Vous êtes sur la page 1sur 6


und Onkologie Original Article

Prognostic Impact of HIF-1α Expression in Patients

with Definitive Radiotherapy for Cervical Cancer
Kathrin Dellas1*, Matthias Bache1*, Steffi U. Pigorsch2, Helge Taubert3, Matthias Kappler1,
Daniel Holzapfel1, 3, Ester Zorn1, 3, Hans-Juergen Holzhausen3, Gabriele Haensgen1

Purpose: To investigate the relationship between the hypoxia-inducible factor-(HIF-)1α expression in tumor tissue, tumor oxygen-
ation and hemoglobin levels in patients with advanced cervical cancers prior to radiotherapy and the effect on clinical outcome.
Patients and Methods: The investigation included 44 patients who underwent definitive radiotherapy for advanced cervical
cancers between May 1995 and March 1999. Tumor biopsies were taken prior to treatment, and HIF-1α expression was determined
by immunohistochemistry. In the same tumor area, tumor tissue oxygenation (pO2) was measured using the Eppendorf device.
Results: The 5-year cancer-specific survival of all patients was 60%. Twelve of 44 tumor specimens were HIF-1α-negative with a
significantly better 5-year survival (92 ± 8%) versus 32 patients who were HIF-1α-positive (45 ± 10%; p < 0.02). There was no
correlation between HIF-1α expression and tumor oxygenation (p = 0.57 both for pO2 median and hypoxic fraction < 5 mmHg vs.
HIF-1α expression). However, patients with hemoglobin levels < 11 g/dl showed elevated HIF-1α expression compared to patients
with hemoglobin levels > 12.5 g/dl (p = 0.04). Furthermore, HIF-1α correlated with vascular endothelial growth factor expression
(p = 0.002). In a multivariate Cox regression model, HIF-1α expression (relative risk [RR] = 7.5; p = 0.05) revealed an increased risk
of tumor-related death.
Conclusion: The study indicates, that endogenous tumor markers such as HIF-1α may serve as prognostic markers of clinical
outcome concerning cervical cancer after primary radiotherapy.
Key Words: Cervical cancer · HIF-1α · Tumor oxygenation · Hemoglobin · Radiotherapy

Strahlenther Onkol 2008;184:169–74

DOI 10.1007/s00066-008-1764-z

Die Bedeutung der HIF-1α-Expression als prognostischer Marker in der Radiotherapie von Patientinnen mit Zervix-
Ziel: Untersuchung möglicher Zusammenhänge zwischen der Expression des hypoxieinduzierbaren Faktors-(HIF-)1α in Tumorge-
webe, der Tumoroxygenierung und dem Hämoglobinwert bei Patientinnen mit fortgeschrittenem Zervixkarzinom vor der Radiothe-
rapie sowie der Auswirkungen dieser Parameter auf den klinischen Verlauf.
Patienten und Methodik: Zwischen Mai 1995 und März 1999 wurden 44 Patientinnen eingeschlossen, die sich einer definitiven
Radiatio eines fortgeschrittenen Zervixkarzinoms unterzogen. Vor Beginn der Therapie wurden Bioptate entnommen und die
HIF-1α-Expression durch Immunhistochemie bestimmt. Außerdem wurde in derselben Tumorregion die Tumoroxygenierung (pO2)
mit dem Eppendorf-Histographen gemessen.
Ergebnisse: Das tumorspezifische 5-Jahres-Überleben betrug 60%. Zwölf von 44 Tumorproben waren HIF-1α-negativ mit einem
signifikant besseren 5-Jahres-Überleben (92 ± 8%) gegenüber 32 Patientinnen mit positivem Nachweis einer HIF-1α-Expression
(45 ± 10%; p < 0,02). Es konnte keine Korrelation zwischen der HIF-1α-Expression und der Tumoroxygenierung nachgewiesen werden
(jeweils p = 0,57 für die Korrelation mit dem medianen pO2 und der hypoxischen Fraktion < 5 mmHg). Jedoch zeigten Patientinnen
mit einem Hämoglobinwert < 11 g/dl eine erhöhte HIF-1α-Expression im Vergleich zu Patientinnen mit einem Hämoglobinwert
> 12,5 g/dl (p = 0,04). Eine Korrelation wurde darüber hinaus zwischen einer HIF-1α- and einer VEGF-(„vascular endothelial growth
factor“-)Expression (p = 0,002) nachgewiesen. Die multivariate Cox-Regression zeigte ein erhöhtes Risiko für einen tumorassoziierten
Tod bei Patientinnen mit Nachweis einer HIF-1α-Expression (relatives Risiko [RR] = 7,5; p = 0,05).
Schlussfolgerung: Die Untersuchungen deuten darauf hin, dass endogene Tumormarker wie HIF-1α als prognostische Parameter
für den klinischen Verlauf bei Patientinnen mit Zervixkarzinom nach einer definitiven Radiotherapie von Bedeutung sein können.
Schlüsselwörter: Zervixkarzinom · HIF-1α · Tumoroxygenierung · Hämoglobin · Radiotherapie

*Both authors contributed equally to the study.

Department of Radiotherapy, Martin Luther University of Halle-Wittenberg, Germany,
Department of Radiotherapy and Radiation Oncology, Technical University of Munich, Germany,
Institute of Pathology, Martin Luther University of Halle-Wittenberg, Germany.

Received: April 30, 2007; accepted: December 13, 2007

Strahlenther Onkol 2008 · No. 3 © Urban & Vogel 169

Dellas K, et al. Prognostic Impact of HIF-1α Expression

Introduction ethics committee, and all of the patients gave written informed
Tumor oxygenation has been recognized as an important pre- consent.
dictive parameter for overall survival, disease-free survival The mean age of the patient population was 58.4 years
and treatment response especially in cervical carcinoma. The (35–76 years). The median follow-up at the time of this analy-
Eppendorf histograph directly measuring the intratumoral sis was 45 months (6–60 months). Patients were clinically cat-
pO2 has been used in clinical studies since the late 1980s in hu- egorized on the basis of clinical examination and CT scan
man cervical carcinomas [11–13, 17, 18, 22, 25, 27, 31, 36, 37, and/or MRI. Nine patients were classified as FIGO IIB, 31
39]. Our own results that examined changes in median pO2 patients as FIGO IIIB, and four patients belonged to stage
values among patients undergoing radiation therapy for ad- IVA disease.
vanced cervical cancer indicated that pO2 level was an adverse
prognostic factor for local control, but only in cases in which Radiotherapy
hypoxia was sustained after a 2-week course of irradiation Irradiation followed a curative concept by combining both
(19.8 Gy) [8, 9, 14]. The measurement with oxygen electrodes external-beam radiotherapy and brachytherapy. In patients
has the critical disadvantages of an invasive procedure and is with stage IIB and IIIB disease, radiotherapy was delivered
restricted to a few accessible tumors. with a linear accelerator using 15-MeV photons and individu-
Endogenous markers have the potential to indicate thera- ally shaped anterior-posterior/posterior-anterior pelvic por-
peutically relevant levels of hypoxia within tumors. The most tals. A dose of 45 Gy was given in daily fractions of 1.8 Gy,
important protein regulating the molecular response on hy- 5 days/week. Additionally, patients received high-dose-rate
poxia is hypoxia-inducible factor-1 (HIF-1), a heterodimeric (HDR) brachytherapy with 192Ir (5 × 7 Gy) with a remote af-
transcription factor consisting of two subunits, HIF-1α and terloading technique once a week (twice in the last week). The
HIF-1β (also known as aryl hydrocarbon receptor nuclear total dose of external-beam radiotherapy and brachytherapy
transporter [ARNT]). Under hypoxic conditions, HIF-1α pro- reached 65–70 Gy in point A (primary cancer) and 52–54 Gy
tein is stabilized and accumulates rapidly allowing dimeriza- in point B (lymph nodes). In stage IVA, generally 54–60 Gy
tion with ARNT and translocation into the nucleus. As a tran- external-beam radiotherapy was given without brachytherapy.
scription factor, HIF-1α can transactivate more than 60 target
genes, and is a master regulator of angiogenesis, cell metabo- Determination of Oxygenation in Tumors
lism as well as genomic stability [35]. In numerous tumors, The measurements of oxygen tension in tumors were per-
overexpression of HIF-1α is a marker of aggressive disease formed polarographically using fine-needle probes by the Ep-
and is associated with poor prognosis and treatment failure pendorf pO2 device as previously described [11]. Four mea-
[15, 35]. The vascular endothelial growth factor (VEGF) regu- surement tracks with at least 100 readings were recorded for
lates proliferation, differentiation, migration and cell metabo- each patient. The relative frequencies of the median tumor
lism resulting in neovascularization and elevated vessel per- pO2 (15 mmHg) and the hypoxic measurements < 5 mmHg
meability. Expression of VEGF is mediated by HIF-1α and by (hypoxic fraction defined as pO2 < 5 mmHg [HF5] and 19% at
hypoxia-induced increased stability of VEGF mRNA [26]. the median, utilizing a cutoff of 50%) were evaluated. In all
Moreover, ischemia in the brain, heart or lung was associated patients, the Hb level was determined by standard procedures
with upregulation of this angiogenic factor [4]. These results prior to pO2 measurement.
may indicate that both HIF-1α and VEGF are important
markers of hypoxia [28]. However, the prognostic value of rel- HIF-1α and VEGF Immunohistochemical Staining
evant tumor hypoxia expression in clinically applicable meth- Immunohistochemical staining was previously described by
ods or as an endogenous marker for assessment is controver- Bache et al. [1, 2]. 4-µm sections of formalin-fixed and paraf-
sial in cervical cancers [29, 38]. We therefore performed an fin-embedded biopsies were placed on poly-l-lysine-coated
immunohistochemical analysis of HIF-1α and VEGF and re- slides and then deparaffinized. After inhibiting endogenous
lated the results to hemoglobin (Hb) level and tumor oxygen- peroxidase (0.1% H2O2 in 95% methanol for > 20 min at 4 °C)
ation, as determined by electrode measurement and clinical and nonspecific binding of the primary antibody (5% of the
outcome. carrier protein solution for 5 min), the incubation with the pri-
mary antibody of HIF-1α (ab463; Abcam, UK) and VEGF
Patients and Methods (Ab2; Calbiochem, Germany) was performed for 24 h at 4 °C,
Patient Population respectively. This process was followed by washing (0.067 M
This report is based on the investigation of 44 patients with phosphate-buffered saline, pH 7.4) and the detection of the
advanced cervical cancer who underwent definitive radiother- primary antibody by a secondary antibody (DCS Detection
apy between May 1995 and March 1999. All 44 patients in- Line; DCS Innovative Diagnostic Systems, Germany). Tissues
cluded in the study only received irradiation without addition- from glioblastoma HIF-1α or thyroids (VEGF) served as a
al chemotherapy. The histology of all tumors was squamous positive control. Nuclei were counterstained using Mayer’s
cell carcinoma. The study was approved by the local medical Hemalaun.

170 Strahlenther Onkol 2008 · No. 3 © Urban & Vogel

Dellas K, et al. Prognostic Impact of HIF-1α Expression

Evaluation of HIF-1α and VEGF Expression Table 1. Prognostic impact of stage and hypoxia parameters on 5-year
Staining was assessed in five high-power fields at × 400 magni- cancer-specific survival determined by Kaplan-Meier analysis. HF5: hy-
poxic fraction < 5 mmHg; HIF-1α: hypoxia-inducible factor-1α.
fication. Expression of HIF-1α and VEGF was determined by
semiquantitatively assessing the percentage of decorated tu- Tabelle 1. Prognostische Bedeutung des klinischen Stadiums und der
Hypoxieparameter in Bezug auf das tumorspezifische 5-Jahres-Über-
mor cells and the staining intensity.
leben (Kaplan-Meier-Analyse). HF5: hypoxische Fraktion < 5 mmHg;
The percentage of positive cells was rated as follows: HIF-1α: hypoxieinduzierbarer Faktor-1α.
1–10% positive cells (1+), 11–50% positive (2+), 51–80% pos-
itive (3+), and > 80% positive cells (4+). Staining intensity was Staging (FIGO) Patients Mean survival p-valuea
scored as no expression (–), weak (1+), moderate (2+), and (n) (4%)
intensive (3+). Points for expression and points of percentage IIB 9 76 ± 15
of positive cells were multiplied to an immunoreactive score IIIB 31 57 ± 10
(IRS) [32]. All slides were examined and scored independent- IVA 4 25 ± 22 0.03
ly by two investigators. For statistical survival analysis, the Hemoglobin
cutoff level for grouping immunohistochemical HIF-1α pro- • < 11 g/dl 10 10 ± 9
tein expression was defined as expression or no expression. • > 12.5 g/dl 34 75 ± 8 0.01
VEGF expression was determined in 39 biopsies. Weak ex- HIF-1α
pression was found in 26% of the samples (10/39 IRS 1–6) and • Expression (–) 12 92 ± 8
• Expression (+) 32 45 ± 10 0.02
strong expression in 74% (29/39 IRS 9–12). Tumor oxygenation
• HF5 > median (19%) 22 62 ± 11
Statistical Analysis • HF5 < median (19%) 22 55 ± 12 0.91
For statistical evaluation, a commercially available software Median pO2
package (SPSS, version 11.0) was used. Survival curves were • > median (15 mmHg) 22 57 ± 13
calculated according to the Kaplan-Meier analysis with log- • < median (15 mmHg) 22 58 ± 11 0.66
rank test and Cox regression model. Differences in the nu- a
p-values refer to log-rank tests
merical data between the two groups were evaluated using an
unpaired t-test or Pearson test for correlation. A p-value Figure 1, an immunohistochemical staining demonstrates a
< 0.05 was considered to be statistically significant. strong HIF-1α expression in tumor tissue.
Evaluating the relationship between HIF-1α expression
Results and tumor oxygenation, we did not find a correlation between
Cervical Cancer Patients and Prognostic Factors HIF-1α expression and the median pO2 (p = 0.57) and the hy-
The 44 patients underwent a definitive radiotherapy following poxic fraction HF5 (p = 0.57) in the 44 invasive squamous cell
a curative concept by combining both external-beam radio- carcinomas (data not shown). HIF-1α correlated with a lower
therapy and HDR brachytherapy. The mean age was 58.4
years (35–76 years), and the median follow-up at the time of
this analysis was 45 months (6–60 months) with an average
5-year cancer-specific survival of 60 ± 7%. The 5-year can-
cer-specific survival was 76% for stage IIB, 57% for stage
IIIB, and 25% for the advanced stage IVA. The prognostic
impact of clinical stage, anemia, HIF-1α, pO2, and HF5 on
5-year cancer-specific survival is shown in Table 1. Whereas
survival significantly correlated with stage (p = 0.03), anemia
(p < 0.01) and HIF-1α (p = 0.02), the hypoxia-related param-
eters pO2 (p = 0.66) or HF5 (p = 0.91) had no impact on sur-
vival according to the log-rank test (Table 1).

HIF-1α Expression and Correlation with Hypoxia-Related

For classification of HIF-1α expression of all 44 biopsies, only
nuclear staining was considered. The tumor tissue of 12/44 pa-
Figure 1. Immunohistochemical staining for HIF-1α expression in ad-
tients (27%) showed no expression of HIF-1α. The distribu- vanced cervical carcinoma (400×). Representation of a tumor cell re-
tion of HIF-1α expression of the other 32/44 patients (73%) gion with strong HIF-1α expression (IRS 9).
was as follows: two tumors with IRS of 1, seven tumors with Abbildung 1. Immunhistochemische Färbung für HIF-1α-Expression
IRS of 2, four tumors with IRS of 3, seven tumors with IRS of bei fortgeschrittenem Zervixkarzinom (400×). Darstellung einer Tu-
4, ten tumors with IRS of 6, and two tumors with IRS of 9. In morzellregion mit starker HIF-1α-Expression (IRS 9).

Strahlenther Onkol 2008 · No. 3 © Urban & Vogel 171

Dellas K, et al. Prognostic Impact of HIF-1α Expression

7.00 1.0

6.00 HIF-1α (–)
95% CI HIF-1α expression


5.00 0.7


3.00 0.4
HIF-1α (+)
1.00 0.1
< 11 11–12.5 > 12.5 0.0
Hemoglobin (g/dl) 0 10 20 30 40 50 60
Time (months)
Figure 2. HIF-1α expression (95% confidence intervals) in correlation to
the pretreatment hemoglobin level (in g/dl) separated into three Figure 3. Kaplan-Meier analysis: overall survival depending on HIF-1α
groups with hemoglobin levels < 11, between 11 and 12.5, and > 12.5 g/dl. expression in cervical cancer. Patients (n = 32) with positive HIF-1α ex-
A significant difference in HIF-1α expression was found between non- pression in tumor tissue [HIF-1α (+)] had a 5-year survival of 45 ± 10%
anemic patients (Hb > 12.5 g/dl) and anemic patients (Hb < 11 g/dl; ( p = in comparison to HIF-1α-negative cancers [HIF-1α (–)] with 92 ± 8%
0.04): the numeric data were evaluated using the Pearson test for cor- (n = 12).
relation. Abbildung 3. Kaplan-Meier-Analyse: Abhängigkeit der Gesamtüberle-
Abbildung 2. HIF-1α-Expression (95%-Konfidenzintervalle) in Korrelati- bensrate von der HIF-1α-Expression beim Zervixkarzinom. 32 Patien-
on zum prätherapeutischen Hämoglobinwert (in g/dl), der in drei Grup- tinnen mit positiver HIF-1α-Expression im Tumorgewebe [HIF-1α (+)]
pen mit einem Hämoglobin < 11, zwischen 11 und 12,5 und > 12,5 g/dl ein- wiesen ein 5-Jahres-Überleben von 45 ± 10% im Vergleich zur Überle-
geteilt wurde. Ein signifikanter Unterschied der HIF-1α-Expression wur- bensrate von 92 ± 8% bei zwölf Patientinnen mit HIF-1α-negativen
de bei nicht-anämischen Patientinnen (Hb > 12,5 g/dl) und anämischen Karzinomen [HIF-1α (–)] auf.
Patientinnen (Hb < 11 g/dl) gefunden (p = 0,04; Pearson-Test).

level of expression in tumors from patients who presented with The Kaplan-Meier analysis revealed a significant correla-
Hb levels > 12.5 g/dl (IRS 2.1 ± 0.4) compared to patients with tion between HIF-1α expression and survival (p = 0.02; Table 1,
Hb levels < 11 g/dl (IRS 3.9 ± 0.9; p = 0.04; see Figure 2). Fur- Figure 3). The 5-year-survival was 92 ± 8% in patients without
thermore, there was a significant correlation between HIF-1α HIF-1α expression compared to only 45 ± 10% in patients with
and VEGF expression (p = 0.002). detectable expression of this marker. The risk of tumor-related
death for patients with HIF-1α expression was 8.0 (p < 0.05),
HIF-1α Expression and Correlation with Clinical 7.0 (p < 0.01) in patients with anemia (patients with Hb levels
Parameters and Survival Analyses < 11 g/dl), and 1.9 (p = 0.4) in patients with high VEGF expres-
Concerning the expression intensity of HIF-1α, no correlation sion, as determined by the univariate Cox regression model. In
with clinical parameters or outcome was observed. We chose a multivariate Cox regression model adjusted to tumor stage,
two groups, group 1 with twelve tumors (IRS 0) and group 2 HIF-1α expression (relative risk [RR] = 7.5; p = 0.05) was asso-
with 32 tumors (IRS 1–9), to analyze the impact of HIF-1α ciated with an increased risk of tumor-related death.
HIF-1α expression did not correlate with tumor stage Discussion
(p = 0.86) or grade of histological differentiation (p = 0.65). Tumor hypoxia is a well-known negative prognostic factor in
However, elevated HIF-1α expression was associated with various tumors including cervical cancer [12, 15, 17, 30, 33, 37,
local tumor progression and an increased rate of metastasis. 38]. However, the role of endogenous hypoxic markers such as
Two of twelve patients (17%) with HIF-1α-negative tumors HIF-1α and its impact as a prognostic marker in patients with
developed progressive disease (one local relapse and one advanced cervical carcinoma has not been comprehensively
distant failure). On the other hand, 19 of 32 patients (60%) investigated.
with HIF-1α-positive tumors suffered from ten local re- In our study of 44 patients, we found no correlation be-
lapses (p = 0.12), and nine patients presented with sustained tween HIF-1α expression and tumor oxygenation parameters
distant metastasis during the follow-up time of 5 years (p = such as HF5 and pO2 level. Similarly, Mayer et al. reported
0.17). that there was no correlation of HIF-1α expression with me-

172 Strahlenther Onkol 2008 · No. 3 © Urban & Vogel

Dellas K, et al. Prognostic Impact of HIF-1α Expression

dian oxygen tension and hypoxic fractions in cervical cancers of 100% [1, 7]. This suggests that the combination of different
[30]. None of the different hypoxia markers including HIF-1α endogenous hypoxia and tumor markers might be a useful ap-
and VEGF expression correlated with Eppendorf oxygen proach to identify prognostically distinct subgroups of patients
electrode measurement in patients with cervical cancer [21, with cervical cancers.
34]. Hutchison et al. found an association between the level of
HIF-1α staining and median pO2 in cervical cancers, but there Conclusion
was a lack of correlation between HIF-1α and HF5 [19]. A Our study indicates, that endogenous tumor markers such as
study of cervical cancer stages FIGO IA–III by Haugland et HIF-1α may serve as prognostic markers of clinical outcome
al. showed strong positivity for 42 tumor specimens and a sig- concerning cervical cancer after primary radiotherapy.
nificant dependence of the HIF-1α expression on the pretreat-
ment tumor oxygen tension HF5 (p < 0.01) using the Eppen- Acknowledgments
dorf device and a computerized image analysis software to We would like to thank Prof. Dr. D.M. Katschinski from the Univer-
score the slides, which determined the HIF-1α-positive tumor sity of Goettingen, Germany, for helpful discussions. We thank our
areas in conjunction with the whole tissue sample [16]. This colleagues from the Department of Radiotherapy and the Institute of
Pathology for contributing to this study and for their continuous sup-
suggests that HIF-1α expression and oxygenation parameters
port. This work was supported by the Deutsche Krebshilfe (grant
analyzed by electrode measurement are determined by two number: 106764/106797) and Wilhelm Roux Program of BMBF/NBL3
different approaches only partially comparable since they de- (FKZ: 14/27).
scribe different oxygen environments.
In our study, elevated HIF-1α expression was associated
with local tumor progression and increased rate of metastasis. References
1. Bache M, Holzapfel D, Kappler M, et al. Survivin protein expression and
This was not significant, although our finding is supported by hypoxia in advanced cervical carcinoma of patients treated by radiotherapy.
the result that HIF-1α expression correlated with disease re- Gynecol Oncol 2007;104:139–44.
currence and distant metastasis but not with local tumor con- 2. Bache M, Reddemann R, Said HM, et al. Immunohistochemical detection of
trol in stage IIIB cervical squamous cell carcinoma after radia- osteopontin in advanced head-and-neck cancer: prognostic role and cor-
relation with oxygen electrode measurements, hypoxia-inducible-factor-
tion therapy [20]. Our data revealed a strong correlation 1alpha-related markers, and hemoglobin levels. Int J Radiat Oncol Biol
between HIF-1α expression and the clinical outcome in lo- Phys 2006;66:1481–7.
cally advanced cervical carcinoma after primary radiotherapy 3. Bachtiary B, Schindl M, Potter R, et al. Overexpression of hypoxia-inducible
factor 1alpha indicates diminished response to radiotherapy and unfavor-
(p < 0.02; Figure 3). Furthermore, immunohistochemical able prognosis in patients receiving radical radiotherapy for cervical cancer.
HIF-1α expression is associated with a 7.5-fold increased risk Clin Cancer Res 2003;9:2234–40.
of tumor-related death independent of the tumor stage (p = 4. Bernaudin M, Tang Y, Reilly M, et al. Brain genomic response following hy-
0.05). In agreement with this, most studies of cervical cancers poxia and re-oxygenation in the neonatal rat. Identification of genes that
might contribute to hypoxia-induced ischemic tolerance. J Biol Chem
detected HIF-1α expression as a negative prognostic factor. 2002;277:39728–38.
Concerning patients with strong HIF-1α expression, Birner et 5. Birner P, Schindl M, Obermair A, et al. Overexpression of hypoxia-inducible
al. demonstrated a significantly shorter overall survival time factor 1alpha is a marker for an unfavorable prognosis in early-stage inva-
sive cervical cancer. Cancer Res 2000;60:4693–6.
(p = 0.03) and even a disease-free survival time (p < 0.0001) as 6. Burri P, Djonov V, Aebersold DM, et al. Significant correlation of hypoxia-in-
compared to those with moderate to absent HIF-1α expres- ducible factor-1alpha with treatment outcome in cervical cancer treated
sion [5]. Different studies detected HIF-1α as a negative prog- with radical radiotherapy. Int J Radiat Oncol Biol Phys 2003;56:494–501.
nostic factor in cervical cancer treated with radical irradiation 7. Capalbo G, Rodel C, Stauber RH, et al. The role of survivin for radiation
therapy. Prognostic and predictive factor and therapeutic target. Strahlen-
[3, 6] or surgery [10]. In the study of Hutchison at al., high ther Onkol 2007;183:593–9.
HIF-1α expression tended to associate with poor outcome 8. Dunst J, Hansgen G, Krause U, et al. A 2-week pretreatment with 13-cis-ret-
only in small cervical tumors treated with irradiation [19]. inoic acid + interferon-alpha-2a prior to definitive radiation improves tu-
mor tissue oxygenation in cervical cancers. Strahlenther Onkol 1998;174:
However, in two studies of cervical cancer patients receiving 571–4.
different treatments including radiotherapy, HIF-1α expres- 9. Dunst J, Kuhnt T, Strauss HG, et al. Anemia in cervical cancers: impact on
sion showed no impact on survival [16, 30]. survival, patterns of relapse, and association with hypoxia and angiogene-
sis. Int J Radiat Oncol Biol Phys 2003;56:778–87.
Our study shows a correlation between HIF-1α expres-
10. Fujimoto J, Alam SM, Jahan I, et al. Plausible linkage of hypoxia inducible
sion and Hb level (p = 0.04) or VEGF expression (p = 0.002). factor-1alpha in uterine cervical cancer. Cancer Sci 2006;97:861–7.
The association of anemia and tumor hypoxia has been dem- 11. Fyles AW, Milosevic M, Pintilie M, et al. Cervix cancer oxygenation measured
onstrated in animal studies, because anemic animals showed a following external radiation therapy. Int J Radiat Oncol Biol Phys 1998;42:
significant correlation to poorly oxygenated tumors [23, 24]. 12. Fyles AW, Milosevic M, Wong R, et al. Oxygenation predicts radiation re-
In this study, the risk of tumor-related death was 7.0 (p < 0.01) sponse and survival in patients with cervix cancer. Radiother Oncol 1998;
in patients with anemia and 1.9 (p = 0.4) in patients with high 48:149–56.
13. Gagel B, Reinartz P, Dimartino E, et al. pO2 polarography versus positron
VEGF expression. Furthermore, patients lacking expression
emission tomography ([18F] fluoromisonidazole, [18F]-2-fluoro-2'-deoxyglu-
of HIF- 1α, or the inhibitor of apoptosis survivin, and without cose). An appraisal of radiotherapeutically relevant hypoxia. Strahlenther
anemia have the best prognosis with a 5-year overall survival Onkol 2004;180:616–22.

Strahlenther Onkol 2008 · No. 3 © Urban & Vogel 173

Dellas K, et al. Prognostic Impact of HIF-1α Expression

14. Haensgen G, Krause U, Becker A, et al. Tumor hypoxia, p53, and prognosis 29. Mayer A, Hockel M, Vaupel P. Endogenous hypoxia markers in locally ad-
in cervical cancers. Int J Radiat Oncol Biol Phys 2001;50:865–72. vanced cancers of the uterine cervix: reality or wishful thinking? Strahlen-
15. Harris AL. Hypoxia – a key regulatory factor in tumour growth. Nat Rev ther Onkol 2006;182:501–10.
Cancer 2002;2:38–47. 30. Mayer A, Wree A, Hockel M, et al. Lack of correlation between expression of
16. Haugland HK, Vukovic V, Pintilie M, et al. Expression of hypoxia-inducible HIF-1alpha protein and oxygenation status in identical tissue areas of squa-
factor-1alpha in cervical carcinomas: correlation with tumor oxygenation. mous cell carcinomas of the uterine cervix. Cancer Res 2004;64:5876–81.
Int J Radiat Oncol Biol Phys 2002;53:854–61. 31. Nordsmark M, Loncaster J, Aquino-Parsons C, et al. The prognostic value of
17. Hockel M, Schlenger K, Aral B, et al. Association between tumor hypoxia pimonidazole and tumour pO2 in human cervix carcinomas after radiation
and malignant progression in advanced cancer of the uterine cervix. Cancer therapy: a prospective international multi-center study. Radiother Oncol
Res 1996;56:4509–15. 2006;80:123–31.
18. Hockel M, Schlenger K, Hockel S, et al. Hypoxic cervical cancers with low 32. Remmele W, Stegner HE. [Recommendation for uniform definition of an
apoptotic index are highly aggressive. Cancer Res 1999;59:4525–8. immunoreactive score (IRS) for immunohistochemical estrogen receptor
19. Hutchison GJ, Valentine HR, Loncaster JA, et al. Hypoxia-inducible factor detection (ER-ICA) in breast cancer tissue.] Pathologe 1987;8:138–40.
1alpha expression as an intrinsic marker of hypoxia: correlation with tumor 33. Rohrer BC, Ohlerth S, Roos M, et al. Influence of pretreatment polaro-
oxygen, pimonidazole measurements, and outcome in locally advanced car- graphically measured oxygenation levels in spontaneous canine tumors
cinoma of the cervix. Clin Cancer Res 2004;10:8405–12. treated with radiation therapy. Strahlenther Onkol 2006;182:518–24.
20. Ishikawa H, Sakurai H, Hasegawa M, et al. Expression of hypoxic-inducible 34. Sakata K, Someya M, Nagakura H, et al. A clinical study of hypoxia using
factor 1alpha predicts metastasis-free survival after radiation therapy alone endogenous hypoxic markers and polarographic oxygen electrodes.
in stage IIIB cervical squamous cell carcinoma. Int J Radiat Oncol Biol Phys Strahlenther Onkol 2006;182:511–7.
2004;60:513–21. 35. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer 2003;
21. Jankovic B, Aquino-Parsons C, Raleigh JA, et al. Comparison between pimo- 3:721–32.
nidazole binding, oxygen electrode measurements, and expression of en- 36. Sundfor K, Lyng H, Trope CG, et al. Treatment outcome in advanced squa-
dogenous hypoxia markers in cancer of the uterine cervix. Cytometry B Clin mous cell carcinoma of the uterine cervix: relationships to pretreatment
Cytom 2006;70:45–55. tumor oxygenation and vascularization. Radiother Oncol 2000;54:101–7.
22. Kallinowski F, Zander R, Hoeckel M, et al. Tumor tissue oxygenation as 37. Vaupel P, Mayer A, Hockel M. Impact of hemoglobin levels on tumor oxygen-
evaluated by computerized-pO2-histography. Int J Radiat Oncol Biol Phys ation: the higher, the better? Strahlenther Onkol 2006;182:63–71.
1990;19:953–61. 38. Vordermark D, Brown JM. Endogenous markers of tumor hypoxia predictors
23. Kelleher DK, Matthiensen U, Thews O, et al. Tumor oxygenation in anemic of clinical radiation resistance? Strahlenther Onkol 2003;179:801–11.
rats: effects of erythropoietin treatment versus red blood cell transfusion. 39. Weitmann HD, Knocke TH, Czembirek C. Prognostische Faktoren für die
Acta Oncol 1995;34:379–84. primäre Radiotherapie des Zervixkarzinoms. Strahlenther Onkol 2005;
24. Kelleher DK, Matthiensen U, Thews O, et al. Blood flow, oxygenation, and 181:17.
bioenergetic status of tumors after erythropoietin treatment in normal and
anemic rats. Cancer Res 1996;56:4728–34.
25. Knocke TH, Weitmann HD, Feldmann HJ, et al. Intratumoral pO2-measure- Address for Correspondence
ments as predictive assay in the treatment of carcinoma of the uterine PD Dr. Gabriele Hänsgen
cervix. Radiother Oncol 1999;53:99–104.
Klinik für Strahlentherapie
26. Koshikawa N, Iyozumi A, Gassmann M, et al. Constitutive upregulation of
hypoxia-inducible factor-1alpha mRNA occurring in highly metastatic lung Martin-Luther-Universität Halle-Wittenberg
carcinoma cells leads to vascular endothelial growth factor overexpression Dryanderstraße 4–7
upon hypoxic exposure. Oncogene 2003;22:6717–24. 06097 Halle
27. Loncaster JA, Harris AL, Davidson SE, et al. Carbonic anhydrase (CA IX) Germany
expression, a potential new intrinsic marker of hypoxia: correlations with Phone (+49/345) 557-3436, Fax -4333
tumor oxygen measurements and prognosis in locally advanced carcinoma
e-mail: gabriele.haensgen@medizin.uni-halle.de
of the cervix. Cancer Res 2001;61:6394–9.
28. Marti HH. Angiogenesis – a self-adapting principle in hypoxia. EXS 2005;

174 Strahlenther Onkol 2008 · No. 3 © Urban & Vogel