Professional Documents
Culture Documents
1471–1478, 2012
Copyright Ó 2012 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/$ - see front matter
doi:10.1016/j.ijrobp.2011.05.009
JIN GAO, M.D.,*yz YA-LAN TAO, M.D.,*y GUO LI, M.D.,*y WEI YI, M.D., PH.D.,*y
y
AND YUN-FEI XIA, M.D.*
*Department of Radiation Oncology, Cancer Center, yState Key Laboratory of Oncology in Southern China, Sun Yat-Sen University,
Guangzhou, People’s Republic of China; and zDepartment of Radiation Oncology, Anhui provincial hospital, Hefei,
People’s Republic of China
Purpose: To investigate the effect of hemoglobin (Hb) concentration and the difference in its decrease during
treatment on outcome of radiotherapy (RT) alone for patients with Stage I and II nasopharyngeal carcinoma.
Methods and Materials: A total of 572 patients with Stage I-II nasopharyngeal carcinoma with RT alone between
January 2001 and December 2004 were retrospectively analyzed. Patient characteristics, tumor variables, and Hb
level, including pre-RT Hb, mid-RT Hb, and dynamic change of Hb between pre- and post- RT and its difference in
decrease (6Hb) were subjected to univariate and multivariable analysis to identify factors that predict disease-
specific survival (DSS), local regional recurrence-free survival (LRFS), and metastases-free survival (MFS).
Results: The 5-year DSS was poorer in the Hb continuous decrease group than in the Hb noncontinuous decrease
group (84% vs. 89%; p = 0.008). There was poorer 5-year DSS in patients with 6Hb of >11.5 g/L than in those with
6Hb of #11.5 g/L (82% vs. 89%; p = 0.001), and poorer LRFS (79% vs. 83%; p = 0.035). Univariate and multi-
variate analysis showed that Hb decrease difference with greater than 11.5 g/L was an independent prognostic
factor for DSS and LRFS.
Conclusions: The difference in decrease of Hb level during the course of radiation treatment appeared as a poor
prognostic factor in Stage I and II nasopharyngeal carcinoma patients. Ó 2012 Elsevier Inc.
Reprint requests to: Yun-Fei Xia, M.D., 651 Dongfeng Road Conflict of interest: none.
East, Guangzhou 510060, People’s Republic of China. Tel: 86- Acknowledgments—We would like to thank Mr. Charles Chen for
20-87343169; Fax: 86-20-87343392; E-mail: xiayf@sysucc.org.cn critically reading and editing this revised manuscript.
Supported by grants from the Hi-Tech Research and Develop- Received Oct 15, 2010, and in revised form April 24, 2011.
ment Program of China (no. 2006AA02Z4B4). Accepted for publication May 10, 2011.
1471
1472 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 4, 2012
Stage I and II NPC treated with RT alone to study the effect Hb level was defined as the average of all Hb values during RT.
of Hb level on the treatment outcome. Up to now, in patients Post-RT Hb level was defined as the Hb value or average Hb value
with Stage I and II NPC treated with RT alone, no published at the last week of RT. Hb continuous decrease was defined as pre-
studies had addressed the prognostic value of Hb level, espe- RT Hb > mid-RT Hb > post-RT Hb. All situations other than Hb
continuous decrease were considered Hb noncontinuous decrease.
cially for the difference in its decrease. The aim of our study
Individual difference value (6Hb) of Hb level from pre-RT to
is to test the hypothesis of whether the difference in the de- post-RT was equal to (pre-RT Hb value) – (post-RT Hb value).
crease of Hb level could be a prognostic factor for patients Figure 1 shows the patients were grouped according to their
with Stage I and II NPC. different Hb levels.
Pre-RT Hb levels
Anemia 86 73 93
No anemia 87 0.51 82 0.16 94 0.65
Mid-RT Hb levels
Anemia 85 74 92
No anemia 87 0.23 82 0.099 93 0.93
Hb dynamic change
Continuous decrease 84 78 92
Noncontinuous decrease 89 0.008 82 0.062 94 0.55
Pre-RT anemia
Continuous decrease 71 56 92
Noncontinuous decrease 93 0.019 80 0.063 93 0.90
Pre-RT no anemia
Continuous decrease 85 78 92
Noncontinuous decrease 89 0.032 82 0.12 94 0.49
6Hb value
6Hb # 11.5 g/L 89 83 94
6Hb > 11.5 g/L 82 0.001 79 0.035 93 0.67
Pre-RT anemia
6Hb # 11.5 g/L 90 76 93
6Hb > 11.5 g/L 70 0.16 56 0.33 92 0.38
Pre-RT no anemia
6Hb # 11.5g/l 89 85 94
6Hb > 11.5 g/L 83 0.002 76 0.028 92 0.56
Continuous decrease
6Hb # 11.5 g/L 89 82 95
6Hb > 11.5 g/L 81 0.036 74 0.065 94 0.53
Noncontinuous decrease
6Hb # 11.5 g/L 90 82 93
6Hb > 11.5 g/L 87 0.46 81 0.83 90 0.59
Abbreviations: DSS = disease-specific survival; Hb = hemoglobin; 6Hb = individual differences between hemoglobin concentrations at
the beginning and end of radiotherapy; LRFS = local regional recurrence-free survival; MFS = metastases-free survival; RT = radiotherapy.
1474 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 4, 2012
Fig 2. Kaplan-Meier curves showing disease-specific survival (DSS) in continuously decreased and noncontinuously
decreased group. (a) The 5-year DSS rates in hemoglobin (Hb) continuously decreased group and noncontinuously
decreased group were 84% and 89% (p = 0.008). (b) The 5-year Local regional recurrence-free survival (LRFS) rates
in Hb continuously decreased group and noncontinuously decreased group were 78% and 82% (p = 0.062). (c) The 5-
year DSS rates in pre-radiation therapy (RT) no anemia patients in the Hb continuously decreased and noncontinuously
decreased group were 85% and 89% (p = 0.032). (d) The 5-year LRFS rates in pre-RT no anemia patients in the Hb
continuously decreased and noncontinuously decreased group were 78% and 82% (p = 0.12).
Fig 3. Kaplan-Meier curves showing disease-specific survival (DSS) and local regional recurrence-free survival (LRFS)
in 6 hemoglobin (Hb) # 11.5 g/L and 6Hb > 11.5 g/L group. (a) The 5-year DSS rates in the 6Hb # 11.5 g/L group and
the 6Hb > 11.5 g/L group were 89% and 82% (p = 0.001). (b) The 5-year LRFS rates in the 6Hb # 11.5 g/L group and
the 6Hb > 11.5 g/L group were 83% and 79% (p = 0.035). (c) The 5-year DSS rates in pre-RT no anemia patients with
6Hb # 11.5 g/L and those with 6Hb > 11.5 g/L were 89% and 83% (p = 0.002). (d) The 5-year LRFS rates in pre-RT no
anemia patients with 6Hb # 11.5 g/L and those with 6Hb > 11.5 g/L were 85% and 76% (p = 0.028). (e) The 5-year DSS
rates in the Hb continuously decreased patients with 6Hb # 11.5 g/L and those with 6Hb > 11.5 g/L were 89% and 81%
(p = 0.036). (f) The 5-year LRFS rates in the Hb continuously decreased patients with 6Hb # 11.5 g/L and those with
6Hb > 11.5 g/L were 82% and 74% (p = 0.065).
1476 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 4, 2012
Table 3. Factors associated with outcome by univariate and multivariate analysis in 572 patients
Multivariate
Univariate
Outcomes and variable p value RR 95%CI p value
Abbreviations: RR = relative risk; CI = confidence interval; 6Hb = individual differences between pre-RT hemoglobin (Hb) and post-RT
Hb; RT = radiotherapy.
mid-RT anemia vs. mid-RT no anemia, continuous decrease The 5-year DSS rates in the Hb continuously decreased
vs. noncontinuous decrease, and 6Hb # 11.5 g/L vs. 6Hb patients with 6Hb # 11.5 g/L and those with 6Hb >
> 11.5 g/L. 11.5 g/L were 89% and 81% (p = 0.036, Fig. 3). The 5-
No significant difference in DSS, LRFS, and MFS rate year LRFS and MFS rates of the two groups didn’t show sig-
was found between anemia group and nonanemia group nificant difference (p = 0.065, p = 0.53).
when patients were stratified according to pre-RT and mid- In patients with Hb non-continuous decrease, no signifi-
RT Hb levels, respectively (Table 2). cant differences in DSS, LRFS, and DMF rates were found
Patients with Hb continuous decrease had significantly between the 6Hb # 11.5 g/L and the 6Hb > 11.5 g/L
poorer DSS than those with non-continuous decrease (84% groups.
vs. 89% for 5-year DSS rate, p = 0.008) (Table 2, Fig. 2),
but their LRFS and MFS didn’t show significant difference. Univariate and multivariate analyses
The subgroups were further analyzed. The 5-year DSS Table 3 summarizes the univariate analysis of other rele-
rates in pre-RT anemic patients in the Hb continuously de- vant prognostic factors. T status, clinical stage, the different
creased and noncontinuously decreased groups were71% combining T and N subgroup (T2N1 vs. non-T2N1), and
and 93% (p = 0.019, Table 2), respectively. The 5-year 6Hb status were significant factors that predicted local re-
DSS rates in pre-RT no anemia patients in the Hb continu- gional recurrence. Age status, 6Hb status, T status, N sta-
ously decreased and noncontinuously decreased group tus, clinical stage, and the different combining T and N
were 85% and 89% (p = 0.032; Table 2, Fig. 2). subgroup (T2N1 vs. non-T2N1) were significant factors
that predicted death. N status and the different combining
Treatment outcomes according to 6Hb T and N subgroup (T2N1 vs. non-T2N1) were significant
The patients with 6Hb # 11.5 g/L had significantly bet- predictors of distant metastasis.
ter DSS (p = 0.001) and LRFS (p = 0.035) compared with The following parameters were included in multivariate
those with 6Hb > 11.5 g/L (Table 2, Fig. 3). The 5-year analysis using a Cox proportional hazards model: gender,
MFS rates in the 6Hb # 11.5 g/L group and the 6Hb > age (#47 years vs. >47 years), T status (T1 vs. T2), clinical
11.5 g/L group were 94% and 93% (p = 0.67), respectively. stage (Stage I vs. II), the different combining T and N sub-
On subgroups analyses, the 5-year DSS, LRFS, and MFS group (T2N1 vs. non T2N1), pre-RT Hb (anemia vs. nonane-
rates did not show significant difference between pre-RT mia), mid-RT Hb (anemia vs. nonanemia), Hb dynamic
anemic patients with 6Hb # 11.5 g/L and those with change (continuous decrease vs. noncontinuous decrease),
6Hb > 11.5 g/L. In contrast, when these comparisons and 6Hb (#11.5 g/L vs. >11.5 g/L). Results were summa-
were made between pre-RT no anemic patients with 6Hb rized in Table 3. 6Hb was found to be independently predic-
# 11.5 g/L and those with 6Hb > 11.5 g/L, significant dif- tive of both DSS and LRFS rates.
ferences were found. The 5-year DSS rates in pre-RT no ane-
mic patients with 6Hb # 11.5 g/L and those with 6Hb >
DISCUSSION
11.5 g/L were 89% and 83% (p = 0.002) (Table 2, Fig. 3).
The 5-year LRFS rates were 85% and 76% (p = 0.028; In this study, we evaluated the relationship between Hb
Table 2, Fig. 3), respectively. concentration, particularly for its difference in decrease,
Hemoglobin and outcome of radiotherapy for nasopharyngeal carcinoma d J. GAO et al. 1477
and prognosis in patients with Stage I and II NPC treated 73%, p = 0.01), disease-free survival (26% vs. 73%, p =
with RT alone. We found that there was no apparent effect 0.005), and survival (35% vs. 83%, p = 0.02). Our study
of pre-RT or mid-RT anemia on treatment outcomes in the supports this explanation through the fact that the decrease
univariate and multivariable analyses. But the decrease in Hb concentration was significantly associated with
6Hb with >11.5 g/L for Stage I and II NPC appeared signif- LRFS.
icantly and independently related to DSS and LRFS in the A limitation of this study was the lack of precise data for
analyzed data set. To our knowledge, this is the first study Karnofsky performance status of patients before and after
to address the prognostic value of Hb concentration in RT, although our patients with Stage I and II NPC were in
patients with Stage I and II NPC treated with RT alone. relatively good general condition. However, the side effect
However, this prognostic significance of decrease 6Hb of RT would impair the general condition of the patients.
was also found in the laryngeal carcinoma reported by Rut- The inferior Karnofsky performance status of patient would
kowski et al. (11). They found that individual change in Hb adversely affect treatment outcome.
concentration during the course of RT appeared as an inde- Because decrease 6Hb might be an independent prog-
pendent prognostic factor for LRFS in the laryngeal cancer nostic factor affecting treatment outcomes, in order to
patients, the patients with decrease 6Hb with greater than correct the anemia and prevent the decline of Hb in the treat-
0 had significantly poorer 5-year LRFS compared with other ment of tumor, active treatment should be considered for
patients (72% vs. 80%, p = 0.01). The reason may be that patients with decreased Hb. Correction of anemia can be
continuous decreased Hb during radiotherapy could lead to achieved with blood transfusion and erythropoietin.
the reduced oxygen carrier capacity in blood and could indi- Although blood transfusion has proven effective in raising
rectly cause tumor tissue hypoxia that can result in radiore- Hb levels and relieving clinical symptoms of anemia, the ad-
sistance, consequently leading to poorer local control (7). verse effect of blood transfusion on prognosis was still re-
However, the result was related to the extent of difference ported. The studies of Madjdpour (17) showed negative
in Hb decrease. Our study showed 5-year LRFS was lower effects of correcting anemia with blood transfusion, such
in the group with DHb of >11.5 g/L, supporting this hypoth- as transfusion transmissible infections, immunological risks,
esis. In addition, in our study on body mass index and NPC and mistransfusion. Therefore, in recent years, erythropoie-
prognostics, we found underweight (poor nutritional status) tin is extensively used clinically. Quirt (18) found epoetin
patients had a significant decrease in the overall survival alfa administered to patients with cancer-related anemia re-
rate (12). So, we surmise that when difference in Hb de- sulted in significantly improved quality of life, increased Hb
crease reaches certain value, it is associated with malnutri- levels, and decreased blood transfusion use. However, the
tion. Malnutrition could increase toxicity and decrease use of erythropoiesis-stimulating agents in patients with
response to treatment. This notion was also supported by head-and-neck cancer has been controversial. Henke (19)
the study results reported by Nitenberg et al. (13) and Isenr- did a multicenter, double-blind, randomized, placebo-
ing et al. (14). The patients with pre-RT anemia (11/572) controlled trial in 351 patients with carcinoma of the oral
and mid-RT anemia (51/572) in this study seldom showed cavity, oropharynx, hypopharynx, or larynx, 82% of patients
continuous decrease DHb of >11.5 g/L, so continuous de- given epoetin beta achieved Hb concentrations higher than
crease in Hb is significant but pre-RT Hb and mid-RT Hb 140 g/L (women) or 150 g/L (men) compared with 15% of
are not. Nordsmark (15) studied the relationship between those given placebo. However, locoregional progression-
pretreatment tumor oxygenation status and prognosis of ad- free survival was poorer with epoetin beta than with placebo.
vanced head-and-neck squamous cell carcinoma. In their Therefore, how to correct anemia in the NPC patients awaits
studies, the majority of tumors were classified as hypoxic further study.
with a 45% locoregional tumor control probability as com-
pared with 90% among well-oxygenated tumors (p = 0.04).
CONCLUSION
Tumor hypoxia not only impairs locoregional tumor control
probability but also impairs overall survival probability. Bri- In our study on the relationship between the Hb level and
zel (16) reported that among 63 patients with head and neck prognosis of patients with Stage I and II NPC, we concluded
cancer receiving RT, median pO2 was 3.8 mm Hg in the that decrease in Hb level, especially for the difference in de-
residual tumor patients and 12.2 mm Hg in the no residual crease of Hb concentration (6Hb) during the course of RT
disease patients (p = 0.002). Tumor median pO2 of <10 mm appears as an independent prognostic factor for DSS and
Hg adversely affected 2-year locoregional control (30% vs. LRFS.
REFERENCES
1. Evans SM, Koch CJ. Prognostic significance of tumor oxygen- 3. Pehlivan B, Zouhair A, Luthi F, et al. Decrease in
ation in humans. Cancer Lett 2003;195:1–16. hemoglobin levels following surgery influences the outcome
2. Vaupel P, Mayer A, Hockel M. Impact of hemoglobin levels on in head and neck cancer patients treated with accelerated
tumor oxygenation: The higher, the better? Strahlenther Onkol postoperative radiotherapy. Ann Surg Oncol 2009;16:1331–
2006;182:63–71. 1336.
1478 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 4, 2012
4. Rades D, Stoehr M, Kazic N, et al. Locally advanced stage IV 12. Hu JY, Yi W, Xia YF, et al. Impact of pretherapy body mass in-
squamous cell carcinoma of the head and neck: Impact of pre- dex on prognosis of nasopharyngeal carcinoma. Chin J Cancer
radiotherapy hemoglobin level and interruptions during 2009;28:1043–1048.
radiotherapy. Int J Radiat Oncol Biol Phys 2008;70:1108–1114. 13. Nitenberg G, Raynard B. Nutritional support of the cancer pa-
5. Fortin A, Wang CS, Vigneault E. Effect of pretreatment anemia tient: Issues and dilemmas. Crit Rev Oncol Hematol 2000;34:
on treatment outcome of concurrent radiochemotherapy in pa- 137–168.
tients with head and neck cancer. Int J Radiat Oncol Biol Phys 14. Isenring EA, Capra S, Bauer JD. Nutrition intervention is ben-
2008;72:255–260. eficial in oncology outpatients receiving radiotherapy to the
6. Schafer U, Micke O, Muller SB, et al. Hemoglobin as an inde- gastrointestinal or head and neck area. Br J Cancer 2004;91:
pendent prognostic factor in the radiotherapy of head and neck 447–452.
tumors. Strahlenther Onkol 2003;179:527–534. 15. Nordsmark M, Overgaard J. A confirmatory prognostic study
7. Chua DT, Sham JS, Choy DT. Prognostic impact of hemoglobin on oxygenation status and loco-regional control in advanced
levels on treatment outcome in patients with nasopharyngeal head and neck squamous cell carcinoma treated by radiation
carcinoma treated with sequential chemoradiotherapy or radio- therapy. Radiother Oncol 2000;57:39–43.
therapy alone. Cancer 2004;101:307–316. 16. Brizel DM, Dodge RK, Clough RW, et al. Oxygenation of head
8. Gao J, Hu JY, Xia YF, et al. Continuous fall in hemoglobin level and neck cancer: Changes during radiotherapy and impact on
is a poor prognostic factor in patients with nasopharyngeal treatment outcome. Radiother Oncol 1999;53:113–117.
carcinoma treated with radiotherapy. Chin J Cancer 2010;29: 17. Madjdpour C, Spahn DR. Allogeneic red blood cell transfu-
561–566. sions: Efficacy, risks, alternatives and indications. Br J Anaesth
9. Greene FL, Page DL, Fleming ID, et al. AJCC cancer staging 2005;95:33–42.
handbook: From the AJCC cancer staging manual. 6th ed.New 18. Quirt I, Robeson C, Lau CY, et al. Epoetin alfa therapy in-
York: Springer; 2002 creases hemoglobin levels and improves quality of life in pa-
10. Greiner M, Pfeiffer D, Smith RD. Principles and practical ap- tients with cancer-related anemia who are not receiving
plication of the receiver-operating characteristic analysis for chemotherapy and patients with anemia who are receiving che-
diagnostic tests. Prev Vet Med 2000;45:23–41. motherapy. J Clin Oncol 2001;19:4126–4134.
11. Rutkowski T, Suwinski R, Idasiak A. The prognostic value of 19. Henke M, Laszig R, Rube C, et al. Erythropoietin to treat head
hemoglobin concentration in postoperative radiotherapy of and neck cancer patients with anaemia undergoing radiother-
835 patients with laryngeal cancer. Int J Radiat Oncol Biol apy: Randomised, double-blind, placebo-controlled trial. Lan-
Phys 2007;69:1018–1023. cet 2003;362:1255–1260.