Download auranofin - Oxford Academic

British Journal of Rheumatology 1997;36:560–572
DISEASE-MODIFYING DRUGS
SERIES EDITOR: T. PULLAR
AURANOFIN
W. F. KEAN, L. HART and W. W. BUCHANAN
Rheumatic Disease Unit, McMaster University, 401 -1 Young Street, Hamilton, Ontario L8N 1T8 , Canada
that rheumatoid disease was an infectious disease
analogous to tuberculosis led him to use gold
thiopropanol sodium sulphonate on 15 patients with
inflammatory rheumatoid disease. The success of this
initial experiment was the seed which steered
researchers over the next 50 yr to investigate both the
beneficial and the toxic effects of anti-arthritic gold
complexes. It is important to note that although
Forestier’s hypothesis was based on a so-called
erroneous assumption, his observations of the histological similarity of ‘reactive’ rheumatoid synovium
and tuberculous nodules are now known to bear more
similarity to scientific truth than to fiction.
HISTORY OF GOLD IN MEDICINE
The presumed magical qualities of gold and its
resemblance to the ‘essence’ of the sun made gold and
gold potions the natural choice of priests, healers and
shaman since the earliest civilizations. Medicinal use of
gold and other precious metals was known to the
Chinese circa 2500  [1], and it is stated that Taoist
philosophers of ancient China circa 600  [2, 3] used
gold and silver to enhance the quality of their
medicines. Probably one of the earliest records
pertaining to the practice of gold alchemy has to be the
dramatic account in Exodus of Moses burning the
golden calf, preparing a potion from the ashes, and
forcing the Children of Israel to ‘... drink of it’ [4]. The
Roman physician, Pliny the Elder [5] and the Greek,
Dioscorides in the first century  [6] record the use of
gold compounds as medicinal agents for multiple
ailments. In the great Persian medical schools founded
by the Nestorian Christians, pharmacist–physicians
such as Yabir, Avicenna and Rhazes all advocated the
use of gold compounds as panaceae. Yabir is reputed
to have discovered and described the formulation of
aqua regia, the mixture of nitric and hydrochloric acid,
necessary for the dissolution of gold. Knowledge of
structured and sophisticated medical practice spread
from the Arab cultures into Europe and the British
Isles by the 11th and 12th centuries. The Franciscan
scholar Roger Bacon (circa 1214–1292 ) was one of
the first known gold alchemists of the British Isles.
He described the synthesis of gold chloride [7] and
advocated its use in the treatment of diseases. The
renaissance pharmacologist Paracelsus proclaimed
gold compounds as a panacea; however, many of his
contemporaries refuted his recondite theories.
The modern use of gold compounds in medicine was
initiated in the early 19th century by the work of
Andre-Jean Chrestien and Pierre Figuier (1765–1817),
two professors at the University of Montpellier [8].
Figuier, a pharmacist, described the chemical formulation for the gold compounds, in particular gold
sodium chloride, which the clinician, Chrestien,
advocated as of value in the treatment of tuberculosis
and syphilis. The observation by Robert Koch [9] that
gold cyanide was bactericidal to tubercle bacilli in vitro,
led European investigators, over the next 40 yr, to
experiment with the use of gold complexes in the
treatment of human and bovine tuberculosis. The
serendipitous assumption by Dr Jacques Forestier [10]
GOLD CHEMISTRY
The element gold is not attacked by oxygen or
sulphur at any temperature, hence its unique stability
[11]. Gold is present at 00.004 p.p.m. in the Earth’s
crust in both metallic and mineral forms. The most
common minerals are the tellurides such as calaverite,
krennerite (different crystalline forms of AuTe2),
montbrayite (Au2Te3) and mixed gold–silver tellurides
such as sylvonite (AuAgTe4) [11]. Mineral forms of
auriferous sulphides also exist in lesser quantities [12].
Metallic gold and silver–gold mixtures such as
electrum, which contains 20% silver, are found in
nugget or particulate form. Gold is a Group IB metal
in the periodic table with the most common oxidation
states as I, II, III and V, although metal–metal bonds
exist in complexes in which it is difficult to assign a
formal oxidation state to the gold atom. Halides which
form the true salts of Au(I) are unstable in the presence
of H2O and disproportionate to Au° (metallic gold) and
Au(III). Au(I) can be stabilized by the formation of
‘soft’ ligand complexes [13] with thiolates and
phosphines. All current anti-arthritic gold complexes
exist as Au(I) thiol or phosphine compounds [14]. The
high cellular toxicity of the Au(III) complexes, such as
chloro-auric acid (HAuCl4), makes them unsuitable for
human ingestion. Studies on laboratory animals
treated with gold sodium thiomalate have shown that
the gold recovered from the tissues and urine exists in
the Au(I) oxidation state (i.e. the same Au(I) oxidation
state in gold sodium thiomalate) and not the Au(III)
oxidation state [15]. Further, if HAu(III)Cl4 is given to
laboratory animals, gold recovered in tissues and urine
exists only in the Au(I) oxidation state. The Au(I)
oxidation state, therefore, appears to be the primary
oxidation state which is stable and exists in the
biological milieu and in vivo. Reduction from Au(III)
Submitted 8 October 1996; accepted 11 October 1996.
= 1997 British Society for Rheumatology
560
KEAN ET AL.: AURANOFIN
to Au(I) is most likely caused by the powerful
sulphydryl-containing reducing enzymes present
in vivo.
In the late 1970s, auranofin emerged as a possible
alternative gold anti-arthritic agent which had the
presumed advantage of oral administration. In order to
obtain a stable gold compound which could be
absorbed from the gastrointestinal tract, a unique
formulation was devised.
Auranofin is a monomeric Au(I) species in which the
triethyl phosphine group stabilizes the gold-thiol
complex. Its chemical name is 2,3,4,6-tetra-o-acetyl-1thio-b--glucopyranosato-S-(triethyl-phosphine) gold.
It is marketed as a 3 mg tablet or capsule which
is administered orally. The compound is a white,
odourless, crystalline solid which is insoluble in water.
The powder is unstable and must be protected from
light and heat. On a weight basis, auranofin contains
029% gold and has a molecular weight of 678.5 with
a melting point of 112–115°C.
AURANOFIN PHARMACOKINETICS
Following oral administration of [195Au]auranofin,
025% of the administered dose is detected in plasma
bound to albumin through bonds to crysteine and
histidine [16]. Peak concentrations of 6–9 mg% are
reached within 1–2 h [31, 32]. The plasma half-life of
auranofin is in the order of 15–25 days with a total
body elimination rate of 55–80 days [17]. Only 01%
of the 195Au is detectable by 180 days, whereas up to
30% of the 195Au from gold sodium thiomalate may be
detected at this time [18].
When auranofin crosses the plasma membrane, the
acetyl groups on the thioglucose moiety are lost and
some dissociation of the compound takes place [19].
Triple label studies with radioactive 195Au, 32P and 35S
show evidence of Au–S and Au–P disruption which
occurs across the plasma membrane [20]. Different
experimental models have shown different results for
radioactive ligand dissociation across the plasma
membrane [19]. Since it is unlikely that the gold moiety
alone is the active species, in vivo study of molecular
gold kinetics has been of little value in the
interpretation of the mechanisms of action of gold
complexes. Proper design of in vitro and ex vivo
experiments must incorporate the expected ligands
which occur following the dissociation of the
auranofin. The application of the intact auranofin
molecule to an in vitro or ex vivo experiment may not
provide results which reflect the in vivo activity of the
dissociated components of the auranofin [21].
AURANOFIN PHARMACODYNAMICS
Chemical structure, biological actions and pharmacokinetic studies indicate that the oral gold compound,
auranofin, differs significantly from the injectable
gold-thiol compounds [22]. Walz et al. [23] showed that
auranofin inhibited carrageenin-induced oedema in
rats in a dose-related fashion at concentrations of 40,
20 and 10 mg/kg with maximum inhibition of 86% at
the highest dose, and a serum gold level of 010 mg/ml.
561
The two basic ligands of auranofin, namely triethylphosphine oxide and 2,3,4,6-tetra-o-acetyl-1-thiob--glucopyranosato were without biological activity
and gold sodium thiomalate, gold thioglucose and
thiomalic acid did not affect rat paw oedema
significantly. Auranofin was shown to suppress
adjuvant arthritis significantly, whereas the ligands
were without any effect. Auranofin inhibited antibodydependent complement lysis, whereas antibody-dependent complement lysis was enhanced in the serum of
rats treated with gold sodium thiomalate [22].
The gold-thiol compounds, particularly gold sodium
thiomalate, are ubiquitous inhibitors of cellular
enzymes, particularly the serine esterase enzymes such
as elastase, cathepsin G and thrombin [24]. No specific
inhibitory effect of auranofin on these enzymes
has been recorded, but auranofin has been shown to
inhibit the release of lysosomal enzymes such as
b-glucuronidase and lysozyme from stimulated polymorphs [22]. Gold thioglucose has no apparent effect
on extracelluar release of lysosomal enzymes. Similarly,
the ligands of the above gold complexes have no effect
on either release or enzyme inhibition. Auranofin is a
potent inhibitor of antibody-dependent cellular cytotoxicity exhibited by polymorphs from adjuvant
arthritic rats. In contrast, gold sodium thiomalate, gold
thioglucose and the ligands of auranofin had no
significant effect on polymorphonuclear antibodydependent cellular cytotoxicity. Although results vary
depending on the system used to study superoxide
production, in general auranofin is a much more potent
inhibitor of superoxide production than gold sodium
thiomalate. In certain systems such as an immune
phagocytosis system, gold sodium thiomalate was
devoid of inhibitory activity at a concentration of 40
times that of auranofin, which had produced marked
inhibition [22]. Walz et al. [22] postulated that
auranofin has also been shown to be more potent than
gold sodium thiomalate in the inhibition of cutaneous
migration, chemotaxis and phagocytosis by peripheral
blood monocytes. Lipsky et al. [25] have shown that
auranofin, like gold sodium thiomalate, inhibits
lymphoblastogenesis in vitro by direct inhibition of
mononuclear phagocyte function, but also has an
inhibitory effect on lymphocyte function not seen
with gold sodium thiomalate. The inhibitory action
on monocytes is achieved with concentrations of
auranofin that are 10- to 20-fold lower than those of
the gold sodium thiomalate. The above findings suggest
that auranofin has immunosuppressive activity. In
general, patients with active rheumatoid disease have a
decreased capacity for either mitogen-stimulated
lymphoblastogenesis or for lymphoblastogenesis produced by the mixed lymphocyte reaction. Although
patients initially treated with gold sodium thiomalate
exhibit a suppression in mitogen-stimulated lymphoblastogenesis, those who eventually respond to the
drug will have normalization of lymphocyte responsiveness in vitro [26]. In contrast, patients who
receive auranofin develop a marked inhibition of
lymphocyte responsiveness [27]. Thus, auranofin has
562
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
powerful immunosuppressant effects in vitro and acts at
an order of magnitude less than the injectable gold
compounds. This may reflect major differences in the
pharmacological properties of the oral compound vs
the injectable gold-thiol compounds. Indeed, it has
been proposed that auranofin could have potential in
cancer therapy [28] because of its potent immunosuppressive properties [29]. To date, there have been no
reports which cast concern on carcinogenicity or other
long-term safety of auranofin treatment with respect to
the dangers of prolonged immunosuppressive therapy.
ADVERSE EFFECTS OF AURANOFIN
Loose stool/diarrhoea
A change in stool pattern with the development of
loose soft stools is the most common side-effect of
auranofin therapy and may occur in over 40% of
treated patients [30, 31]. The frequency of loose stools
is highest in the first month of treatment, but the lower
incidence of altered stool pattern in later months may
be directly related to a pre-selected drop-out of those
patients susceptible to the diarrhoea. The development
of frank watery diarrhoea occurs in 2–5% of patients
and is dose related, but some patients can be totally
intolerant of even 3 mg/day. The precise mechanism
underlying gold-induced diarrhoea is not well defined.
The major part of solute and water absorption occurs
in the small intestine, with the large intestine,
particularly the distal colon, playing the role of a final
modifier. This mechanism is similar to that observed in
the renal tubule where obligatory reabsorption of salts,
solutes and water occurs in the proximal segments, with
the distal tubule and collecting duct responsible for
facultative reabsorption. By analogy with the renal
tubule collecting duct, the reabsorptive capacity of the
distal colon is limited. Decreased reabsorption of
solutes and water in the proximal segments can
overwhelm the reabsorptive capacity of the distal
segment and result in diuresis in the case of the kidney,
and diarrhoea in the case of the gut [32].
Clinical and experimental studies have suggested
that auranofin alters the intestinal transport process to
produce diarrhoea.
Van Riel et al. [30] studied the occurrence of
diarrhoea during the course of a long-term clinical trial.
Eleven patients developed diarrhoea and, in eight who
were studied, K+ ion and dry weight were reduced, but
faecal Na+ concentration was elevated. The authors
suggested that auranofin could have a direct effect on
the absorption of salt and water by the colon [33]. In
six patients taking auranofin, Behrens et al. [34] found
that intestinal transit time was decreased, the
concentration of Na+ in faecal water was increased
with a reduction in bicarbonate, but no significant
changes in levels of Cl− or K+ were recorded. They also
noted that the excretion of 51Cr-EDTA and lactulose
was increased with little increase in the excretion of
mannitol. This argued for an increase in intestinal
permeability. Whether this implied destruction of the
absorptive surface was not clear, but experimental
studies in the rat produced evidence of mucosal
damage. In contrast to these results suggesting that
gold compounds inhibit absorption, a study of a single
patient given gold thioglucose suggested that they
could produce a secretory type of diarrhoea [35].
Experimental studies have provided evidence that
both anti-absorptive and secretory mechanisms may be
operative. Hardcastle et al. [36] used everted rat colon
sacs, and showed that auranofin in the mucosal
solution reduced fluid and Na+ absorption, whereas it
was ineffective from the serosal side. The drug also
inhibited the activity of Na+/K+ATP-ase. Similar
observations were made in dogs where auranofin
caused significant elevations in efferent volume,
osmolarity and Na+ concentration with significant
decreases in K+ [37]. These results led to the hypothesis
that auranofin and other gold compounds, such as gold
sodium thiomalate and chloroauric acid, reduced
absorptive processes by inhibiting the Na+ pump
activity of enterocytes. In contrast, Ammon et al. [38]
found that auranofin induced a net fluid and electrolyte
secretion across the in vivo perfused rat jejunum. Since
gold sodium thiomalate did not produce similar effects,
they suggested that it was not the gold moiety but the
carrier molecule that was responsible for the effects.
Later, these authors showed that prolonged perfusion
with auranofin produced mucosal injury, whereas gold
sodium thiomalate did not. It should be noted that the
structural differences amongst gold compounds, especially between the oral compound auranofin and the
injectable compounds, suggest that the gold moiety
may not be the active species and may not be
chemically available for biochemical interaction with
other ligands, except in conjunction with another
molecule such as sulphur or with a complete ligand
(see gold chemistry).
The colonic epithelium has an extensive neural
network and it is possible that gold compounds could
alter colonic transport by modulating enteric nerve
activity. This was investigated [39] by using two
different in vitro preparations from the canine colon: (a)
mucosa which had the muscularis mucosa and
attendant submucous plexuses present and (b) an
epithelium that was functionally nerve free [40]. Both
preparations responded to auranofin from the luminal
and from the contraluminal (serosal) sides. However,
the concentrations required to elicit responses from the
luminal side were 100-fold greater. With the mucosal
preparation, responses to auranofin were significantly
reduced by tetrodotoxin which is a neurotoxin. These
results suggest that the effects of auranofin on colonic
secretion have a significant neural component, as well
as the direct effects noted above. In addition, it appears
that auranofin traverses the epithelial lining to
modulate colonic activity from the serosal side. The
ionic gold compound, sodium aurothiosulphate,
altered colonic transport only on serosal application,
but had no effects on luminal application. Thus, two
chemically distinct gold compounds, auranofin and
sodium aurothiosulphate, produced similar effects on
serosal application. It should be noted that this
suggests that a Au–S complex or Au–ligand or the
KEAN ET AL.: AURANOFIN
ligand alone is the active species. It is not appropriate
to assume that ‘free’ gold is the active species since it
could not exist in these biological conditions, but
would have to be bound to some other species (see gold
chemistry). The precise molecular mechanisms responsible for the above effects are not clear.
Snyder et al. [41] suggested that both the therapeutic
and toxic effects of auranofin were due to the
interactions of the drug with -SH groups, and that
auranofin and other gold complexes stimulate the
activity of phospholipase C. This leads to the
production of secondary messengers such as diacylglycerol (DAG) and inositol 1,4,5 triphosphate (IP3).
DAG is hydrolysed to produce arachidonic acid which
could, in turn, produce a variety of eicosanoids
through the cyclo-oxygenase and lipoxygenase pathways. The prostanoids and leukotrienes have pronounced effects on intestinal transport [42].
Intracellular Ca2+ is released by IP3 and can stimulate
Cl− secretion by enterocytes [43]. Snyder et al. [41]
also stated that lipid peroxidation could lead to the
production of superoxide (O−
2 ) and hydroxyl radicals
(OH), as well as hydrogen peroxide, with disruption of
membrane fluidity. It is not known which of these
mechanisms is responsible for the effects on the
epithelial cells and the enteric nerves, but the activation
of phospholipase C does appear to be a possibility.
The above results suggest that gold compounds can
produce diarrhoea by inhibition of absorption and/or
stimulation of secretion. Altered motility may play a
role. In the clinical studies of Behrens et al. [34], the
mean transit time decreased from 71.0 h (off auranofin)
to 40.5 h (on auranofin) and the authors suggested that
much of this could be due to more rapid transit
through the colon. However, Fondacaro et al. [37]
found that auranofin had no effect on the basal tension
of isolated colonic smooth muscle strips and did not
alter KCl-induced contractions. Since the responses of
the muscle to neurotransmitters or hormones could
have been altered, the evidence for an effect on gut
motility is inconclusive at present.
Although loose bowel movement or frank diarrhoea
are the most common gastrointestinal side-effects of
auranofin, non-specific digestive system complaints
account for 020% of adverse reactions and 2% of all
withdrawals from auranofin therapy [44].
Rash/stomatitis/conjunctivitis
Rash is a common side-effect of auranofin therapy
and occurs in up to 20% of subjects. Half of these
patients also experience pruritus. Rash is most
common in the first year of therapy, but can occur at
any time. When rash develops, the drug should be
withheld until the condition resolves. Approximately
2–3% of patients have to stop treatment because of
severe skin rash [45]. It is our opinion that clinicians
engaged in a research protocol might be more likely to
discontinue a patient from gold therapy because of rash
compared to a clinician who observes a mild to
moderate rash in a clinical non-research setting in
which the patient is achieving benefit.
563
Stomatitis occurs in 1–12% of patients and may
occur concomitantly with skin rash. While occurrence
is greatest in the first month, stomatitis can occur at
any time [45]. The oral ulceration may or may not be
painful. The ulcers resemble aphthous ulcers in the
mucous membrane and in the vestibule of the mouth.
Occasionally, the lesions are present on the tongue or
the hard palate. The development of a mouth ulcer is
a definite contraindication to gold therapy until it has
resolved. Oral ulceration can precede the development
of pemphigoid-like bullous skin lesions. The causes of
rash and stomatitis are unknown.
Proteinuria
The wide variations in the definition or lack of
definition of proteinuria in the literature have resulted
in a considerable difference in the 0–40% reported
incidence for injectable gold compounds [45–49].
Proteinuria occurs in up to 5% of patients treated with
auranofin. The drug should be withheld and assessments of renal function made in a manner similar to
that for renal toxicity due to injectable gold
compounds. Proteinuria in urine in amounts of 1 + by
dipstick over 2–3 weeks warrants a 24 h urine protein
estimation. If proteinuria is Q500 mg/24 h, the drug
should be continued. Between 500 and 3000 mg/24 h,
auranofin therapy should be withheld until it is
established that renal function is normal. Patients with
proteinuria q3000 mg/ 24 h should have the auranofin
stopped until the proteinuria resolves. There are no
well-documented cases of long-term serious or permanent renal damage due to gold-induced proteinuria.
The most common renal histological lesion is
membranous glomerulonephritis with deposition of
IgG, IgM and C3 on the glomerulus [50], although
heavy metal tubular damage does occur during
injectable gold therapy [51].
One postulated mechanism of the proteinuria is that
the gold or a gold complex damages the renal tubule,
analogous to heavy metal damage, and results in
release of a ‘neo antigen’ which complexes with
immunoglobulin and deposits on the glomerulus to
result in a glomerulonephritis. The tubular damage is
characterized by the presence of b2-microglobulin in the
urine. Electron microscopy has demonstrated the
presence of fibrillar particulate gold-containing material within the proximal tubular cells and also the
presence of mitochondrial degeneration [51]. Clinical
studies have suggested an association between HLA
DR3 and the presence of proteinuria [52].
Haematuria should not be considered as a recognized side-effect of auranofin therapy. When microscopic haematuria develops, auranofin should be
stopped immediately and a cause for the haematuria
sought. Once haematuria has resolved, auranofin may
be reinstituted at a reduced dosage.
Thrombocytopenia/bone marrow suppression
Thrombocytopenia and related bone marrow suppression may present as rare side-effects of auranofin
treatment. This type of auranofin-induced thrombocy-
564
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
topenia (unlike the more common type seen with gold
sodium thiomalate) [53] does not express platelet
surface-associated autoantibodies. Physicians monitoring auranofin therapy should observe a platelet count
of Q200 000/mm3 as an indication to discontinue gold
therapy temporarily. Most laboratories record a value
of 150 000/mm3 as the lower level of normal for the
platelet count. A falling platelet count, even within the
normal range, can be a signal of early thrombocytopenia. A sudden change in a platelet count which has
been steady at e.g. 0400 000/mm3 weekly to a value of
e.g. 210 000/mm3 behoves a physician to withhold
auranofin therapy until a repeat platelet count confirms
a stable value q200 000/mm3 on at least two occasions
1 week apart. When a fall in platelet count results in
a value which is persistently Q200 000/mm3, extreme
caution is advised. Close observation of changes in
platelet count even within the normal range may result
in earlier identification of some patients who may
develop a sudden thrombocytopenia. The treatment of
choice is immediate withdrawal of drug therapy. The
development of thrombocytopenia is an absolute
contraindication to therapy.
Bone marrow suppression due to auranofin is rare,
but is sufficiently serious to warrant strict monitoring
of blood parameters on a weekly basis [46]. The
marrow suppression secondary to injectable gold
therapy is postulated to be due to a direct action of the
drug on the marrow cells [54]. A fall in either platelet
count as recorded above, or a fall in haemoglobin
below 10 gram%, and/or a fall in total white count
below 4000/mm3, requires immediate discontinuation
of therapy until cause and effect have been established.
A reversal of the white cell differential ratio and/or a
rise in monocyte count above 10% are also indications
for immediate discontinuation of the gold drug until at
least two normal values, 1 week apart, have been
recorded. If any of the above indicators of haemopoiesis remain abnormal, a bone marrow examination
and investigation for autoantibodies to white cells, red
cells and platelets are essential before auranofin
therapy can be reintroduced. Bone marrow suppression
secondary to auranofin is an absolute contraindication
to further continuation of therapy.
While auranofin-associated bone marrow suppression could be a direct effect of the gold compound,
an alternative postulate is that there is an alteration
in the marrow feedback system from monocytes in
the peripheral blood due to an action of the
auranofin components on these cells or their ‘cell
messengers’.
In spite of the potential for gold compounds to
induce marrow suppression, successful treatment of
Felty’s syndrome has been carried out in a controlled
situation where assessment of bone marrow tissue
and peripheral blood counts were available [55].
However, one case of Felty’s syndrome, treated
with auranofin 6 mg/day for 1 yr, failed to respond.
This patient also received prednisone which may
have negated the potential effectiveness of the
auranofin [56].
Eye problems
Conjunctivitis has been reported from collective
studies in 010% of patients treated with auranofin,
and occurred with equal frequency at each 3 month
time period over 3 yr of treatment. Withdrawal from
therapy because of conjunctivitis is rare [57].
Gold in pregnancy and lactation
Injectable gold compounds have been implicated in
the cause of fetal abnormalities in humans [58, 59], but
this has not been substantiated due to the small
numbers studied [60, 61]. Since gold levels have been
found in the blood of newborns [60, 62, 63], it would
seem wise to avoid gold therapy during pregnancy if
possible. Usually, rheumatoid arthritis (RA) will
decrease in severity or even remit during pregnancy. A
small number of patients who received auranofin
during pregnancy have been observed with no effect to
mother or child [64]. Since auranofin has immunosuppressive-like properties and contains triethylphosphine,
it would seem wise to avoid this medication during
pregnancy [65].
Trace amounts of gold derived from gold thioglucose
therapy have been detected in breast milk [66, 67], but
since gold is not readily absorbed from the gastrointestinal tract, it is unlikely to be a significant problem.
To date, there are no data on the concentration of
auranofin in breast milk. As noted above, because of
its immunosuppressive properties, auranofin should be
avoided in breast-feeding mothers.
Adverse effects—comment
Despite the apparent overall lower number of
side-effects related to auranofin compared to injectable
gold compounds, auranofin should not be considered
a benign drug. A strict monitoring system, as for
injectable gold compounds, should be undertaken for
each patient. Insufficient data are available at present
to determine whether there will be any long-term
side-effects related to auranofin therapy. Particular
attention should be paid to effects on immune functions
related to long-term therapy.
A common misconception among members of the
medical profession is the idea that -penicillamine may
be used as a chelating agent for gold if a toxic reaction
occurs during, or following, the administration of the
anti-arthritic gold complexes. There is no theoretical or
biochemical evidence that -penicillamine chelates
Au(I) in vivo. Au(I) compounds normally have a linear
geometry in which the Au(I) atom is attached to only
two ligand atoms (X) such that the X–Au(I)–X angle
is 180°. Less frequently, and only for rather specific
ligands, Au(I) binds to three or four ligand atoms.
Thus, one would expect -penicillamine to bind to
Au(I) only through the sulphydryl site. As supported
by the studies of Davis and Barraclough [68], there is
no theoretical reason why -penicillamine should
chelate Au(I) [69]. -Penicillamine and other chelating
agents should NOT be used for the treatment of
auranofin (gold)-induced toxicity.
KEAN ET AL.: AURANOFIN
CLINICAL OVERVIEW
Methodology
By definition, the disease-modifying anti-rheumatic
drugs (DMARDs) are thought to impact on the
progression of RA by (i) favourably altering the
(short-term) natural history of the disease and (ii)
slowing the development of radiologically demonstrated joint erosions or joint destruction [70, 71].
However, because of the often unpredictable course of
RA, it is often difficult to determine the real (and
unconfounded) impact of DMARDs, or other interventions, on disease outcome.
When evaluating the effectiveness of a specific
therapeutic agent, much depends on the robustness of
the instruments that are used to measure outcome.
While it is generally agreed that multiple measures of
clinical outcome are helpful in determining the
effectiveness of a particular intervention, it needs to be
acknowledged that such measures may be duplicative
and vary in their relative sensitivity to change [72].
Moreover, although radiological assessments can
satisfactorily demonstrate the extent of joint damage at
any given time, there is little correlation between the
radiological progression of disease and functional
impairment [73].
What therefore becomes apparent is that any
scrutiny of the available literature on one or another of
the DMARDs needs to include a meticulous appraisal
of the potential imprecisions of clinical outcome
measurement and radiological interpretation. In
addition, the broader context of study design and
general methodological robustness also need to be
considered carefully. There is an expectation that
rigorously conducted studies and their reports should
lead to better and more realistic estimates of treatment
effects, more accurate and reproducible predictions of
treatment efficacy, and greater acceptance of their
results within the health care community [74, 75].
The quality of evidence in published research can be
assessed according to various critical appraisal criteria,
but for the specific purposes of this overview, the
general guidelines developed by Sackett et al. [76] on
the effectiveness of therapy provide the basis for
judging the merits of particular studies. In evaluating
each report, the following questions were addressed:
1. In randomized controlled studies (RCTs) was the
assignment of patients really randomized?
2. Were clinically important outcomes assessed
objectively?
3. Was there at least 80% follow-up of subjects?
4. Were both statistical and clinical significance
considered?
5. If the study was negative, was power assessed?
As a further stage in the ‘quality filter’, a level of
evidence can be attributed to the selected studies, and
a hierarchy for grading recommendations (for application of the research conclusions) can be assigned [74].
The relationship between the levels of evidence and
grades of recommendation is summarized in Table I
565
[77]. This framework, although less rigorous than a
precursor advocated by Sackett [78], is a useful medium
for categorizing treatment recommendations.
This overview will focus primarily on studies
demonstrating Grade I evidence (according to the
criteria in Table I) but, to provide context and
clarification, a small number of selected reports,
reflecting Grade II evidence, will also be included.
To retrieve the relevant studies on the use of
auranofin in the management of RA in adults, a
MEDLINE search was conducted on the available
English language literature for the period between 1980
and 1996. Additional studies were identified through
reference lists contained in some of the searched reports
and those contained in textbooks of the rheumatic
diseases.
The notion of a possible oral gold preparation took
root in 1972 with a report by Sutton et al. [79] that
orally administered alkylphosphine gold co-ordination
complexes exhibited anti-inflammatory properties
when administered to adjuvant arthritic rats. In the
same year, it was demonstrated that triethylphosphine
gold chloride was equipotent to parenterally administered gold sodium thiomalate for suppression of
inflammatory lesions in these rats [80]. Because triethyl
gold chloride is extremely toxic in humans, further
studies of its potential usefulness were not attempted.
However, a related compound, 2,3,4,6-tetra-o-acetyl-1thio-b--glucopyranosato-S-(triethylphosphine) gold,
later known as auranofin, was also shown to exhibit
anti-arthritic properties and was considerably less toxic
than its precursor [81].
Grade II evidence favouring the use of auranofin in
humans began to emerge following open studies by
Finkelstein et al. [82–84] and Bernhard [85]. In the first
clinical study on auranofin, Finkelstein et al. [82]
described eight patients with RA treated for 3 months
followed by a 3 month period on placebo. During the
treatment period, the total number of active joints fell
from 60 to 17 at week 12 and to nine at week 15.
Clinical improvement was recorded at 5 weeks and in
general the drug was well tolerated. Rheumatoid factor
titre, IgG levels and a2-macroglobulin levels fell during
the treatment period. During the following 3 month
placebo period, IgG levels rose and patients experienced a flare-up in disease activity, suggestive of a
TABLE I
Categories for quality of evidence on which treatment recommendations are made
Grade
Definition
I
Evidence from at least one properly randomized,
controlled trial
Evidence from at least one well-designed clinical trial
without randomization, from cohort or case-controlled
analytic studies, preferably from more than one centre,
from multiple time series, or from dramatic results in
uncontrolled experiments
Evidence from opinions of respected authorities on the
basis of clinical experiences, descriptive studies, or reports
of expert committees
II
III
566
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
cause and effect action of the drug. In addition to
establishing efficacy and toxicity profiles for oral gold,
the early studies focused on determining the optimal
dosage of the new preparation. In one larger study,
Calin et al. [86] were able to demonstrate that
auranofin at a dose of 1 mg/day was insufficient to
exert therapeutic benefit, while the 9 mg/day dose was
efficacious but provoked toxic side-effects that necessitated dose reductions in a significant number of study
patients. At the first 3 month period, 060% of the
1 mg group and 33% of the 9 mg group broke the code
because of insufficient therapeutic effect. It was also
noted that reduction in immunoglobulins IgM and IgG
and reduction in erythrocyte sedimentation rate (ESR)
were greater for patients receiving the 9 mg dose.
Conclusions from this interim assessment were that
1 mg was insufficient for therapeutic effect, but 9 mg
caused sufficient diarrhoea to make this dosage
unsuitable. In a separate study that compared different
dosage schedules [87], a 9 mg dose was again found to
cause excessive toxicity while 1 mg (or 2 mg) per day
appeared to be inadequate. Since the publication of
these data, most investigators opted for dosages of
3–6 mg/day; but even within this range differences were
evident, with generally greater benefit at the higher
dosage that was offset by higher incidences of adverse
effects (primarily diarrhoea and skin rashes) that often
necessitated a reduction to the lower dosage schedule.
Using multivariate analysis, Champion et al. [88]
demonstrated that little was to be gained by increasing
auranofin dosage above 6 mg/day.
These early studies were followed by several RCTs
of generally high quality that (i) compared auranofin to
placebo, (ii) examined similarities and differences
between auranofin and i.m. gold injections, and (iii)
tried to identify a specific place for auranofin in the
overall management of RA by comparing its
properties, effectiveness and side-effect profile with
various other traditionally used DMARDs.
Auranofin vs placebo
Grade I evidence favouring the preference of
auranofin over placebo has been provided by several
RCTs. While variations in study design and differences
in the selection of outcome measures have precluded
any firm consensus on the effectiveness of oral gold,
trends inferred from each of the studies have, in
general, demonstrated the superiority of auranofin over
placebo.
Under the aegis of the Cooperating Systemic Studies
of Rheumatic Diseases (CSSRD), Ward et al. [45]
compared auranofin, gold sodium thiomalate (GST)
and placebo in 193 patients with RA. Results after 20
weeks of treatment showed a statistically significant
improvement in pain/tenderness scores for patients on
either auranofin or GST (but not for those on placebo).
No significant differences were detectable for joint
swelling, patient assessment or physician assessment
[45].
Three separate studies have compared auranofin
with placebo over a 6 month time period. In 1982, Katz
et al. [89] reported a randomized double-blind
controlled study of 242 patients who received 3 months
of therapy and 144 patients who received 6 months of
therapy with either 3 mg b.i.d. of auranofin or placebo.
Significant improvement in the treated group was
recorded for the number of tender joints at 3 and 6
months, and for the number of swollen joints and
increase in grip strength at 6 months. The investigators’
global assessment of efficacy was recorded as
significantly improved at 3 and 6 months in the
auranofin group for those patients recorded as having
a marked improvement. The authors concluded that
the addition of auranofin to non-steroidal anti-inflammatory drug (NSAID) therapy added to the benefit
derived from the latter in the treatment of RA [89]. In
the 340 patients studied by Wenger et al. [90],
significant improvements were demonstrated in grip
strength, early morning stiffness, number of swollen
joints, number of tender joints and physicians’
assessments of global response. Bombardier and the
Auranofin Cooperating Group [91] utilized a comprehensive spectrum of clinical, functional, pain and
global impression measures in their follow-up of 303
patients. Significantly greater improvement in the
auranofin group (vs placebo) was recorded for the
composite scores computed for the clinical, functional
and global impression measures. Improvements were
demonstrated for all of the clinical variables (i.e.
number of swollen joints, number of tender joints, 50
feet walking time, early morning stiffness and grip
strength).
One 24 month follow-up study has been reported.
Borg et al. [92] followed 138 patients with ‘early RA’
(as defined by joint symptoms for Q 2 yr). Improvements at the end of this time period were recorded for
number of swollen joints, the Keitel Functional Test,
the Stanford Health Assessment Questionnaire and the
Beck Depression Scale. However, no differences were
detectable for grip strength, early morning stiffness or
number of tender joints.
Auranofin vs i.m. gold injections
Grade II evidence from two open comparative
studies [93, 94] suggested that auranofin was less
effective than (i.m.) GST over a 2 yr period. Also,
except for a higher frequency of gastrointestinal
side-effects (especially diarrhoea), the authors concluded that GST was more toxic than the oral gold
preparation, based on the number of adverse responses
noted.
Various RCTs (providing Grade I evidence) have
compared the effectiveness and relative toxicities of
auranofin and GST. In 1983, Ward and colleagues who
comprised the Cooperative Systematic Studies of
Rheumatic Diseases group [45] reported a prospective,
controlled, double-blind multicentre trial which compared placebo, auranofin and gold sodium thiomalate.
Of the 208 patients who fulfilled entry criteria, 193 were
eligible for study. A total of 161 patients completed
20 weeks of therapy. When gold sodium thiomalate
was compared to placebo, there was a significant
KEAN ET AL.: AURANOFIN
improvement over placebo for number of tender joints,
joint tenderness score, joint swelling score, increase in
haemoglobin, fall in ESR and fall in platelet count.
Skin rash was the most common cause for withdrawal
from therapy (10%), followed by stomatitis (5%),
nitritoid reactions (4%), abnormal liver enzymes (4%),
thrombocytopenia (2%), proteinuria (2%) and individual patients with diarrhoea, leucopenia and
pneumonitis. Three patients had rash plus either
thrombocytopenia, leucopenia or stomatitis.
When auranofin was compared to placebo, significant improvement was recorded for number of tender
joints, pain tenderness score, physician’s global
assessment of disease activity and decrease in ESR.
Adverse reactions accounted for the withdrawal of 6%
of patients from the auranofin group due to one
each of rash, diarrhoea, stomatitis, eosinophilia and
leucopenia.
A comparison of auranofin vs gold sodium
thiomalate in the above study by Ward et al.
demonstrated that the injectable GST was superior to
the oral gold drug for improvement in anaemia and
thrombocytosis. Both auranofin and GST were
superior to placebo for improvement in number of
tender joints, joint pain/tenderness score, physician’s
overall assessment and ESR. Although statistical
significance was not achieved, the authors indicated
that GST produced a 12% greater improvement in
joint pain/tenderness score and a 32% advantage
relative to joint swelling score. The authors concluded
that Type II error was possible with the small sample
sizes, thus masking a significant indicator of benefit of
GST over auranofin for these variables. Ward et al.
concluded that GST does have a therapeutic advantage
over auranofin, although the oral gold preparation has
less side-effects leading to cessation of therapy. An
overall assessment of trials of auranofin vs placebo and
auranofin vs injectable gold compounds would support
these conclusions by Ward et al. In many patients
treated with injectable gold, individual observers will
discontinue the drug following the development of
rash, mouth ulcer or proteinuria Q1000 mg/24 h. In
contrast, many observers will merely temporarily
withhold the gold drug until the side-effect has cleared
or modified and then re-introduce the drug. Thus,
inter-observer variations and clinical variations have to
be considered, especially in multicentre studies of
toxicity and when two or more trials are compared. It
may, therefore, be a false conclusion that there is an
increase in the number of adverse effects resulting in
withdrawal in the GST-treated patients over the
auranofin patients. Following a 12 month open study
that was initiated after completion of the RCT by
Ward et al., it was suggested that auranofin had a
long-term use profile that was similar to other
DMARDs. Moreover, it appeared that auranofin
could sustain an initial response to GST [95].
Davis et al. [96] compared the therapeutic benefits
and toxicity profiles of auranofin and GST in 120
patients with RA who were followed for 12 months. A
similar number of patients (060%) remained on each
567
of the two preparations until the termination of the
study and showed a similar statistically significant
improvement in all of the measured clinical variables.
No significant differences between the two drugs were
detected. However, withdrawal from the study because
of lack of therapeutic benefit was twice as frequent in
the group of patients receiving auranofin, while
attrition resulting from an adverse drug reaction was
twice as common and potentially more serious in those
receiving GST.
In a 12 month comparative study on 122 patients,
Rau et al. [97] did not detect any differences in clinical
outcome between patients on auranofin and GST,
although the effects on grip strength, articular index
and ESR were more pronounced in the GST group.
Withdrawals because of adverse effects occurred more
frequently in patients on GST.
Auranofin vs other DMARDs
Grade II evidence, based on two small single-blind
trials comparing auranofin, GST and -penicillamine
[98], and auranofin and -penicillamine [99], respectively, demonstrated a trend towards efficacy of all of
the target DMARDs, with auranofin producing the
least toxicity in both studies.
Auranofin and -penicillamine have also been
compared in an RCT that has provided Grade I
evidence [100]. In a 12 month study, 90 patients with
RA were randomized to receive auranofin, 6 mg/day or
a gradually increasing dose of -penicillamine (to a
maximum of 750 mg/day). While patients in both
groups completed the study with significant improvements in most measures of efficacy, those in the
-penicillamine group were considered to have done
better in terms of physician global assessments, swollen
joint counts, swollen joint scores and grip strength.
Thrombocytopenia and proteinuria occurred with
greater frequency in the -penicillamine group than in
patients on auranofin and, in general, more patients
had to withdraw from the study because of adverse
effects from -penicillamine than was evident with
auranofin. These results inferred that in the dosage
ranges that were used, -penicillamine was more
effective than auranofin, but was also somewhat more
toxic.
An open study of sulphasalazine and auranofin in
patients with RA (yielding Grade II evidence) has
demonstrated comparable efficacy and a similar
frequency of adverse effects for the two drugs [101].
When comparing methotrexate and auranofin,
convincing Grade I evidence favouring the use of
weekly treatment with low-dose methotrexate has been
provided by a 36 week RCT [102]. While patients in
both groups improved over the study period, responses
in those on methotrexate tended to occur earlier and
were consistently greater than in those who were on
auranofin. Moreover, adverse reactions requiring
withdrawal from the study were more common in the
auranofin group. Data from this study prompted the
conclusion that methotrexate was more effective and
less toxic than auranofin. In a further RCT, comparing
568
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
methotrexate, auranofin and a combination of the two
[103], no statistically significant differences in clinical or
laboratory outcome were detected for the three groups,
although patients on methotrexate or combination
therapy tended to respond more quickly than those on
auranofin alone. The frequency of adverse effects was
similar for the three groups, but slightly more common
in those taking both drugs.
Meta-analyses examining the comparative efficacy
and toxicity of DMARDs have been reported [104].
For the efficacy study, tender joint counts, ESRs and
grip strength evaluations constituted the target set of
outcomes. Sixty-six clinical trials containing 117
treatment groups of interest were identified. For each
of the three outcomes, auranofin tended to exert less
benefit than injectable gold, -penicillamine, sulphasalazine, antimalarials and methotrexate. For the
purposes of the toxicity studies, 71 trials incorporating
129 treatment groups were examined. Average
proportions of study drop-outs and the proportions
who were withdrawn because of drug side-effects were
calculated for each DMARD. Using this approach, it
was found that injectable gold had significantly higher
toxicity rates (P Q 0.05) and higher total drop-out rates
than any of the other drugs (P Q 0.01). Thirty per cent
of patients treated with injectable gold dropped out
because of side-effects as compared to 15% of all trial
patients.
The impact of auranofin on the radiographic manifestations of joint destruction
Whether one or another DMARD actually slows the
radiological progression of joint damage is a matter of
ongoing debate. The interpretation of most of the
studies that have addressed this important issue has
been complicated by a pervasive lack of conformity in
criteria used to read and report the radiological lesions.
As a result, there has been little agreement on the
impact of these drugs [70].
Various studies on auranofin have examined possible
correlations between drug efficacy and slowing of
radiological progression of joint lesions. Grade II
evidence by Gofton and O’Brien [105] and Gofton
et al. [106], and Grade I evidence by Borg et al. [92]
have suggested that auranofin may reduce the
progression of articular erosions. In contrast, studies
by Caruso et al. [107] and Lopez-Mendez et al. [108]
provide Grade II and Grade I evidence, respectively,
that auranofin has no effect on the retardation of
adverse joint changes.
Auranofin use in children and teenagers
In an open study by Giannini et al. in 1983 [109], 21
children aged 1–17 yr were treated with 0.1–0.2 mg/kg/
day of auranofin. The authors claimed a significant
clinical improvement (25%) in more than half of the
children. This included number and severity of joints
with swelling, pain on motion and tenderness. The
beneficial response was greater in those children taking
the higher doses. Only 2/21 had to discontinue therapy:
one because of headaches and one because of
haematuria, anaemia and a flare in disease activity.
Three other children had side-effects which required
dose reduction. These were proteinuria, diarrhoea, and
one with haematuria and anaemia.
CONCLUSIONS
Choosing auranofin or i.m. gold
Injectable gold compound 50 mg i.m. weekly is
similar in cost to auranofin 3 mg twice daily. When
maintenance therapy is started at 05 or 6 months, the
injectable gold is usually given as 50 mg i.m. monthly,
whereas maintenance with auranofin is continued at
3 mg once or twice daily. Thus, oral auranofin will have
a more expensive maintenance program. Fortunately,
for the majority of patients, cost is not a major issue.
Effectiveness is probably one of the principal factors
influencing decision making in drug administration. If
given in high enough dosage of 9 mg/day, the
immunosuppressive actions of auranofin compound
should result in a faster onset of action over the
injectable gold compounds, but toxicity in the form of
diarrhoea generally precludes the use of this dosage [32,
86]. Auranofin in a dosage of 3–6 mg/day does not
appear to have a faster onset of action than injectable
gold compounds [33, 45] and, indeed, the reverse is
true: that a faster more meaningful and clinically
measured response is achieved with the injectable gold
compounds when compared to the oral auranofin [33,
45, 97]. The convenience of oral medication is a much
discussed topic, especially in relation to compliance,
but the compliance issue is best reserved for
non-symptom-driven usage, such as blood pressure
control or oral contraception. Patients with pain are
usually compliant, those who are not will never be,
irrespective of the route of administration. The
convenience of oral auranofin is a small concession
which is outweighed by the benefits accrued from
injectable gold therapy. The process of the patient
proceeding to the ‘gold clinic’ to receive the injection
increases the insurance that someone with medical
expertise will question the patient regarding drug
benefit and adverse effects and will test blood and urine
for potential abnormalities. Thus, the patient is more
likely to have a regular ‘hands-on’ management of their
treatment and will perhaps also benefit from the
additional placebo effect recorded by the late Dr N.
Thomas Fraser [110]. Patient compliance with gold
injection treatment is reinforced by physicians and staff
keeping a weekly record of the drug administration,
clinical response, and blood and urine measurements.
The patient is provided with a copy of these results and
will thus be more intimately involved in their own care,
leading to increase in compliance. Toxicity monitoring
with regards to auranofin is usually carried out only on
a monthly basis.
Clinical studies have documented a greater frequency of adverse reactions for injectable gold
compounds compared to auranofin [33, 34, 97], but the
quality and relevance of these side-effects has to be
considered. The most common side-effect of injectable
gold is rash, and in the context of clinical trials this is
KEAN ET AL.: AURANOFIN
recorded in 030–50% of those treated. However, in
clinical practice, the majority of cases of rash are
extremely mild and patients merely require a reduction
in dosage or a temporary discontinuation of therapy
without major interference or termination of treatment.
The most common side-effect of auranofin is diarrhoea
[32, 33, 45, 86, 97] which is also dose related. Patients
with severe RA, like colchicine-treated gout sufferers,
do not thank their clinicians for the diarrhoea
side-effect of auranofin. Reduction in the dose of
auranofin to reduce the potential for diarrhoea reduces
efficacy [86]. Haematological side-effects are uncommon with both injectable compounds and with
auranofin. The total quantity of adverse haematological effects is probably less with auranofin than with
the injectable gold compounds. The mechanism of
auranofin-induced thrombocytopenia is probably central, whereas the most common type of injectable
gold-induced thrombocytopenia is peripheral and
associated with IgG surface autoantibodies [54]. Bone
marrow-induced thrombocytopenia from injectable
gold compounds is commonly a precursor to the onset
of aplastic anaemia and is fortunately rare. The renal
side-effects of auranofin are less common than those of
injectable compounds [33, 45, 97], but since the
consequences of gold compound-induced proteinuria
are almost always of no serious outcome, the difference
in side-effect numbers is not relevant.
When the decision is made to treat a RA patient with
a gold compound, the greater effectiveness of the
injectable compound may therefore take precedence
over the quantity of side-effects recorded in favour of
auranofin.
Inferences from the clinical studies
Although there is general agreement that high-quality RCTs usually confer satisfactory rigour when
determining drug efficacy and toxicity, and while it is
encouraging when cumulative Grade I evidence
provides consistent results on outcomes of interest,
there nonetheless needs to be caution in the application
of even the most robust data. When embarking on a
treatment programme for patients with RA, it needs to
be acknowledged that despite an expanding armamentarium of NSAIDs, DMARDs and newer biological
agents, the overall prognosis for a majority of patients
remains less than optimal. RA is still a rather severe
and progressive chronic disease that has a generally
poor prognosis [110].
Pincus has drawn attention to the paradox in the
rheumatology literature where apparently effective
therapies in clinical trials do not necessarily translate
into successful outcomes when longer-term observational studies are scrutinized [111]. Although most
clinical trials in RA are completed within a 24 month
time frame, their results not only direct long-term
therapy, but also dictate the place of specific drugs
within the therapeutic hierarchy. In reality, all studies
in which patients with RA have been followed for 5 yr
or more have demonstrated significant morbidity and
accelerated mortality rates, regardless of treatment.
569
Treatment with auranofin should be commenced as
early as possible after the definite diagnosis of
rheumatoid disease, once it has been established that
the disease symptoms are not responsive to adequate
treatment with non-steroidal anti-inflammatory agents.
In view of the potential for toxicity, a strict monitoring
system should be applied. Treatment and management
should be conducted as outlined above with the use of
a flexible regimen. Oral gold therapy in children should
be under the supervision of a physician with in-depth
knowledge of both paediatrics and gold therapy.
The available evidence, based on relatively shortterm trials, seems to infer that auranofin (i) is
efficacious when compared to placebo, (ii) is less
effective than other DMARDs, (iii) causes fewer
serious side-effects than other DMARDs and (iv) its
impact on the radiological progression of disease has
not been unequivocally determined. A more compelling
question, of course, is whether auranofin, alone or in
combination with other DMARDs, has any positive
impact on disease prognosis in the long term. Only
large cohort prospective studies will answer the
question as to whether DMARDs do modify the
outcome of the arthritis over the long term. To date,
this question remains unanswered.
R
1. Wiegleb JC. Historisch-Kritische Untersuchung der
Alchemie. Unveranderter Nachdruck der Originalausgable 1965;1777:185. Leipsig: Zentral-Antiquarit der
Deutschen Demokratischen Republik.
2. Waley A. The way and its power: a study of the Tao
teaching and its place in Chinese thought. London:
Allen and Unwin, 1956.
3. Fung YL. A history of Chinese philosophy. Translated
by Derk Bodde. 1952:Chapter 8. Princeton: Princeton
University Press.
4. Exodus. The holy bible. Toronto: Oxford University
Press, Chapter 32, Verse 20.
5. Pliny GP. Historia naturalis, book 33, as cited by
Garrison FH. Introduction to the history of medicine,
4th edn. Philadelphia: WB Saunders, 1966:112.
6. Dioscorides P. De materia medica, cited by Garrison
FH. Introduction to the history of medicine, 4th edn.
Philadelphia: WB Saunders, 109–10.
7. Pattison Muir MM. Roger Bacon. his relations to
alchemy and chemistry. In: Little AG, ed. Roger Bacon
essays. New York: Russell and Russell, 1972:285–320.
8. Hunt LB. The Figuiers of Montpeller. Ambix 1979;
26:221–3.
9. Koch R. Uber Bacteriologische Forschung. Tenth
International Medical Congress, Berlin. Dtsch Med
Wochenschr 1980;16:756–7.
10. Forestier J. L’Aurothiopie dans les rhumatisme
chronique. Bull Mem Soc Med Hop Paris, 1929;53:
323–7.
11. Puddephatt RJ. The chemistry of gold. New York:
Elsevier, 1978:1.
12. Mohide TP. Gold. Ontario Ministry of Natural
Resources, Mineral Policy Background Paper No. 12,
1981;246–52.
13. Pearson RG. Hard and soft acids and bases. J Am
Chem Soc 1973;85:3533–9.
14. Sadler PJ. The comparative evaluation of the physical
570
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
and chemical properties of gold compounds. J
Rheumatol 1982;9:71–8.
Elder RC, Eidsness MK, Heeg MJ, Tepperman KG,
Shaw CF, Schaeffer N. Gold-based antiarthritic drugs
and metabolities. Am Chem Soc Symp Ser 1983;
209:385–90.
Shaw III FC. The mammalian biochemistry of gold: An
inorganic prospective of chrysotherapy. Inorganic
Perspect Biol Med 1979;2:287–355.
Blocka AK. Auranofin versus injectable gold. Comparison of pharmacokinetic properties. Am J Med
1983;75:114–22.
Gottlieb NL, Smith PM, Smith EM. Pharmacodynamics of 195Au-labelled aurothiomalate in blood. Arthritis
Rheum 1974;17:171–83.
Tepperman K, Finer R, Donovan S et al. Intestinal
uptake and metabolism of auranofin, a new oral
gold-based antiarthritis drug. Science 1984;225:430–2.
Intoccia AP, Flanagan TL, Walz DT et al. Pharmacokinetics of auranofin in animals. J Rheumatol 1982;9:
90–8.
Kean WF, Lock CJL, Howard-Lock H. Gold complex
research in medical science. Difficulties with experimental design. Inflammopharmacology 1991;1:103–14.
Walz ET, Dimartino MJ, Griswold DE, Intoccia AP,
Flanagan TL. Biologic actions and pharmacokinetic
studies of auranofin. Am J Med 1983;75:90–108.
Walz DT, DiMartino MJ, Juch JH, Zuccarello W.
Adjuvant-induced arthritis in rats. I. Temporal
relationship of physiological, biochemical and haematological parameters. Proc Soc Exp Biol Med
1971;136:907–10.
Kean WF, Kassam YB, Lock CJL et al. Antithrombin
activity of gold sodium thiomalate. Clin Pharmacol
Ther 1984;35:627–32.
Salmerton G, Lipsky PE. Effect of gold compounds on
human mononuclear phagocyte function. In: Schattenkirchner M, Muller W, eds. Modern aspects of gold
therapy. Basel: Karger, 1983:63–74.
Percy JS, Davis P, Russell AS, Brisson E. A longitudinal
study of in vitro tests for lymphocyte function in
rheumatoid arthritis. Ann Rheum Dis 1978;37:416–20.
Lorber A, Simon T, Leeb J et al. Chrysotherapy.
Suppression of immunoglobulin synthesis. Arthritis
Rheum 1978;21:785–91.
Simon TM, Kunishmia DH, Biert GJ, Lorber A.
Inhibitory effects of a new oral gold compound on
HeLa cells. Cancer 1978;44:1865–75.
Crooke ST. A comparison on the molecular pharmacology of gold and platinum complexes. J Rheumatol
1982;9:61–70.
Van Riel PL, Gribnau FW, Van De Putte LB, Yap SH.
Loose stools during auranofin treatment: clinical study
and some pathogenic possibilities. J Rheumatol
1983;10:222–6.
Bandilla KK, Delattre M, Rahn B, Missler B.
Long-term treatment of rheumatoid arthritis with
auranofin: clinical results with auranofin from German
and international studies which included a comparison
of once and twice daily treatment. In: Capell HA, Cole
DS, Manghani KK, Morris RW, eds. Auranofin.
Amsterdam: Excerpta Medica, 1983:97–14.
Rangachari PK, Kean WF. Gold and D-penicillamine
and the gastrointestinal tract. Clin Rheumatol 1989:
3:411–23.
Van Riel PLCM, Van de Putte LBA, Gribnau FWJ
et al. A single-blind comparative study of auranofin and
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
gold thioglucose in patients with rheumatoid arthritis.
In: Capell HA, Cole DS, Manghoni KK, Morris RW,
eds. Auranofin. Excerpta Medica, 1983:135–46.
Behrens R, Devereaux M, Hazeleman B et al.
Investigation of auranofin-induced diarrhoea. Gut
1986;27:59–65.
Nagler J, Paget SA. Non-exudative diarrhoea after gold
salt therapy: Case report and review of the literature.
Am J Gastroenterol 1983;78:12–4.
Hardcastle J, Hardcastle PT, Kelleher DK, Fondacaro
JD. The effect of auranofin on the colonic transport of
Na+ and fluid in the rat. J Pharm Pharmacol
1986;38:466–8.
Fondacaro JD, Henderson LS, Hardcastle PT et al.
Effects of auranofin (SKF 39162) on water and
electrolyte flux in canine small bowel: A possible
diarrhoegenic mechanism. J Rheumatol 1986;13:541–6.
Ammon HV, Fowle SA, Cunningham JA, Komorowski
RA, Loeffler RF. Effects of auranofin and myochrysine
on intestinal transport and morphology in the rat. Gut
1987;28:829–34.
De Beaux A, Keenan CM, Tytgat K, Rangachari PK.
Responses of the colonic epithelium to auranofin:
evidence for involvement of enteric nerves. J Rheumatol
1990;9:1137–41.
Rangachari PK, McWade D. Epithelial and mucosal
preparations of canine proximal colon in Ussing
chambers: comparison of responses. Life Sci 1986;
38:1641–52.
Snyder RG, Traeger C, Kelly L. Gold therapy in
arthritis: observations in 100 cases treated with gold
sodium thiosulphate and aurocein. Ann Intern Med
1939;12:1672–81.
Whittle BJR, Vane JR. Prostanoids as regulators of
gastrointestinal function. In: Johnson LR, ed. Physiology of the gastrointestinal tract, 2nd edn. New York:
Raven Press, 1987:143–80.
Powell DW. Intestinal water and electrolyte transport.
In: Johnson LR, ed. Physiology of the gastrointestinal
tract, 2nd edn. New York: Raven Press, 1987:1267–305.
Morris RW, Cole DS, Horton J, Hever MA, Pietrusko
RG. Worldwide clinical experience with auranofin. Clin
Rheumatol 1984;3(suppl. 10):105–12.
Ward JR, Williams NJ, Egger MJ et al. Comparison of
auranofin, gold sodium thiomalate, and placebo in the
treatment of rheumatoid arthritis. Arthritis Rheum
1983;26:1303–15.
Kean WF, Anastassiades TP. Long term chrysotherapy.
Incidence of toxicity and efficacy during sequential time
periods. Arthritis Rheum 1979;22:495–501.
Kean WF, Bellamy N, Brooks PM. Gold therapy in the
elderly rheumatoid patient. Arthritis Rheum 1983;26:
705–11.
Co-operating Clinics Committee. A controlled trial of
gold salt therapy in rheumatoid arthritis. Arthritis
Rheum 1973;16:353–8.
Furst D, Levine S, Srinivasan R et al. A double-blind
trial of high versus conventional dosages of gold salts
for rheumatoid arthritis. Arthritis Rheum 1977;20:
1473–80.
Tornrot T, Skrifvars B. Gold nephropathy prototype of
membranous glomerulonephritis. Am J Pathol 1974;
75:573–84.
Merle LJ, Reidenberg MM, Camacho MT, Jones BR,
Drayer DE. Renal injury in patients with rheumatoid
KEAN ET AL.: AURANOFIN
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
arthritis treated with gold. Clin Pharmacol Ther
1980;28:216–22.
Panayi GS, Wooley P, Batchelor JR. Genetic basis of
rheumatoid disease: HLA antigens, disease manifestations, and toxic reactions to drugs. Br Med J
1978;2:1326–8.
Kelton JG, Carter CJ, Rodger C. The relationship
between platelet-associated IgG, platelet life-form and
reticuloendothelial cell function. Blood 1984;63:1434–8.
Howell A, Gumpel JM, Watts RWE. Depression of
bone marrow colony formation in gold induced
neutropenia. Br Med J 1975;1:432–4.
Bellelli A, Veneziani M, Tumiati B. Felty’s syndrome:
long term followup after treatment with auranofin.
Arthritis Rheum 1987;30:1057–61.
Job-Deslandre C, Menkes CJ. Treatment of Felty’s
syndrome with auranofin and methylprednisolone.
Arthritis Rheum 1989;32:1188–9.
Heur MA, Morris RW. Smith Kline & French
worldwide clinical experience with auranofin: A review.
In: Capell HA, Cole DS, Manghani KK et al., eds.
Auranofin: proceedings of a Smith Kline & French
international symposium. Amsterdam: Excerpta
Medica, 1983:474–503.
Miyamato TS et al. Gold therapy in bronchial
asthma—special emphasis on blood levels of gold and
its teratogenicity. J Jpn Soc Intern Med 1974;63:1190–7.
Rogers JG et al. Possible teratogenic effect of gold. Aust
Paediatr J 1980;16:194–5.
Cohen DL, Orzel J, Taylor A. Infants of mothers
receiving gold therapy. Arthritis Rheum 1981;24:104–5.
Tarp U, Graudal H. A follow up study of children
exposed to gold in utero. Arthritis Rheum 1985;28:
235–6.
Rocker I, Hendeson MJH. Transfer of gold from
mother to foetus. Lancet 1976;2:1246.
Richards, AJ. Transfer of gold from mother to foetus.
Lancet 1977;1:99.
Ostensen M, Husby G. Anti-rheumatic drug treatment
during pregnancy and lactation. Scand J Rheumatol
1985;14:1–7.
Brooks PM, Kean WF, Buchanan WW. The clinical
pharmacology of anti-inflammatory agents. London:
Taylor and Francis, 1986:91–7.
Blau S. Metabolism of gold during lactation. Arthritis
Rheum 1973;16:777–8.
Ostensen M, Skavdal K, Myklebust G et al. Excretion
of gold in human breast milk. Eur J Clin Pharmacol
1986;31:261.
Davis P, Barraclough D. Interaction of D-penicillamine
with gold salts. Arthritis Rheum 1977;20:1413–8.
Kean WF, Lock CJL. Penicillamine does not chelate
gold (I). J Rheumatol 1983;10:527–30.
Hart LE, Tugwell P. The use of disease modifying
antirheumatic drugs in the management of rheumatoid
arthritis. Postgrad Med J 1989;65:905–12.
Iannuzi L, Dawson N, Zein N et al. Does drug therapy
slow radiographic deterioration in rheumatoid arthritis?
N Engl J Med 1983;309:1023–8.
Anderson J, Felson DT, Meenan RF, Williams HJ.
Which traditional measures should be used in
rheumatoid arthritis clinical trials? Arthritis Rheum
1989;32:1093–9.
Scott DL, Coulton BL, Bacon PA, Popert AJ. Methods
of x-ray assessment in rheumatoid arthritis: A
re-evaluation. Br J Rheumatol 1985;24:31–9.
Hart LE. Arthritis. In: Basmajian JV, Banerjee SN, eds.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
571
Clinical decision making in rehabilitation: efficacy and
outcomes. Churchill Livingstone, New York. 1996:203–
22.
Moher D, Jadad AR, Nichol G et al. Assessing the
quality of randomized controlled trials: an annotated
bibliography of scales and checklists. Controlled Clin
Trials 1995;16:62–73.
Sackett DL, Haynes RB, Guyatt GH, Tugwell P (eds)
Clinical epidemiology. A basic science for clinical
medicine, 2nd edn. Boston: Little Brown, 1991.
MacPherson DW. Evidence-based medicine. Can
Commun Dis Rep 1994;20:145–7.
Sackett DL. Rules of evidence and clinical recommendations. Can J Cardiol 1993:9:487–9.
Sutton BM, McGustry E, Waltz DT, Dimartino MJ.
Oral gold. Antiarthritic properties of alkylphosphine
gold coordination complexes. J Med Chem 1972;
15:1095–8.
Walz DT, Dimartino MJ, Sutton BM, Misher A. S K
& F 36914—an agent for oral chrysotherapy. J
Pharmacol Exp Ther 1972:181:292–7.
Walz DT, Cimartino MJ, Chakrin LW, Sutton BM,
Misher A. Antiarthritic properties and unique pharmacologic profile of a potential chrysotherapeutic agent: S
K & F D-39162. J Pharmacol Exp Ther 1976;197:
142–52.
Finkelstein AE, Walz DT, Batista V et al. Auranofin.
New oral gold compound for the treatment of
rheumatoid arthritis. Ann Rheum Dis 1976;35:251–7.
Homma M, Abe T, Akizuki M et al. Interim results of
a multicenter open study with auranofin in Japan. J
Rheumatol 1982;9:160–8.
Bandilla K, Gross D, Gross W et al. Oral gold therapy
with auranofin (SK&F 39162): A multicenter open
study in patients with rheumatoid arthritis. J Rheumatol 1982;9:154–9.
Bernhard GC. Auranofin treatment for adult rheumatoid arthritis: Comparison of 2 mg and 6 mg daily
dose. J Rheumatol 1982;9:149–53.
Calin A, Saunders D, Bennett R et al. Auranofin: 1 mg
or 9 mg? The search for the appropriate dose. J
Rheumatol 1982;9:146–8.
Weiss TE. Auranofin: Dose-related risk to benefit. Am
J Med 1983;75:128–32.
Champion GD, Cairns DR, Bieri D et al. Dose response
studies and longterm evaluation of auranofin in
rheumatoid arthritis. J Rheumatol 1988;15:28–34.
Katz WA, Alexander S, Bland JH et al. The efficacy and
safety of auranofin compared to placebo in rheumatoid
arthritis. J Rheumatol 1982;9:173–8.
Wenger ME, Alexander S, Bland JH, Blechman WJ.
Auranofin versus placebo in the treatment of rheumatoid arthritis. Am J Med 1983;75(6A):123–7.
Bombardier C, Ware J, Russell IJ et al. Auranofin
therapy and quality of life in patients with rheumatoid
arthritis. Am J Med 1986;81:565–78.
Borg G, Allander E, Lund B et al. Auranofin improves
outcome in early rheumatoid arthritis: Results from a
2-year, double blind, placebo controlled study. J
Rheumatol 1988;15:1747–54.
Harth M, Davis P, Thompson JM et al. Comparison
between sodium aurothiomalate and auranofin in
rheumatoid arthritis: Results of a two-year open
randomized study. Scand J Rheumatol 1987;16:177–84.
Smith PR, Brown GMM, Meyers OL. An open
comparative study of auranofin vs. gold sodium
thiomalate. J Rheumatol 1982;9:190–6.
572
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 5
95. Williams HJ, Dahl SL, Ward JR et al. One-year
experience in patients treated with auranofin following
completion of a parallel, controlled trial comparing
auranofin, gold sodium thiomalate, and placebo.
Arthritis Rheum 1988;31:9–14.
96. Davis P, Menard H, Thompson J et al. One-year
comparative study of gold sodium thiomalate and
auranofin in the treatment of rheumatoid arthritis. J
Rheumatol 1985;12:60–7.
97. Rau R, Schattenkirchner M, Muller-Fassbender H et al.
A three year comparative study of auranofin and gold
sodium thiomalate in rheumatoid arthritis. Clin
Rheumatol 1990;9:461–74.
98. Barraclough D, Brook A, Brooks P et al. A
comparative study of auranofin, gold sodium thiomalate, and d-penicillamine in rheumatoid arthritis: A
progress report. J Rheumatol 1982;9:197–200.
99. Felix-Davies DD, Stewart AM, Wilkinson BR et al. A
12-month comparative trial of auranofin and d-penicillamine in rheumatoid arthritis. Am J Med 1983;
75(6A):138–41.
100. Hochberg MC. Auranofin or d-penicillamine in the
treatment of rheumatoid arthritis. Ann Intern Med
1986;105:528–35.
101. Porter D, Madhok R, Hunter JA, Capell HA.
Prospective trial comparing the use of sulphasalazine
and auranofin as second line drugs in patients with
rheumatoid arthritis. Ann Rheum Dis 1992;51:461–4.
102. Weinblatt ME, Kaplan H, Germain BF et al. Low-dose
methotrexate compared with auranofin in adult
rheumatoid arthritis: A thirty-six-week, double-blind
trial. Arthritis Rheum 1990;33:330–8.
103. Williams HJ, Ward JR, Reading JC et al. Comparison
of auranofin, methotrexate, and the combination of
104.
105.
106.
107.
108.
109.
110.
111.
both in the treatment of rheumatoid arthritis: A
controlled clinical trial. Arthritis Rheum 1992;35:
259–69.
Felson DR, Anderson JJ, Meenan RF. The comparative
efficacy and toxicity of second-line drugs in rheumatoid
arthritis. Arthritis Rheum 1990;33:1449–61.
Gofton JP, O’Brien WM. Effects of auranofin on the
radiological progression of joint erosion in rheumatoid
arthritis. J Rheumatol 1982;9:169–72.
Gofton JP, O’Brien WM, Hurley JN, Scheffler BJ.
Radiographic evaluation of erosion in rheumatoid
arthritis: Double blind study of auranofin vs placebo.
J Rheumatol 1984;11:768–71.
Caruso J, Santandrea S, Sarzi Puttini P et al. Prevention
of appearance of radiological lesions in early rheumatoid arthritis: A randomized, single-blind study
comparing intra-articular rifamycin with auranofin. J
Int Med Res 1992;20:61–77.
Lopez-Mendez A, Daniel WW, Reading JC et al.
Radiographic assessment of disease progression in
rheumatoid arthritis patients enrolled in the cooperative
systematic studies of the rheumatic disease program
randomized clinical trial of methotrexate, auranofin, or
a combination of the two. Arthritis Rheum 1993;
36:1364–9.
Giannini EH, Brewer EJ, Person DA. Auranofin in the
treatment of juvenile rheumatoid arthritis. J Rheumatol
1983;102:138–41.
Fraser TN. Gold treatment in rheumatoid arthritis. Ann
Rheum Dis 1945;4:71–5.
Pincus T. Rheumatoid arthritis: Disappointing longterm outcomes despite successful short-term clinical
trials. J Clin Epidemiol 1988;41:1037–41.