566
cultured with CsA, but can be partially protected from the toxic
effects of the drug by either removing Ca2’ from the growth
medium or co-treating the cultured cells with calcium channel
blockers (7).
A unifying hypothesis is that activation of Na+/K+-
ATPase by the Ca2+-regulated phosphatase calcineurin is an
important mechanism for regulating intracellular concentra-
tions of Ca2+.Inhibition of calcineurin by CsA would lead to
an overall decrease of Na+/K+ pump activity, and thus in-
crease the intracellular concentration of Ca2+, leading to the
pathologic activation of calpain, a Ca2+-dependentproteolytic
enzyme, which in turn would lead to cell death. Any cell type
in joints or bones that had a Na’/K+-ATPase regulated by
calcineurin would be affected by CsA. Furthermore, given this
potential link between calcium-mediated bone pain and nephro-
toxicity, it would be interesting to know whether transplant
patients treated with calcium channel blockers have a de-
creased incidence of nephrotoxicity.
It is fortunate for individuals requiring treatment with
CsA that the mechanism-based toxicities involve different roles
for calcineurin in non-T cells (Na’ /KC-ATPase-dependent
Ca2+ regulation) and T cells (calcium-dependent transcription
factor regulation). This allows treatment of a toxicity (bone
pain) without mitigation of the clinically desirable effect
(inhibition of T cell activation). Calcineurin has also been
implicated as a factor regulating nitric oxide synthase (8) and
dynamin I GTPase in nerve cells (9), and so it might also play
a role in the neurotoxicity associated with CsA treatment.
Edward A. O’Neill, PhD
Merck Research Laboratories
Rahway, NJ
Victor S. Sloan, MD
University of Medicine and Dentist? of New Jersey
New Brunswick, NJ
1. Stone JH, Peterfy CG, Sack KE. Bone pain in a transplant patient.
Arthritis Rheum 1997;40:1361-3.
2. Gauthier VI, Barbosa LM. Bone pain in transplant recipients
responsive to calcium channel blockers. Ann Intern Med 1994;
121536335,
3. Naredo Sanchez E, Balsa Criado A, Sanz Guajardo A, Pantoja
Zarza L, Martin Mola E, Gijon Banos J. Leg bone pain syndrome
due to cyclosporine in a renal transplant patient. Clin Exp Rheu-
matol 1994;12:653-6.
4. Barbosa LM, Gauthier VJ, Davis CL. Bone pain that responds to
calcium channel blockers: a retrospective and prospective study of
transplant recipients. Transplantation 1995;59:541-4.
5. Kincaid RL, O’Keefe SJ. Calcineurin and immunosuppression: a
calmodulin-dependent protein phosphatasc acts as the “gate-
keeper” to interleukin-2 gene transcription. Adv Protein Phospha-
tase 1993;7:543-83.
6. Crabtree GR. Contingent genetic regulatory events in T lymphocyte
activation. Science 1989;243:355-61.
7. Wilson PD. Hartz PA. Mechanisms of cyclosporin A toxicity in
defined culture of renal tubule epithelia: a role for cysteine
proteases. Cell Biol Int Rep 1991;15:1243-58.
8. Dawson TM, Steiner JP, Dawson VL, Dinerman JL, Uhl GR,
Snyder SH. Immunosuppressant FK506 enhances phosphorylation
of nitric oxide synthase and protects against glutamate neurotoxi-
city. Proc Natl Acad Sci U S A 1993;90:9808-12.
9. Liu J-P, Sim ATR, Robinson PJ. Calcineurin inhibition of dynamin
LETTERS
I GTPase activity coupled to nerve terminal depolarization. Science
1994:265:970-3.
Limitations on the usefulness of procalcitonin as a
marker of infection in patients with systemic auto-
immune disease: comment on the article by Eberhard
et a1
To the Editor:
In a recent article in Arthritis & Rheumatim, Eberhard
et a1 (I) reported that procalcitonin (PCT), the 116-amino
acid precursor of calcitonin, is a useful marker for differenti-
ating between systemic autoimmune disease activity and bac-
terial infection. Although we mainly concur with this finding,
we would like to point out that patients with very highly active
underlying disease, but without infection, sometimes also have
markedly elevated serum PCT levels. In our studies on a group
of 26 patients with active generalized Wegener’s granuloma-
tosis, we found 3 patients with PCT serum levels ranging from
0.7 ng/ml to 3.5 ng/ml (Figure l), considerably above the
normal value (upper limit 0.5 ngiml, PCT-LumiTest; Brahms-
Diagnostica, Berlin, Germany). None of the patients had any
signs of infection at the time of serum collection. Initiation of
immunosuppressive therapy brought clinical improvement and
reduced PCT levels in these patients (2). Further limiting the
value of PCT as a marker of infection is the lack of increased
serum levels of PCT in patients with viral infections (3),which
are at least as important as opportunistic bacterial complica-
tions in immunocompromised patients (e.g., cytomegalovirus).
In conclusion, we believe that PCT might be useful as
a marker for inflammation of bacterial origin, but that it can be
PCT serum levels in WG
41
3-
-
E
\
F 2-
c
0
Q
+
1- n = 26
0
active WG inactive WG
Figure 1. Serum procalcitonin (PCT) levels in 26 patients with We-
gener’s granulomatosis (WG). One sample from each patient was
obtained during active disease and 1 during inactive disease. Boxes
show the mean values; bars above boxes show the SD. The difference
between the mean value for active disease and the mean value for
inactive disease was statistically significant (P = 0.02, Wilcoxon
matched pairs test).
LETTERS
misleading in some cases of highly active underlying auto-
immune disease. In our opinion, final diagnosis still must be
based on a combination of laboratory data, including routine
studies and special immunologic tests, clinical findings, and
imaging results.
F. Moosig, MD
E. Reinhold-Keller, MD
E. Csernok, PhD
W. L. Gross, MD
University of Liibeck
Bad Brarnstedt, Germany
Eberhard OK, Haubitz M, Brunkhorst FM, Kliem V, Koch KM,
Brunkhorst R. Usefulness of procalcitonin for differentiation be-
tween activity of systemic autoimmune disease (systemic lupus
erythematosusisystemic antineutrophil cytoplasmic antibody-
associated vasculitis) and invasive bacterial infection. Arthritis
Rheum 1997;40:1250-6.
Moosig F, Reinhold-KellerE, Csernok E, Gross WL. Procalcitonin
as discriminating marker for infections in Wegener’s granulomato-
sis? Submitted for publication.
Assicot M, Gendrel D, Carsin H, Raymond J, Guilbaud J, Bohuon
C. High serum procalcitonin concentrations in patients with sepsis
and infection. Lancet 1993;341:515-8.
Use of procalcitonin as a diagnostic test: comment on
the article by Eberhard et a1
To the Editor:
We read with interest the article by Eberhard and
colleagues on the use of serum determinations of PCT levels to
distinguish between active systemic lupus erythematosus
(SLE) or vasculitis and invasive bacterial infections (1). .A-
though such determinations are a novel approach to solving an
often vexing clinical problem, there are important flaws in the
article that challenge the authors’ conclusions.
The objective of this study was to “investigate. . .
whether serum [PCT] measurements could help differentiate
between systemic infection and activity of underlying disease.”
To accomplish this, the authors studied 35 patients with
“antineutrophil cytoplasmic antibody (ANCA)-associated vas-
culitis” and 18 patients with SLE. However, systemic infection
was present among only 11 of the patients with “ANCA-
associated vasculitis” and among none of the patients with
SLE. Thus, it is not clear how Eberhard et al could tell that
measurements of serum PCT “may serve as a useful marker for
the detection of systemic bacterial infection in patients with
systemic autoimmune disease,” as they claim, since no infec-
tions were present in their SLE patients.
We also have a number of objections related to the
design and analysis of the study. This is a study conducted on
serum samples, not patients. As stated above, there were only
53 patients. But the authors conducted analyses, and show
results, on 397 serum samples, o r an average of 7.4 samples per
patient. Among the 18 SLE patients, 226 samples were studied,
or 12.5 per patient, while among the 35 vasculitis patients,
there were 171 samples, or 4.8 per patient. Eberhard and
colleagues do not specify how many observations were made
on each individual patient, and yet they calculate correlation
567
coefficients and other statistical tests on all of the samples as a
whole, as if each measurement were an independent observa-
tion. Thus, the patients with the most measurements contrib-
ute disproportionately to the results, biasing the findings in a
direction that depends on the characteristics of the patients
who had the highest number of PCT determinations made.
The report also has some basic omissions and format-
ting flaws that could easily have been avoided: statistical tests
that were not mentioned in the Patients and Methods section
are reported in the Results, the type of correlation coefficients
employed is not described, results are presented in the Discus-
sion section, and data are not shown in a way that would permit
confirmation of the results. At one point, the authors state that
the “chi-square test was performed to determine the sensitivity
and specificity” of the variables measured. However, chi-
square is not necessary to calculate sensitivity and specificity,
and the report contains no categorical comparisons for which
the chi-square test would be indicated. Furthermore, data are
not presented in standard 2 X 2 tables that would permit
readers to substantiate the 100% sensitivity in detecting inva-
sive bacterial infections that the authors claim for measure-
ment of serum PCT.
PCT measurements may in fact be of use in distin-
guishing severe systemic bacterial infection from other types of
inflammation (2). However. the article by Eberhard and col-
leagues does not allow that conclusion, because the authors
haven’t shown that PCT levels rise in SLE patients with
infection. The danger of erroneously concluding that serum
PCT determinations can distinguish between bacterial infec-
tion and active autoimmune disease has potential repercus-
sions that go from the individual patient to the population at
large. At the individual patient level, clinicians reading Eher-
hard and colleagues’ finding that serum PCT levels have 100%
sensitivity in identifying bacterial infection could be misled
into failing to treat an infection in an SLE patient who has low
PCT levels. At the population level, widespread use of a
diagnostic test of unproven performance goes against current
efforts to contain rising health care costs.
The ultimate use of serum PCT determinations is as a
diagnostic test to help confirm or exclude infection and should
be studied as such (3). Some ways in which the diagnostic
usefulness of serum PCT measurement could be evaluated
more accurately would be by comparing levels in SLE patients
with and those without infection, in a matched case-control
study. The resulting data could then be used to calculate the
sensitivity and specificity of the test at various levels of
positivity or negativity, based on a single determination of
serum PCT on each subject. This approach could be used to
establish cutoff values to distinguish between infection and
absence of infection (4), and would give a measure of the
accuracy of the test, by calculating the area under the receiver
operating characteristic curve plotted with the data (5). Alter-
natively, a study with a longitudinal design could show whether
serum PCT levels fluctuate before, during, and after an
episode of bacterial infection.
Until such studies are performcd, the usefulness of
measurement of serum PCT in diagnosing bacterial infection
in patients with systemic autoimmunity remains unproven. We