAIDs) on articular cartilage.1,2 There are no convincing data that show an effect of NSAIDs on cartilage metabolism in man and extrapolation from laboratory studies may not be appropriate. The availability of more sensitive diagnostic methods such as quantitative microfocal radiography3 and magnetic resonance imaging (MRI),4 together with new immunoassaysS that can detect cartilage breakdown products should overcome these difficulties. Osteoarthritis (OA) leads to changes not only in cartilage, but also in subchondral bone, the joint capsule, and the synovium. Suitable laboratory animal models that simulate the human disorder within a time-scale acceptable for practical investigation are therefore difficult to find. However, OA is also a common complication after meniscectomy and anterior cruciate ligament rupture, and research has shown that when these injuries are reproduced experimentally in knee joints of large animals, such as the sheep6 and dog,7,8 biochemical and histological changes occur in the cartilage, subchondral bone, and synovium which are analogous to those seen in the early stages of human OA. We believe that these animal models qualify as suitable analogues of human OA, albeit of mechanical origin. Over the past 10 years, we have adopted the meniscectomy model to evaluate the effects of postoperative immobilisation,8 exerciseand therapeutic measures on the progression of cartilage and bone damage after induction of joint
3%, respectively). The clinical significance of the findings reported in our table 11 is obvious in that memory impairment observed with objective tests is consistent with the striking episodes of memory impairment/ amnesia for daytime activities reported by the patients. For further appreciation of these data we refer to the many clinical reports and controlled studies cited in
our
paper, to
a recent
letter
on
the
potential for pilot error due to amnesia associated with triazolam,2 and to Ayd’s review and editorial comments3,4 in which he indicates that triazolam may have considerably greater toxicity at lower doses than do other benzodiazepines. Girard feels that "it would be premature to withdraw triazolam from the market". Does this mean that he is unconcerned about the drug’s profile of side-effects and narrow margin of safety? Does he not agree with us that, at the very least, "the labelling for triazolam should be strengthened to warn patients that its adverse effects are common and severe", especially since serious cognitive impairments such as amnesia are frequent, yet the drug is promoted for situations where next-day "alertness" is desired? On the basis of the compelling evidence that has now accumulated, the scientific
Proteoglycan fragments D
=
Control group,
drug therapy. *p
