Mass Screening for Neuroblastoma Necessity for
Rediscussion
Motoi Nishi
Motoi
Nishi, Department of Fundamental Health Sciences, Health Sciences University of
Hokkaido, Kanazawa 1757 Tobetsu, Hokkaido 061-0293, Japan
Correspondence
to: Motoi Nishi, Department of Fundamental Health Sciences, Health
Sciences University of Hokkaido, Kanazawa 1757 Tobetsu, Hokkaido 061-0293,
Japan
Email:
motoi@hoku-iryo-u.ac.jp
Telephone: +81-133-23-1211 ext.3642 Fax: +81-133-23-1669
Received: July 16, 2013
Revised: September 1, 2013
Accepted:
September 6, 2013
Published
online: December 18, 2013
ABSTRACT
In this article, the history of mass screening for neuroblastoma (NBMS)
targeting 6-month-old infants is briefly described, and the change in the
mortality after stopping NBMS is reported. Several cohort studies performed to
evaluate NBMS are summarized, as well. NBMS was previously performed on a
nationwide scale in Japan targeting 6-month-old infants, and after 1992 all the
specimens were measured by high performance liquid chromatography (HPLC). But
NBMS stopped in most communities by the spring 2004. For the purpose of
estimating the influence of stopping NBMS, we investigated the mortality rates
of the children who died of neoplasms of the adrenal glands at the ages of 1 to
4 years by birth cohorts from the introduction of HPLC to the period after
stopping NBMS. The mortality rate in the birth cohort from 1992 to 2003, most
of whose members participated in NBMS with HPLC, was significantly lower than the
mortality rate in the cohort from 1989 to 1991 (P<0.01). The mortality rate
in the cohort from 2004 to 2006, most of whose members were not screened, was
the highest among the cohorts after 1992, and the significant difference from
the mortality rate in the cohort from 1989 to 1991 disappeared. Considering the
increasing mortality of the neoplasms of the adrenal glands in Japan and the
recent results of the cohort studies revealing the effectiveness of NBMS, it is
necessary to discuss its restart on a nationwide scale.
© 2013 ACT. All rights reserved.
Key words: Child;
Epidemiology; Mass Screening; Mortality; Neuroblastoma
Nishi M. Mass
Screening for Neuroblastoma Necessity for Rediscussion. Journal of Tumor
2013; 1(7): 62-65 Available from: URL: http://www.ghrnet.org/index.php/JT/article/view/543
INTRODUCTION
Neuroblastoma (NB)
is a frequent solid tumor during childhood. In Japan, before the start of mass
screening for neuroblastoma (NBMS), NB was the third most frequent malignancy
in children 0-14 years of age, after leukemia and brain tumors[1].
In many cases of NB, the original site is the adrenal glands. Though
therapeutic methods have become better, even now its prognosis has not improved
much. The 5-year survival rate of cases in stage Ⅳ is about 30-40%[2].
One feature of NB is the
excretion of large amounts of vanillylmandelic acid (VMA) and homovanillic acid
(HVA) in urine. Taking advantage of this feature, NBMS was previously performed
on a nationwide scale in Japan, but this is no longer done except in a few
areas. In this article, the history of NBMS in Japan is briefly described and
the change in the mortality rate after stopping NBMS is reported, employing the
data of vital statistics.
HISTORY OF MASS SCREENING FOR NEUROBLASTOMA
NBMS was started in
Kyoto Prefecture, Japan, by Dr. Tadashi Sawada in 1973[3,4]. After
1985, NBMS gradually spread to other areas of Japan. Its subjects were
6-month-old infants. Urine from babies was sampled on filter paper and levels
of urinary VMA and HVA were measured using high performance liquid
chromatography (HPLC)[5,6]. HPLC is a quantitative method. At first,
these levels were estimated by qualitative methods, e.g., the spot test, dip
test, and thin layer chromatography (TLC). However, after it became known that
the efficacy of HPLC was far better than those of qualitative methods, it was
gradually introduced and all the specimens were measured by HPLC after 1992
throughout Japan. In fact, there are large differences between the results for
NBMS obtained using HPLC and those from non-HPLC methods[7,8].
After several years,
however, it turned out that the incidence of true positives in NBMS with HPLC
was higher (1 case per 5,000 screened) than the natural incidence previously
reported (1 case in 10,000 live births), and more than a few false negatives
(patients who appear to have clinical symptoms after having negative results in
MS) were detected (1 in 20,000 screened). The sensitivity was, therefore, about
80%[7].
Therefore, some researchers began
to suggest that NBMS found only those whose tumors regressed spontaneously
(spontaneous regression cases) without finding those whose symptoms appeared at
older ages (proper cases)[9]. In fact, several true positives
detected by NBMS were followed without any treatment and, in some cases, tumor
markers such as VMA or HVA, became normal[10].
In 2000 the study group of
Hisashige et al[11], which conducted a large-scale
retrospective cohort study, concluded that NBMS with HPLC was effective in
reducing mortality. Their subjects were 6-month-old infants from the
introduction of HPLC to 1997 (children screened, 19,520,000 person-years;
children not screened, 3,340,000 person-years). The relative risk of mortality
was 0.547 (95% CI; 0.306-0.976). From 2003 to 2004, Hayashi et al[12,13]
reported the results of their large-scale retrospective cohort study in which
screened and unscreened children born from 1995 to 1999 were compared. The
relative risk of mortality was 0.269 (95% CI; 0.170-0.428).
In 2003, however, groups in
Canada[14] and Germany[15] published their findings that
NBMS was not effective. That year the Japanese Government decided to stop the
nationwide NBMS. By the spring 2004, NBMS stopped in most communities in Japan.
As stated above, there is an
opinion that NBMS identifies only "spontaneous regression cases", and
misses "proper cases", and that it is therefore ineffective in
reducing mortality. If this opinion is correct, after NBMS starts, the
mortality rate must increase, because some of those who are true positives
detected by NBMS will die of the side effects of chemotherapy or of sequelae of
operations, and their mortality rate will be added to that of "proper
cases". If this opinion is correct, after NBMS is stopped the mortality
rate should decrease because there would be no more deaths of those with
"spontaneous regression". The reality is, however, quite the
opposite.
CHANGE IN MORTALITY RATE OF NEOPLASMS OF THE ADRENAL GLANDS
Mortality rates at 1-4 years of age
For the purpose of
estimating the influence of stopping NBMS, we investigated the mortality rates
of the children who died of neoplasms of the adrenal glands at the ages of 1 to
4 years by birth cohorts from 1989 to 2006, i.e., from the introduction of HPLC
to the period after stopping NBMS[16].
In the vital statistics
there are no codes that stand for NB. The codes 194.0 (ICD 9) and C74 (ICD 10)
stand for neoplasms of the adrenal glands. Every year the Japanese Government
publishes the numbers of deaths at the ages of 0, 1, 2, 3, 4, 5-9, 10-14, etc.,
up to 95-99, and 100 years of age or older[17]. Using these data,
birth cohorts can be made. For example, those who died at 0 years of age in
2000, at 1 year of age in 2001, at 2 years of age in 2002, at 3 years of age in
2003, and at 4 years of age in 2004 belonged to a birth cohort whose birth year
was 1999 or 2000 (1999-00 cohort). The number of live births of this cohort was
estimated to be "the number of live births in 1999×0.5 + that in
2000×0.5". Using the numbers of deaths and live births, the mortality rate
(per 100,000 live births) was calculated. Multiplying the mortality rate of the
"1989-91 cohort" (the birth cohort before all the babies were
screened by HPLC) to the number of live births of each cohort, the expected
number of deaths was calculated. Employing the number of deaths and the
expected number, chi-square test was performed (Table 1).
The mean participation rate
for NBMS in the 1990s was 86%, according to the report of Hiyama et al[18]
That of the babies born from 1995 to 2000 reported by Hayashi et al[13]
was 87.4%. Therefore, about 90% of the "2001-03 cohort" participated
in NBMS. There were no large differences in the mortality rates among the
"1992-94", "1995-97", "1998-00" and
"2001-03" cohorts. The total number of deaths in these cohorts from
1992 to 2003 was 223. Since the expected number of deaths was 286.8, the number
of deaths was significantly small (P<0.01). The mortality rate of the
"2004-06 cohort", most of whose members were not screened, was the
highest among the cohorts after 1992-94, and the significant difference from
the mortality rate of the "1989-91 cohort" disappeared.
Mortality rates at 5-9 years of age
As stated above, for
the number of deaths at the ages of 5 years or older, only the data by 5-year
age-groups (5-9 years, 10-14 years, etc.) are available. Therefore, it is
impossible to make birth cohorts. Figure 1 shows the changes in the mortality
rates (per 100,000) of children whose age at death was 5-9 years. In the group
aged 5-9 years before 1996, there were few children who underwent NBMS with
HPLC, since it was first in 1992 that all 6-month-old babies had the chance to
be screened by HPLC. The mortality rate had no tendency to decrease. In 1997,
however, when the number of children who were screened by HPLC began to
increase, the mortality rate started to decrease. The mortality rates from 2001
to 2008, when all the children of this age group had the chance of being
screened by HPLC, were the lowest. However, in 2009, when the number of
children who were not screened began to increase within this age group, the
mortality rate started to increase.
Effects of mass screening in reducing
the mortality of neuroblastoma
As reported before[19],
Hanawa et al[20] published several reports, using all cases
with NB. According to their data, there were 842 deaths at 0-4 years of age
from 1979 to 1992. On the other hand, according to the vital statistics during
the same period, there were 771 children (91.6% of the cases reported by Hanawa
et al) who died of neoplasms of the adrenal glands. It is possible that
among the 771 cases there were several cases with adrenal carcinoma or
pheochromocytoma, but these are very rare during childhood. It is not
unreasonable, therefore, to conclude that NBMS with HPLC reduced the mortality
rate of children with NB whose original site was the adrenal glands. In 2008
Hiyama et al[18] published a report of a large-scale
retrospective cohort study. They compared the screened cohort and the one not
screened. Their conclusion was that the relative risk of mortality for the
children screened by HPLC was about 0.53 (0.42-0.63). The present data compare
the group whose members all had the chance to be screened by HPLC and the one
whose members did not. Though the subjects in the study of Hiyama et al[18]
and the present one are different in character, the conclusions of the two
studies are similar.
As pointed out previously[21],
the study in Canada[14] employed TLC, the efficacy of which is far
worse than that of HPLC. The German study[15] combined false
negatives with true positives to calculate the incidence of the "screened
group", but since many cases with spontaneous regression were included as
true positives, this method inflated the incidence of the "screened
group" and led to underestimation of the effectiveness of NBMS, indicating
that it could not reduce the incidence of NB. Furthermore, they employed
subjects in a "control area" as controls, not "non-participants"
for whom biases would be fewer because of being in the same area as the
participants. The incidence of NB among the subjects in the "control
area" was about 25% smaller than that of "non-participants".
This also led to underestimation of the effectiveness of NBMS.
Influence of prolongation of survival
period
The mortality rate
of the "2004-06 cohort" was a little lower than that of the
"1989-91 cohort" (Table 1). This might partially have been due to
complete cure and/or prolongation of life in some cases as a result of the
progress of therapeutic methods. In France, where virtually no NBMS has been
performed, there was a slight decrease in the mortality rate of the cases with
neoplasms of the adrenal glands at 1-4 years of age from the 1980s to 1990s
(Table 2)[22]. However, the mortality rate at 5-9 years of age
increased, and there was no significant change in the total mortality rate for
1-9 years of age, suggesting that prolongation of the survival period due to
the progress of therapy brought about this change.
Since the time interval of
the present study was about 20 years, the contribution of progress in
therapeutic methods cannot be ignored. In Japan, too, many children who would
have died at 1-4 years of age died at 5-9 years of age due to prolongation of
the survival period. It is probable that this made the mortality rate at 1-4
years of age in the "2004-06 cohort" lower (Table 1). The mortality
rate at 5-9 years of age had a tendency to increase by 1996 (Figure 1). This suggests that there
were already cases whose survival period was prolonged.
Influence of NBMS targeting
18-month-old children in several areas
There are several
areas that continue NBMS, including the City of Sapporo[23], Niigata
Prefecture, Kyoto Prefecture[24], and Osaka Prefecture[25].
Now, NBMS in all of these areas targets 18-month-old children, not 6-month-old
babies. One of the aims of this change in the age targeted is to reduce the
detection of cases of "spontaneous regression"[26].
In Kyoto[24] and
Osaka[25], the incidence rates of true positives were 1 of 8,060
screened, and 1 of 8,556, respectively, which are lower than that of NBMS
targeting 6-month-old infants (about 1 in 5,000), though it was higher in
Sapporo (1 of 2,965 screened)[26].
In Sapporo[26], 9
cases were identified from April 2006 to December 2008. In Kyoto[24],
4 cases were identified from 2004 to 2009. In Osaka[25], 8 cases
were identified from April 2005 to March 2008. Since most of these cases
belonged to the "2004-06 cohort" (Table 1), the mortality rate of
this cohort was decreased by saving the lives of these children.
Biological features of detected cases
in NBMS
Biological
characterization is one of the important factors which influence the prognosis
of NB cases.
In Sapporo City, the NBMS
targeting 6-month-old infants were performed from April 1981. On April 1991,
the second NBMS was started, targeting 14-month-old children. The 6-month NBMS
was stopped on March 2004. On April 2006, the age targeted was changed from 14
months to 18. That is, the 6-month NBMS was performed for 23 years, the
14-month NBMS, for 15 years, and the NBMS targeting 18-month-old children have
been performed since April 2006. Hanai et al[23,27] reported
biological aspects of the cases detected in these 3 mass screenings (Table 3).
Only 1 case detected in the 6-month MS was in INSS stage 3 (2%). But in the
18-month MS, 4 cases (33%) were in this stage. 50% of the cases detected in the
18-month MS had the unfavorable type of Shimada's classification. Thus, in the
18-month MS, though no cases had amplification of MYCN, so far as Shimada's
classification and clinical stages are concerned, more cases with unfavorable
prognosis are detected than in the 6- and 14-month MS.
Considering the increasing
mortality of the neoplasms of the adrenal glands in Japan and the recent
results of the cohort studies revealing the effectiveness of NBMS, it is
necessary to discuss its restart on a nationwide scale.
REFERENCES
1 Bessho F. Epidemiology of solid tumors in
children. Jpn J Pediatr Med 1986; 18: 321-329
2 Kaneko M. Advances and future
perspectives of neuroblastoma treatment. Jpn J Pediatr Hematol/Oncol
2012; 49: 3-10
3 Sawada T, Todo S, Fujita K, Iino
S, Imashuku S, Kusunoki T. Mass screening of neuroblastoma in infancy. Am J
Dis Child 1982; 136: 710-712
4 Sawada T, Hirayama M, Nakata T et
al. Mass screening for neuroblastoma in infants in Japan. Interim report of a
mass screening study group. Lancet 1984; 2: 271-273
5 Sato
Y, Hanai J, Takasugi N, Takeda T. Determination of urinary
vanillylmandelic acid and homovanillic acid by high performance liquid
chromatography for mass screening of neuroblastoma. Tohoku J Exp Med 1986; 150(2):169-174
6 Hanai J, Kawai T, Sato Y, Takasugi N,
Nishi M, Takeda T. Simple liquid-chromatographic measurement of
vanillylmandelic acid and homovanillic acid in urine on filter paper for mass
screening of neuroblastoma in infants. Clin Chem 1987; 33(11):
2043-2046
7 Nishi
M, Miyake H, Takeda T et al. Cases of spontaneous regression and true patients detected in mass
screening for neuroblastoma. Int J Pediatr Hematol/Oncol 1995; 1:
557-563
8 Nishi M, Miyake H, Hanai J, Fujita
K, Hatae Y, Takeda T. Mass screening for neuroblastoma by HPLC and by non-HPLC:
a review. Oncol Rep 1998; 5:
1399-1402
9 Yamamoto K, Hanada R, Kikuchi A et
al. Spontaneous regression of
localized neuroblastoma detected by
mass screening. J Clin Oncol 1998; 16: 1265-1269
10 Kigasawa H. Results of long-term
follow-up in Kanagawa Children's Medical Center. Jpn J Pediatr Hematol/Oncol
2013; 50: 1-5
11 Hisashige A, Hayashi K, Mikasa H.
Evaluation of the mass screening for neuroblastoma - 1. Mortality. Tokyo:
Ministry of Health and Welfare, 2000.
http://www.niph.go.jp/wadai/mhlw/1999/h1105001.pdf
12 Hayashi K, Fujita T, Katanoda H et al. Effectiveness of mass-screening program on
neuroblastoma mortality in 1995-2000 birth cohort of Japan: Nationwide
neuroblastoma mortality study. In:
De Bernardi B, Garaventa A eds. Advanced in Neuroblastoma Research 11th
Conference. 2004, Genova, Italy: p93.
13 Hayashi K, Fujita T, Katanoda H et
al. Effectiveness of
mass-screening program on neuroblastoma mortality.
http://mhlw-grants.niph.go.jp/niph/search/NIDD00.do?resrchNum=200300840A#selectHokoku/200300840A0007.pdf
14 Woods WG, Gao RN, Shuster JJ,
Robison LL, Bernstein M, Weitzman S, Bunin G, Levy I, Brossard J, Dougherty G,
Tuchman M, Lemieux B. Screening of infants and mortality due to neuroblastoma. N
Engl J Med 2002; 346:
1041-1046
15 Schilling FH, Spix C, Berthold F,
Erttmann R, Fehse N, Hero B, Klein G, Sander J, Schwarz K, Treuner J, Zorn U,
Michaelis J. Neuroblastoma screening
at one year of age. N Engl J Med
2002; 346: 1047-1053
16 Nishi M, Hanai J, Mitsuhashi Y.
Mortality of neuroblastoma in birth cohorts in Japan. J Jpn Mass Screen
Assoc 2012; 22: 228-232.
17 Vital
statistics of Japan. http://www.e-stat.go.jp/SG1/estat/GL08020101.do?_toGL08020101_&tstatCode=000001028897&requestSender=dsearch
18 Hiyama
E, Iehara T, Sugimoto T, Fukuzawa M, Hayashi Y, Sasaki F, Sugiyama M,
Kondo S, Yoneda A, Yamaoka H, Tajiri T, Akazawa K, Ohtaki M. Effectiveness of
screening for neuroblastoma at 6 months of age: a retrospective
population-based cohort study. Lancet 2008; 371: 1173-1180
19 Nishi M, Takeda T, Hatae Y et al. Contribution of the HPLC mass screening for
neuroblastoma to the decrease in mortality. J Jpn Mass Screen Assoc
2001;11: 45-49
20 Hanawa
Y, Sawada T, Tsunoda A. Decrease in childhood neuroblastoma death in
Japan. Med
Pediatr Oncol 1990; 18: 472-475
21 Nishi
M, Hanai J, Fujita K et al. Is the mass screening for neuroblastoma ineffective? J Exp Clin
Cancer Res 2003; 22(4): 673-676
22 Nishi M, Satgé D, Haupt R. Change
in mortality of neuroblastoma in Japan, France and Italy. J Jpn Mass Screen
Assoc 2007; 17: 37-41
23 Hanai
J, Oota Y, Takahashi H et al. Precise assessment of
the mass screening for
neuroblastoma; 18-month mass screening in Sapporo City. J Jpn Ass Cancer
Detection Diagnosis 2011; 18: 228-232
24 Iehara
T, Numata K, Matsumura J, Sugimoto T, Sawada T, Hosoi H. Precise
assessment of the mass screening for neuroblastoma; 18-month new mass screening
system in Kyoto Prefecture. J Jpn Ass Cancer Detection Diagnosis 2010; 18:
27
25 Nakayama
M, Takeshima K, Irie A et al. 18-month new mass screening system in
Osaka Prefecture. J Jpn Ass Cancer Detection Diagnosis 2011; 18(3):
257-262
26 Hanai
J, Oota Y, Tagami Y et al. Results of screening for neuroblastoma at 18 months of age in Sapporo
City. J Jpn Mass Screen Assoc 2010; 20: 17-20
27 Hanai
J, Tagami Y, Abe A et al. Results of screening for neuroblastoma at 18
months of age in Sapporo City. Jpn J Pediatr Surg 2008; 40:
1069-1073
Peer reviewer: Birgit Geoerger, MD, PhD,
Gustave Roussy, Pediatric and Adolescent Oncology, CNRS UMR8203 - Pediatric
Group, 114 Rue Edouard Vaillant, 94805 Villejuif cedex, France; Claudia Spix,
Stellv. Leiterin Deutsches Kinderkrebsregister,
UNIVERSITÄTSMEDIZIN, der Johannes Gutenberg-Universität Mainz, Deutsches
Kinderkrebsregister am, Institut für Medizinische Biometrie, Epidemiologie und
Informatik(IMBEI), Postanschrift: 55101 Mainz, Haus- und Lieferanschrift: Obere
Zahlbacher Str. 69, 55131 Mainz, Germany.
Refbacks
- There are currently no refbacks.