DrugScan is a review of the international literature on therapeutic drugs. Leading pharmacy practitioners scan
the literature and present information on major clinical trials, important pharmacoepidemiology studies, pharmacoeconomic research and other related materials in a succinct format. Interested readers are encouraged to
explore the original publications in greater detail.
Developments in Oncology—2002
Michael J Cain
When asked by a lay person what advances have been
made in the treatment of cancer in the last year the reply
could be summarised as ‘we know more about how cancer works, but save for rare examples this has not yet
translated into substantial change’. There is a short list
of cancers against which substantial improvements have
been reported, although this result is still poor if measured against what we would wish for. Some new drugs
have found a place in cancer treatment and some old
agents have been reworked with improved effect, reduced
toxicity, and sometimes both.
If sheer publication volume can be taken as a reasonable measure, the unravelling of cancer cell biology
is the guiding force of anticancer research. The emerging information is a mosaic, quite incomplete and to those
at the patient end of the business looks something like a
partly solved puzzle of uncertain dimension. Nonetheless a decade of basic science is bearing fruit, with imatinib (Glivec) representing the foremost therapeutic
example to date.1-4 Rituximab (Mabthera), likewise a therapy born of clever science, has further cemented a position in therapy. 5,6 Ultimately, the greater benefit of
understanding cancer biology will be the ability to
choose not to treat those individual patients who will
not respond to particular therapies. As yet, practical examples of this approach have not emerged.
The other common feature of cancer-related literature is the apparently boundless reports of phase I and
II studies initiated to determine the next direction in treating some aspect of cancer by applying new agents or
rearranging available therapies. Phase II studies have
particular impact when they present outcomes substantially better than the rest. The reports detailed in this
review on the activity of imatinib in chronic myeloid leukemia (CML) and on the use of weekly cisplatin in relapsed
ovarian cancer are good examples of this type of research.1-4,7
Phase III studies provide a cornerstone of proof but
do not always lend clarity to decision making. The recently reported four-arm study of newer cytotoxics
against non-small cell lung cancer found all of them equally unimpressive.8 Conversely, irinotecan plus cisplatin
Michael J Cain, BPharm, GradDipPharm, FPS, Senior Pharmacist
(Oncology), Sir Charles Gairdner Hospital, Perth, Western Australia
Address for correspondence: Michael Cain, Pharmacy Department, Sir
Charles Gairdner Hospital, Perth WA 6000
E-mail: [email protected]
appears to have a clear benefit over cisplatin plus etoposide against small cell lung cancer.9 Bleomycin, however,
has been proven to be non-contributory in the treatment
of advanced cervical cancer.10
Imatinib—convincing against late chronic phase
Preliminary results from a landmark phase II study reveal an unprecedented level of activity against CML
after the failure of interferon.1 After 18 months of follow-up 40% of patients have become cytogenetically
negative for the Philadelphia chromosome, 89% have
not progressed to accelerated or blast phase, and 95%
are still alive.
CML is a myeloproliferative disease characterised
by three distinct phases: a chronic phase lasting three to
six years, an accelerated phase lasting less than six
months, and finally the more aggressive blast phase.
From a treatment and research perspective CML is unique
in that an oncogene very intimately linked to the genesis
of the disease was identified a long time ago. The Philadelphia chromosome is a fusion gene brought about by
a genetic translocation that places the BCR and ABL
genes together. The BCR-ABL gene product is a tyrosine kinase which is ‘always on’. Imatinib mesylate
(Glivec), previously known as STI-571, is a tyrosine kinase inhibitor specifically designed to inhibit BCR-ABL.
A large multicentre phase II study tested the effectiveness of imatinib in chronic phase CML resistant to
interferon, or where the patient was intolerant of interferon. Strict definitions for chronic phase were set and a
central panel independently confirmed the classification
of each case. Of 532 patients enrolled, 484 met the predetermined criteria. There was no upper age limit but patients were required to have reasonable performance
status as well as good renal, liver and cardiac function.
A daily oral 400 mg dose of imatinib was given and
researchers were given the option of increasing the dose
to 400 mg twice daily if a patient did not respond after
three months. The primary end point for this study was
cytogenetic response, as determined by the degree of
clearance of the Philadelphia chromosome from bone
marrow. Other end points included complete haematological response, time to progression, and survival.
After a median follow up of 18 months the outcomes,
particularly cytogenetic response, were better than that
normally achieved with first-line interferon. Almost all
patients (95%) achieved a complete normalisation of
blood counts, 60% had a major or complete cytogenetic
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
response, 5% had a minor response and 11% had minimal response. Responses were largely durable with an
estimated 89% of patients alive without progressing disease after 18 months. Serious (grade III/IV) haematological adverse effects included anaemia (7%), thrombocytopenia (20%) and neutropenia (35%). Non-haematological side effects (oedema, nausea, muscle cramp, rash)
were common but generally mild. Treatment was discontinued in only 2% of patients because of side effects.
The combination of a long natural history for chronic phase CML and a possible extended benefit from imatinib means that mature results for this study will be a
long time coming. In the meantime the authors have used
cytogenetic response as an indicator of longer-term benefit. With up-front interferon therapy cytogenetic response does correlate with better survival. Is this also
true for imatinib? This remains to be demonstrated. Practically, however, for patients who don’t respond to interferon or who cannot tolerate it there are precious few
alternative treatments. Imatinib in the short term offers
very good activity. Does this indicate that imatinib should
be used as initial therapy when CML presents? The answer for most individuals at this stage is no. Interferon
induction and allogeneic stem cell transplantation can
be offered with some prospect of cure to the 30% of
individuals who have a compatible donor. The definitive
answers will be known once early data from a phase III
comparison of imatinib with interferon plus cytarabine
in early CML sees publication.
Imatinib—useful also for accelerated and blast crisis
Two studies demonstrate substantial activity of imatinib against late stage CML. The outcomes against accelerated and blast phase CML were not as good nor
as persistent as against chronic phase CML but are
nonetheless better than would be expected from the limited alternative treatment options. Results were considerably better when dosage was increased from 400
mg to 600 mg per day. An unexpected feature was a
high incidence of severe myelosuppression.2,3
Accelerated phase CML is associated with raised
blood cell counts, splenomegaly, and constitutional
symptoms as the disease displays an increasing array of
karyotypic and genetic changes. Blast crisis, which is
the extension of this, is usually defined as greater than
30% of cells in bone marrow being the more primitive
blast cells. Blast phase generally follows within six
months of diagnosis of accelerated phase and runs a
course before death that is often measured in weeks.
Treatments available for these patients are limited and
there is a lack of consensus on what standard therapy
might be.
Two recently reported multicentre phase II studies
of similar design have characterised the efficacy of imatinib against accelerated and blast phase CML.2,3 In both
studies a central review panel was used to reassess the
classification of each case against strict criteria, ensuring that the results obtained reliably represented the true
activity of imatinib in these groups. The accelerated
phase study enrolled 235 patients from 18 centres, and
181 were confirmed as eligible for use in the analysis of
imatinib efficacy.2 The blast phase study enrolled 260
patients, 229 were confirmed as eligible. In both studies
a protocol change was introduced to escalate the daily
dose used from 400 mg per day to 600 mg.
In the study of accelerated phase, 62 patients received 400 mg per day of imatinib and 119 were given 600
mg per day. The overall response rates (defined as complete response, marrow response or a return to chronic
phase lasting at least four weeks) were 65% and 71%
respectively. Cytogenetic responses were also more likely
with the higher dose, 28% compared with 16%. After
median follow up of 11 months, the benefits of higher
dose appear more durable. The estimated 12-month response duration exceeds 57% and 79% respectively for
the 400 mg and 600 mg dosages (though at this early
stage the 95% confidence levels overlap).
The study of patients with blast phase CML revealed
an overall response rate of 52%. In only 31% of patients
was there a response that persisted for at least four
weeks. The median survival times for non-responders
and those with only a transient response were 3 months
and 4.7 months respectively, while for those with a durable response median survival was 19 months. Though
only 14% of patients started with the lower 400 mg daily
dose, an interesting dose effect was noted. Amongst
patients started on 400 mg per day, 6% had a major cytogenetic response, while with 600 mg daily the rate was
As in other studies non-haematological side effects
were common but rarely problematic. Haematological
toxicity was significant but manageable for the most part.
Approximately half the treated patients had a reduction
in dose or a delay in therapy at some time. At the time of
publication, 85% of enrolled patients had withdrawn from
imatinib therapy; 58% because of inadequate or lost
effect and 9% because of adverse events.
This was a relatively large, closely regulated phase
II trial that produced results substantially better than
those published elsewhere. There is no standard therapy and as such it is difficult to see at this point how a
phase III study could be contemplated. Consequently
the results published from the MD Anderson Cancer
Centre provide useful additional commentary.4 These
researchers have collated outcome data for 75 patients
treated at their centre as part of three phase I or II studies, and compared this with data from 133 historical controls treated with cytarabine-containing chemotherapy
between 1972 and 2000. Response rates and median survival were almost twice as high with imatinib.
Weekly cisplatin plus daily etoposide—surprising
activity against relapsed ovarian cancer
The report of a phase II study using weekly cisplatin at
50–70 mg/m2 in conjunction with daily oral etoposide
to treat relapsed ovarian cancer reveals a surprisingly
high level of activity. The report is particularly noteworthy because of the high level of activity noted in
patients who had relapsed very quickly after their initial course of chemotherapy—classically this is a group
that responds poorly to second line therapies.7
Cisplatin and its analogue carboplatin have played
a pivotal role in the treatment of epithelial ovarian cancer
for over a decade. Cisplatin-cyclophosphamide combinations were used extensively in the early to mid 1990s
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
as first line therapy until being displaced by cisplatinpaclitaxel or carboplatin-paclitaxel combinations. For
second line therapies, the time between the last treatment and disease recurrence is a major predeterminant
of likely response. Disease that recurs within four months
of platinum-based therapy is considered to have a poor
prognosis while disease that relapses more than 12
months later characteristically responds better to second
line therapies and indeed to retreatment with cisplatin.
Researchers at a Dutch institution have investigated the worth of intensified cisplatin therapy in conjunction with a synergistic agent (oral etoposide) against
relapsed ovarian cancer. Individual doses of cisplatin
cannot be escalated much above 100 mg/m2 without causing substantial renal, otic and neurological toxicity. Normally such doses would be repeated every 21 days.
These researchers have established previously that cisplatin at 50–70 mg/m2 given weekly for six weeks with
oral etoposide 50 mg/m2 on days 1–15 and again on days
29–43 is as tolerable as conventional therapy, and now
report the results of treating 107 patients (98 of whom
were evaluable for response).
The 38 ‘good prognosis’ patients (who had relapsed
more than 12 months after the last of their initial therapy)
were treated with 50 mg/m2 of cisplatin, while the ‘intermediate risk’ patients (n = 32, relapse 4–12 months posttreatment) and ‘poor risk’ patients (n = 28, relapse <4
months post-treatment) received 70 mg/m2. Patients who
had stable disease or maintained a response at the end
of therapy were maintained on 50 mg/m2 of oral etoposide per day for 21 days in each 4-week period, repeated
for 6–9 cycles.
For the good, intermediate and poor risk cohorts
respectively the reported response rates were 92%, 91%,
and 46%. The complete response rates were 63%, 31%
and 29% respectively and median survival 26 months, 16
months and 13 months respectively. Toxicity reported
was modest. Serious leukopenia was noted at some time
in up to 16% of patients, though there was only one case
of neutropenic fever. Substantial thrombocytopenia affected 25% of patients treated with 70 mg/m2 of cisplatin.
There were no grade III/IV non-haematological toxicities. Grade I neurotoxicity was common and persistent.
Renal impairment was mild and apparently only affected
4% of patients.
The response figures are impressive when compared
to outcomes with other second line therapy. For example, the recently published phase III comparison between
topotecan and liposomal doxorubicin reported response
rates of 40% and 47%, complete response rates of 11%
and 9%, and median survivals of 13 and 14 months respectively.11 The response rates in the poor risk group
are particularly notable. For example, in the study just
mentioned, the response rates for topotecan and liposomal doxorubicin in poor risk cases were 8% and 16%
respectively, with 1% in each having a complete response. Conversely, a second set of researchers have
applied the weekly cisplatin-etoposide protocol to 47
patients with similar responses in poor risk patients.12
Irinotecan improves survival with small cell lung
A study comparing an irinotecan-cisplatin combination to conventional cisplatin-etoposide in patients with
advanced small cell lung cancer was terminated early
because of a strikingly better survival rate in the test
Small cell lung cancer (SCLC) behaves and is treated
very differently from the more common non-small cell
lung cancer (NSCLC). It is an aggressive malignancy,
causing death within six to twelve weeks if left untreated. It spreads early and surgery has little if any role. It is,
however, very sensitive (at least initially) to a number of
cytotoxic agents. Complete responses to chemotherapy
are not uncommon, even with advanced disease, and
survival is extended out to a median of 7–11 months and
12–18 months respectively depending on whether disease is advanced or limited. While a variety of cytotoxic
combinations have been applied, improvements in survival have been small.
A Japanese multicentre trial comparing combined
cisplatin-irinotecan to a conventional cisplatin-etoposide protocol in the treatment of advanced SCLC was
terminated early when an interim analysis revealed a statistically significant benefit with the cisplatin-irinotecan
combination. Of the planned 250 patients, 154 were enrolled and randomised using a technique that balanced
the treatment arms for patients’ performance status and
the treating institution.
To be eligible patients needed to have previously
untreated, extensive SCLC, an expected life span of more
than three months, good organ function, age less than
70 years and a reasonable performance status. Planned
treatment consisted of four cycles of chemotherapy.
Subjects in the control arm received cisplatin 80 mg/m2
and etoposide 100 mg/m2 (both on day 1) with etoposide
repeated on the following two days. Cycles were repeated every three weeks. Those in the treatment arm received cisplatin 80 mg/m2 on day 1 and irinotecan 60 mg/
m2 on days 1, 8, and 15—all repeated four-weekly.
The groups were equally balanced for possible confounding factors, with 77 predominantly male patients
with a mean age of 63 years in each group. Grade III or IV
myelosuppression was more common in the cisplatinetoposide arm, but severe diarrhoea affected 16% of the
cisplatin-irinotecan treated group while none of the control group was affected. There were no other significant
differences in toxicity.
The overall response rate was 84% with cisplatinirinotecan treatment and 67% in the control group
(though interestingly there were more complete responses, 9.1% versus 2.6%, in the control arm). The one-year
survival rate was 58% in the cisplatin-irinotecan arm,
compared with 38% in the control arm. The two-year
survival rates were 19% and 5.2%. A quality of life analysis planned for this study was abandoned because of
poor compliance. The median period of time spent without cancer progressing was 6.9 months with cisplatinirinotecan and 4.8 months with the control arm. Six months
after randomisation 65% of irinotecan-treated patients
were free of progression, after one year the figure was
12%. Corresponding values in the control arm were 36%
and 7.9% respectively.
The impetus for this trial came from response rates
seen in earlier phase II trials against untreated and treated SCLC. Similar phase III studies are currently underway and will hopefully confirm the newly demonstrated
benefits. There remain some obvious questions that will
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
need to be answered. Will there be a similar or better
increase in survival with less advanced disease? How
might irinotecan be added to or otherwise incorporated
with the combined radiotherapy-chemotherapy protocols already in use? Once the answers are forthcoming
the place for this new treatment will be much clearer.
Bleomycin adds little in advanced cervical cancer
A study comparing cisplatin and ifosfamide with and
without bleomycin removes one of the uncertainties
about what represents best therapy for advanced cervical carcinoma.10
Carcinoma of the cervix that has progressed past
locoregional management approaches (surgery or combined radiotherapy-chemotherapy) has a bleak prognosis. A variety of cytotoxic agents and combinations have
reported activity in disease control though none has a
demonstrated survival benefit. Median survival in these
circumstances is approximately six months. Published
response rates from small phase II studies have been
50–60% in some studies, and occasionally higher. The
Gynecologic Oncology Group (GOG) has systematically
been assessing the most promising of these candidate
therapies in phase III comparisons. From a previously
reported phase III study a combination of cisplatin and
ifosfamide proved superior.13 With high response rates
reported with the addition of bleomycin to cisplatin-ifosfamide,14,15 the GOG initiated a recent study.
The trial enrolled 303 women (287 evaluable cases)
using a design powered to detect a 15% or greater response rate with the bleomycin-cisplatin-ifosfamide arm.
Women with stage IVB, recurrent, or persistent squamous cell carcinoma of the cervix but not suitable for
surgery or other definitive local therapy were eligible.
Patients had at least a moderate performance status and
adequate renal, hepatic, lung and bone marrow function.
All patients received cisplatin 50 mg/m2 with ifosfamide
5 g/m2 over 24 hours starting the same day. Those randomised to the test arm had, preceding cisplatin-ifosfamide, a 24-hour infusion of bleomycin 30 000 international
units. Cycles were repeated every three weeks up to a
maximum of six. All ifosfamide administration was supported with prophylactic mesna.
There was no meaningful difference in disease control between the groups, with a response rate of 32% in
each arm. Median duration of response was 5.8 months
with bleomycin and 6.2 months without. Median survival was 8.5 and 8.4 months respectively. Additional toxicity amongst bleomycin-treated patients was minimal.
There was a 13% increase in lung toxicity with bleomycin, though this represented largely asymptomatic
changes detected with lung function testing.
The GOG is systematically bringing order to the treatment options proposed for advanced cervical cancers.
Another GOG study, GOG-169, comparing cisplatin to
cisplatin-paclitaxel, closed in 1999. Preliminary results
have been released in abstract form only at this point.
The cisplatin-paclitaxel combination had the superior
response rate but again with no improvement in survival.16 Other comparative trials are in progress.
1. Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C, GambacortiPasserini C, et al. Hematologic and cytogenetic responses to imatinib mesylate
in chronic myelogenous leukemia. N Engl J Med 2002; 346: 645-52.
2. Talpaz M, Silver RT, Druker BJ, Goldman JM, Gambacorti-Passerini C, Guilhot F, et al. Imatinib induces durable hematologic and cytogenetic responses in
patients with accelerated phase chronic myeloid leukemia: results of a phase 2
study. Blood 2002; 99: 1928-37.
3. Sawyers CL, Hochhaus A, Feldman E, Goldman JM, Miller CB, Ottmann OG,
et al. Imatinib induces hematologic and cytogenetic responses in patients with
chronic myelogenous leukemia in myeloid blast crisis: results of a phase II
study. Blood 2002; 99: 3530-9.
4. Kantarjian HM, Cortes J, O’Brien S, Giles FJ, Albitar M, Rios MB, et al.
Imatinib mesylate (STI571) therapy for Philadelphia chromosome-positive
chronic myelogenous leukemia in blast phase. Blood 2002; 99: 3547-53.
5. Davis TA, Grillo-Lopez AJ, White CA, McLaughlin P, Czuczman MS, Link
BK, et al. Rituximab anti-CD20 monoclonal antibody therapy in non-Hodgkin’s lymphoma: safety and efficacy of re-treatment. J Clin Oncol 2000; 18:
6. Coiffier B, Lepage E, Briere J, Herbrecht R, Tilly H, Bouabdallah R, et al.
CHOP chemotherapy plus rituximab compared with CHOP alone in elderly
patients with diffuse large-B-cell lymphoma. N Engl J Med 2002; 346: 235-42.
7. van der Burg ME, de Wit R, van Putten WL, Logmans A, Kruit WH, Stoter
G, et al. Weekly cisplatin and daily oral etoposide is highly effective in platinum pretreated ovarian cancer. Br J Cancer 2002; 86: 19-25.
8. Schiller JH, Harrington D, Belani CP, Langer C, Sandler A, Krook J, et al.
Comparison of four chemotherapy regimens for advanced non-small-cell lung
cancer. N Engl J Med 2002; 346: 92-8.
9. Noda K, Nishiwaki Y, Kawahara M, Negoro S, Sugiura T, Yokoyama A, et al.
Irinotecan plus cisplatin compared with etoposide plus cisplatin for extensive
small-cell lung cancer. N Engl J Med 2002; 346: 85-91.
10. Bloss JD, Blessing JA, Behrens BC, Mannel RS, Rader JS, Sood AK, et al.
Randomized trial of cisplatin and ifosfamide with or without bleomycin in squamous carcinoma of the cervix: a gynecologic oncology group study. J Clin
Oncol 2002; 20: 1832-7.
11. Gordon AN, Fleagle JT, Guthrie D, Parkin DE, Gore ME, Lacave AJ. Recurrent epithelial ovarian carcinoma: a randomized phase III study of pegylated
liposomal doxorubicin versus topotecan. J Clin Oncol 2001; 19: 3312-22.
12. Meyer T, Nelstrop AE, Mahmoudi M, Rustin GJ. Weekly cisplatin and oral
etoposide as treatment for relapsed epithelial ovarian cancer. Ann Oncol 2001;
12: 1705-9.
13. Omura GA, Blessing JA, Vaccarello L, Berman ML, Clarke-Pearson DL,
Mutch DG, et al. Randomized trial of cisplatin versus cisplatin plus mitolactol
versus cisplatin plus ifosfamide in advanced squamous carcinoma of the cervix:
a Gynecologic Oncology Group study. J Clin Oncol 1997; 15: 165-71.
14. Murad AM, Triginelli SA, Ribalta JC. Phase II trial of bleomycin, ifosfamide,
and carboplatin in metastatic cervical cancer. J Clin Oncol 1994; 12: 55-9.
15. Tay SK, Lai FM, Soh LT, Ho TH, Ang PT, Au E. Combined chemotherapy
using cisplatin, ifosfamide and bleomycin (PIB) in the treatment of advanced
and recurrent cervical carcinoma. Aust N Z J Obstet Gynaecol 1992; 32: 2636.
16. Moore DH, McQuellon RP, Blessing JA. A randomized phase III study of
cisplatin versus cisplatin plus paclitaxel in stage IVB, recurrent or persistent
squamous cell carcinoma of the cervix [abstract]. Proc Soc Gynecol Oncol 1991;
32: 21.
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
Developments in Rheumatology–2002
David G Cosh
This review offers the reader three suggestions for consideration, all of which may have relevance to pharmacists who wish to work in the area of clinical therapeutics.
Musculoskeletal disease is the template upon which the
suggestions are offered, but they have equal relevance
to other areas of therapeutics.
The first suggestion is not to forget the past in a
rush to embrace the new. The example chosen to support this exhortation is the new evidence for prednisolone in rheumatoid arthritis that adds to the clinical story
made possible by Philip Hench and those who shared
the Nobel Prize with him 52 years ago.1
The second message is to never dismiss out of hand
an improbable association between two events. Australian evidence of the importance of this axiom includes the
acceptance, albeit unduly delayed by many, of the relationship between peptic ulcer disease and Helicobacter
pylori. Warren and Marshall did not dismiss what might
have first appeared to be an improbable association, and
their persistence has resulted in a paradigm shift in the
treatment of peptic ulcer disease. With this in mind, reviewed here is the possible relationship between the
COX-2 specific inhibitors and risk of cardiovascular disease. It so happens that in this case the ‘risk’ referred to
is cardiovascular but the implicated drugs fit comfortably within the field of musculoskeletal medicine. With
this example, the reverse to the ulcer story applies, in
that there appears there is not an association when one
was first thought to exist, although it remains early days.
The last suggestion is to resist the temptation to
become overly excited with the prospect that evidencebased medicine will convert the practice of therapeutics
into a set of tidy algorithms, allowing one to simply follow the dotted lines and—Hey Presto—arrive at the right
answer! Wouldn’t domiciliary medication management
and residential care medication reviews be so easy if this
were to be the case? Unfortunately a book of algorithms
will not act as a substitute for critical thought based on
a sound knowledge of drug and disease. To illustrate
this point this review examines an intriguing scenario
involving a group of Australian experts who were asked
to arrive at a position statement on the clinical use of the
COX-2 specific inhibitors. The outcome of their endeavours would have been of interest, but not as interesting,
if the group had not agreed to publish the reasons why
all could not agree on a joint position statement. The
fact that they have aired their differences in the pages of
the Medical Journal of Australia provides a very useful
and cogent insight into how we all are likely to find ourselves dealing with the evidence surrounding drug use.
David G Cosh, MPharm, FSHP, FPS, BCPS, Clinical Pharmacist, Private
Practice, Netherby, South Australia
Address for correspondence: David Cosh, 9 Smith-Dorrien St, Netherby
SA 5062, E-mail: [email protected]
Low dose oral steroids are disease-modifying in
rheumatoid arthritis
The risks associated with the use of oral corticosteroids are well known and a decision to use these drugs
on a continuing basis is not usually taken lightly. The
emphasis in treating inflammatory rheumatoid arthritis has been to use one of the so-called disease modifying antirheumatic drugs (DMARDs), sometimes in
combination with each other and often with non-steroidal anti-inflammatory drugs (NSAIDs) and or corticosteroids (regular or intermittent). The question of
whether oral prednisolone is a DMARD has been asked
before, but results from studies designed to answer the
question have been confounded by concurrent use of
DMARDs, short study duration and methodological
shortcomings. The study by van Everdingen et al. is
important because it demonstrates a disease modifying
role for low-dose prednisone given on a regular basis,
and quantifies some of the risks associated with the use
of steroids in this clinical setting.2
From October 1992 until October 1995, 81 eligible
outpatients with evidence of untreated inflammatory
rheumatoid arthritis and naive to any DMARDs were
consecutively selected from those attending the outpatient clinic of two tertiary referral centres in the Netherlands. Those selected needed to have evidence of
disease activity for at least one year, including symptoms and objective markers of inflammation.
A history of an inability to tolerate prednisone or
NSAIDs led to exclusion, as did a history of active gastrointestinal (GI) problems, severe hypertension, the presence of a bleeding diathesis, current treatment with
immunosuppressive drugs, a history of mental illness or
drug or alcohol abuse.
In double blind fashion, 41 patients received 10 mg
of oral prednisone daily, and 40 were given placebo. Randomisation was in blocks of 10 and the planned duration
of the study was two years. Both groups were allowed
to take NSAIDs of their choice. After six months, sulfasalazine (up to 2000 mg daily) and intra-articular corticosteroids were allowed at the discretion of the treating
rheumatologist. Each patient took 500 mg of elemental
calcium at night and all patients recorded their use of
NSAIDs, simple analgesics and physiotherapy services.
The groups were well matched at baseline; 50 patients (24 in the prednisone group and 26 in the placebo
group) had non-erosive disease, and the mean total score
for the chosen radiologic outcome measure (the combination of the scores for joint space narrowing and erosions) was slightly higher for the placebo group than for
the prednisone group, although the difference was not
statistically significant. Disability was assessed using
the Health Assessment Questionnaire (HAQ), and radiological assessment of joint destruction (the 44 joints in
hands and feet were used for comparison) was performed
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
at six-monthly intervals.
Adverse effects were documented at baseline and
at three-monthly intervals. Infections (both site and incidence) were recorded and a post-prandial serum glucose of 11 mmol/L or above was defined as the cut-off
point for a diagnosis of hyperglycaemia. GI bleeding and
the presence of peptic ulcers were monitored for, as were
the standard blood tests for biochemical and haematological abnormalities. Body weight was recorded
throughout the study and neuropsychiatric status was
evaluated. Five radiographs of the spine were taken
throughout the study (baseline and six-monthly thereafter) to assess for the presence of osteoporosis.
There was a significant increase in weight of the
order of 3 kg in the prednisone group compared with no
increase in those receiving placebo. Serum glucose increased to the point where hyperglycaemia occurred in
two patients on steroid and one on placebo. Five patients in the prednisone group developed new vertebral
fractures compared with two in the placebo group, this
difference not reaching significance. There was no significant difference in the rate or type of infections, GI
disorders or the development of hypertension.
Ten patients (four treated with prednisone and six
receiving placebo) withdrew from the study but their
last measurements were carried forward with the exception of the radiographic score. Mean differences in changes from baseline between the two groups for radiographic
scores, disability and grip strength were tested after 24
months with two-sided t-tests or the Mann-Whitney Utest, where appropriate. None of the 10 patients who
withdrew from the study did so because of adverse effects associated with the study medication. Of the 71
who completed the study, 39 (20 receiving placebo and
19 receiving prednisone) required supplemental sulphasalazine. After 24 months, 30% of patients in the placebo group and 22% in the prednisone group still had
non-erosive disease.
For most clinical variables, the changes from baseline favoured the prednisone group after 12 and 24
months, but did not differ significantly between the two
groups. Exceptions were grip strength and the 28-joint
score for tenderness, where a larger, statistically significant improvement was seen in the prednisone group
compared with the placebo group. A statistically significant interaction of time and medication was seen for
three clinical variables: pain, 28-joint score for tenderness, and grip strength. This was due to more rapid improvement during the first six months in the prednisone
group than in the placebo group. Individual patient improvement was 33% in the prednisone group and 24% in
the placebo group after 12 months, and 30% and 22%
respectively after 24 months.
There was a marked difference in the need for intraarticular steroid injections, so that after 24 months the
total number of injections given to the prednisone group
was 40% lower than the number needed by those taking
placebo. A reduction of similar magnitude (49%) in the
number of paracetamol tablets ingested was seen. There
was no statistical difference between the two groups in
radiographic evidence of joint erosions at baseline, but
after 12 months radiographs showed significantly less
joint erosions in the prednisone group.
It is now known that not to use DMARDs in patients with inflammatory arthritis results in a greater rate
of disease progression, and both the authors and Pincus et al. in an associated editorial agree that to seek to
perform this study in 2002 would be unethical because
of the risk to those allocated placebo.3 Nevertheless, the
results do provide important information that would not
have been possible to obtain without a placebo group in
place. Prednisone at a daily dose of 10 mg does have
some disease modifying activity and importantly this
was demonstrated in a group that had not been given
DMARDs prior to entering the study. This benefit is
parallelled by objective symptomatic improvement. The
authors advocate combination of oral corticosteroids
with other DMARDs and a more aggressive prophylaxis
of osteoporosis than they provided to those enrolled in
their study.
There are many patients in Australia, especially
amongst the elderly, whose rheumatoid arthritis is managed with low doses of prednisolone without the concurrent use of other DMARDs. There may well be good
reasons for doing so, such as comfort care or intolerance of DMARDs, but for some prescribers the decision
not to use DMARDs will have been taken on the basis
of regimen simplicity, a fear of side effects, or a reluctance to embark on a course of treatment that requires
constant monitoring. Any reservations about denying
disease modification to this group appears to have been
somewhat exaggerated (although there are obviously
different levels of disease modification, hence van Everdingen’s advice to combine agents). There is now good
evidence that this approach does provide some disease
modification and this should reawaken interest in the
use of low doses of prednisone or prednisolone in younger patients with evidence of inflammatory disease in an
era when aggressive pharmacological protection against
osteoporosis is available.
Evidence to support pro-thrombotic effect of celecoxib
is lacking
Data from the celecoxib long-term arthritis safety study
(CLASS study) have been examined to test the hypothesis that celecoxib is more likely to be associated with
thromboembolic cardiovascular events than the conventional comparator NSAIDs ibuprofen and diclofenac.
The authors conclude that the study provides no evidence to support the hypothesis.4,5
Subjects with a diagnosis of rheumatoid or osteoarthritis that had been evident for at least three months
and who were aged 18 years or older and considered to
need continuous ongoing treatment with an NSAID were
randomised in a double-blind fashion to receive celecoxib 400 mg twice daily, (celecoxib group—CG) ibuprofen
800 mg three times daily (IG) or diclofenac 75 mg twice
daily (DG). Three hundred and eighty-six centres in the
USA and Canada participated in the study, which began
recruiting patients in December 1998. Patients were reviewed at baseline and then again 4, 13, and 26 weeks
later, and thereafter at 13-week intervals. Cardiovascular
protective doses of aspirin up to 325 mg per day were
permitted. At each visit subjects were asked to describe
any untoward event not associated with their arthritis,
and cardiovascular events were categorised into cardiac
events, cerebrovascular events and peripheral vascular
events. This population is that enrolled in the CLASS
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
study, which sought to evaluate the comparative gastrointestinal toxicity of celecoxib compared with two conventional NSAIDs.4
The average age of subjects in the study population
was 60 years. The CG contained 3987 subjects and 3981
subjects received either ibuprofen or diclofenac. Cardiovascular prophylactic doses of aspirin were taken by
22% of all patients and the average blood pressure (again
for all patients) was 133/80 mm Hg. Forty per cent had a
history of cardiovascular disease, the most common being hypertension. All patients who received at least one
dose of study medication were included in the analysis.
Baseline characteristics, aspirin use and risk factors for
a thromboembolic event were compared using 2-way
ANOVA or Pearson’s chi-square test. Crude event rates
and time to event analysis were conducted. Given that
the CLASS study was powered for a total of 40 GI ulcer
complications it was that end point that determined study
cessation and the oversight committee terminated the
study when 38 events had been obtained.
There was no difference in the incidence of a cardiovascular events requiring hospitalisation between the
groups. The relative risk of an event in those taking
celecoxib versus ibuprofen or diclofenac was 1.1 (95%
CI 0.7–1.6). No difference was observed when the events
were divided into the three aforementioned categories
(cardiac, cerebral or peripheral). Subgroup analysis of
those patients not taking aspirin showed a lower rate of
events than would be expected of a group with no known
reason to be on low-dose aspirin. Again, no differences
in event rates were found and the relative risk of a cardiovascular event in those taking celecoxib was 1.1 (95%
CI 0.6–1.9). There was no difference in ‘time to an event’
amongst the groups, with a Kaplan-Meier rate for a combination of the 3 categories of events being CG 1.4%, DG
1.6% and IG 0.7% respectively. There was a high-risk
subgroup of 371 patients who were not taking aspirin
but had an indication to do so (secondary prevention,
post-myocardial infarction). With myocardial infarction
as an end point there was no significant difference in
event rate between the groups.
This study also sought to establish the rate of potential risk factors for cardiovascular events—hypertension, oedema and congestive cardiac failure. A lower rate
of oedema (peripheral and generalised) was seen in the
CG group and there was less hypertension in the CG
group compared with the IG group, but no difference
between the CG and the combined NSAID group.
It was as a result of the analysis of the VIGOR study
where the potential for upper gastrointestinal toxicity of
rofecoxib was compared with naproxen that the observation of less cardiovascular events in the naproxen
group was made.6 This prompted the study described
above and there have been others. Shapiro and colleagues from Merck Laboratories have reviewed eight
phase II and IIIb trials of rofecoxib in the treatment of
osteoarthritis and have not found an association between cardiovascular events and use of rofecoxib.7 Fleming has commented on the propensity of two COX-2
specific inhibitors on the Australian market to cause
hypertension, and argues for a dose-dependent effect
of rofecoxib, but adds a note of caution on the practice
of dredging data for answers to questions that studies
were not designed to answer.8 The debate will continue
and the following publication advises us to accept the
postulate that all COX-2 specific inhibitors can cause
fluid and salt retention, which in turn can increase blood
pressure and worsen cardiac failure. This is eminently
sensible advice.
Position statements are not necessarily easy even in
the era of evidence-based medicine
An Australian working group comprising 31 clinicians
and others has produced a position statement on the
safe prescribing and use of COX-2 specific inhibitors.9
The findings are of interest but some will undoubtedly
be modified as further experience with the class accrues. What may better stand the test of time are the
reasons given for the failure of the group to arrive at
consensus. Most of us are likely to be able to agree on
the quality of evidence on offer; however, to agree on
the interpretation and description of such evidence may
be more difficult, as Edmonds et al. show in an editorial
in the same edition of the MJA.10 The factors at play are
applicable to any area of therapeutics and suggest that
debates over therapeutic options will become no less
interesting despite the increasing sophistication with
which evidence for or against treatment modalities is
made available to those needing to make prescribing
Revisiting White et al.’s group (see above) who arrived at the conclusion that celecoxib, in doses higher
than those usually used in practice, poses no greater
cardiovascular risk than ibuprofen or diclofenac (the
doses of these drugs were a ‘healthy’ 2400 mg and 150
mg daily respectively), we should note that the Australian prescribing group (APG) state that ‘whether any
COX-2 specific inhibitor poses a risk to cardiovascular
safety remains a subject to debate’. They did restrict
their search to papers published no later than May 2001;
however, there is no reason to assume that they would
change their position if given the opportunity to review
more recent evidence, a point that becomes obvious
when one reads the editorial by Edmonds et al.10
The APG comprised rheumatologists on the medical
advisory boards of Pharmacia/Pfizer and Merck, Sharp
& Dohme, and experts in the fields of general practice,
clinical epidemiology, gastroenterology, cardiology, nephrology and clinical pharmacology. Representatives
from the Arthritis Foundation of Australia, the Australian Rheumatology Association and the National Prescribing Service were included. The group has published
guidelines that have majority but not unanimous support, and in doing so took the unusual step of giving all
members of the group the opportunity to:
• sign off on the final draft;
• decline to sign off and remain anonymous;
• decline to sign off and add their name without expressing a reason;
• decline to sign off and add both their name and
Those who did not agree with the group’s findings
all exercised the final option.
The final position statement addresses a range of
issues. On the matter of efficacy equivalence of the COX2 specific inhibitors with conventional NSAIDs, the group
found in favour of equivalence. When asked to decide
whether the COX-2 specific inhibitors are disease modi-
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
fying in any of the musculoskeletal diseases for which
they have been used, the answer was no. In regard to
comparative gastrointestinal toxicity, the APG opinion is
that when compared with the NSAIDs the COX-2 specific inhibitors are associated with considerably fewer peptic ulcers, slightly fewer upper GI symptoms and fewer
serious upper GI complications, notably bleeding.
With respect to effects on renal function and blood
pressure, equivalent effect is suggested. As mentioned
previously, the matter of cardiovascular risk remains open
to debate. Prescribers are advised to take into account
comorbidities and co-prescribed drugs when prescribing drugs from either class, and the group advocates
regular patient review irrespective of which class is used.
In arriving at the position statement, eight of the 31
members of the group decided to not sign off on the
statement and as stated above accepted the option that
allowed them to publish the reasons for their decision.
Four of the eight worked for Pfizer/ Pharmacia, and the
others were rheumatologists (two with expertise in epidemiology). The problem in reaching consensus developed because of differing opinions on how to both
interpret the evidence, and to express in writing the interpretations. The reader should be cautioned against
jumping to the conclusion that any bias (if any existed)
of those working for the industry is obvious compared
with biases that others might hold. Edmonds et al. remind us that there are equally relevant biases associated
with knowledge, expertise and affiliations outside those
deemed ‘commercial’. Those with an interest in either of
the drugs in question, or the broader question of how
we use evidence in our daily practice may find both articles of interest.
1. Shampo MA, Kyle R. Philp S. Hench-1950 Nobel Laureate. Mayo Clin Proc
2001; 76: 1073-5.
2. van Everdingen AA, Jacobs JWG, Siewertsz van Reesema DR, Bijlsma JWJ.
Low-dose prednisone therapy for patients with early active rheumatoid arthritis: clinical efficacy, disease-modifying properties and side effects. A randomised
double-blind placebo controlled clinical trial. Ann Intern Med 2002; 136: 112.
3. Pincus T, Sokka T, Stein CM. Are long-term very low doses of prednisone for
patients with rheumatoid arthritis as helpful as high doses are harmful? [editorial]. Ann Intern Med 2002; 136: 76-8.
4. Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, et al.
Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory
drugs for osteoarthritis and rheumatoid arthritis - The CLASS study: a randomized controlled trial. JAMA 2000; 284: 1247-55.
5. White WB, Faich G, Whelton A, Maurath MS, Ridge N, Nancy J, et al. Comparison of thromboembolic events in patients treated with celecoxib, a cyclooxygenase-2 specific inhibitor, versus ibuprofen or diclofenac. Am J Cardiol 2002;
89: 425-30.
6. Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis B, et al.
Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000; 343: 1520-8.
7. Reicin AS, Shapiro D, Sperling RS, Barr E, Yu Q. Comparison of cardiovascular thrombotic events in patients with osteoarthritis treated with rofecoxib
versus nonselective non-steroidal anti-inflammatory drugs (ibuprofen, diclofenac, and nabumetone) Am J Cardiol 2002; 89: 204-9.
8. Fleming M. Cardiovascular events and COX-2 inhibitors. JAMA 2001; 286:
9. The Australian COX-2-specific inhibitor (CSI) prescribing group. Considerations for the safe prescribing and use of COX-2-specific inhibitors. Med J Aust
2002; 176: 328-31.
10. Edmonds JP, Day RO, Bertouch JV. The road to consensus: considerations
for the safe use and prescribing of COX-2-specific inhibitors [editorial]. Med J
Aust 2002; 176: 332-4.
Breaking News
In this section, we present concise overviews of important clinical trials and observational studies, new insights into the safety of
drug therapies, highlights in pharmacoepidemiology and pharmacoeconomics, details of particularly noteworthy case reports,
and other brief items of appeal to healthcare practitioners with an interest in safe and effective drug treatment.
Psychiatry—Chris Alderman
Over seven million people use cannabis products weekly or more often in the United States alone, and with
interest in marijuana regulations running high, researchers have set out to ascertain the answers to fundamental
questions that should inform the debates that are ongoing. In particular, recent research has been designed to
determine whether frequent, heavy or prolonged use of
cannabis products leads to a deterioration in cognitive
function that persists beyond acute intoxication. A threeyear study in the USA compared 102 near-daily cannabis users (51 long-term users: mean duration of use 23.9
years; 51 shorter-term users: mean duration 10.2 years of
use) with 33 controls who were non-users. Measures
from standard neuropsychological tests assessed attention, memory, and executive functioning, and were administered prior to entry to a treatment program and
following a median 17-hour abstinence. Long-term users
performed significantly less well than shorter-term users
and controls on tests of memory and attention, but there
was no difference between shorter-term users and controls. Long-term users also demonstrated impaired learning, retention, and retrieval, and both groups of cannabis
users performed poorly on a time estimation task. Performance measures often correlated significantly with
the duration of cannabis use, being worse with increasing years of use. (Solowij N et al. JAMA 2002; 287: 112331.)
A recent consensus conference on the use of placebo in
mood disorder studies included expert presentations
on bioethics, biostatistics, unipolar depression, and bipolar disorder. There was consensus that placebo controls should be used in mood disorder studies. The group
considered that a finding of equivalence between a new
drug and standard treatment in an active control study
is not evidence of efficacy unless the new drug is also
significantly more effective than placebo. It was stated
that mood disorders in elderly and paediatric patients
are understudied, and properly designed trials were
thought to be urgently needed. The group identified
the need to limit risks of medication-free intervals and
to facilitate post-study treatment, and suggested that
patients must be fully informed about the risks involved
in research. This consensus statement provides useful
guidance for institutional ethics committees that must
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
grapple with this contentious issue. (Charney DS et al.
Arch Gen Psychiatry 2002; 59: 262-70.)
British researchers have recently explored the putative
relationship between measles, mumps, and rubella
(MMR) vaccination and bowel problems and developmental regression in children with autism, looking for
evidence of a ‘new variant’ form of autism. 278 children
with core autism and 195 with atypical autism, born between 1979 and 1998, were the subjects. The proportion
of children with developmental regression (25% overall)
or bowel symptoms (17%) did not change significantly
during the 20 years from 1979, a period which included
the introduction of MMR vaccination in October 1988. A
possible association between non-specific bowel problems and regression in children with autism was seen
but this was unrelated to MMR vaccination. This comforting research provides no support for an MMR-associated ‘new variant’ form of autism with developmental
regression and bowel problems, and further evidence
against involvement of MMR vaccine in the aetiology
of autism. (Taylor B et al. BMJ 2002; 324; 393-6.)
A recent six-week double-blind, randomised, placebocontrolled trial examined the efficacy of combined therapy with olanzapine and either valproate or lithium,
compared with valproate or lithium alone in treating
acute manic or mixed bipolar episodes. Olanzapine (520 mg daily) or placebo was added to ongoing moodstabiliser therapy. Concurrent treatment with olanzapine produced significantly more improvement than
placebo and clinical response rates were significantly
higher with the combination therapy. Extrapyramidal
symptoms were not significantly changed from baseline to end point in either treatment group, but treatment-emergent symptoms that were significantly higher
for the olanzapine group included somnolence, dry
mouth, weight gain, increased appetite, tremor, and
slurred speech. Researchers conclude that compared
with the use of valproate or lithium alone, the addition
of olanzapine provided superior efficacy in the treatment of manic and mixed bipolar episodes. (Tohen M et
al. Arch Gen Psychiatry 2002; 59: 62-9.)
Pharmacy Practice—Sandra Mangion
A recent review of literature dealing with medication errors provides an insightful discussion on system errors,
from latent conditions in the system that make an error
more likely to active failures where the system breaks
down. The review is peppered with discussions on error
analysis in the aviation and nuclear industries, and the
need for a ‘systems approach’ when reviewing medication incidents is emphasised. The authors have comprehensively discussed the literature on medication errors
in relation to medical specialty, stage of drug delivery
process, drug class, time of day, staff experience and
other factors. There was a particular focus on transfusion medicine due to the often-complex drug protocols
and regimens involved. This review also critiques medication error studies in general and highlights the essential elements that are required to allow these to provide
useful information in establishing cause and determining solutions. The authors point out the limited value of
error counting and instead highlight the need to trace
errors through the system and to understand the circumstances under which error recovery takes place. (Allard J et al. Br J Haematol 2002; 116: 255-65.)
Although medication errors in hospitals are a significant problem, less than 20% of these errors in US hospitals are related to distribution or dispensing problems.
A large study evaluating medication errors and their
relationship to clinical pharmacy services and hospital staffing in US hospitals sought to identify the particular aspects of pharmaceutical services to hospital
that were associated with increased or decreased medication errors. The researchers found that increased
medication errors that adversely affected patient care
outcomes were associated with increased pharmacy
administrator staffing per occupied bed and increased
dispensing pharmacist staffing per occupied bed. On
the other hand, the two most important factors for reducing medication errors were pharmacist-conducted
drug histories and increased staffing levels for clinical
pharmacists. Other factors associated with decreased
medication errors included drug information services,
clinical research, adverse drug reaction management,
drug therapy monitoring, drug protocol management,
drug counselling, and medical rounds participation.
In a current climate of pharmacist shortages perhaps
the results of this study can help guide decisions on
where best to invest pharmacist services. (Bond CA et
al. Pharmacotherapy 2002; 22: 134-47.)
The Audit Commission for the National Health Service
(NHS) in England and Wales published its review of
medicines management in NHS hospitals, entitled ‘A
spoonful of sugar’, in December 2001. The two aims of
this report were to raise the profile of medicines management in hospitals and to make a case for providing adequate investment to enable standards to be raised. A
somewhat scathing assessment of the report suggests
that the Audit Commission’s conclusions are based on
uncertain analysis of unreliable data and that their advocacy of clinical pharmacy is not enough to remedy the
problems they have identified relating to medicines management in hospitals. The author purports that although
good clinical pharmacy services are essential, so are other
elements such as doctors skilled in specific aspects of
the therapeutic process. His commentary makes for some
thought-provoking reading. ‘A spoonful of sugar’ can
be accessed via the Internet at http://www.auditcommission.gov.uk/publications/pdf/nrspoonfulsugar.
pdf. (Ferner RE. QJM 2002; 95: 181-4.)
Pharmacists have been able to substitute brand name
medications with less expensive, generic bioequivalent
medications for some time now. Whilst many patients
value the savings generic substitution provides, pharmacists need to be cognisant that dispensing errors are
less likely to be noticed by patients as they may assume
that the product they don’t recognise as their regular
medication is a generic equivalent. In addition, pharmacists may change the generic brands they are stocking and different pharmacies may stock different generic
brands, therefore patients may receive a variety of generically substituted medications. A recent letter highlighting this problem describes the ordeal of a patient
who was dispensed glibenclamide instead of metopro-
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
lol. She was admitted to two hospitals with hypoglycaemia, subjected to numerous investigations and escaped unnecessary surgery only through some clever
clinical detective work. A warning to all who dispense
to be vigilant with generic substitution and to ensure
that the patient is adequately counselled! (Barzel US.
Mayo Clin Proc 2002; 77: 296-7.)
attributed to the use of NSAIDs and the mortality figures for UGIB in North Jutland are quoted as 8–14%.
The findings are consistent with those from similar cohort studies conducted elsewhere and demonstrate the
value of comprehensive databases constructed within
publicly funded healthcare systems. (Mellemkjær L et
al. Br J Clin Pharmacol 2002; 53: 173-81.)
Gastroenterology—Dave Cosh
Investigators at the Prince of Wales hospital in Hong
Kong recently studied 100 patients proven to be Helicobacter pylori positive and deemed to need long-term
non-steroidal anti-inflammatory drug (NSAID) treatment.
In 51 cases the patients were randomly assigned to a H.
pylori eradication treatment group (7 days of treatment
with omeprazole, amoxycillin and clarithromycin). Fortynine patients received omeprazole and placebo instead
of antibiotics. All patients were then given 100 mg of
slow release diclofenac daily and after six months an
endoscopy was performed (earlier if clinically indicated). The groups were well matched; nine subjects in the
eradication group were taking low dose aspirin compared
with eight in the placebo group, and the number of subjects with a history of previous ulcer disease were seven
and nine in the respective cohorts. Eradication was
achieved in 90% of the antibiotic group and in 6% of the
placebo group. Ulcers were detected in five of those in
the eradication group and in 15 of those receiving placebo. Two of the patients with ulcers in the eradication
group had symptoms but none bled. Ulcers were larger
in the placebo group—nine patients were symptomatic
and three required hospitalisation for gastrointestinal
bleeding. The authors suggest that screening for H. pylori with an intention to offer eradication therapy is worthwhile in those with a need for long-term NSAID treatment.
(Chan FK et al. Lancet 2002; 359: 9-13.)
Perhaps not surprisingly, the reason that three patients
with inflammatory bowel disease (IBD) enrolled in a
study of 10 subjects using thalidomide as a moderator
of inflammation withdrew was because of sedation—
the drug was widely used as a sedative before its worldwide withdrawal in the 1960s because of teratogenicity.
Inhibitors of tumour necrosis factor alpha (TNF-α) such
as infliximab have been shown to reduce disease activity in IBD. Other pro-inflammatory cytokines, for example interleukin (IL)-1α and IL-6 are produced by
intestinal lamina propria mononuclear cells and peripheral blood monocytes. IL-12 has also been demonstrated in patients with active Crohn’s disease. All patients
in the study had symptomatic and objective evidence of
active disease despite treatment with corticosteroids and
azathioprine. After baseline colonoscopy and establishment of an index score of disease severity, subjects received 300 mg of thalidomide at bedtime. Patients were
reviewed at intervals two, four, eight and 12 weeks after
commencing treatment (with repeat colonoscopy at two
and 12 weeks). At each review laboratory tests and a
physical examination were performed, and index scores
of severity were calculated. In-vitro studies using tissue
obtained at biopsy was carried out concurrently, and
confirmed the ability of the drug to reduce the levels of
inflammatory cytokines. Four of the remaining six patients with Crohn’s disease (one had ulcerative colitis)
experienced remission with a significant decline in disease activity scores. Three were able to stop their steroid treatment, and in one case prednisolone was reduced
from 20 mg to 10 mg daily. Significant improvement was
also seen in the one patient with ulcerative colitis. One
patient developed peripheral neuropathy (a known adverse effect of thalidomide) that resolved on drug cessation. The authors suggest that lower doses may be
equally effective and that analogues of thalidomide without teratogenic effects are under development. (Bauditz
J et al. Gut 2002; 50: 196-200.)
The existence of a population pharmacoepidemiologic
database in the North Jutland county of Denmark (pop.
490 000) that can be linked to other databases constructed under the auspices of the Danish National
Health Service has allowed Danish investigators to identify (within the time frame 1991–95) approximately
164 000 users of non-aspirin non-steroidal anti-inflammatory drugs (NSAIDs). Prescription records were then
linked to the National Hospital Discharge Database to
determine the incidence of hospitalisation for upper
gastrointestinal bleeding (UGIB) in this group. After
excluding those with conditions that predispose to
UGIB, about 156 000 remained. The risk of an UGIB
was compared with the incidence in the background
population not exposed to NSAIDs. There were 515
UGIBs observed for NSAID users, compared to the expected figure of 124 (yielding an odds versus expected
ratio (O/E) of 4.1—95% CI 3.8–4.5). Given that there
are other commonly used drugs that increase the risk of
an UGIB (e.g. aspirin and anticoagulants), the risks of
such combinations were assessed. The majority (98%)
of the users had periods when only NSAIDs were taken
and in this group the incidence of UGIB was 3.6 times
more than expected (95% CI 3.3–4.0). The risk was
higher when NSAIDs were used in combination with
other drugs such as glucocorticoids, anticoagulants
and aspirin. The authors conclude that about 15% of
hospital admissions for UGIB in North Jutland can be
Sixteen subjects (nine male) who suffered from chronic
heartburn at least three times a week entered a recent
single-blind, three-way crossover study and were allocated in sequential fashion to receive omeprazole 20
mg at 8 a.m. and placebo at 10–11 p.m. (OP arm), omeprazole 20 mg at 8 a.m. and ranitidine 75 mg at 10-11
p.m. (OR), or omeprazole 20 mg twice daily (OBD).
Gastric and oesophageal pH data were measured at
baseline and then at six days after treatment. There was
a 7–10 day washout period between treatments. The
use of 24-hour pH monitoring allowed researchers to
record a reading for integrated acidity, a parameter
that allows for a determination of cumulative acid concentration for any time period. Before treatment there
was a significant increase in nocturnal gastric acidity
(10 p.m. to 4 a.m.), and while OP decreased both integrated gastric and oesophageal acidity it did not eliminate the nocturnal increase in gastric acid release.
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
Both OR and OBD nearly abolished nocturnal acid secretion. The OP group had evidence of nocturnal gastric acidity but were not troubled by nocturnal reflux,
leading to the conclusion that once daily omeprazole
would suffice for this pattern of disease. However, for
those patients on a daily proton pump inhibitor (PPI)
and yet still troubled by nocturnal reflux there is now
objective evidence that the addition of a small night
time dose of ranitidine will be effective (the dose of 75
mg was chosen in the study on the basis that this dose is
contained in non-prescription products available in
the USA). A second dose of a PPI will also provide
relief but at greater cost and reduced ease of availability. It always has been the case that pharmacists seeing
patients on a PPI/H2 receptor antagonist combination
should not assume that the prescriber has forgotten to
stop the latter on starting the former. There is now objective evidence that the agents from both drug classes
used together can be more effective than a single daily
dose of PPI for patients troubled by nocturnal acidity.
(Robinson M et al. Dig Dis Sci 2002; 47: 265-73.)
Infectious Diseases—Jeff Hughes
Diarrhoea associated with Clostridium difficile contributes significantly to the length of hospital stay (LOS)
and healthcare costs. In a recent study of 271 patients
admitted with infections, 40 developed nosocomial C.
difficile-associated diarrhoea. Adjusted hospital costs
were 54% higher (95% CI 17%–105%) for this group compared to those whose admission was not complicated
by C. difficile-associated diarrhoea. The length of stay
for patients with C. difficile-associated diarrhoea was
3.6 days longer (95% CI 1.5–6.2 days) than for those
without. Multivariate analysis demonstrated that C. difficile-associated diarrhoea was an independent predictor
of extended length of stay. Forty-eight percent of patients with nosocomial C. difficile-associated diarrhoea
died compared with 22% whose course was not complicated by C. difficile-associated diarrhoea, but after adjusting for age, comorbidity and severity of illness, C.
difficile-associated diarrhoea was not an independent
predictor of death. The authors estimated that the annual cost of C. difficile-associated diarrhoea in the USA
exceeds US $1.1 billion. (Kyne L et al. Clin Infect Dis
2002; 34: 346-53.)
Single dose gatifloxacin offers an alternative for the
management of uncomplicated urinary tract infections
(UTIs) in women. In a multicentre study, women with
uncomplicated UTIs were randomly assigned to receive
either single dose gatifloxacin (400 mg), gatifloxacin
200 mg daily for three days, or ciprofloxacin (100mg
twice daily for three days). Patients were assessed in
the 48 hours before the initiation of the study medication, at the end of treatment (by telephone contact on
day three), and in person 5–9 days after completion of
treatment (test-of-cure visit). Patients who achieved
eradication of their infection were followed up again
at a time 29–42 days after treatment. The bacterial eradication rate for the single-dose gatifloxacin, three-day
gatifloxacin, and three-day ciprofloxacin groups was
90%, 95%, and 89%, respectively; and the clinical efficacy rate was 93%, 95%, and 93% respectively for
microbiologically assessable patients at the test-of-cure
visit. Both gatifloxacin and ciprofloxacin eradicated
common uropathogens, including Escherichia coli, Klebsiella pneumoniae and Proteus mirabilis. The authors
concluded that single dose gatifloxacin offers advantages of decreasing secondary use of medical resources
and improving patient compliance. (Richards GA et al.
Urology 2002; 59: 344-9.)
Selective serotonin reuptake inhibitors (SSRIs) exhibit
fungicidal properties against Aspergillus species in vitro. Using a microdilution broth, researchers tested the
fungicidal activity of fluoxetine, citalopram, sertraline,
paroxetine and reboxetine against clinical isolates of A.
fumigatus (n = 11), A. flavus (n = 9), A. terreus (n = 10)
and Candida parapsilosis. All drugs showed time-dependent and dose-dependent effects and were fungicidal against all fungi tested. Sertraline and paroxetine were
the most active drugs. The SSRIs also demonstrated a
concentration-dependent post-antibiotic effect. The authors suggest that animal studies are needed to evaluate
the potential role of these psychotropic drugs in the
management of fungal infections. (Lass-Flörl C et al. J
Antimicrob Chemother 2001; 48: 775-9.)
Prior use of amoxycillin-clavulanic acid (co-amoxiclav)
has been shown to increase the risk of isolation of a coamoxiclav-resistant Eschericia coli in patients with urinary tract infections (UTIs). In a study involving 108
patients with E. coli UTIs, exposure to co-amoxiclav in
the four weeks prior to the diagnosis of a UTI increased
the risk of having a co-amoxiclav-resistant isolate by a
factor of 4.36 (95% CI 1.97–9.65; p < 0.001). Patients
with co-amoxiclav-resistant E. coli infections represented only 11% of those studied and were mostly elderly (≥ 60 years old) or babies ( ≤ 6 months old). The
authors pointed out that these two population categories are commonly prescribed co-amoxiclav for sinusitis, otitis media and lower respiratory tract infections.
They highlight the need to question patients with a UTI
about recent treatment with co-amoxiclav in order to
avoid another course, since these patients have a significantly higher risk of co-amoxiclav-resistant E. coli
infections. (Leflon-Guibout V et al. J Antimicrob Chemother 2002; 49: 367-71.)
Respiratory Medicine—Karin Nyfort-Hansen
One of the main goals of bronchodilator therapy in chronic
obstructive pulmonary disease (COPD) is to decrease
airflow limitation and as a consequence improve dyspnoea and exercise tolerance. The authors of a recent systematic review of the effects of bronchodilators on
exercise capacity in COPD have concluded that benefits
are limited, although a number of methodological problems in the 33 studies included in the review were identified. Anticholinergic agents had significant effects in
the majority of studies, with a trend toward an improved
response with higher doses (>80 mcg ipratropium per
day). Shorter-acting β-agonists had favourable effects
in more than two-thirds of the studies, but studies with
salmeterol did not show significant effects. The majority
of findings with theophylline products provided negative results. The search strategy in this review was limited to Medline, and did not include unpublished studies,
which suggests that the identified benefits may be over-
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
estimated. The paper includes a short description and
comparison of the various types of exercise tests used
in research studies, which may be of additional interest
to some readers. (Liesker JJ et al. Chest 2002; 121: 597608.)
The systematic review by Liesker et al (above) found no
studies of the effects of tiotropium bromide on exercise
capacity. This new, once-daily inhaled anticholinergic
agent with prolonged antagonism of M3 muscarinic receptors is expected to reach the Australian market this
year. Earlier studies have examined the effects of this
agent with short-term use in COPD, but now the first
reports to confirm the benefits of tiotropium bromide
therapy over twelve months in patients with COPD have
appeared. An accompanying editorial suggests that
when tiotropium becomes generally available, this drug
is likely to replace ipratropium as standard anticholinergic therapy in COPD. Benefits may include more convenient administration with improved compliance, a
reduction in the number of COPD exacerbations and
improved health-related quality of life. (Rees PJ. Eur
Respir J 2002; 19: 205-6.)
A recent case report is a reminder that HMG-CoA reductase inhibitors (statins) can be associated with interstitial lung disorders. The 51-year-old patient had been
taking simvastatin for six years when he was admitted to
hospital with fever, polymyalgia, cough and progressive
dyspnoea over one month. Pulmonary function tests
revealed a restrictive ventilatory pattern with decreased
diffusing capacity, and computed tomography of the
lungs showed bilateral interstitial infiltrates with groundglass opacities predominating in the lower lobes. An
open lung biopsy showed histological features consistent with a fibrotic, non-specific interstitial pneumonia.
Simvastatin was ceased and the patient made a progressive recovery on corticosteroid therapy over six months.
One month later pravastatin was inadvertently introduced and an increase in dyspnoea, myalgia and new
infiltrates on chest radiography were observed. This case
differs from earlier cases of statin-related lung disease,
where biopsies have revealed a pattern of hypersensitivity pneumonitis with granulomas. (Lantuejoul S et al.
Eur Respir J 2002; 19: 577-80.)
A small placebo-controlled, randomised, prospective
study has investigated the influence of smoking on the
efficacy of inhaled corticosteroids in mild asthma. Steroid-naive patients (n = 38) were randomised to receive
either fluticasone 1000 µg daily or placebo for three
weeks. In non-smoking patients fluticasone produced
expected improvements in morning peak expiratory flow
(PEF), spirometric values, bronchial hyper-reactivity
and sputum eosinophilia, whereas no significant changes were found in smokers who received fluticasone. A
possible explanation for these results is that the smoking patients had mild COPD, not asthma, although the
investigators believe this was not the case. Cigarette
smoking is common amongst patients with asthma and
this study is the first randomised, placebo-controlled
study examining this issue. Larger studies are needed
to confirm these results, and to investigate whether increasing the dose of inhaled steroids can restore responsiveness. Until this happens the results further
reinforce the need for smoking cessation in all patients
with asthma. (Chalmers GW et al. Thorax 2002; 57:
Neurology—Phil Nairn
Raised plasma homocysteine levels are a risk factor for
the development of dementia and Alzheimer’s disease
(AD). A recent study followed 1092 dementia-free subjects from the Framingham study, with a mean age of 76
years. Plasma homocysteine levels, measured at baseline and eight years later, were correlated with the incidence of new cases of AD and dementia. Over a median
period of eight years, 111 individuals developed dementia (83 with AD). The relative risk of dementia was 1.4
(95% CI 1.1–1.9) for each increase of one standard deviation (SD) of the log-transformed homocysteine level.
The relative risk for AD was 1.8 (95% CI 1.3-2.5) for each
increase of one SD in homocysteine level. When the
plasma homocysteine level exceeded 14 mmol/L the risk
of AD doubled. This intriguing study links AD to raised
homocysteine levels; however, it does not demonstrate
that reducing homocysteine with vitamin B12, B6, and
folic acid will prevent or treat AD. The study was weakened by the fact that homocysteine levels were not measured in fasting subjects, and that vitamin levels were not
determined for all subjects. (Seshadri S et al. N Engl J
Med 2002; 346: 476-83.)
Use of a blood pressure lowering regimen based on
perindopril and indapamide reduced the risk of stroke
in both hypertensive and non-hypertensive subjects with
a history of stroke or transient ischaemic attack (TIA).
The Progress Study compared perindopril 4 mg daily
to perindopril 4 mg with indapamide 2.5 mg, with a
third arm allocated to receive placebo. For the combined treatment groups there was a 28% relative risk
reduction (absolute risk reduction 4%) in the primary
outcome measure (stroke) relative to placebo. The relative risk reduction (43%) and mean blood pressure
reduction (12/5 mm Hg) were greatest in the combined
drug treatment arm. The single agent perindopril reduced BP by only 5/3 mm Hg and had no effect on stroke
risk. The cut-off point selected for normotensive individuals (160/90 mmHg) was higher than would be considered normal. It is important to consider absolute
risk reduction, and number needed to treat to avoid an
incident, rather than relative risk reduction. (Progress
Collaborative Group. Lancet 2001; 358: 1033-41.)
Treatment with valaciclovir did not reduce the formation
of active lesions in patients with relapsing-remitting
multiple sclerosis examined in a recent trial. The Scandinavian multicentre, double-blind, placebo-controlled
magnetic resonance imaging (MRI) study compared valaciclovir 1000 mg three times daily to placebo. The study
followed 70 patients for 24 weeks. The primary outcome
measure—formation of new active lesions on MRI—and
other secondary clinical measures, were not different between the groups. However, a subgroup analysis of individuals with high levels of disease activity in the
treatment group suggested a reduced number of new
active lesions (28% of scans free of lesions compared to
5% in placebo group). Given the short trial duration and
small number of participants, a positive clinical benefit
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
was not expected. Overall, valaciclovir did not reduce
the formation of MRI lesions; however, the subgroup
analysis is intriguing. A higher dose or another antiviral
drug may be more effective, as some of the suspected
viruses are not susceptible to this drug. (Bech E et al.
Neurology 2002; 58: 31-6.)
Interferon β -1b (IFNβ-1b) is effective in delaying the
progression of secondary progressive multiple sclerosis (SPMS). The final analysis of the European multicentre trial of IFN β -1b in SPMS demonstrated a
significant delay in the time to a one-point progression
on the Expanded Disability Status Scale (EDSS), when
compared to placebo (p = 0.007). The proportion of
patients with a two-point progression in EDSS was 27%
lower in the treatment group. Subgroup analysis suggested that patients with higher levels of pre-study disease, (more than two relapses or EDSS progression by
greater than one point, or both) seemed to have more
pronounced benefit. This study supports the use of IFNβ1b in the secondary progressive form of multiple sclerosis. (Kappos L et al. Neurology 2001; 57: 1969-75.)
Cardiovascular Medicine—Neil Cottrell
The early use of aspirin (prior to thrombolysis) reduces
mortality for patients experiencing an acute myocardial
infarction. The Argatroban in Acute Myocardial Infarction study (ARGAMI-2) was a three arm, parallel, randomised trial comparing two dosages of argatroban with
heparin in patients receiving alteplase or streptokinase
after acute myocardial infarction. A subgroup of patients
(30% of 12 000) enrolled in the study) received aspirin
prior to randomisation and thrombolysis, with the remaining receiving aspirin within one hour of thrombolysis. In the early aspirin group the mortality was lower at
seven days (2.5% vs 5.6%, p = 0.001), and this effect was
maintained at 30 days (3.3% vs 7.3%, p = 0.008) and at
one year (5% vs 10.6%, p = 0.002). A shorter time to
thrombolysis, use of β-blockers, and a higher incidence
of re-ischaemia were all higher in the early aspirin users.
The analysis of early aspirin users in the ARGAMI-2
study adds further evidence that early use of aspirin in
myocardial infarction is beneficial. (Freimark D et al. Am
J Cardiol 2002; 89: 381-5.)
Gemfibrozil is a cost-effective therapy for patients with
low HDL cholesterol and normal LDL cholesterol. An
economic analysis of the US Department of Veterans
Affairs (VA) Co-operative Studies Program HDL-C Intervention Trial (VA-HIT) demonstrated that gemfibrozil was cost-effective therapy in this patient population.
VA-HIT investigated the effect of gemfibrozil 1200 mg
daily compared with placebo in patients with coronary heart disease and low HDL-cholesterol and normal LDL-cholesterol levels, and revealed a reduction
in major cardiovascular events. The analysis was based
on that used in the Scandinavian Simvastatin Survival
Study. As age increased the cost-effectiveness ratios rose
from $6300 per year to $17 000 per year of life saved.
The net lifetime cost saving is cost-effective as long as
the purchase price of the gemfibrozil remains below
US$100 per patient per year, and the VA currently purchases for less than this price. This analysis of the VAHIT is the first to assess the cost-effectiveness of raising
HDL-cholesterol and lowering triglycerides in patients
without an effect on LDL-cholesterol levels. (Nyman JA
et al. Arch Intern Med 2002; 162: 177-82.)
Losartan, an angiotensin II type I receptor blocker, appears to be more effective than atenolol in reducing cardiovascular morbidity and mortality in patients with signs
of left ventricular hypertrophy. The recently published
Losartan Intervention For Endpoint reduction in hypertension (LIFE) study was a multicentre trial comparing
losartan with atenolol in patients with essential hypertension. A target blood pressure of 140/80 mmHg was
set and drugs were titrated to achieve this. Evidence of
left ventricular hypertrophy was also noted through ECG
observations. The primary end points were cardiovascular morbidity and mortality (stroke, myocardial infarction and other cardiovascular death). 9222 patients
entered the study and were followed over 4 years. There
was a 13% relative risk reduction in the primary end point
with losartan compared with atenolol. Losartan also produced a 25% reduction in the incidence of stroke. Blood
pressure reduction was not achieved in 3% of the atenolol group and there was an increased incidence of the
occurrence of diabetes (25%) amongst those treated with
the β-blocker. The study provides further evidence that
inhibition of the angiotensin system may have additional benefits in the management of hypertension. (Dahlöf
B et al. Lancet 2002; 359: 995-1003.)
A subgroup of patients in the LIFE study who had diabetes on entry to the study was also analysed. As for
others, medication for these subjects was titrated to a
target blood pressure of 140/80 mmHg or less. Patient
characteristics were similar in losartan and atenolol
groups, with a slightly higher proportion (3%) of patients in the atenolol group receiving a biguanide. 1195
patients were included in this analysis. The primary
composite end point was lower in the losartan group
(relative risk 0.76, p = 0.031). This trend was also repeated when individual outcomes were analysed: cardiovascular death (relative risk 0.63, p = 0.028), stroke
(relative risk 0.79, p = 0.204) and myocardial infarction (relative risk 0.83, p = 0.373) appeared lower,
and the average blood pressure was lower in the losartan group than the atenolol group (146/79 mm Hg vs
148/79 mm Hg). Albuminuria was reported significantly less frequently in those patients treated with losartan (p = 0.002). The results of this subgroup analysis
suggest that patients with diabetes, hypertension and
left ventricular hypertrophy may benefit more from losartan than atenolol. (Lindholm LH et al. Lancet 2002;
359: 1004-10.)
Geriatrics—Bruce Williamson
A multicentre, double-blind, placebo-controlled trial that
enrolled 592 patients with probable vascular dementia or
Alzheimer’s disease showed that after six months, those
randomly assigned to receive 24 mg of galantamine daily
had statistically significantly improved cognitive function relative to baseline and compared to placebo (p <
0.0001 in both cases). The primary efficacy measures
were the standard 11-item ADAS-cog instrument (ADAScog/11) which assesses cognitive ability, and the clinician’s interview based impression of change plus
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.
caregiver input (CIBIC-plus). The neuropsychiatric inventory (NPI) indicated significant beneficial effects of
galantamine upon behavioural symptoms compared to
placebo (p = 0.016) with significant improvements in anxiety and apathy (both p < 0.0001). (Erkinjuntti T et al.
Lancet 2002; 359: 1283-90.)
In a case-control study of psychiatric patients with hyponatraemia, researchers identified that 29 subjects
who had a serum sodium in the range 104 to 130 mmol/
L were treated with antidepressants (22 using selective
serotonin reuptake inhibitor antidepressants—SSRIs).
Those using SSRIs were nearly four times more likely
(odds ratio OR 3.9; 95% CI 1.2–13.1) to develop hyponatraemia compared to those treated with drugs from
other antidepressant groups. Patients who were 65
years of age and older were markedly at risk for hyponatraemia (OR 6.3; 95% CI 1.0–41). A synergistic
effect for the concurrent use of diuretics was evident
(OR 8.4; 95% CI 2.1–34) and this effect was even more
pronounced in those 65 years and over (OR 13.5; 95%
CI 1.8–101). Over half (55%) of all the prescriptions
were for paroxetine. (Movig KL et al. Br J Clin Pharmacol 2002; 53: 363-9.)
A cross-sectional study of 802 patients (502 females)
with a mean age of 63 years, showed that renal impair-
Journal of Pharmacy Practice and Research Volume 32, No. 2, 2002.