Radiotherapy-Associated Anemia:

The Scope of the Problem

LOUIS B. HARRISON, DANIEL SHASHA, CAROL WHITE, BEVERLY RAMDEEN
Department of Radiation Oncology, Beth Israel Medical Center, St. Luke’s-Roosevelt Hospital Center,
New York, New York, USA

Key Words. Radiotherapy · Anemia · Quality of life · Tumor hypoxia · Locoregional control

A BSTRACT
The impact of anemia on cancer patients undergo- seem to be associated with tumor hypoxia and poorer
ing chemotherapy is well established, but only recently outcomes of radiotherapy in a number of patient popu-

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

has the prevalence of anemia in patients receiving radio- lations, ignoring even modest anemia can result in
therapy received much attention. Many cancer patients decreased locoregional control, overall survival, and
present with anemia prior to radiotherapy, and even quality of life (QOL). Because increasing hemoglobin
more experience anemia or a worsening of anemia at levels 1-2 g/dl is usually easily accomplished, there exists
some point during treatment. However, the problem of the potential for improving outcomes by paying greater
anemia is often ignored because patients may experi- attention to this problem. This article focuses on the
ence only functional anemia, defined as a hemoglobin prevalence of anemia, particularly functional anemia,
level less than 12 g/dl. Unless physiologic anemia (hemo- and discusses the impact of anemia on locoregional con-
globin = 8 g/dl) is discovered, efforts to correct anemia trol, overall survival, and QOL. The Oncologist
are often not made. Because hemoglobin levels <12 g/dl 2000;5(suppl 2): 1-7

INTRODUCTION RADIOTHERAPY-RELATED ANEMIA: THE BETH

Continuing advances in oncology care, including ISRAEL/ST. LUKE’S-ROOSEVELT EXPERIENCE
new radiotherapy technologies and recent developments
with multimodal therapies, have significantly improved Although anemia is a recognized cancer-related disor-
clinical outcomes and quality of life (QOL) for many der, recent studies have focused on its impact rather than its
patients with cancer. However, the prevalence and impact prevalence among patients undergoing radiotherapy [1, 3,
of cancer-associated anemia is not widely appreciated. 4]. The radiation oncology community, in general, has not
Emerging data demonstrate that cancer-related anemia is placed a significant effort into the global problem of anemia
unexpectedly common in cancer patients receiving radio- in our patients, or its correction. Therefore, we are currently
therapy and/or chemotherapy and that even modest anemia conducting a retrospective assessment of patient charts for
(hemoglobin <12 g/dl) substantially impairs QOL. In addi- patients seen at Beth Israel Medical Center and St. Luke’s-
tion, a growing body of literature suggests that hemoglobin Roosevelt Hospital Center between December 1996 and
concentrations of less than 12-13 g/dl during cancer therapy March 1999. To date, 202 randomly selected patients who
are associated with reduced local tumor control and were referred for radiation oncology evaluation, and who
decreased survival [1, 2]. Although anemia of this magni- ultimately received radiotherapy, were included. The
tude can be easily managed, it is often overlooked or con- results show that 48% of these patients presented to our
sidered clinically insignificant. However, restoration of department with anemia (defined as hemoglobin <12 g/dl)
normal hemoglobin concentrations in patients undergoing and a total of 57% ultimately became anemic by the end of
radiotherapy has the potential to improve local tumor con- therapy. When stratified by sex, 32% of men and 57% of
trol and survival and produce sizable improvements in women presented with anemia, which increased to 51% of
patient QOL. men and 64% of women by the end of treatment. These

Correspondence: Louis B. Harrison, M.D., Department of Radiation Oncology, Beth Israel Medical Center, St. Luke’s-
Roosevelt Hospital Center, 10 Union Square East, New York, New York 10003, USA. Telephone: 212-844-8087; Fax: 212-
844-8086; e-mail: lharrison@bethisraelny.org Received April 13, 2000; accepted for publication April 21, 2000.
©AlphaMed Press 1083-7159/2000/$5.00/0
The Oncologist 2000;5:(suppl 2):1-7 www.TheOncologist.com
2 Radiotherapy-Associated Anemia

numbers indicate that anemia is a common problem for a among patients immediately before undergoing radio-
majority of patients undergoing radiotherapy. Given the therapy. In our study, 28 breast cancer patients received
impact of anemia on local control and QOL, this is a prob- chemotherapy prior to radiotherapy. Of these, nearly
lem that needs to be addressed in a more comprehensive 68% presented with anemia prior to radiotherapy, most
manner. (57%) having only modest anemia (hemoglobin level
To assess the prevalence of anemia according to tumor between 11.0 and 11.9 g/dl). Most of the 202 patients
type, we stratified the data according to anatomical site. we reviewed so far had modest, or functional, anemia,
The data show that for patients with breast, prostate, col- which is often not treated. Very few had physiologic ane-
orectal, lung and bronchus, and uterine-cervix cancers, the mia (hemoglobin of 8 g/dl or less), which is more apt to
percentage experiencing anemia increased during the receive attention.
course of radiotherapy (Fig. 1). The most notable increase
occurred in the prostate cancer group. Only 5% of patients UNDERSTANDING THE IMPACT OF ANEMIA
with prostate cancer presented with anemia, but roughly In 1951 Hollaender et al. published their study on the
one-third (32%) were anemic by the end of their treatment. radiosensitivity of aerobically and anaerobically cultured

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

More uterine-cervix cancer patients presented with anemia Escherichia coli [5]. Their study helped form the founda-
(67%) than any other group, and by the end of radiotherapy tion of current thinking on tumor hypoxia and its effect on
82% had experienced anemia. This number will more than radiotherapy. Under hypoxic conditions, tumors appear to
likely increase as more uterine-cervix cancer patients be less susceptible to radiation, and clinical outcomes are
receive combination therapy. poorer [6]. A recent study investigated the relationship
When we examined the data for uterine-cervix cancer between tumor hypoxia and tumor control and survival in
patients with anemia, we discovered that for most patients, patients with cervical cancer undergoing radiotherapy. The
the severity of anemia before and during therapy was results of tumor oxygenation measurements taken prior to
modest (Fig. 2). Of the 67% of patients who were radiotherapy by an Eppendorf oxygen electrode and mea-
anemic upon presentation, 44% had hemoglobin levels sured by the Eppendorf pO2 histograph indicate that tumor
between 10 g/dl and 11.9 g/dl. Of the 82% who were ane- hypoxia is associated with an increased risk of relapse
mic at some point during treatment, over half (59%) had and death, particularly in patients with bulky hypoxic
hemoglobin levels between 10 g/dl and 11.9 g/dl. For tumors. Tumor oxygenation was significantly associated
both groups, 11% had hemoglobin levels between 9 g/dl with disease-free survival (p = 0.02), as was tumor size
and 9.9 g/dl. Only a few in either group had levels below (p = 0.0003), stage (p = 0.006), and pretreatment hemoglo-
8.9 g/dl. Data on colorectal and prostate (Fig. 3), lung, bin level (p = 0.001) [7].
and breast cancer patients (Table 1) yielded similar infor- The association between anemia and tumor hypoxia is
mation. The current practice of treating breast can- not fully understood, but it is well established that low
cer patients with lumpectomy and adjuvant chemotherapy hemoglobin levels independently predict poorer survival
prior to radiotherapy results in a high incidence of anemia and relapse [8-10]. However, it is not clear whether low
Percent of patients with anemia

100 Anemia Before and During RT (by site)

82
80
67 67
63
60 53 53
41 44
40 32
20
5
Figure 1. Results of Beth Israel Medical 0
Center/St. Luke’s-Roosevelt Hospital Breast Colon/ Lung/ Prostate Uterine-
Center retrospective chart review assessing (n = 63) Rectum Bronchus (n = 19) Cervix
the incidence of anemia, defined as hemo- (n = 36) (n = 10) (n = 27)
globin concentration <12 g/dl, before and
during radiation therapy by site. The data Baseline
provided are for the 202 patients reviewed to During therapy Note: Anemia = Hb < 12 g/dl
date.
Harrison, Shasha, White et al. 3

Table 1. Distribution of anemia before and during radiation therapy (RT) in patients with breast cancer (n = 63) in the Beth Israel Medical
Center/St. Luke’s-Roosevelt Hospital Center retrospective chart review

Patients who received

Baseline During RT pre-RT chemotherapy (n = 28)
n (%) n (%) n (%)
All patients with anemia 26 (41.2) 28 (44.4) 19 (67.9)
Hemoglobin concentrations
6.0–6.9 g/dl 0 0 0
7.0–7.9 g/dl 0 0 0
8.0–8.9 g/dl 2 (3.2) 0 0
9.0–9.9 g/dl 3 (4.8) 4 (6.3) 1 (3.6)
10.0–10.9 g/dl 5 (7.9) 5 (7.9) 2 (7.1)
11.0–11.9 g/dl 16 (25.4) 19 (30.2) 16 (57.1)

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

100 Anemia in Patients with Cervical Cancer Receiving RT by Hb Levels (n = 27)
Percent of patients with anemia

81.5
80
66.7
60

40 37 37

22.2
20
7.4 11.1 11.1 7.4
3.7 3.7 3.7 3.7
0
0
All 6 g/dl– 7 g/dl– 8 g/dl– 9 g/dl– 10 g/dl– 11 g/dl–
Patients 6.9 g/dl 7.9 g/dl 8.9 g/dl 9.9 g/dl 10.9 g/dl 11.9 g/dl

Baseline

During therapy Note: Anemia = Hb < 12 g/dl

Figure 2. Results of Beth Israel Medical Center/St. Luke’s-Roosevelt Hospital Center retrospective chart review assessing the inci-
dence of anemia, defined as hemoglobin concentration <12 g/dl, before and during radiation therapy in cervix cancer patients. The
data, grouped by hemoglobin levels, are for 27 patients with cervix cancer.

hemoglobin levels are associated with poor survival with a hypoxic cell sensitizer, etanidazole, or radiation ther-
because they indicate advanced disease or because they apy alone. The survival rate at five years for patients with
indicate poor tumor oxygenation. Current thinking, how- anemia — defined as hemoglobin concentration <14.5 g/dl
ever, recognizes anemia as a likely contributor to tumor for men and <13 g/dl for women — was 22%. For nonane-
hypoxia [8, 11] with a focus on anemia and locoregional mic patients the rate at five years was 36%, a significant
failure. increase compared with anemic patients (p = 0.0016).
The impact of anemia on survival and relapse after Treatment with etanidazole did not significantly improve
radiotherapy has been investigated for a number of tumor survival among any group of patients. Locoregional failure
types [8, 9]. A recent study looked at the association at five years differed significantly between patients with or
between anemia and overall survival and local tumor con- without anemia. Patients with normal hemoglobin levels
trol in patients with locally advanced head and neck cancer experienced locoregional failure at a rate of 52% compared
undergoing radiation therapy. Patients were stratified by with 68% for anemic patients (p = 0.00028) [3]. Again, the
hemoglobin levels and were given either radiation therapy addition of etanidazole did not significantly affect locore-
4 Radiotherapy-Associated Anemia

100 Anemia in Patients with Prostate Cancer Receiving RT by Hb Levels (n = 19)

Percent of patients with anemia

80

60

40
31.6

20
26.3
5.3 5.3 5.3
0 0 0 0 0 0 0 0 0 0
All 6 g/dl– 7 g/dl– 8 g/dl– 9 g/dl– 10 g/dl– 11 g/dl–
Patients 6.9 g/dl 7.9 g/dl 8.9 g/dl 9.9 g/dl 10.9 g/dl 11.9 g/dl

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

Baseline
During Therapy Note: Anemia = Hb < 12 g/dl

Figure 3. Results of Beth Israel Medical Center/St. Luke’s-Roosevelt Hospital Center retrospective chart review assessing the inci-
dence of anemia, defined as hemoglobin concentration < 12 g/dl, before and during radiation therapy in patients with prostate can-
cer. The data, grouped by hemoglobin levels, are for 19 patients with prostate cancer.

gional control among any group of patients. Anemia was presented with hemoglobin values >13.0 g/dl had signifi-
significantly related to decreased survival and increased cantly higher two-year rates of locoregional tumor control
relapse rates. (95% versus 66%, p = 0.0018) and survival (88% versus
A strong correlation between hemoglobin levels, local 46%, p < 0.001) as compared with patients with hemoglo-
control, and survival was also observed in a study of 109 bin values <13.0 g/dl.
patients with T1-T2 squamous cell carcinoma of the glottic Improving the efficacy of radiation therapy and,
larynx treated with definitive radiotherapy [9]. Patients who therefore, locoregional response and survival, may be
achieved by addressing anemia and tumor hypoxia. Recent
studies have investigated strategies that reduce anemia
100 and tumor hypoxia such as hypoxic cell sensitizers, fluosol,
Local recurrence-free survival (%)

p = 0.002
carbogen breathing, hyperbaric oxygen, red blood cell
80 transfusions, and recombinant human erythropoietin
(epoetin alfa). In a study of patients with head and neck
60 cancer given radiation and mitomycin C, a hypoxic cell
sensitizer, or radiation alone, the data showed that the addi-
40 tion of mitomycin C significantly improved local recur-
rence-free survival and cause-specific survival (Figs. 4 and
RT alone 5) [12]. In a preliminary study, advanced head and neck
20
RT mito
cancer patients treated with daily chemoradiotherapy
0 received carbogen breathing, blood transfusions, or ery-
0 1 2 3 4 5 6 7 8 9 10 thropoietin to increase their hemoglobin levels to nonane-
Years mic levels (≥ 13 g/dl) in order to overcome tumor hypoxia.
A 100% complete response rate was observed. At 18
Figure 4. Results from the trial reported by Haffty et al. [12]. months, 6% had local recurrence and 14% had distant
Local recurrence-free survival in patients with head and neck metastasis. Carbogen breathing and anemia correction
cancer receiving radiation plus a radiosensitizer or radiation resulted in improvements in local control, cause-specific
alone as adjuvant radiotherapy after surgery. Reprinted with per-
survival, and overall survival (Fig. 6) [13].
mission.
Harrison, Shasha, White et al. 5

nificant improvements in clinical outcomes. In addition to

100
the head and neck data just discussed [9, 13], opportunities
Cause-specific survival (%)

p = 0.005
exist in other sites. In patients with cervical cancer under-
80
going radiotherapy with or without chemotherapy, those
who maintained an average weekly nadir hemo-
60
globin level (ANWH) above 12 g/dl experienced a
decrease in pelvic and distant recurrence (p < 0.0001 and
40
p < 0.0005, respectively) compared with those whose
RT alone ANWH fell below 12 g/dl. In this study, patients whose
20 RT mito hemoglobin fell below 10 g/dl, 11 g/dl, or 12 g/dl, depend-
ing upon the study site, received blood transfusions in
0
order to maintain ANWH above these levels [2]. Finally,
0 1 2 3 4 5 6 7 8 9 10
in patients with advanced cervical cancer, those patients
Years who maintained hemoglobin levels above 10 g/dl during

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

radiotherapy had improved locoregional tumor control
Figure 5. Cause-specific survival in patients with head and neck
cancer receiving radiation plus a radiosensitizer or radiation
compared with those patients whose on-therapy values fell
alone [12]. Reprinted with permission. below 10 g/dl (p < 0.01) [11].
Professional awareness of the human and economic
costs of anemia in patients with cancer has progressed sub-
Aggressive management of anemia, when incorpor- stantially in the last decade. Most recently, research com-
ated into the overall management strategy for advanced pleted by The Fatigue Coalition has revealed the profound
head and neck cancer, has the potential to provide sig- burden that fatigue imposes on patients with cancer and

Actuarial local control and survival rates

100
91 91
Percentage of patients

80 75 75
69
62 62
60 55
50

40

20

0
Local control Cause-specific Overall survival
survival

Carbogen patients at 18 months (n = 36)

Carbogen patients at 3 years (n = 36)*
Noncarbogen patients at 18 months (n = 36)†

*Estimated probabilities
† Comparison group: received some chemotherapy/RT without carbogen [13]

Figure 6. Actuarial local control and survival rates for patients treated with chemoradiotherapy. Data for carbogen patients at three
years are estimated probabilities. Patients not receiving carbogen received the same chemoradiotherapy regimen as the carbogen
group [13].
6 Radiotherapy-Associated Anemia

their caregivers [4, 14]. In turn, recent studies have shown women—had hemoglobin >10 g/dl [17]. Because most of
that modest improvements in hemoglobin concentrations our patients also had anemia that was modest and “func-
can result in significant improvements in patient QOL. A tional,” we believe the same opportunities exist to improve
recent study of epoetin alfa use in patients receiving outcomes.
chemotherapy demonstrated that increases in hemoglobin
of less than 2 g/dl resulted in significantly improved QOL CONCLUSIONS
according to a questionnaire (Functional Assessment of Greater awareness of the prevalence and impact of
Cancer Therapy-Anemia; FACT-An) and linear analog anemia in patients with cancer is an important goal for all
scale assessment (LASA) [15]. QOL improvements were clinicians. Although raising hemoglobin levels by 1 or 2
observed independent of response to chemotherapy. g/dl to 12 g/dl is easily done, historically, radiation oncolo-
Another study of erythropoietin use in patients undergoing gists have not given ample attention to treating moderate,
chemotherapy demonstrated significant improvements in or functional, anemia. The data we are collecting at Beth
LASA scores, indicating improvement in QOL. The mean Israel Medical Center and St. Luke’s-Roosevelt Hospital
increase in hemoglobin in this study was only 1.8 g/dl [16]. Center highlight the prevalence of anemia in cancer

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018

Even modest anemia has a detrimental effect on patients. We are seeing similar data for a number of tumor
clinical outcomes. Dubray [17] reported a study of head sites. Increasing attention is being given to the detrimental
and neck cancer patients receiving radiotherapy and impact anemia has on QOL, local tumor control, and over-
demonstrated that high pretreatment hemoglobin signifi- all survival. Therefore, managing anemia, even mild-to-
cantly related to higher survival rates. In this study, those moderate anemia, with an approach that minimizes risk to
with anemia had functional, not physiologic, anemia. All the patient and reduces undesirable side effects, is impera-
but 3 of 63 patients with anemia—defined as hemoglobin tive if patients are to preserve QOL and improve outcomes
<13.5 g/dl for men and hemoglobin <12.0 g/dl for of radiotherapy.

R EFERENCES
1 Cella D. Factors influencing quality of life in cancer patients: 9 Fein DA, Lee R, Hanlon AL et al. Pretreatment hemoglobin
anemia and fatigue. Semin Oncol 1998;25:43-46. level influences local control and survival of T1-T2 squa-
mous cell carcinomas of the glottic larynx. J Clin Oncol
2 Grogan M, Thomas GM, Melamed I et al. The importance of
1995;13:2077-2083.
hemoglobin levels during radiotherapy for carcinoma of the
cervix. Cancer 1999;86:1528-1536. 10 Warde P, O’Sullivan B, Bristow RG et al. T1/T2 glottic can-
3 Lee WR, Berkey B, Marcial V et al. Anemia is associated cer managed by external beam radiotherapy: the influence of
with decreased survival and increased locoregional failure in pretreatment hemoglobin on local control. Int J Radiat Oncol
patients with locally advanced head and neck carcinoma: a Biol Phys 1998;41:347-353.
secondary analysis of RTOG 85-27. Int J Radiat Oncol Biol 11 Girinski T, Pejovic-Lenfant MH, Bourhis J et al. Prognostic
Phys 1998;42:1069-1075. value of hemoglobin concentrations and blood transfusions in
4 Vogelzang NJ, Breitbart W, Cella D et al. Patient, caregiver, advanced carcinoma of the cervix treated by radiation ther-
and oncologist perceptions of cancer-related fatigue: results apy: results of a retrospective study of 386 patients. Int J
of a tripart assessment survey. Int J Radiat Oncol Biol Phys Radiat Oncol Biol Phys 1989;16:37-42.
1997;34:4-12. 12 Haffty BG, Son YH, Papac R et al. Chemotherapy as
5 Hollaender A, Stapleton GE, Martin FL. X-ray sensitivity of an adjunct to radiation in the treatment of squamous cell
E. coli as modified by oxygen tension. Nature 1951;167:103- carcinoma of the head and neck: results of the Yale
104. Mitomycin Randomized Trials. J Clin Oncol 1997;15:268-
276.
6 Fleming ID, Brady LW, Mieszkalski GB et al. Basis for
major current therapies for cancer. In: Murphy GP, Lawrence 13 Martinez J, Cabezon M, Fuentes C et al. Hyperfractionated
W, Lenhard RE, eds. American Cancer Society Textbook of chemoradiotherapy with carbogen breathing for advanced
Clinical Oncology. Atlanta: American Cancer Society, head and neck cancer. Int J Radiat Oncol Biol Phys
1995:96-134. 1999;45(suppl):377.

7 Fyles AW, Milosevic M, Wong R et al. Oxygenation predicts 14 Curt GA, Breitbart W, Cella DF et al. Impact of cancer-
radiation response and survival in patients with cervix cancer. related fatigue on the lives of patients. Proc Am Soc Clin
Radiother Oncol 1998;48:149-156. Oncol 1999;18:573A.

8 Bush RS. The significance of anemia in clinical radiation 15 Demetri GD, Kris M, Wade J et al. Quality-of-life benefit in
therapy. Int J Radiat Oncol Biol Phys 1986;12:2047-2050. chemotherapy patients treated with epoetin alfa is indepen-
Harrison, Shasha, White et al. 7

dent of disease response or tumor type: results from a community oncology practice. J Clin Oncol 1997;15:1218-
prospective community oncology study. J Clin Oncol 1234.
1998;16:3412-3425.
17 Dubray B, Mosseri V, Brunin F et al. Anemia is associated
16 Glaspy J, Bukowski R, Steinberg D et al. Impact of therapy with lower local-regional control and survival after radiation
with epoetin alfa on clinical outcomes in patients with therapy for head and neck cancer: a prospective study.
nonmyeloid malignancies during cancer chemotherapy in Radiology 1996;201:553-558.

Downloaded from http://theoncologist.alphamedpress.org/ by guest on July 20, 2018