|
 |
|
| |
|
|
| Post-treatment detection of Helicobacter
pylori |
The main characteristics of the 147
patients evaluated with UBTs after eradication of H. pylori
are summarized in Table 2. In this population, 31 patients were
post-treatment verifications form the pretreatment population
at our lab. The remaining 116 patients had been diagnosed and
treated for H. pylori at other hospitals, and were referred
for follow-up examination more than 1 month after finishing
their H. pylori eradication treatment. Four patients did not
meet all enrolment criteria (1 on proton-pump inhibitor during
the week before UBT; 2 on antibiotics; 1 did not fast for 6
h) and were excluded from the main analysis.
The H. pylori-eradication success rate was 85%after the first
treatment (125/147 patients). The success rate following the
second and third treatments was not assessed in this study.
The spread of the Heliprobe cpm values in the H. pylori-positive
and -negative groups are shown in Figure 2. Again, the difference
was highly significant (median difference 532 cpm, CI 452-569
cpm; P<0.0001). The minimum, median and maximum Heliprobe
cpm values were 58, 537 and 824 cpm, respectively, for the H.
pylori-positive group, and 0, 1 and 25 cpm, respectively, for
the H. pylori-negative group. |
| |
 |
| |
| Scatter plot between the Heliprobe
counts per minute (cpm) values (Heliprobe) and the conventional
14C urea breath test (UBT) disintegrations per minute
(dpm)values at 10 min (conventional UBT at 10 min). (a) Scatter
for the pretreatment population; (b) scatter for the post-treatment
population. |
|
| Table2 Summary of the study population
for post-treatment detection of Helicobacter pylori. |
| |
Per protocol |
| Patient characteristic |
H. pylori-positive |
H. pylori-positive |
All |
| Number of subjects (N) |
22 |
125 |
147 |
| Male/Female (N/N) |
10/12 |
62/63 |
74/73 |
| Age (years; mean ±
SD) |
51 ± 17 |
53 ± 16 |
53 ± 16 |
| Weight (kg; mean ±
SD) |
72 ± 11 |
72 ± 14 |
72 ± 14 |
| Months after treatment
(median, range) |
2(1-24) |
2(1-24) |
2(1-24) |
|
| |
|
top |
| Sensitivity and specificity of the
Heliprobe system |
| Since there was complete separation
between the two groups in Figure 2, the Heliprobe follow-up
sensitivity and specificity were 100% when choosing a cut-off
level in the gap between 25 and 58 cpm (CI 85-100% and 97-100%,
respectively). |
| |
| Association between the Heliprobe
system and the conventional urea breath test |
| Figure 3 shows the correlation analysis
for the post-treatment population. As for the pretreatment detection
population, the correlation was highly significant (P<0.0001),
but due to the high number of H. pylori-negative patients, the
correlation coefficient of 0.48 (CI 0.34-0.59) was not a distinct
as for the pretreatment population. The ? statistics did, however
reveal an equally strong association at follow-up as that seen
for the pretreatment population (? = 1, CI 0.84-1.16). |
| |
|
| |
| Influence of protocol violations |
Including the four patients violating
the enrolment criteria in the follow-up population did not change
the range of the Heliprobe cpm values. One patient was classified
as positive by the conventional UBT (Heliprobe value 439 cpm)
and the remaining three as negative (Heliprobe values 0, 4 and
11 cpm). These values were within the per-protocol distribution,
and the sensitivity and specificity remained at 100%.
One of the 11 protocol violations (1 patient on proton pump
inhibitor) in the detection population raised the maximum Heliprobe
value in the H. pylori-negative group from 25 to 43 cpm. The
remaining ten patients did not impose any changes on the ranges
in either group. Since the minimum Heliprobe value in the H.
pylori-positive group remained at 41 cpm, this meant that one
patient had to be misclassified, either false positive or false
negative, depending on how the cutoff was set. The optimal cut-off
range that allowed only one to two erroneous patient classifications
was set between 26 and 47 cpm. |
| |
|
|
top |
| Optimal cut-off value for the Heliprobe
system |
| Ideally, the same Heliprobe cut-off
level would be used for both the pre- and post-treatment populations.
That is possible if the optimal cut-off ranges for the two groups
overlap. As shown in Table 3, a cut-off value between 25 and
41 cpm fulfils this criterion, and any value within this range
would be appropriate to use. Another sensible alternative is
to set the range where the result is presented as equivocal.
A practical range for equivocal result would be between 25 and
50 cpm, which is the default setting of the Heliprobe system. |
| |
|
| |
| The Heliprobe system with encapsulated
14C-urea |
The characteristics of 37 patients undergoing
UBT with an encapsulated form of 14C-urea in combination with
the Heliprobe system are summarized in Table 4. In this group,
21 patients were tested for detection of H. pylori before, and
16 for detection after, eradication treatment. Due to the small
sample size, both populations were analysed together.
The H. pylori eradication success rate was 88% (14/16 patients)
after treatment. The minimum, median and maximum cpm values
in the H. pylori-positive group were 47, 131, 618 cpm, respectively;
in the H. pylori negative group, the corresponding values were
0, 1 and 28 cpm. The difference was highly significant (median
difference 130 cpm, CI 76-356 cpm; P<0.0001). |
|
| Table 3 Summary of cut-off values
for pre- and post-treatment detection of Helicobacter pylori
infection |
| |
| |
Pre-treatment
|
Post-treatment |
| Patient category |
Per protocol |
All patients |
Per protocol |
All patients |
| Maximum value of H. pylori-negative
group (cpm) |
25 |
25 |
25 |
25 |
| Minimum value of H. pylori-positive
group (cpm) |
41 |
47 |
58 |
58 |
| Misclassification |
0 |
1 |
0 |
0 |
|
| |
| Table 4 Summary
of the study group for detection of Helicobacter pylori using
encapsulated urea |
| |
| |
Per protocol |
| Patient Characteristic |
H. pylori-positive |
H. pylori-negative |
All |
| Number of subjects (N) |
11 |
26 |
37 |
| Male/Female (N/N) |
6/5 |
13/13 |
19/18 |
| Age (years; mean ±
SD) |
54 ± 15 |
54 ± 15 |
54 ± 15 |
| Weight (kg; mean ±
SD) |
73 ± 15 |
70 ± 15 |
71 ± 15 |
|
| |
As the maximal value in the H. pylori-negative
group was 28 cpm and the minimal value in the H. pylori-positive
group was 47 cpm, any value between these two values might be
used as a cut-off between negative and positive responses.
Eleven true positives and 26 true negatives were found, while
no false positives or false negatives were found. In comparison
with the conventional UBT, the sensitivity using encapsulated
urea was calculated to be 100% (CI 72-100%) and the specificity
100% (CI 87-100%).
As for the liquid urea group, the correlation was highly significant
with a correlation coefficient of 0.71 (CI 0.50-0.84; P<0.0001).
The ? statistics revealed a strong agreement between the conventional
UBT and Heliprobe system using encapsulated urea (? = 1, CI
0.68-1.32; P<0.0001). |
| |
|
|
top |
| Discussion |
All the biological test results obtained
with the Heliprobe system were confined to a 95% confidence
interval. The test results with the Heliprobe system were indistinguishable
form those of our conventional UBT. In terms of both sensitivity
and specificity, no significant differences between the two
tests were detectable, with data approaching equality. Thus
full concordance between the two tests seems to prevail. This
outcome permits us to draw the conclusion that the two systems
are equi-efficacious in diagnosing H. pylori status. The advantages
of the Heliprobe system are speed and simplicity. With no aid
required from external facilities or expertise, the diagnostician
will through the use of this novel system, have access to the
test result within 15-20 min after the patient has swallowed
the urea. This increases the options for when and where to perform
the analysis. The small size and handiness of the Breath Card
also make handling very easy. In cases where mailing might be
needed, a regular envelop can be used for mailing the test to
the analyzer. Due to the stability of the chemical binding of
CO2 to the LiOH in the BreathCard, later reanalysis
of a specific sample is possible; even years after the test
was first carried out. The swap in technology has also made
the Heliprobe Analyzer comparably cheaper to produce, with an
estimated cost of about one-tenth of that of a ß-scintillator.
The present investigation showed excellent concordance between
the conventional 14C-UBT and the novel Heliprobe
UBT. In order to simplify the test and interpretation process,
the expression of results in recovery standard units (% dose
mmol CO2-1 kg) has been abandoned in the Heliprobe
system in favour of simply basing cut-off levels directly on
measured cpm values. As reviewed elsewhere, several groups have
argued that it is illogical to make allowance for endogenous
CO2 production by incorporating a ‘Fudge factor’
involving the patient’s body weight. Indeed, most groups
no longer express their results as a recovery of administered
dose adjusted for weight, preferring instead to use radioactive
cpm or dpm, since the correlation between the two measures is
excellent [11]. This is also what is to be expected due to the
apparent dependency of the two correlation factors evaluated.
Hence, dpm from our conventional UBT could be used and further
correlated to the cpm as given by the Heliprobe Analyzer for
validation of this system.
With the use of a urea cocktail for administration of 14C
and then detection of 14CO2 by simultaneously
using the conventional UBT and Heliprobe systems, we found a
few restrictions that have to be taken on consideration when
performing the UBT. First, careful tooth brushing seems important
for obtaining conditions representative for the H. pylori status
in the stomach, not being blurred by the patient’s oral
microflora and microbial conditions. As a further development
of this method, we are aiming to produce an encapsulated form
of the urea/citric acid composition needed to achieve a standardized
and reliable test with stable outcomes. Second, acid suppression
was detrimental for the outcome of the test. We therefore decided
to withhold potent acid-inhibitory drugs, such as proton-pump
inhibitors, for 7 days before the test was carried out, while
less potent acid inhibitors such as H2-receptor antagonists,
were stopped 24 h before the test was carried out. Preliminary
reports indicate that the addition of citric acid to the urea
solution / capsule diminishes the effect of acid-inhibitory
drugs on the accuracy of the test [12]. Further verification
is needed, however, before we can recommend continued drug use
with proton-pump inhibitors or H2-receptor blockers in conjunction
with UBT. Third, drugs known to retain binding capacity to different
substances, such as antacids and sucralfate, were withheld for
24 h before the UBT. Fourth, antibiotics or bismuth treatment
were not allowed during the month preceding the UBT. By keeping
a tight hand over these rules, we were able to optimize the
diagnostic procedure with a minimum of radioactivity (1 µCi,
37 ?Bq) known to be effective in order to achieve accurate test
results [4, 11]. Thus, both the conventional UBT as well as
the Heliprobe UBT were carried out with 1-µCi 14C dose
per test.
Concern with 14C usually arises because of its long
half-life, but this is less important for organic compounds
such as CO2 and urea, which are excreted rapidly.
In the 14C-UBT, urea either undergoes hydrolysis,
being exhaled as 14CO2, or is eliminated
unchanged in urine. Because the biological half-life of urea
is short, the cumulated radiation dose from each breath test
is small and far below variations in natural radiation. According
to data reported by Munster et al. [13], approximately 90% of
the 14C from a UBT is eliminated as CO2
in breath or as urea in urine. This would mean that after 3
days, the amount of isotope retained in the body is negligible.
The cumulative lifetime radiation exposure from this test has
been calculated to be not more than 0.3 mrem/µCi [14],
considered to equal the background radiation a person is exposed
to in 1 day [15]. Due to the very low level of radioactive exposure,
the 1- µCi 14C dose has been permitted for
general use in UBTs in the UA (Nuclear Radioactive Committee,
USA, 10CFR § 30.21 Radioactive drug: Capsules containing
carbon-14 urea for diagnostic use in humans). We therefore consider
the radioactive bioburden on each person to be very limited,
even not precluding repeated tests in the same person. Some
reports even conclude that there is no restriction on repeated
tests in the same person. Some reports even conclude that there
is no restriction on repeated investigations in whole families,
including children [16]. We do, however, fully accept that in
children and pregnant women, it is preferable to use the 13C-UBT.
For the evaluation of the conventional UBT with the Heliprobe
system, we used duplicate samples for the conventional test,
whereas single samples were taken for the Heliprobe test. A
detailed discussion has appeared previously of the relative
merits of multiple as opposed to single samples for the UBT
[11]. In accordance with our findings showing high concordance
between results using either the conventional or Heliprobe UBTs,
we consider one sample at a single time interval after administration
of the labeled urea to be acceptable for most non-research purposes
in the clinic. Using the 14C-UBT, even a baseline
sample seems unnecessary, as there should be no detectable 14C
in breath under basal conditions. Furthermore, recommendations
have been given to roll the patient over and on the sides in
an attempt to get the tracer distributed evenly over the gastric
lining. There is, however, no evidence that moving about should
increase the sensitivity or specificity of the test, therefore
this recommendation should be abandoned.
The cut-off for H. pylori positivity was chosen to not overlook
any cases of true H. pylori infection. Rather, false positives
were considered acceptable, as this would lead only to another
antibiotic treatment in a few cases with suspected infection.
This approach to the diagnostic performance of the UBT would
not leave any patient without treatment for a potentially ulcerogenic
infection.
With the introduction of the Heliprobe system, broader applications
of the UBT are at hand without compromising accuracy, as is
the case for serology-based tests. With the portability of the
equipment, it may be well be used for epidemiological studies,
especially in elderly people in whom serology tests may not
be reliable. The UBT also has the advantage over serology in
testing the current infection status, as it is known that serology
for H. pylori may remain positive for several years in a significant
percentage of patients whose infections have been eradicated
[17]. In addition, with our current evaluation of the Heliprobe
system, we have the possibility of not only detecting H. pylori
in antibiotics-naïve subjects but also of retesting in
post-treatment patients followed up at least 1 month after termination
of anti-H.pylori treatment.
We conclude that the Heliprobe system is a rapid and reliable
UBT for pre- and post-treatment follow-up detection of H. pylori.
The Heliprobe system offers the first on-site fully accurate
diagnostic system for detection of H. pylori directly in the
doctor’s surgery within minutes after oral intake of the
urea tracer. |
| |
|
|
top |
| Acknowledgements |
| We thank Ms. Ulrika Mjörnell and
MS Christina Prior for expert patient care and running the test
samples. |
| |
|
| |
| References |
| 1. |
Graham DY,
Klein PD, Evans DJJ, Evans DG, Alpart LC, Opekun AP et
al. Campylobacter pylori detected noninvasively by the
13C-urea breath test. Lancet 1987;1:1174-1177. |
| 2. |
Langenberg ML, Tytgat GN,
Schnipper MEI. Campylobacter-like organisms in the stomach
of patients and healthy individuals. Lancet 1984; 1:12?? |
| 3. |
Marshall BJ, LAngton SR.
Urea hydrolysis in patients with Campylobacter pyloridis
infection. Lancet 1986;1:965-966. |
| 4. |
Rehnberg AS, Bengtsson
C, Befrits R, Granstrom M, Hellström PM. Refinement
of the 14C-urea breath test for detection of Helicobacter
pylori. Scand J Gastroenterol 2001;36:822-6. |
| 5. |
Lee A. The nature of Helicobacter
plyori. Scand J Gastroenterol 19??.:214(Suppl):5-8. |
| 6. |
Atherton JC. Non-endoscopic
tests in the diagnosis of Helicobacter pylori infection.
Aliment Pharmacol Ther 1997;11(Suppl 1):11-20. |
| 7. |
Logan RP, Polson RJ, Misiewicz
JJ, Rao G, Karim NQ, Newell D, et al. Simplified single
sample 13Carbon urea breath test for Helicobacter pylori.
Comparison with histology, culture, and ELISA serology.
Gut 1991;32:1461-1464. |
| 8. |
Braden B, Duan LP, Caspary
WF, Lembcke B, More convenient 13C-urea breath test modifications
still meet the criteria for valid diagnosis of Helicobacter
pylori infection. Z Gastroenterol 1994;32:198-202. |
| 9. |
Rollan A, Giancaspero R,
Arrese M, Figueroa C, Vollrath V, Schultz M, et. Al. Accuracy
of invasive and noninvasive tests to diagnose Helicobacter
pylori infection after antibiotic treatment. Am J Gastroenterol
1997;92:1268-1274. |
| 10. |
Mowat C, Murray L, Hilditch
TE, Kelman A, Oien K, McColl KE. Comparison of helisal
rapid blood test and 14C-urea breath test in determining
Helicobacter pylori status and predicting ulcer disease
in dyspeptic patients. Am J Gastroenterol 1998;93:20-25. |
| 11. |
Bell GD, Weil J. Detection
of Helicobacter pylori by the 14C-urea breath test. In:
Rathbone BJ, Heatly RV, editors. Helicobacter pylori and
Gastroduodenal Disease. Oxford:Backwell Scientific Publciations:
1992.pp.74-87. |
| 12. |
Hamlet A, Stage L, Lönroth
H, Cahlin C, Nyström C, Pettersson A. A novel tablet-based
13C-urea breath test for Helicobacter pylori with enhanced
performance during acid suppression. Scand J Gastroenterol
1999; 34:367-374. |
| 13. |
Munster DJ, Chapman BA,
Burt MJ, Dobbs BR, Allardyce RA, Bagshaw PF, et al. The
fate of ingested 14C-urea in the urea breath test for
Helicobacter pylori infection. Scand J Gastroenterol 1993;
28:661-666. |
| 14. |
Stubbs JB, Marshall BJ,
Radiation dose estimates for the carbon – 14-labeled
urea breath test. J Nucl Med 1993;34:821-825. |
| 15. |
Goddard AF, Logan RP. Urea
breath tests for detecting Helicobacter pylori. Aliment
Pharmacol Ther 1997;34:821-825. |
| 16. |
Leide-Svegborn S, Stenström
K, Olofsson M, Mattsson S, Nilsson LE, Nosslin B. et al.
Biokinectics and radiation doses for carbon-14 urea in
adults and children undergoing Helicobacter pylori breath
test. Eur J Nucl Med 1999;26:573-580. |
| 17. |
Logan RPH. Detection of
Helicobacter pylori by the 13C-urea breath test. In: Rathbone
BJ, Heatley RV, editors. Helicobacter pylori and gastroduodenal
disease. Oxford: Blackwell Scientific Publications; 1992.pp.
88-106. |
|
| |
|
|
|
top |
|
