1 23

Cancer Causes & ControlAn International Journal of Studies ofCancer in Human Populations ISSN 0957-5243Volume 24Number 6 Cancer Causes Control (2013)24:1111-1118DOI 10.1007/s10552-013-0188-y

Childhood leukemia and lymphoma: timetrends and factors affecting survival in fiveSouthern and Eastern European CancerRegistries

Eleni Th. Petridou, Nadya Dimitrova,Sultan Eser, Denis Kachanov, HulyaKarakilinc, Svetlana Varfolomeeva,Maria Belechri, et al.

1 23

Your article is protected by copyright and all

rights are held exclusively by Springer Science

+Business Media Dordrecht. This e-offprint

is for personal use only and shall not be self-

archived in electronic repositories. If you wish

to self-archive your article, please use the

accepted manuscript version for posting on

your own website. You may further deposit

the accepted manuscript version in any

repository, provided it is only made publicly

available 12 months after official publication

or later and provided acknowledgement is

given to the original source of publication

and a link is inserted to the published article

on Springer's website. The link must be

accompanied by the following text: "The final

publication is available at link.springer.com”.

ORIGINAL PAPER

Childhood leukemia and lymphoma: time trends and factorsaffecting survival in five Southern and Eastern European CancerRegistries

Eleni Th. Petridou • Nadya Dimitrova • Sultan Eser • Denis Kachanov •

Hulya Karakilinc • Svetlana Varfolomeeva • Maria Belechri • Margarita Baka •

Maria Moschovi • Sofia Polychronopoulou • Fani Athanasiadou-Piperopoulou •

Ioanna Fragandrea • Eftichia Stiakaki • Apostolos Pourtsidis • Christodoulos Stefanadis •

Jan-Willem Coebergh • Theodoros N. Sergentanis

Received: 14 December 2012 / Accepted: 14 March 2013 / Published online: 26 March 2013

� Springer Science+Business Media Dordrecht 2013

Abstract

Purpose Within Europe, incidence and mortality rates of

childhood leukemia and lymphoma are rather heteroge-

neous. The present study comprising data from five

Southern and Eastern European Cancer Registries aims to

compare time trends and examine whether sociodemo-

graphic variables, clinical parameters, and proxies of effi-

cient care affect survival.

Methods Data spanning 1996–2010 were obtained for a

total of 3,041 newly diagnosed childhood leukemia and

1,183 lymphoma cases reported by the Greek Nationwide

Registry for Childhood Hematological Malignancies, Bul-

garian National Cancer Registry, Moscow Region and

Turkey (Antalya and Izmir) Cancer Registries. Poisson

modeling for the evaluation of time trends and multivariate

Cox regression analysis for the assessment of prognostic

factors were performed.

Results The incidence of leukemia was increasing in all

cases, with Bulgaria and Greece presenting statistically

significant annual changes (?3.5, and ?1.7 %, respec-

tively), followed by marginally increasing trends in Izmir

and Moscow; by contrast, there was a remarkable, statis-

tically significant, decreasing mortality trend for leukemia.

Rates for lymphoma remained flat. Greece experienced

almost twofold better survival rates for both leukemia and

lymphoma, probably due to its higher socioeconomic status

during the study period. Overall, patients with leukemia

living in rural areas had a 28 % lower prognosis (RR: 1.28,

95 % CI 1.03–1.59), pointing to effects of remoteness,

when the most privileged country (Greece) was excluded

from the analysis.

Conclusions The favorable mortality trends highlight the

progress in Southern–Eastern European countries alongElectronic supplementary material The online version of thisarticle (doi:10.1007/s10552-013-0188-y) contains supplementarymaterial, which is available to authorized users.

E. Th. Petridou (&) � M. Belechri � T. N. Sergentanis

Department of Hygiene, Epidemiology and Medical Statistics,

Athens University Medical School, 75 M. Asias Str.

Goudi, 115 27 Athens, Greece

e-mail: epetrid@med.uoa.gr

N. Dimitrova

Bulgarian National Cancer Registry, National Oncology

Hospital, Sofia, Bulgaria

S. Eser

Cancer Control Department, Izmir Public Health Institute,

Izmir, Turkey

D. Kachanov � S. Varfolomeeva

Department of Clinical Oncology, Federal Clinical Research

Center of Pediatric Hematology, Oncology and Immunology,

Moscow, Russian Federation

H. Karakilinc

Provincial Health Directory of Antalya, Antalya Cancer

Registry, Antalya, Turkey

M. Baka � A. Pourtsidis

Department of Pediatric Haematology–Oncology,

‘‘Pan. & Agl. Kyriakou’’ Children’s Hospital, Athens, Greece

M. Moschovi

Haematology–Oncology Unit, First Department of Pediatrics,

Athens University Medical School, ‘‘Aghia Sophia’’

Children’s Hospital, Athens, Greece

S. Polychronopoulou

Department of Pediatric Haematology–Oncology,

‘‘Aghia Sophia’’ Children’s Hospital, Athens, Greece

123

Cancer Causes Control (2013) 24:1111–1118

DOI 10.1007/s10552-013-0188-y

Author's personal copy

their trajectory to converge with Northern–Western EU

counterpart states. Socioeconomic status may act as a

multipotent factor underlying the study findings.

Keywords Childhood leukemia � Childhood lymphoma �Survival � Incidence � Mortality � Time trends �Socioeconomic status

Introduction

Childhood leukemia, at least in the developed countries,

seems to exhibit a two-faced pattern regarding incidence

and survival rates, in the sense that survival rates seem to

be impressively and constantly improving, whereas in the

same period increasing incidence trends have been widely

reported [1–3]. Longer survival is reflected upon sizeable

decrease in mortality rates, as reported in the USA [4],

Europe [5], Latin America, Asia, and Oceania [6].

The globally reported impact of survival rates upon

mortality may not seem surprising, given the considerable

therapeutic progress. Indeed, the 5-year survival rates have

climbed from\5 % before the early 1960s [7, 8] to close to

90 % [9] for acute lymphoblastic leukemia (ALL) and

approximately 50 % for acute myeloid leukemia (AML)

[10]. Regarding childhood lymphoma, 5-year survival for

non-Hodgkin lymphoma (NHL) has also increased, reach-

ing 88 % in the USA, whereas steady improvements in 5-

and 10-year survival rates for Hodgkin’s lymphoma (HL)

are approaching figures as high as 95 % [11]. In the

European Union, however, the impressive declines in

mortality rates for childhood leukemia from 1.7 to 0.9/

100,000 for boys and from 1.3 to 0.7/100,000 for girls do

not seem to be equally shared, as the respective figures

were considerably higher in Southern or Eastern European

countries [5].

The present study focuses on the rather less privileged

region of Southern and Eastern Europe, aiming to compar-

atively assess incidence and mortality time trends of child-

hood leukemia and lymphoma in five Registries of four

South-Eastern and Eastern European countries and to

examine whether inherent sociodemographic and clinical

parameters, notably age, gender, country, subtype of leuke-

mia/lymphoma, and place of residence, affect survival in the

cross-national study setting. Specifically, data gathered from

two nationwide Cancer Registries in two adjacent EU

member states [Nationwide Registry for Childhood Hema-

tological Malignancies (NARECHEM) in Greece and the

Bulgarian National Cancer Registry (BNCR)] as well as

from three Regional Registries located in countries sharing

borders with EU, namely Russia (Moscow Region Cancer

Registry) and Turkey (Antalya and Izmir Cancer Registries),

during a period spanning circa 1996–2010 were examined.

Materials and methods

Country (Greece, Bulgaria, Turkey, and Russia) and Reg-

istry (NARECHEM, BNCR, Antalya, Izmir, and Moscow

Region Cancer Registries) characteristics are provided in

the Supplemental Methods file for all five Registries.

Specifically, during a 15-year study period for which data

were available, a total of 1,215 incident leukemia cases and

422 lymphoma cases were retrieved from NARECHEM

(1996–2010); 675 leukemias and 292 lymphomas from

BNCR (1996–2009); 403 leukemias and 158 lymphomas

from the Moscow Region Registry (2000–2010); 323 leu-

kemias and 121 lymphomas from the Antalya Registry

(1998–2008); and 425 leukemias and 190 lymphomas from

the Izmir Cancer Registry (1996–2007) and contributed to

the study dataset.

Data analysis

Details regarding the analysis of incidence and mortality

rates by country as well as the respective time trends are

presented in the Supplemental Methods.

Survival (vital status) data by age, gender, place of resi-

dence at diagnosis, and location of Registry as well by specific

subtypes of the disease (ALL, AML and other leukemia type;

HL, NHL, Burkitt and other lymphoma type) were compared;

the chi-square test was used to identify any differences. In the

survival analyses, only incident cases diagnosed in the Reg-

istries during the study periods were included.

Finally, survival of leukemia or lymphoma was exam-

ined, by modeling the data through Cox proportional hazards

multivariate regression analysis. Independent variables

included age, gender, location of the Registry, disease sub-

type, and place of living. Data for the latter variable were not

F. Athanasiadou-Piperopoulou

2nd Department of Pediatrics, Aristotelian University

of Thessaloniki, AHEPA General Hospital, Thessaloniki, Greece

I. Fragandrea

Pediatric Oncology Department, Hippokration

Hospital, Thessaloniki, Greece

E. Stiakaki

Department of Pediatric Hematology-Oncology,

University of Crete, Heraklion, Crete, Greece

C. Stefanadis

1st Department of Cardiology, Athens University

Medical School, Athens, Greece

J.-W. Coebergh

Eindhoven Cancer Registry of Comprehensive

Cancer Centre South, Eindhoven, The Netherlands

1112 Cancer Causes Control (2013) 24:1111–1118

123

Author's personal copy

available for Moscow Registry, whereas there were some

missing values in the Greek and the Antalya Registry; con-

sequently, it was introduced as an additional variable to the

initial model. Missing cases were due to administrative

reasons and were not considered to introduce any bias in the

results. Follow-up was active for NARECHEM, Izmir, and

Antalya, passive for the BNCR, and both active and passive

for the Moscow Region Registry, with 31 December 2010

being the censoring date. Statistical analyses were carried

out using SAS software, version 9 for Windows (SAS

Institute Inc., Cary, NC, USA).

Results

Table 1 presents the calculated average age-adjusted and

crude incidence rates for the initial and last 3 years as well

as the percentage of the annual change in each Registry.

The incidence of leukemia was increasing in all Registries,

with Bulgaria and Greece presenting statistically signifi-

cant annual changes, followed by Izmir and Moscow

Region with marginally increasing trends (Fig. 1a). Con-

cerning mortality rates, there was a remarkable, decreasing

trend in all Registries with Moscow presenting the highest

annual change (-9.9 %), followed by Greece and Bulgaria

(Fig. 1b). The AMRs in Bulgaria were considerably higher

Table 1 Average first and last 3 years (circa 1996–2010), age-adjusted and crude incidence and mortality rates (AIR, CIR and AMR, CMR) per

1,000,000 in the 5 Cancer Registries, along with corresponding annual change, 95 % confidence intervals (95 % CI), and p values

Country Greece Bulgaria Turkey Turkey Russia

Cancer Registry NARECHEM BNCR Izmir CR Antalya CR Moscow Region CR

Years 3-first 3-last 3-first 3-last 3-first 3-last 3-first 3-last 3-first 3-last

Leukemia

AIR 49.1 56.6 38.7 48.6 46.9 58.2 52.6 60.1 45.7 43.2

CIR 43.7 54.4 35.2 47.2 45.0 53.5 48.2 54.9 37.5 41.8

% annual change

(95 % CI), p value

1.7 (0.4, 3.0), 0.01 3.5 (1.6, 5.5), 0.0002 2.7 (-0.1, 5.6), 0.06 0.9 (-8.7, 11.5), 0.86 2.7 (-0.3, 5.9), 0.08

AMR 12.1 5.6 21.7 12.0 NA NA 16.7 4.0

CMR 11.6 5.6 20.8 12.0 NA NA 14.3 3.8

% annual change

(95 % CI), p value

-4.5 (-8.0, -0.9),

0.02

-3.7 (-6.6, -0.8),

0.01

NA NA -9.9 (-15.5, -4.0),

0.001

Lymphoma

AIR 16.7 17.1 17.0 14.0 17.0 24.8 18.4 17.3 14.6 13.0

Crude IR 18.1 17.8 17.9 14.6 17.9 25.2 17.9 17.0 17.8 13.4

% annual change

(95 % CI), p value

-0.5 (-2.7, 1.7), 0.64 -2.1 (-4.8, 0.8), 0.15 3.7 (-0.5, 8.1), 0.08 2.3 (-12.6, 19.8),

0.78

-3.0 (-7.6, 1.8), 0.22

Average AMR 0.8 2.9 5.2 3.5 NA NA 2.1 2.2

Crude MR 0.7 2.9 5.2 3.3 NA NA 2.9 2.3

% annual change

(95 % CI), p value

8.2 (0.2, 16.9), 0.05 -5.3 (11.0, 0.8), 0.09 NA NA -4.6 (-16.6, 9.2), 0.49

NA Not available

2000-2010

1996-2007

1998-2008

1996-2009

1996-2010

AIR for leukemia -Annual increase (%, 95%CIs) in each Registry

+2.7 (-0.3, +5.9)

+2.7 (-0.1, +5.6)

+0.9 (-8.7, +11.5)

+3.5 (+1.6, +5.5)

+1.7 (+0.4, +3.0)

AMR for leukemia -Annual decrease (%, 95%CIs) in each Registry

-9.9 (-15.5, -4.0) Moscow

Not available

-3.7 (-6.6, -0.8)

-4.5 (-8.0, -0.9)

Izmir /

Antalya

Bulgaria

Greece

2000-2010

1996-2009

1996-2010

a

b

Fig. 1 a AIR for leukemia. Annual increase (%, 95 % CIs) in each

Registry. b AMR for leukemia. Annual decrease (%, 95 % CIs) in

each Registry

Cancer Causes Control (2013) 24:1111–1118 1113

123

Author's personal copy

than the corresponding AMRs for Greece (p = 0.002) and

Moscow Region (p \ 0.001), especially during the last

period.

In contrast to leukemia, results about lymphoma showed no

significant trends in AIRs; Moscow Region displayed the

lowest AIRs (Moscow vs. Izmir p = 0.0015). Concerning

mortality, Bulgaria appeared to have a considerably higher

AMR than the other two Registries with available data for the

initial 3 years, reaching significance versus Greece

(p \ 0.001); the sizably lower starting mortality rate (first

3-year period) in Greece might have resulted in the increasing

trend observed in the most recent period (p = 0.05). Sup-

plemental Figs. 1 and 2 present the comparisons between

countries regarding incidence and mortality rates.

In Table 2, the distribution of the sociodemographic and

disease characteristics among the studied regions is dis-

played for the total of 3,041 incident childhood leukemia and

1,183 lymphoma cases. The known preponderance of males

was noted; considerable differentials pertaining to age were

observed (p = 0.003), with infant leukemia occurrence

being higher in Turkey and leukemia in children 10? year

old being more common in Bulgaria. Semi-urban place of

living was the predominant category for leukemia cases

reported from Greece and Bulgaria, whereas in Izmir and

Antalya, the majority of cases were living in urban areas

(p = 0.0001 for place of residence). Of note is the very low

proportion of other than ALL or AML leukemias in Greece,

which should also be interpreted along with the favorable

survival in this country (82.1 %).

As expected, lymphomas were more common among

boys. The high proportion of Burkitt lymphoma in Greece

and Turkey ([21 %) and the disparities in survival from

87 % in Greece and Moscow Region to 70 % in Bulgaria

seem notable.

In Supplemental Table 1, survival (vital) status by soci-

odemographic and disease characteristics are presented;

these results serve descriptive purposes only, as they are not

mutually adjusted. Table 3 quantifies these results by age:

indeed, infants (B1 year old) were more than 2 times more

likely to die from leukemia compared with standard age risk

children (1–9 years), the latter having also better survival

compared with older children. Likewise, the risk of death

was more than twofold higher among children diagnosed

with AML and ‘‘other leukemia’’ subtypes, compared with

ALL. It is worth noting that an established risk factor, namely

male gender, did not seem to exert a sizeable role; moreover,

the risk of death in the other four participating sites was

1.5–2.5 times higher compared with that in Greece. Lastly,

rural place of residence seemed to increase marginally the

risk of dying by 21 % (p = 0.06); nevertheless, at a post hoc

sensitivity analysis excluding the best performing country

(Greece), area remoteness reached formal significance as

both semi-urban (RR = 1.28, p = 0.03) and rural (RR =

1.17, p = 0.04) places of residence were associated with

poorer survival (Supplemental Table 2). As a rule, the sub-

analysis on ALL, which comprised over 80 % of the total

leukemias, replicated the overall results (Supplemental

Table 3).

Concerning the survival analysis for lymphoma, a 27 %

detrimental effect of gender was noted that did not reach

nominal statistical significance (p value: 0.10). The out-

come by disease subtype was better for HL, as expected.

The country of origin seemed to follow the pattern noted in

childhood leukemia. Finally, lymphoma cases living in

rural areas had 28 % increased death rates than those

residing in urban areas, although this result did not reach

significance.

Discussion

The main messages of this study pertain to the uniform

improvement in mortality rates for childhood leukemia,

together with increasing incidence rates, predominantly for

Greece and Bulgaria. On the contrary, regarding childhood

lymphoma, no sizable, consistent patterns were observed

either at the level of mortality or at the level of incidence.

Importantly, predictors modifying survival from childhood

leukemia in this cross-national setting included also rural

place of living and country, apart from the rather well-

established effects of age and leukemia type.

The present, favorable findings regarding the evolution

of mortality rates seem in accordance with the conclusions

of the most recent EU-wide study, pointing to room for

further improvement in the Southern-Eastern European

countries [5] along their trajectory to convergence with

their Northern and Western EU counterpart states. The

background reasons for the globally observed improvement

in mortality and survival trends may encompass the

implementation and administration of multidrug chemo-

therapy protocols, the development of supportive measures

limiting the effects of toxicity, the progress in bone marrow

transplantation, and improved diagnostic methods [6].

The increasing rates in the incidence of childhood leukemia

may be inscribed into the wider context of upward incidence

figures reported in the USA (1992–2004) [3] and the Europe-

wide Automated Childhood Cancer Information System

(ACCIS) project (1978–1997) [1]. Similar data have appeared

from country-specific studies over various periods, such as the

UK (1971–2000) [8], Spain (1983–2002) [12], Germany

(1987–2004) [13], Italy (1967–2001) [14]; nevertheless, the

Nordic countries (1982–2001) [15] and Canada (1999–2006)

[16] have reported stable childhood leukemia rates during the

most recent years. Interestingly, the present study, encom-

passing four additional years, when compared to the previous

NARECHEM Greek study (1996–2006) [17], has allowed the

1114 Cancer Causes Control (2013) 24:1111–1118

123

Author's personal copy

detection of the significant upward trend, most possibly due to

larger statistical power and longer study period. At a patho-

physiological level, the underlying explanation remains elu-

sive; changes in lifestyle habits, decreased crowding index in

the family [18] and day care [13], modifications in maternal

diet [19], decreasing exposure to infectious agents [20] or

underlying trends of atopic disease [21], or even growingly

larger proportion of otherwise infertile women giving birth

Table 2 Distribution of the 3,041 leukemia and 1,183 lymphoma incident cases recorded circa 1996–2010 in the 5 Cancer Registries by study

variables and survival status as of 31 December 2010

Greece Bulgaria Izmir-Turkey Antalya Turkey Moscow Region p value§

N % N % N % N % N %

Leukemia

Age (years) 0.003

\1 47 3.9 31 4.6 30 7.1 19 5.9 14 3.5

1–9 939 77.3 473 70.1 298 70.1 242 74.9 305 75.7

10? 229 18.8 171 25.3 97 22.8 62 19.2 84 20.8

Sex 0.26

Male 688 56.6 365 54.1 220 51.8 191 59.1 224 55.6

Female 527 43.4 310 45.9 205 48.2 132 40.9 179 44.4

Place of residence 0.0001

Rural \10,000 341 29.9 188 27.9 41 9.6 74 23.9 MA MA

Semi-urban 553 48.5 275 40.7 36 8.5 107 34.5

Urban 100,000? 247 21.6 212 31.4 348 81.9 129 41.6

Type 0.0001

ALL 1056 86.9 530 78.5 328 77.2 246 76.2 320 79.4

AML 143 11.8 90 13.3 72 16.9 50 15.5 61 15.1

Other 16 1.3 55 8.2 25 5.9 27 8.3 22 5.5

Status 0.0001

Dead 217 17.9 273 40.4 135 31.8 116 35.9 93 23.1

Alive 998 82.1 402 59.6 290 68.2 207 64.1 310 76.9

Lymphoma

Age (years) 0.0001

\10 215 50.9 149 51.0 111 58.4 85 70.3 53 33.5

10? 207 49.1 143 49.0 79 41.6 36 29.7 105 66.5

Sex 0.01

Male 302 71.6 200 68.5 129 67.9 66 54.6 101 63.9

Female 120 28.4 92 31.5 61 32.1 55 45.4 57 36.1

Place of residence 0.0001

Rural \10,000 96 29.5 83 28.4 19 10.0 37 32.5 MA MA

Semi-urban 179 55.1 111 38.0 18 9.5 35 30.7

Urban 100,000? 50 15.4 98 33.6 153 80.5 42 36.8

Type 0.0001

HL 173 41.0 160 54.8 73 38.4 39 32.2 84 53.2

NHL, non-Burkitt 139 32.9 98 33.5 74 38.9 50 41.3 48 30.4

Burkitt 110 26.1 25 8.6 40 21.1 25 20.7 22 13.9

Other 0 0.0 9 3.1 3 1.6 7 5.8 4 2.5

Status 0.0001

Dead 52 12.3 87 29.8 38 20.0 22 18.2 21 13.3

Alive 370 87.7 205 70.2 152 80.0 99 81.8 137 86.7

Place of residence is missing for 490 leukemia cases (74 for Greece, 13 for Antalya, and 403 for Moscow Region) and for 262 lymphoma cases

(97 for Greece, 7 for Antalya, and 158 for Moscow Region)§ p values derived from chi-square test

Cancer Causes Control (2013) 24:1111–1118 1115

123

Author's personal copy

with the help of assisted reproduction technologies [1, 22]

have been implicated, along with the delivery of higher-birth-

weight newborns.

Poorer survival was noted in all countries for either

childhood leukemia or lymphoma when compared to the

reference category, that is, Greece; this finding may not seem

surprising, as registry site may only represent a proxy of the

underlying socioeconomic status. Specifically, according to

the World Bank Data, the per capita income in Greece was

consistently and considerably higher than that in Bulgaria,

Turkey, or Russia during the study period [23]. Indeed, lower

socioeconomic status is a well-established, global detri-

mental factor for survival from childhood leukemia, as evi-

denced from studies conducted in numerous high- and

middle-income countries, such as the UK [24], USA [25],

Greece [26], Latin America [27], and Asia [28].

Rural place of residence also emerged as another poor

prognostic factor for survival from childhood leukemia, after

the exclusion of the most privileged country, namely Greece.

Indeed, the most recent study in Greece has highlighted that

newly established, charitable family settlements nearby the

treating hospitals, as well as extensive improvements in

motorway infrastructures linked with ATHENS 2004

Olympic Games, have counterbalanced the effects of area

remoteness in this relatively small country [26]. The detri-

mental role of area remoteness in the remaining countries

seems in line with a recent Australian study, according to

which remoteness was associated with shorter survival rates

for childhood cancer and leukemia, most probably due to

reduced access to diagnostic, treatment, and support services

[29]. Nevertheless, rural residence may also entail an aspect

relevant to the aforementioned socioeconomic status notions,

as more remote areas tend also to be afflicted by lower

socioeconomic indicators [30]. The association between

poorer survival and rural place of living did not reach signif-

icance concerning childhood lymphoma most probably due to

smaller numbers, as the effect estimate was numerically equal

to that for childhood leukemia (RR = 1.28).

Apart from the associations discussed above, the survival

analysis confirmed the bimodal association of age with

prognosis of childhood leukemia [31]. Similarly, the better

survival rates achieved for ALL among leukemia subtypes

[9, 10] and HL among lymphoma subtypes [11] were con-

firmed. The fact that female gender was not associated with

poorer prognosis from childhood leukemia may not seem

surprising; indeed, it seems that improved risk stratification

Table 3 Death rate ratios (RR) along with 95 % confidence intervals (95 % CIs) and p values derived from proportional hazards modeling

among children with leukemia or lymphoma by sociodemographic and clinical variables

Variable Category or increment Leukemia Lymphoma

RR 95 % CIs p value RR 95 % CIs p value

Age \1 years 2.49 1.94 3.19 0.0001

1–9 (\10§) Baseline Baseline

10? 1.57 1.34 1.83 0.0001 0.93 0.70 1.23 0.59

Gender Male Baseline Baseline

Female 1.01 0.87 1.15 0.98 1.27 0.96 1.68 0.10

Country/region Greece Baseline Baseline

Bulgaria 2.53 2.11 3.03 0.0001 2.91 2.05 4.14 0.0001

Izmir 1.83 1.47 2.27 0.0001 1.77 1.17 2.70 0.01

Antalya 2.44 1.95 3.07 0.0001 1.71 1.02 2.85 0.04

Moscow 1.64 1.28 2.09 0.0001 1.52 0.91 2.54 0.11

Leukemia type ALL Baseline

AML 2.83 2.40 3.32 0.0001

Other 2.72 2.13 3.46 0.0001

Lymphoma type HL Baseline

NHL 3.12 2.27 4.30 0.0001

Burkitt 0.69 0.47 1.00 0.05

Other 5.84 3.00 11.35 0.0001

Additionally introduced variable available for 2,551 leukemia and 921 lymphoma cases

Place of residence Rural(\10,000) 1.21 0.99 1.48 0.06 1.28 0.87 1.89 0.21

Semi-urban 1.11 0.92 1.34 0.30 0.91 0.62 1.35 0.64

Urban (100,000?) Baseline Baseline

§ Age was treated \1 versus 1–9 and 10? versus 1–9 for leukemia; 10? versus \10 years for lymphoma

1116 Cancer Causes Control (2013) 24:1111–1118

123

Author's personal copy

and newer chemotherapy regimens have eliminated the

previous independent prognostic significance of gender, as

observed in the most recent Nordic Society of Paediatric

Haematology and Oncology (NOPHO) data [32].

Regarding limitations of this study, the fact that the survival

analysis comprised data from all five Registries, whereas the

figures on mortality rates did not include data from Izmir and

Antalya, may not seem mutually contradictory. Specifically,

the survival analysis included data from incident cases during

the study period, whereas mortality data were based on cases

dying during the study period; inconsistencies regarding the

follow-up of earlier cases in the Turkish Registries did not

allow the inclusion of the latter in the analysis of mortality

trends. Other limitations of the present effort include the fact

that more sophisticated models, such as joint-point regression

analysis, were not implemented, whereas the fact that the

study periods for each Registry did not absolutely coincide

should also be declared. Moreover, time trends in the validity

of the registration process could theoretically represent a key

limitation affecting the observed trends; nevertheless, the fact

that all five included Registries provide data to International

Association for Research on Cancer (IARC)-related databases

may underline the validity and quality of the included Registries.

In conclusion, this study points to a uniform improve-

ment in mortality rates from childhood leukemia and

lymphoma in Southern–Eastern Europe, although of dif-

ferent size. The increasing incidence trends follow soci-

odemographic trends seen elsewhere, whereas the nature of

the underlying pathophysiological mechanisms remains to

be explored. Socioeconomic status consistently as a mul-

tipotent factor underlies the study findings about inequality

in survival between and within countries, which seemed

also of special interest, as constantly changing financial

conditions and local economic crises in the region may

modify the picture portrayed herein in the future.

Acknowledgments This study has been supported in part by the

National and Kapodistrian University of Athens. Special thanks are

due to Nick Dessypris for his contribution in the collection and

analysis of the data.

References

1. Kaatsch P, Steliarova-Foucher E, Crocetti E, Magnani C, Spix C,

Zambon P (2006) Time trends of cancer incidence in European

children (1978–1997): report from the Automated Childhood

Cancer Information System project. Eur J Cancer 42:1961–1971

2. Coebergh JW, Reedijk AM, de Vries E, Martos C, Jakab Z,

Steliarova-Foucher E, Kamps WA (2006) Leukaemia incidence

and survival in children and adolescents in Europe during

1978–1997. Report from the Automated Childhood Cancer

Information System project. Eur J Cancer 42:2019–2036

3. Linabery AM, Ross JA (2008) Trends in childhood cancer inci-

dence in the U.S. (1992–2004). Cancer 112:416–432

4. Centers for Disease Control and Prevention (CDC) (2007) Trends

in childhood cancer mortality–United States, 1990–2004.

MMWR Morb Mortal Wkly Rep 56:1257–1261

5. Bosetti C, Bertuccio P, Chatenoud L, Negri E, Levi F, La Vecchia

C (2010) Childhood cancer mortality in Europe, 1970–2007. Eur

J Cancer 46:384–394

6. Chatenoud L, Bertuccio P, Bosetti C, Levi F, Negri E, La Vecchia

C (2010) Childhood cancer mortality in America, Asia, and

Oceania, 1970 through 2007. Cancer 116:5063–5074

7. Thames Cancer Registry (1994) Cancer in South East England,

1991: cancer incidence, prevalence and survival in residents of

the district health authorities in South East England. Thames

Cancer Registry, Sutton, p 43

8. Shah A, Coleman MP (2007) Increasing incidence of childhood

leukaemia: a controversy re-examined. Br J Cancer 97:1009–1012

9. Pui CH, Mullighan CG, Evans WE, Relling MV (2012) Pediatric

acute lymphoblastic leukemia: where are we going and how do

we get there? Blood 120:1165–1174

10. Colby-Graham MF, Chordas C (2003) The childhood leukemias.

J Pediatr Nurs 18:87–95

11. Pulte D, Gondos A, Brenner H (2008) Trends in 5- and 10-year

survival after diagnosis with childhood hematologic malignancies

in the United States, 1990–2004. J Natl Cancer Inst 100:

1301–1309

12. Peris-Bonet R, Salmeron D, Martinez-Beneito MA, Galceran J,

Marcos-Gragera R, Felipe S, Gonzalez V, Sanchez de Toledo

Codina J (2010) Childhood cancer incidence and survival in

Spain. Ann Oncol 21(Suppl 3):iii103–iii110

13. Spix C, Eletr D, Blettner M, Kaatsch P (2008) Temporal trends in

the incidence rate of childhood cancer in Germany 1987–2004.

Int J Cancer 122:1859–1867

14. Dalmasso P, Pastore G, Zuccolo L, Maule MM, Pearce N, Mer-

letti F, Magnani C (2005) Temporal trends in the incidence of

childhood leukemia, lymphomas and solid tumors in north-west

Italy, 1967–2001. A report of the Childhood Cancer Registry of

Piedmont. Haematologica 90:1197–1204

15. Hjalgrim LL, Rostgaard K, Schmiegelow K, Soderhall S, Kol-

mannskog S, Vettenranta K, Kristinsson J, Clausen N, Melbye M,

Hjalgrim H, Gustafsson G (2003) Age- and sex-specific incidence

of childhood leukemia by immunophenotype in the Nordic

countries. J Natl Cancer Inst 95:1539–1544

16. Mitra D, Shaw AK, Hutchings K (2012) Trends in incidence of

childhood cancer in Canada, 1992–2006. Chronic Dis Inj Can

32:131–139

17. Petridou ET, Pourtsidis A, Dessypris N, Katsiardanis K, Baka M,

Moschovi M, Polychronopoulou S, Koliouskas D, Sidi V, Atha-

nasiadou-Piperopoulou F, Kalmanti M, Belechri M, La Vecchia

C, Curado MP, Skalkidis I (2008) Childhood leukaemias and

lymphomas in Greece (1996–2006): a nationwide registration

study. Arch Dis Child 93:1027–1032

18. Petridou E, Trichopoulos D, Kalapothaki V, Pourtsidis A, Ko-

gevinas M, Kalmanti M, Koliouskas D, Kosmidis H, Panagiotou

JP, Piperopoulou F, Tzortzatou F (1997) The risk profile of

childhood leukaemia in Greece: a nationwide case–control study.

Br J Cancer 76:1241–1247

19. Petridou E, Ntouvelis E, Dessypris N, Terzidis A, Trichopoulos D

(2005) Maternal diet and acute lymphoblastic leukemia in young

children. Cancer Epidemiol Biomarkers Prev 14:1935–1939

20. Greaves MF, Alexander FE (1993) An infectious etiology for

common acute lymphoblastic leukemia in childhood? Leukemia

7:349–360

21. Schuz J, Morgan G, Bohler E, Kaatsch P, Michaelis J (2003)

Atopic disease and childhood acute lymphoblastic leukemia. Int J

Cancer 105:255–260

22. Petridou ET, Sergentanis TN, Panagopoulou P, Moschovi M,

Polychronopoulou S, Baka M, Pourtsidis A, Athanassiadou F,

Cancer Causes Control (2013) 24:1111–1118 1117

123

Author's personal copy

Kalmanti M, Sidi V, Dessypris N, Frangakis C, Matsoukis IL,

Stefanadis C, Skalkidou A, Stephansson O, Adami HO, Kieler H

(2012) In vitro fertilization and risk of childhood leukemia in

Greece and Sweden. Pediatr Blood Cancer 58:930–936

23. The World Bank. GNI per capita, Atlas method (current US$)

24. Lightfoot TJ, Johnston WT, Simpson J, Smith AG, Ansell P,

Crouch S, Roman E, Kinsey SE (2012) Survival from childhood

acute lymphoblastic leukaemia: the impact of social inequality in

the United Kingdom. Eur J Cancer 48:263–269

25. Kent EE, Sender LS, Largent JA, Anton-Culver H (2009) Leu-

kemia survival in children, adolescents, and young adults: influ-

ence of socioeconomic status and other demographic factors.

Cancer Causes Control 20:1409–1420

26. Sergentanis TN, Dessypris N, Kanavidis P, Skalkidis I, Baka M,

Polychronopoulou S, Athanassiadou F, Stiakaki E, Fragandrea I,

Moschovi M, Petridou ET (2012) Socio-economic status, area

remoteness and survival from childhood leukemia: results from

the Nationwide Registry for Childhood Hematological Malig-

nancies (NARECHEM) in Greece. Eur J Cancer Prev 12 Dec

2012 [Epub ahead of print]

27. Bonilla M, Gupta S, Vasquez R, Fuentes SL, deReyes G, Ribeiro

R, Sung L (2010) Predictors of outcome and methodological

issues in children with acute lymphoblastic leukaemia in El

Salvador. Eur J Cancer 46:3280–3286

28. Tang Y, Xu X, Song H, Yang S, Shi S, Wei J (2008) Long-term

outcome of childhood acute lymphoblastic leukemia treated in

China. Pediatr Blood Cancer 51:380–386

29. Youlden DR, Baade PD, Valery PC, Ward LJ, Green AC, Aitken

JF (2011) Differentials in survival for childhood cancer in Aus-

tralia by remoteness of residence and area disadvantage. Cancer

Epidemiol Biomarkers Prev 20:1649–1656

30. Eurostat. Regional gross domestic product (PPS per inhabitant)

by NUTS 2 regions

31. Donadieu J, Auclerc MF, Baruchel A, Perel Y, Bordigoni P,

Landman-Parker J, Leblanc T, Cornu G, Sommelet D, Leverger

G, Schaison G, Hill C (2000) Prognostic study of continuous

variables (white blood cell count, peripheral blast cell count,

haemoglobin level, platelet count and age) in childhood acute

lymphoblastic leukaemia. Analysis of a population of 1545

children treated by the French Acute Lymphoblastic Leukaemia

Group (FRALLE). Br J Cancer 83:1617–1622

32. Schmiegelow K, Forestier E, Hellebostad M, Heyman M, Kris-

tinsson J, Soderhall S, Taskinen M (2010) Long-term results of

NOPHO ALL-92 and ALL-2000 studies of childhood acute

lymphoblastic leukemia. Leukemia 24:345–354

1118 Cancer Causes Control (2013) 24:1111–1118

123

Author's personal copy