- Email: [email protected]
Hepatitis-E-associated fulminant hepatic failure in non-pregnant young women
THE LANCET
advantages, giving rise to a new model of pharmacoeconomic analysis.5 Michael T Isaac, Maria B Tome Suite 6, Lewisham Hospital, London SE13 6LH, UK
1
2
3
4
5
Perez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97. Bermann RM, Darnell AM, Miller HL, Anand A, Charney DS. Effect of pindolol in hastening response to fluoxetine in the treatment of major depression: a doubleblind placebo-controlled trial. Am J Psych 1997; 154: 37–43. Tome MB, Isaac MT, Harte R, Holland C. Paroxetine and pindolol: a randomised trial of serotonergic autoreceptor blockade in the reduction of antidepressant latency. Int Clin Psychopharmacol 1997; 12: 81–89. Tome M, Cloninger C, Watson J, Isaac M. Serotonergic autoreceptor blockade in the reduction of antidepressant latency personality variables and response to SSRI antidepressant and ‘augmenting agent’. J Affect Dis 1997; 44: 101–09. Tome M, Isaac M. Cost-benefit and costeffectiveness analysis of rapid onset of antidepressant and ‘augmenting agent’. Int J Psychiatry Med (in press).
SIR—Victor Pérez and co-workers1 report that the addition of pindolol to fluoxetine antidepressant treatment increases the efficacy of fluoxetine therapy. A difference in the cumulative percentage of patients with a sustained response was seen by day 9, with a significantly greater proportion of patients in the pindolol group. More surprising, this difference was maintained until the end of the 6-week study, when the percentage of patients with sustained response was 69% in the pindolol group versus 48% in the placebo group. The rationale for the use of 5-HT1A antagonists in depression has been lent support by some experimental evidence, reviewed by Artigas and colleagues.2 Both electrophysiological and microdialysis experiments have suggested that pindolol accelerates the effects of serotonin reuptake inhibitors (SSRIs) by blocking the serotonin negative feedback of 5-HT1A somatodendritic receptors until desensitisation of these receptors takes place. This rationale was clinically illustrated in two open studies that have shown a shorter time to improvement of depressive symptoms by paroxetine associated with pindolol.3,4 Our group has done a randomised, double-blind, placebo-controlled trial, comparing pindolol (15 mg/day) with placebo, in association with paroxetine for the first 3 weeks of a 6-month treatment of major depressive episode (unpublished). Intermediate analysis of results for the first month of the first 100 patients revealed that the number of patients who remitted (17-item
Vol 350 • July 26, 1997
Hamilton score 聿10) was higher in the pindolol group by day 10, but that the difference between the two groups disappeared at day 15 and thereafter. The scores for depression severity scales were lower in the pindolol group at day 5 and day 10, but differences disappeared at day 15 and thereafter. At the end of the first month of treatment, there was no difference in response to paroxetine between the pindolol and placebo groups. These data are compatible with the delay necessary to the 5-HT1A autoreceptor desensitisation. On the other hand, the results presented by Pérez et al cannot be explained only by this hypothesis. If a kinetic interaction is excluded, how can we account for the prolonged higher response rate in the pindolol group without a difference in time-to-onset of clinical improvement when only patients with sustained response are considered? It would be of interest to know if experimental evidence supports such prolonged enhancement of SSRI effects by pindolol and how long this enhancement could be maintained in the treatment of depression. *Régis Bordet, Pierre Thomas, Bernard Dupuis, on behalf of the REEP Pharmacologie Centre, Hospitalier Régional, Faculté de Médecine, 59045 Lille, France
1
2
3
4
Pérez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97. Artigas F, Romero L, de Montigny C, Blier P. Acceleration of the effect of selected antidepressant drugs in major depression by 5-HT1A antagonists. Trends Neurosci 1996; 19: 378–83. Artigas F, Pérez V, Alvarez E. Pindolol induces a rapid improvement of depressed patients treated with serotonin reuptake inhibitors. Arch Gen Psychiatry 1994; 51: 248–51. Blier P, Bergeron R. Effectiveness of pindolol with selected antidepressant drugs in the treatment of major depression. J Clin Psychopharmacol 1995; 15: 217–22.
SIR—In Victor Pérez and colleagues’ study,1 the number of days to reach a sustained response was lower in the fluoxetine and pindolol group than in the fluoxetine and placebo group, but there was no difference when only those patients who responded were in the analysis. Thus, there is a discrepancy about the time-to-onset of clinical improvement. Pérez et al suggest that a longer study period may increase the number of responders and resolve the discrepancy because their study period (6 weeks) is not the maximum time-limit for a response to antidepressants. I believe that another possibility is that pindolol cannot hasten the clinical response to fluoxetine but can convert fluoxetine
non-responders into responders, which might account for the discrepancy. Takeshi Terao Department of Psychiatry, University of Occupational and Environmental Health, Yahatanishi-ku, Kitakyushu 807, Japan
1
Pérez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97.
Hepatitis-E-associated fulminant hepatic failure in non-pregnant young women SIR—A Mas and J Rodés’ seminar on fulminant hepatic failure (April 12, p 1081)1 states that hepatitis-Eassociated fulminant hepatic failure occurs predominantly in pregnant women in endemic countries. This statement is based on reports of fulminant hepatic failure in developing countries where hepatitis E is endemic but laboratory facilities are scant and serological confirmation of the cause of the hepatitis is difficult.2 Tsega and colleagues3 found that among patients with fulminant hepatitis in whom hepatitis E was confirmed there was an over-representation of pregnant women in the third trimester; though this finding could be the effect of their selection of patients. Fulminant hepatic failure has also been described in women who are not pregnant. We have reported two nonpregnant, young women with fulminant hepatic failure and serologically proven hepatitis E.4,5 Both patients had visited rural endemic areas in India. The first patient developed grade III encephalopathy and dialysis-dependent acute renal failure but made a full recovery with supportive therapy.4 The second patient developed grade IV encephalopathy and required emergency liver transplantation.5 The increasing popularity of travel to countries where hepatitis E is endemic could lead to this potentially lethal complication becoming more common, even in those women who are not pregnant. The widely held view that hepatitisE-induced fulminant hepatic failure only occurs in pregnant women is confusing and potentially harmful. If hepatitis E is overlooked as a possible cause of fulminant hepatic failure, acute liver transplantation may be carried out too early in the course of the disease. In developed countries, the prognosis of hepatitis E is similar to that of hepatitis. There may be a more favourable chance of spontaneous recovery despite the development of fulminant hepatic failure, compared with other causes of viral hepatitis, such
289
THE LANCET
as hepatitis C and other non-A, non-B hepatitis. An early decision on liver transplantation in hepatitis-Eassociated fulminant hepatic failure may therefore be unjustified. The question of whether the medication that both our patients were taking (oestrogen-containing oral contraceptives and antimalarials) contributed to the development of fulminant hepatic failure is intriguing but speculative *J G Zijlstra, E B Haagsma, J E Tulleken, T S van der Werf Intensive and Respiratory Care Unit and Division of Gastroenterology and Hepatology, Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, Netherlands
1 2
3
4
5
Mas A, Rodés J. Fulminant hepatic failure. Lancet 1997; 349: 1081–85. Khuroo MS, Teli MR, Skidmore S, Sofi Ma, Khuroo MI. Incidence and severity of viral hepatitis in pregnancy. Am J Med 1981; 70: 252–55. Tsega E, Hansson BG, Krawczynski K, Nordenfelt E. Acute sporadic viral hepatitis in Ethiopia: causes, risk factors, an effect on pregnancy. Clin Infect Dis 1992; 14: 961–65. Verschuuren EAM, Haagsma EB, Zijlstra JG, Stegeman CA. Non-oliguric acute renal failure associated with hepatitis E. Nephrol Dial Transpl 1997; 12: 799–801. Verschuuren EAM, Haagsma EB, Zijlstra JG. Fulminant leverfalen door hepatitis E en de indicatie voor levertransplantatie. Intensive Care Rev 1996; 11: 109–12.
EMEA and third-generation oral contraceptives SIR—Jan P Vandenbroucke and Frits R Rosendaal (April 19, p 1113)1 refer to the cross-sectional study by Rosing et al2 which reported an association between the use of oral contraceptives (OCs) and functional resistance to the natural anticoagulant, activated protein C (APC); they found greater APC resistance among women using “third generation” OCs than among those who used second-generation OCs. According to Vandenbroucke and Rosendaal, those results “fit admirably with the epidemiological data”, showing that the incidence of deepvenous thrombosis and pulmonary embolism (VTE) is about twice as high in users of third-generation OCs as in second-generation OC users. They wonder why the results of the epidemiological studies have apparently been ignored by the European Agency for the Evaluation of Medicinal Products (EMEA); they suggest that EMEA experts have focused on irrelevant arguments made redundant by further analyses; and they seek the involvement of “independent consumer boards or patient platforms in drugsafety regulation” who might be able to resist pressure better.
290
The comparative safety of secondgeneration and third-generation OCs was addressed by the scientific committee of EMEA (the CPMP) in October, 1995, and again, after the publication of the first four epidemiological studies, in December, 1995, and in January, 1996. CPMP set up a working group (including experts in pharmacoepidemiology, clinical pharmacology, obstetrics and gynaecology, cardiology, haemostasis, and biostatistics) which continues to hold regular meetings to analyse all new data. CPMP’s evaluation of the epidemiological evidence was set out on Jan 22, 1997. At that time, seven studies had been presented to CPMP. The risk of VTE was higher in women who used desogestrel or gestodene containing OCs (third generation) than in women using OCs containing levonorgestrel (the majority), lynoestrol, or norethisterone. The increased risk was about two-fold in the first four studies (and significant in three). In the next three studies, the risk difference was less; one was reported as an interim analysis,3 one reached statistical significance,4 and the other did not.5 Because of possible confounding biases, CPMP asked for further data and analyses from the investigators and from the marketing authorisation holders. However, not all the relevant information was available to the investigators so the impact of bias and confounding could not be fully evaluated. Nonetheless, the difference in risk decreased with increasing control for factors that could constitute confounders and introduce bias. The apparent absolute increase in risk of VTE is low even in third-generation OC users, and the case fatality rate is also low. The emergence of a new possible biological explanation (ie, decreased APC sensitivity) in no way permits “immediate separation of epidemiological chaff from wheat”. A mechanism for an association between a medicinal product and a risk strengthens the probability of a causal relation but does not prove it. The new method for measuring thrombin formation needs further validation and a more analytical approach than a cross-sectional study. The risk of VTE is linked to the balance between activation of the coagulation system and the fibrinolytic system, so simultaneous measurement of fibrinolytic activity is needed. One such study planned, at CPMP’s initiative, will include measurement of APC resistance (by the current method and by Rosing’s new method) at the same time in the menstrual cycle. The study will be done by the three manufacturers of third-generation OCs.
An assessment of the risks and benefits of OCs must take into account arterial complications such as myocardial infarction and stroke. Third-generation OCs might have an advantage here. CPMP has asked investigators to pool the WHO and the Transnational Study data to achieve better statistical power. In general, it would be valuable to know more about what risks patients are willing to take to attain certain benefits, once these have been reasonably well established. However, the regulatory process would probably benefit more from such knowledge than from direct involvement of “consumer boards”. CPMP and its ad hoc group will continue to assess the data and will recommend such regulatory measures as may be necessary. *J M Alexandre, K Strandberg *CPMP, EMEA, 7 Westferry Circus, London E14 4HB, UK; and Ad hoc Expert Group on Third Generation Oral Contraceptives and Cardiovascular Risks, CPMP
1
2
3
4
5
Vandenbroucke JP, Rosendaal FR. End of the line for “third-generation-pill” controversy? Lancet 1997; 349: 1113–14. Rosing J, Tans G, Nicolaus GA, et al. Oral contraceptives and venous thrombosis: different sensitivities to activated protein C in women using second- and thirdgeneration oral contraceptives. Br J Haematol 1997; 97: 233–38. Lidegaard O. Oral contraceptives and venous thromboembolism: an epidemiological review. Eur J Contracept Reprod Health Care 1996; 1: 13–20. Farmer RD, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Populationbased study of risk of venous thromboembolism associated with various oral contraceptives. Lancet 1997; 349: 83–88. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995; 15: 195–200.
Home blood-pressure control in Japanese hypertensive population SIR—Giuseppe Mancia and colleagues (Feb 15, p 454)1 suggest that among the Italian hypertensive population, the number of patients with inadequate blood-pressure control is high not only when assessed in the clinic but also when assessed at home. We have investigated blood-pressure control at home among hypertensive patients living in Kahoku, a rural Japanese town.2 Our study comprised 2582 adults (1017 aged 25–64 years, 1565 aged ⭓65 years), which was 52% of the eligible adult population of Kahoku. We defined home blood pressure as the mean value of 20 home measurements
Vol 350 • July 26, 1997
advantages, giving rise to a new model of pharmacoeconomic analysis.5 Michael T Isaac, Maria B Tome Suite 6, Lewisham Hospital, London SE13 6LH, UK
1
2
3
4
5
Perez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97. Bermann RM, Darnell AM, Miller HL, Anand A, Charney DS. Effect of pindolol in hastening response to fluoxetine in the treatment of major depression: a doubleblind placebo-controlled trial. Am J Psych 1997; 154: 37–43. Tome MB, Isaac MT, Harte R, Holland C. Paroxetine and pindolol: a randomised trial of serotonergic autoreceptor blockade in the reduction of antidepressant latency. Int Clin Psychopharmacol 1997; 12: 81–89. Tome M, Cloninger C, Watson J, Isaac M. Serotonergic autoreceptor blockade in the reduction of antidepressant latency personality variables and response to SSRI antidepressant and ‘augmenting agent’. J Affect Dis 1997; 44: 101–09. Tome M, Isaac M. Cost-benefit and costeffectiveness analysis of rapid onset of antidepressant and ‘augmenting agent’. Int J Psychiatry Med (in press).
SIR—Victor Pérez and co-workers1 report that the addition of pindolol to fluoxetine antidepressant treatment increases the efficacy of fluoxetine therapy. A difference in the cumulative percentage of patients with a sustained response was seen by day 9, with a significantly greater proportion of patients in the pindolol group. More surprising, this difference was maintained until the end of the 6-week study, when the percentage of patients with sustained response was 69% in the pindolol group versus 48% in the placebo group. The rationale for the use of 5-HT1A antagonists in depression has been lent support by some experimental evidence, reviewed by Artigas and colleagues.2 Both electrophysiological and microdialysis experiments have suggested that pindolol accelerates the effects of serotonin reuptake inhibitors (SSRIs) by blocking the serotonin negative feedback of 5-HT1A somatodendritic receptors until desensitisation of these receptors takes place. This rationale was clinically illustrated in two open studies that have shown a shorter time to improvement of depressive symptoms by paroxetine associated with pindolol.3,4 Our group has done a randomised, double-blind, placebo-controlled trial, comparing pindolol (15 mg/day) with placebo, in association with paroxetine for the first 3 weeks of a 6-month treatment of major depressive episode (unpublished). Intermediate analysis of results for the first month of the first 100 patients revealed that the number of patients who remitted (17-item
Vol 350 • July 26, 1997
Hamilton score 聿10) was higher in the pindolol group by day 10, but that the difference between the two groups disappeared at day 15 and thereafter. The scores for depression severity scales were lower in the pindolol group at day 5 and day 10, but differences disappeared at day 15 and thereafter. At the end of the first month of treatment, there was no difference in response to paroxetine between the pindolol and placebo groups. These data are compatible with the delay necessary to the 5-HT1A autoreceptor desensitisation. On the other hand, the results presented by Pérez et al cannot be explained only by this hypothesis. If a kinetic interaction is excluded, how can we account for the prolonged higher response rate in the pindolol group without a difference in time-to-onset of clinical improvement when only patients with sustained response are considered? It would be of interest to know if experimental evidence supports such prolonged enhancement of SSRI effects by pindolol and how long this enhancement could be maintained in the treatment of depression. *Régis Bordet, Pierre Thomas, Bernard Dupuis, on behalf of the REEP Pharmacologie Centre, Hospitalier Régional, Faculté de Médecine, 59045 Lille, France
1
2
3
4
Pérez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97. Artigas F, Romero L, de Montigny C, Blier P. Acceleration of the effect of selected antidepressant drugs in major depression by 5-HT1A antagonists. Trends Neurosci 1996; 19: 378–83. Artigas F, Pérez V, Alvarez E. Pindolol induces a rapid improvement of depressed patients treated with serotonin reuptake inhibitors. Arch Gen Psychiatry 1994; 51: 248–51. Blier P, Bergeron R. Effectiveness of pindolol with selected antidepressant drugs in the treatment of major depression. J Clin Psychopharmacol 1995; 15: 217–22.
SIR—In Victor Pérez and colleagues’ study,1 the number of days to reach a sustained response was lower in the fluoxetine and pindolol group than in the fluoxetine and placebo group, but there was no difference when only those patients who responded were in the analysis. Thus, there is a discrepancy about the time-to-onset of clinical improvement. Pérez et al suggest that a longer study period may increase the number of responders and resolve the discrepancy because their study period (6 weeks) is not the maximum time-limit for a response to antidepressants. I believe that another possibility is that pindolol cannot hasten the clinical response to fluoxetine but can convert fluoxetine
non-responders into responders, which might account for the discrepancy. Takeshi Terao Department of Psychiatry, University of Occupational and Environmental Health, Yahatanishi-ku, Kitakyushu 807, Japan
1
Pérez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment. Lancet 1997; 349: 1594–97.
Hepatitis-E-associated fulminant hepatic failure in non-pregnant young women SIR—A Mas and J Rodés’ seminar on fulminant hepatic failure (April 12, p 1081)1 states that hepatitis-Eassociated fulminant hepatic failure occurs predominantly in pregnant women in endemic countries. This statement is based on reports of fulminant hepatic failure in developing countries where hepatitis E is endemic but laboratory facilities are scant and serological confirmation of the cause of the hepatitis is difficult.2 Tsega and colleagues3 found that among patients with fulminant hepatitis in whom hepatitis E was confirmed there was an over-representation of pregnant women in the third trimester; though this finding could be the effect of their selection of patients. Fulminant hepatic failure has also been described in women who are not pregnant. We have reported two nonpregnant, young women with fulminant hepatic failure and serologically proven hepatitis E.4,5 Both patients had visited rural endemic areas in India. The first patient developed grade III encephalopathy and dialysis-dependent acute renal failure but made a full recovery with supportive therapy.4 The second patient developed grade IV encephalopathy and required emergency liver transplantation.5 The increasing popularity of travel to countries where hepatitis E is endemic could lead to this potentially lethal complication becoming more common, even in those women who are not pregnant. The widely held view that hepatitisE-induced fulminant hepatic failure only occurs in pregnant women is confusing and potentially harmful. If hepatitis E is overlooked as a possible cause of fulminant hepatic failure, acute liver transplantation may be carried out too early in the course of the disease. In developed countries, the prognosis of hepatitis E is similar to that of hepatitis. There may be a more favourable chance of spontaneous recovery despite the development of fulminant hepatic failure, compared with other causes of viral hepatitis, such
289
THE LANCET
as hepatitis C and other non-A, non-B hepatitis. An early decision on liver transplantation in hepatitis-Eassociated fulminant hepatic failure may therefore be unjustified. The question of whether the medication that both our patients were taking (oestrogen-containing oral contraceptives and antimalarials) contributed to the development of fulminant hepatic failure is intriguing but speculative *J G Zijlstra, E B Haagsma, J E Tulleken, T S van der Werf Intensive and Respiratory Care Unit and Division of Gastroenterology and Hepatology, Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, Netherlands
1 2
3
4
5
Mas A, Rodés J. Fulminant hepatic failure. Lancet 1997; 349: 1081–85. Khuroo MS, Teli MR, Skidmore S, Sofi Ma, Khuroo MI. Incidence and severity of viral hepatitis in pregnancy. Am J Med 1981; 70: 252–55. Tsega E, Hansson BG, Krawczynski K, Nordenfelt E. Acute sporadic viral hepatitis in Ethiopia: causes, risk factors, an effect on pregnancy. Clin Infect Dis 1992; 14: 961–65. Verschuuren EAM, Haagsma EB, Zijlstra JG, Stegeman CA. Non-oliguric acute renal failure associated with hepatitis E. Nephrol Dial Transpl 1997; 12: 799–801. Verschuuren EAM, Haagsma EB, Zijlstra JG. Fulminant leverfalen door hepatitis E en de indicatie voor levertransplantatie. Intensive Care Rev 1996; 11: 109–12.
EMEA and third-generation oral contraceptives SIR—Jan P Vandenbroucke and Frits R Rosendaal (April 19, p 1113)1 refer to the cross-sectional study by Rosing et al2 which reported an association between the use of oral contraceptives (OCs) and functional resistance to the natural anticoagulant, activated protein C (APC); they found greater APC resistance among women using “third generation” OCs than among those who used second-generation OCs. According to Vandenbroucke and Rosendaal, those results “fit admirably with the epidemiological data”, showing that the incidence of deepvenous thrombosis and pulmonary embolism (VTE) is about twice as high in users of third-generation OCs as in second-generation OC users. They wonder why the results of the epidemiological studies have apparently been ignored by the European Agency for the Evaluation of Medicinal Products (EMEA); they suggest that EMEA experts have focused on irrelevant arguments made redundant by further analyses; and they seek the involvement of “independent consumer boards or patient platforms in drugsafety regulation” who might be able to resist pressure better.
290
The comparative safety of secondgeneration and third-generation OCs was addressed by the scientific committee of EMEA (the CPMP) in October, 1995, and again, after the publication of the first four epidemiological studies, in December, 1995, and in January, 1996. CPMP set up a working group (including experts in pharmacoepidemiology, clinical pharmacology, obstetrics and gynaecology, cardiology, haemostasis, and biostatistics) which continues to hold regular meetings to analyse all new data. CPMP’s evaluation of the epidemiological evidence was set out on Jan 22, 1997. At that time, seven studies had been presented to CPMP. The risk of VTE was higher in women who used desogestrel or gestodene containing OCs (third generation) than in women using OCs containing levonorgestrel (the majority), lynoestrol, or norethisterone. The increased risk was about two-fold in the first four studies (and significant in three). In the next three studies, the risk difference was less; one was reported as an interim analysis,3 one reached statistical significance,4 and the other did not.5 Because of possible confounding biases, CPMP asked for further data and analyses from the investigators and from the marketing authorisation holders. However, not all the relevant information was available to the investigators so the impact of bias and confounding could not be fully evaluated. Nonetheless, the difference in risk decreased with increasing control for factors that could constitute confounders and introduce bias. The apparent absolute increase in risk of VTE is low even in third-generation OC users, and the case fatality rate is also low. The emergence of a new possible biological explanation (ie, decreased APC sensitivity) in no way permits “immediate separation of epidemiological chaff from wheat”. A mechanism for an association between a medicinal product and a risk strengthens the probability of a causal relation but does not prove it. The new method for measuring thrombin formation needs further validation and a more analytical approach than a cross-sectional study. The risk of VTE is linked to the balance between activation of the coagulation system and the fibrinolytic system, so simultaneous measurement of fibrinolytic activity is needed. One such study planned, at CPMP’s initiative, will include measurement of APC resistance (by the current method and by Rosing’s new method) at the same time in the menstrual cycle. The study will be done by the three manufacturers of third-generation OCs.
An assessment of the risks and benefits of OCs must take into account arterial complications such as myocardial infarction and stroke. Third-generation OCs might have an advantage here. CPMP has asked investigators to pool the WHO and the Transnational Study data to achieve better statistical power. In general, it would be valuable to know more about what risks patients are willing to take to attain certain benefits, once these have been reasonably well established. However, the regulatory process would probably benefit more from such knowledge than from direct involvement of “consumer boards”. CPMP and its ad hoc group will continue to assess the data and will recommend such regulatory measures as may be necessary. *J M Alexandre, K Strandberg *CPMP, EMEA, 7 Westferry Circus, London E14 4HB, UK; and Ad hoc Expert Group on Third Generation Oral Contraceptives and Cardiovascular Risks, CPMP
1
2
3
4
5
Vandenbroucke JP, Rosendaal FR. End of the line for “third-generation-pill” controversy? Lancet 1997; 349: 1113–14. Rosing J, Tans G, Nicolaus GA, et al. Oral contraceptives and venous thrombosis: different sensitivities to activated protein C in women using second- and thirdgeneration oral contraceptives. Br J Haematol 1997; 97: 233–38. Lidegaard O. Oral contraceptives and venous thromboembolism: an epidemiological review. Eur J Contracept Reprod Health Care 1996; 1: 13–20. Farmer RD, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Populationbased study of risk of venous thromboembolism associated with various oral contraceptives. Lancet 1997; 349: 83–88. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995; 15: 195–200.
Home blood-pressure control in Japanese hypertensive population SIR—Giuseppe Mancia and colleagues (Feb 15, p 454)1 suggest that among the Italian hypertensive population, the number of patients with inadequate blood-pressure control is high not only when assessed in the clinic but also when assessed at home. We have investigated blood-pressure control at home among hypertensive patients living in Kahoku, a rural Japanese town.2 Our study comprised 2582 adults (1017 aged 25–64 years, 1565 aged ⭓65 years), which was 52% of the eligible adult population of Kahoku. We defined home blood pressure as the mean value of 20 home measurements
Vol 350 • July 26, 1997