Anemia Associated With Androgen Deprivation (AAAD)

Quick Search


by S. B. Strum, J. E. McDermed, M. C. Scholz, H. Johnson and G. Tisman
This article originally appeared as a booklet, last edited May, 1999.

An Overview And Description Of This Adverse Effect Commonly Seen In Prostate Cancer Patients Receiving Combination Hormone Blockade


To describe the incidence, time to onset and extent of anemia occurring in patients with prostate cancer receiving combined hormone blockade (CHB) and the timing and extent of recovery from anemia in those patients who discontinued CHB.

Patients And Methods
Patients with prostate cancer were evaluated prospectively by physical examination and laboratory tests at baseline and at routine intervals while receiving CHB. Of 142 patients, 133 were assessable for their anemia. CHB was discontinued in 76 patients, of whom 64 were assessable for recovery from their anemia.

Hemoglobin levels declined significantly in all patients from a mean baseline of 149 g/L to means of 139 g/L, 132 g/L and 131 g/L at 1, 2 and 3 months, respectively. Hemoglobins continued to decline during CHB to a mean nadir of 123 g/L at a mean of 5.6 months of CHB, representing a mean absolute hemoglobin decline at nadir of 25.4 g/L. In 120 of the 133 (90%) patients, the relative decline in hemoglobin at nadir was 10% or greater and was 25% or greater in 17 (13%) others, representing a mean absolute hemoglobin decline in this subset of 42.7 g/L. Significant symptoms related to anemia occurred in 17 patients (13%). Anemia and symptoms in these patients were easily corrected with the subcutaneous administration of recombinant human erythropoietin.

The anemia associated with androgen deprivation is significant and occurs routinely in men receiving CHB. It is normochromic, normocytic, temporally related to the initiation of androgen blockade and usually resolves after CHB is discontinued. We suggest that patients receiving CHB undergo hematological testing at baseline, 1-2 months after initiating CHB and periodically thereafter. Patients developing anemia should be questioned about symptoms reflecting physiologic compromise (e.g., angina, dyspnea on exertion). In the absence of other causative factors, CHB should be suspected in the development of anemia in patients receiving this treatment.

Anemia, androgen deprivation, combined hormone blockade, prostate cancer, hemoglobin



Androgen blockade has long been recognized as a mainstay in the management of advanced stage prostate cancer1. Labrie, et. al.2 reported markedly improved survival in patients with advanced prostate cancer treated with CHB using orchiectomy or an LHRH-agonist plus an antiandrogen when compared with historical control patients receiving hormone monotherapy. In subsequent controlled trials, Crawford et. al.3 and others4 showed the superiority of CHB compared with orchiectomy or LHRH- agonist therapy alone, with a median survival advantage that exceeded 20 months in patients of high performance status and minimal stage D2 disease4. Recent trials are attempting to expand benefits of CHB to patients with earlier stages of prostate cancer. The Southwest Oncology Group (SWOG) is evaluating the feasibility of intermittent hormone blockade (IHB), and recent reports have described IHB5,6 and CHB using a neoadjuvant approach in clinically localized prostate cance7,8.

In these clinical settings, it is important to recognize any adverse effects of CHB and distinguish them from symptoms characteristically seen in patients with prostate cancer. In early trials comparing CHB with monotherapy2,3, the overall incidence of anemia was reported to be 6% in patients with stage D2 disease randomized to receive flutamide plus medical or surgical castration. These studies did not discuss the degree and timing of anemia or whether it was a side-effect of CHB treatment or a manifestation of metastatic prostate cancer.


Review Of The Literature

As early as 1949 Hamilton9 described the physiological and metabolic changes that occur following orchiectomy in six incarcerated men who involuntarily underwent castration. In that study, serum testosterone fell to castrate levels within 10 days. Hemoglobin levels also declined, reaching a mean decline of 10 g/L after 40 days.

In 1991, Weber, et. al., described an anemia that resulted from the use of an LHRH-agonist (nafarelin acetate) in 7 men treated for BPH10. Hemoglobin levels declined from a mean baseline of 152 g/L to a mean nadir of 141 g/L after 6 months of treatment, representing a statistically significant mean fall of 11 g/L, (a 7.3% decline compared to baseline, P < 0.05). Hemoglobin and serum testosterone levels returned to baseline values within 6 months after androgen suppression was stopped. The levels of serum erythropoietin levels during nafarelin treatment were normal. In-vitro studies of normal human bone marrow cells with and without nafarelin showed no evidence of toxicity to the erythroid cell line. The authors concluded that hemoglobin changed in a parallel but delayed manner with respect to serum testosterone levels.

In 1994, Eli, et. al.11 administered an LHRH-agonist (leuprolide acetate 3.75 mg per month) to 50 men with BPH (the recommended leuprolide dose to treat prostate cancer in the USA is 7.5 mg per month). After 6 months, the mean hemoglobin level decreased by 8 g/L in leuprolide treated patients compared with those who received placebo injections. Hemoglobin levels returned to baseline in treated patients 6 months after androgen blockade was stopped.

In 1994, we reported an anemia in patients with prostate cancer who received CHB; the anemia was temporally related to initiation of hormone blockade and was more severe than that reported with either orchiectomy or LHRH-agonist monotherapy12,13. Asbell, et al.14 also reported an unexplained anemia that developed in 144 patients who began neoadjuvant CHB 2 months before EBRT. As EBRT alone had not been reported to cause anemia, Asbell et. al. also concluded that this anemia was attributable to CHB.

Monotherapy with flutamide or other non-steroidal antiandrogens is not associated with significant anemia15-17. In view of the greater degree of anemia observed in our patients, we proposed that an additive anemia-producing effect occurs when methods which both lower testosterone and block androgen receptors are used simultaneously. We here elaborate further on the nature and clinical relevance of this anemia, which we term ‘anemia associated with androgen deprivation’ (AAAD).

Patients And Methods

Patients were referred to our community-based oncology practice which specialises in the treatment of prostate cancer. Patients were eligible for the study if they had a histological diagnosis of prostate cancer, no prior chemotherapy, an ECOG performance status of 0-2, a normal white blood cell and platelet count, a normal hemoglobin level and renal and liver function tests that were less than twice the upper limits of normal.

At least 2 months must have elapsed since the patients had undergone biopsies, surgery or EBRT. Patients receiving prior CHB but who had therapy discontinued for at least 6 months could enter the study. Those who elected to receive additional finasteride (e.g., 3-drug androgen blockade) were considered eligible. Those failing hormone monotherapy (e.g., orchiectomy, LHRH-agonist alone) entered the study once antiandrogen treatment began. Any other drugs known to affect the pituitary-adrenal-gonadal axis must have been discontinued at least 1 month before the initiation of CHB.


Conduct Of The Study

Patients were evaluated for age, clinical stage of disease, prior treatment(s) (EBRT to the prostatic fossa, radical prostatectomy (RP) with or without EBRT, prior hormone monotherapy), addition of finasteride, LHRH-agonists used, antiandrogens used and baseline hemoglobin and testosterone levels.

The baseline evaluation included full laboratory tests, complete blood cell count (CBC) with differential and red blood cell (RBC) indices, urine analysis, determination of serum PSA and PAP levels, stool guaic, serum testosterone level, a chest radiograph and ECG; the monthly follow-up included a history, system review and physical examination. Evaluation of the performance status, CBC with differential and RBC indices, PSA and full laboratory tests were repeated monthly for the first 3 months, then every 2-3 months thereafter.

CHB was initiated when all the staging evaluations were completed. As this was not a randomized trial, several combinations of testosterone-lowering and antiandrogen treatments were used. Most patients (121/133, 91%) received leuprolide acetate (7.5 mg intramuscularly once monthly) and 116/133 patients (87%) received the antiandrogen flutamide (250 mg orally every 8 hours). To prevent potential tumour flare, the antiandrogen was begun at least 48 hours before the LHRH-agonist when possible18. Forty-eight patients elected to add the 5 alpha-reductase inhibitor finasteride (5 mg orally twice a day) to their CHB19.

AAAD Evaluation Criteria

The baseline was considered to be the date when patients began both testosterone-lowering and antiandrogen therapy. The lowest measured hemoglobin on CHB was defined as the nadir. The month of CHB treatment when the hemoglobin nadir occurred was defined as the nadir month. Months following this nadir were referred to as N + # month (e.g., 3 months post nadir = N + 3 months). The relative degree of hemoglobin decline at nadir compared with baseline was calculated as [(baseline hemoglobin - nadir hemoglobin)/baseline hemoglobin] x 100. Relative hemoglobin and relative testosterone recoveries after CHB was stopped were defined as the ratio of hemoglobin or testosterone off therapy to baseline levels. Patients developing symptomatic anemia were treated with recombinant human erythropoietin (RHE).

For categorical variables (e.g., prior treatment), outcomes were compared using cross-tabulations; the chi-square test was used to evaluate the statistical significance of observed differences. For continuous variables, subgroups were compared using unpaired Students’ t-tests. Correlations between continuous variables were assessed using linear regression.

Of 142 consecutive patients evaluated for inclusion, 133 were assessable for anemia. Five patients were excluded from the study as they lacked baseline and/or sufficient follow-up hemoglobins in the intervals before or after office consultation. Two patients required palliative EBRT while receiving CHB and were considered inevaluable.

Clinical stage of prostate cancer at CHB initiation included 93 patients with early stage disease (T1c = 22, T2a-c = 61, T3a-c = 10) and 40 with biochemical or measurable evidence of metastatic disease (D0 = 2, D1 = 14, D2 = 24).

Table 1 Summary of Patients in Study

Testosterone And Hemoglobin Changes During CHB

Table 1 summarises the 133 evaluable patients by clinical stage of disease, age and prior treatment(s) with respect to mean baseline serum testosterone and hemoglobin levels in each patient subset. No correlation between baseline testosterone and hemoglobin was observed within the entire patient sample or within any patient subset (r = 0.27, P > 0.10).

Once patients began CHB, serum testosterone declined significantly from a mean (CI) baseline level of 4114 (240) ng/L to means of 195 (56) ng/L, 172 (42) ng/L and 131 (48) ng/L after 1, 2 and 3 months, respectively (all P < 0.001). Hemoglobins also fell significantly from a mean baseline of 149 (1.8) g/L to means of 139 (2.1) g/L, 132 (2.2) g/L and 131 (1.8) g/L after 1, 2 and 3 months, respectively (all P < 0.001). Hemoglobins continued to decrease for the next several months, reaching a mean nadir of 123 (1.5) g/L after a mean of 5.7 (0.6) months of CHB. Despite continued androgen suppression, hemoglobins recovered slightly but significantly at N + 3 months and N + 6 months to means of 133 (1.8) g/L and 135 (2.1) g/L, respectively.

The mean (CI) absolute hemoglobin decline at nadir in the patient population was 25.5 (1.8) g/L, representing a mean relative hemoglobin decline of 17.0 (1.1)% compared to baseline. The relative hemoglobin decline at nadir was 10% or greater in 120 (90%) patients. Seventeen patients (13%) had a 25% or greater fall in their hemoglobin at nadir, equivalent to a mean 42.7 (2.2) g/L decline in hemoglobin in this patient subset.

Effects Of Age And Prior Treatment Upon AAAD

The effect of patient age at CHB initiation upon the relative degree of AAAD was examined and is depicted in Figure 1. The mean (CI) age of the entire patient population was 66.8 (1.4) years; hemoglobin levels were compared between the 32 patients who were 73 years of age and older (more than 2 standard deviations above the mean) with the 101 patients who were 73 years of age and younger. Patients in the older subset had a significantly lower mean hemoglobin level at baseline and following CHB initiation compared to the younger subset. The mean relative hemoglobin decline at nadir did not differ significantly between the former (18.1 [2.3]%) compared to the latter (17.2 [1.1]%) subset (P = 0.50).

Figure 1 - Effects of Patient Age upon AAAD
Figure 1: Effects of Patient Age Upon AAAD
Legend for Figure 1:

Patients 73 years of age and older had significantly lower mean hemoglobin levels at baseline, at 1 month, at nadir, at N+3 months and at N+6 months following CHB initiation when compared to patients 73 years of age and younger.

Figure 2 illustrates the AAAD in (2A) the 122 patients who had not received prior hormone therapy and in (2B) the 11 patients who failed hormone monotherapy. Not unexpectedly, the former subset had a significantly lower mean (CI) baseline hemoglobin (140 [4.0]) g/L compared to the latter subset who were not androgen suppressed at the time CHB was initiated (149 [1.9]) g/L, P < 0.001). Once CHB was begun, mean hemoglobin levels declined in each patient subset to mean hemoglobin nadirs which were not significantly different (123 [1.8] g/L vs 127 [4.0] g/L, respectively, P = 0.11).

Figure 2A - Patients with No Prior Hormone Monotherapy
Figure 2A: AAAD in Patients Who Had Not Received Prior Hormone Monotherapy

Figure 2B - Patients Who Had Failed Prior Hormone Monotherapy
Figure 2B: AAAD in Patients Who Had Failed Prior Hormone Monotherapy

Legend to Figures 2A & 2B:
Patients who had not received prior hormone therapy (2.A.) had a significantly higher mean hemoglobin level at baseline compared to patients who failed prior hormone monotherapy (2.B, P < 0.05). The mean hemoglobin level at nadir was significantly different than the mean hemoglobin level at baseline in patients who failed prior hormone monotherapy (P < 0.001). The mean relative hemoglobin decline at nadir was significantly smaller in patients who failed prior hormone monotherapy compared to that observed in patients who had not received prior hormone therapy (P < 0.001).

The effects of other prior treatments upon the relative degree of AAAD were also examined, however, no significant differences were observed in mean serum testosterone or hemoglobin levels with respect to prior RP, EBRT or CHB (all P > 0.05).

Patients who failed hormone monotherapy were significantly older (71.9 [3.5] years) compared to patients who were not androgen suppressed at the time of CHB was initiated (66.4 [1.4] years) (P < 0.05). In order to determine whether age or prior treatment were responsible for the lower hemoglobin levels observed, each subset was analyzed for age. Age significantly influenced hemoglobin levels independent of the types of prior treatments with the exception of patients who failed hormone monotherapy, who were found to have lower mean hemoglobin levels regardless of their age.


Effects Of Antiandrogen Used And Other Treatment Variables Upon AAAD

The 122 patients who had not failed hormone monotherapy were separately examined to determine there were significant effects on AAAD with respect to the antiandrogen used (Figure 3). Although the mean baseline hemoglobin levels in these two subsets were not significantly different (P = 0.28), once patients began CHB, mean hemoglobin levels were consistently lower in the 107 patients who received flutamide compared to the 16 patients who received bicalutamide.

Figure 3: AAAD with CHB Using Flutamide or Bicalutamide

Figure 3 - AAAD with CHB Using Flutamide or Bicalutamide
Legend for Figure 3:

Patients receiving CHB with flutamide had significantly lower mean hemoglobin levels all measured time intervals after baseline compared to patients receiving CHB with bicalutamide. Mean (CI) relative hemoglobin declines for each subset were significantly different (flutamide patients = 18.8% [1.1]% compared to bicalutamide patients = 12.8 [1.7]%, P < 0.0001).
To determine the effects of added finasteride upon AAAD, an analysis of the 107 patients who received flutamide with (n = 34) or without finasteride (n = 73) was done. There were no significant differences in baseline or on-CHB hemoglobin levels between these 2 patient subsets. We were unable to detect significant differences in the degree of AAAD between patients receiving different testosterone-lowering modalities due to small numbers of patients who received goserelin (n = 6) or underwent orchiectomy (n = 5).

Patients requiring RHE for symptomatic anemia on CHB

Significant symptoms due to anemia (e.g., angina, dyspnoea, syncope, severe weakness) developed in 17 (13%) patients after a mean of 5.3 months (range 1-12 months) of CHB. Pretreatment and treatment-related characteristics that were significantly different between patients who did and did not require RHE included clinical stage D2 disease (P < 0.05), mean [CI] baseline hemoglobin (142 [4.7] vs 150 [1.9] g/L, p < 0.01) and hemoglobin at 2 months of CHB (119 [4.5] vs 134 [2.1] g/L, p < 0.001). Anemia and its associated symptoms responded to RHE within the first 1 to 2 months. The dosage of RHE was titrated with the goal to maintain the hemoglobin within a 125 to 135 g/L range.

Hemoglobin Changes Following CHB Discontinuation

Of 76 patients who discontinued CHB, 64 were assessable for recovery from AAAD. At the time CHB was discontinued, the mean (CI) hemoglobin was 129 (2.0) g/L, representing a 14.5% decline compared to the mean baseline hemoglobin of 151(2.4) g/L. Hemoglobin levels recovered off CHB to means of 140 (3.1) g/L, 143 (2.8) g/L, 146 (2.9) g/L and 147 (2.9) g/L at 3, 6, 12 and 24 months, respectively (all P < 0.05). These represented mean relative hemoglobin recoveries versus baseline of 93.5% (2.0%), 94.7% (2.0%), 97.0% (1.9%) and 98.1% (2.3%), respectively.

Patient age had a significant impact upon the depth of hemoglobin decline on androgen blockade and on the recovery of hemoglobin after discontinuation of CHB. Patients 68 years of age and older, the median age of this sample, had a mean hemoglobin of 126 (2.3) g/L at the time CHB was stopped compared to 132 (1.9) g/L in patients 68 years of age and younger (P < 0.005). Subsequent levels were significantly lower in the older subset as well, with mean hemoglobins of 136 (3.7) g/L, 138 (3.6) g/L, 142 (3.6) g/L and 142 (3.8) g/L at 3, 6, 12 and 24 months after CHB was discontinued as opposed to means of 145 (4.5) g/L, 147 (3.7) g/L, 150 (3.8) g/L and 151 (2.8) g/L in the younger subset, respectively (all P < 0.05).

The duration of CHB significantly affected both rate and the extent of hemoglobin recovery after therapy was discontinued. Although the mean hemoglobin when CHB was discontinued was not significantly different in patients treated > 1 year (129 [2.5] g/L) compared to those patients treated < 1 year (130 [3.1) g/L, P = 0.59), patients receiving androgen blockade > 1 year had a much slower rate of hemoglobin recovery. Patients treated < year had mean hemoglobins at 3, 6, 12 and 24 months of 144 (3.6) g/L, 146 (3.8) g/L, 150 (4.0) g/L and 149 (3.7) g/L, respectively. Patients treated > 1 year had mean hemoglobins at these respective time intervals of 136 (3.7) g/L, 140 (3.8) g/L, 144 (4.0) g/L and 146 (4.0) g/L, respectively. These differences were significant at 3 months (P < 0.05), at 6 months (P < 0.01) and at 12 months (P < 0.05).

The rate and extent of hemoglobin recovery were examined with respect to both patient age at the time of CHB initiation and the duration of CHB (Figure 4). The subset of 17 men < 68 years of age who were treated < 1 year had full hemoglobin recovery (mean 98.6% [4.5%] of baseline) 3 months after CHB was discontinued. All of the 31 patients who received CHB > 1 year showed a significant delay in hemoglobin recovery. This was most apparent in the 19 men > 68 years who were treated > 1 year; full hemoglobin recovery (mean > 95% of baseline) did not occur until 12 months after CHB was discontinued.

Figure 4A - Effects of Age - Androgen Blockade 1 Year
Figure 4A: Effects of Age Upon Hemoglobin Recovery Following CHB Discontinuation in Patients Receiving Androgen Blockade < 1 year

Figure 4B - Effects of Age, Androgen Blockade 1 Year
Figure 4B: Effects of Age Upon Hemoglobin Recovery Following CHB Discontinuation in Patients Receiving Androgen Blockade >1 year

Legend for Figures 4A and 4B:
Mean hemoglobin levels in patients > 68 were significantly lower when CHB was discontinued compared to patients < 68 years of age. Mean hemoglobin levels in the older subset remained significantly lower than baseline levels and lower compared to those in the younger subset up to 24 months after CHB was stopped. The duration of CHB significantly impacted hemoglobin recovery in the older subset; patients < 68 years of age receiving CHB > 1 year had >95% hemoglobin recovery 3 months after CHB was discontinued.

Serum Testosterone Changes Following CHB Discontinuation

Following CHB discontinuation, serum testosterone also recovered, increasing to mean levels of (CI) of 2560 (768) ng/L, 2780 (686) ng/L, 3060 (672) ng/L and 3210 (630) ng/L at 3, 6, 12 and 24 months after CHB was discontinued, respectively (all P < 0.05). When compared to baseline, these represented mean relative testosterone recoveries of 55.7%, 60.9%, 68.4% and 70.3%, respectively (all P < 0.05).

The rate and extent of serum testosterone recovery were also affected by patient age at CHB initiation and the duration of CHB treatment (Figure 5). These effects were most apparent when the subset of 17 men less than 68 years of age who were treated with CHB for less than 1 year were compared to the subset of 19 men 68 years of age and older who received CHB longer than 1 year. In the former subset, 50% testosterone recovery was observed after a mean (CI) of 5.3 (2.5) months, but was not seen until after a mean of 14.9 (3.1) months off androgen blockade in the latter subset (P < 0.05). Mean maximum relative testosterone recovery in the former subset was 86.1% (13.2%), in contrast to a mean of only 64.9% (16.1%) in the latter subset (P < 0.05).

Figure 5 - Serum Testosterone Recovery vs. Age and CHB Duration
Figure 5: Serum Testosterone Recovery with Respect to Patient Age and CHB Duration

Legend for Figure 5:
The rate and extent of testosterone recovery varied with respect to patient age at CHB initiation and the duration of CHB administration. Of the 14 patients 68 years of age and older who had levels drawn at 24 months off androgen blockade, 8 (57%) had serum testosterone levels < 3000 ng/L; in 5 of these 8, serum levels were < 2000 ng/L.

Laboratory evaluation of patients’ peripheral blood smears revealed an anemia that was normochromic and normocytic and smears stained with Wright-Giemsa showed no evidence of leukoerythroblastic findings20. Stool guaic testing performed serially while patients received CHB remained negative and serum lactic dehydrogenase (LDH) and bilirubin levels remained normal.


In this study, 133 patients treated with medical or surgical castration plus an antiandrogen developed a significant degree of anemia that became apparent within the first three months of CHB and peaked after a mean of six months of androgen blockade. The anemia was unrelated to renal insufficiency, bone metastases, blood loss or hemolysis. Finasteride had no significant effect upon the timing or extent of AAAD when added to CHB using a LHRH agonist plus flutamide. Bicalutamide, when compared to flutamide, was associated with a significantly lower degree of AAAD at the recommended dose of 50 mg/day. In 64 assessable patients in whom CHB was discontinued for > 1 year, recovery of hemoglobin and serum testosterone levels was delayed and often incomplete, particularly in patients > 68 years of age and older treated with CHB > 1 year.

Previous published reports of anemia secondary to androgen withdrawal therapy have been limited to studies using LHRH agonists for 6 months to treat BPH10,11. In these trials, anemia occurred in all patients within the same mean period, but with less than half the severity observed in the present patients receiving CHB. These authors also reported that serum testosterone and hemoglobin became normal within a mean of 4 and 6 months, respectively after LHRH agonist therapy was discontinued. In the present study, the degree of anemia was additive when both testosterone-lowering and androgen receptor blocking modalities were used simultaneously, and resolved more slowly with the CHB duration required for the treatment of prostate cancer.

The relationship between androgens and erythropoiesis has been well documented in many previous studies. Modder et. al.21 demonstrated that testosterone and 5 beta-DHT stimulated the production and maturation of early erythroid precursors in laboratory animals, whereas 5 alpha-DHT had no such effect. However, testosterone and 5 alpha-DHT both stimulated erythropoietin production. Thus, it is not surprising that our patients who experienced symptomatic AAAD responded to subcutaneous administration of RHE with rapid normalization of their hemoglobins.

In the present study, 11 patients entered the trial having failed orchiectomy or LHRH agonist alone after a mean treatment period of 21.3 months. As expected, these patients had significantly lower mean baseline hemoglobin than the others but their mean nadir hemoglobin was not significantly different (Figure 2). Although the mean relative decline of hemoglobin at nadir of 9.5% in these patients was the smallest of all subsets, it represented a significant fall when compared to baseline hemoglobin (P < 0.001). Therefore, it appears that the addition of an antiandrogen to patients who have failed hormone monotherapy and are already anemic can result in further decline in hemoglobin, thus illustrating the additive anemia-producing effect of CHB. The exact mechanism of AAAD is not known. Whether it is caused solely or partly by androgen deprivation or by an as yet unknown direct or indirect suppressive effect upon erythropoiesis by antiandrogens or other factors relating to red blood cell survival remains to be determined. Importantly, only 31 (50%) of the 71 patients receiving CHB > 1 year reached a hemoglobin nadir by their sixth month of androgen blockade (mean 6.9 months, range 1-15 months). To monitor any possible delayed nadir, we recommend the determination of hemoglobin every 2 to 3 months while patients receive CHB.

The persistence of testosterone suppression in our older patients as long as 24 months off CHB was an unexpected finding. We have obtained serum LH levels in several of these patients in order to ascertain possible etiologies for this and found them all to be normal or only minimally elevated. This observation is inconsistent with testicular atrophy, where one would expect to see markedly elevated LH levels in response to low serum testosterone. This finding does suggest that prolonged androgen blockade may alter the LHRH receptor, which should respond to low serum testosterone with a secondary rise in levels of LH.

With CHB now being used to treat patients with earlier stages of prostate cancer, the development of anemia in otherwise healthy males could be misinterpreted as gastrointestinal bleeding, hemolysis or metastatic disease to the bone marrow. Many patients described by Asbell, et. al.14 and several in our practice who began CHB elsewhere had undergone exhaustive evaluations by gastroenterologists and hematologists to determine the cause of recent anemia.

In conclusion, AAAD is a normochromic, normocytic anemia that is temporally related to the initiation of CHB and occurs in virtually all patients. Hemoglobins declined and reached a minimum after approximately 6 months of androgen blockade in a parallel but delayed manner with respect to the fall in serum testosterone levels. In 100 of 133 (75%) patients, the mean hemoglobin at nadir represented an absolute decline of > 10% compared with baseline and in 17 (13%), others the mean nadir was > 25% lower than their baseline hemoglobin. In 17 of 133 (12%) of patients, the symptoms of AAAD were severe enough to require treatment. In all such patients, anemia responded to the subcutaneous administration of RHE.

Hemoglobin levels recovered in all patients when CHB was discontinued, but the recovery period was protracted such that full recovery not seen in some patients until 24 months after CHB was discontinued. Serum testosterone recovered, albeit more gradually than did hemoglobin levels, but was incomplete in patients > 68 years of age who were treated with CHB > 1 year.

We recommend patients receiving CHB be monitored for symptoms which may indicate significant AAAD and that they be screened using hemoglobins at 1 to 2 months after the initiation of CHB and periodically thereafter.


1. Beck JR, Kattan MV, Miles BJ: A critique of the decision analysis for clinically localized prostate cancer. J Urol 152:1894-1899, 1994.

2. Labrie F, Dupont A, Belanger A, et al: Combination therapy with flutamide and castration (LHRH agonist or orchiectomy) in advanced prostate cancer: a marked improvement in response and survival. J Steroid Biochem 123:833-841, 1985.
3. Crawford ED, Blumenstein BA, Goodman PJ, et al: Leuprolide with and without flutamide in advanced prostate cancer. Cancer 66(Suppl 1):1039-1044, 1990.

4. Eisenberger M, Crawford ED, McLeod D, et al: A comparison of leuprolide and flutamide vs leuprolide alone in newly diagnosed stage D2 prostate cancer: Prognostic and therapeutic importance of the minimal disease subset. Proc Am Soc Clin Oncol 11:201, 1992 (abstract).

5. Strum SB, Scholz MC, Strum M: Prolonged non-detectable PSA (NDPSA) in PC patients treated with androgen deprivation may allow for discontinuation of hormone blockade (CHB). International Symposium on Recent Advances in Diagnosis and Treatment of Prostate Cancer, Quebec City, September 21-23, 1995 poster presentation).

6. Higano CS, Ellis WJ, Lange PH: Intermittent androgen suppression for treatment of prostate cancer: The Seattle pilot experience. Proc Am Soc Clin Oncol 15:243, 1996 (abstract).

7. Labrie F, Dupont A, Gomez J, et al: Downstaging of localized prostate cancer by neoadjuvant therapy with flutamide and Lupron: the first controlled and randomized trial. Clin Invest Med 16:499-509, 1993.

8. Kocheril PG, Shamsa F, Forman JD: The significance and pattern of change of PSA in patients treated with neo-adjuvant hormonal therapy and radiotherapy. Proc Am Soc Clin Oncol 15:243, 1996 (abstract).

9. Hamilton JB: The role of testosterone secretion as indicated by the effects of castration in man and by studies of pathological conditions and the short lifespan associated with maleness. Recent Prog Horm Res 3:257-289, 1948.

10. Weber JP, Walsh PC, Peters CA, Spivak JL: Effect of reversible androgen deprivation on and serum immunoreactive erythropoietin in men. Am J Hemat 36:190-194, 1991.

11. Eri LM, Tveter KJ: Safety, side effects and patient acceptance of the luteinizing hormone releasing hormone agonist leuprolide in the treatment of benign prostatic hyperplasia. J Urol 152:448-452, 1994.
12. Strum SB, Tisman G: Anemia of androgen deprivation (AAD) in patients receiving combination hormonal blockade: response to erythropoietin. Presented at the 6th International Symposium on Supportive Care in Cancer, Abstract O-1, New Orleans, LA, March 2-5, 1994.

13. Strum SB, Tisman G: Anemia of androgen deprivation (AAD) in patients receiving combination hormone blockade: Response to erythropoietin. Proc Am Soc Clin Oncol 13:238, 1994.

14. Asbell SO, Tester WJ: The incidence of anemia in prostate cancer patients treated with androgen suppression and pelvic radiotherapy. Proc Am Soc Clin Oncol 13:241 (abstract).

15. Sogani PC, Vagaiwala MR, Whitmore WF: Experience with flutamide in patients with advanced prostate cancer without prior hormone therapy. Cancer 54:744-750, 1984.

16. Prout GR, Keating MA, Griffin PP, et al: Long-term experience with flutamide in patients with prostatic carcinoma. Urology 34(Suppl 4):37-45, 1989.

17. Chang A, Yeap B, Davis T, Blum R, et al: Double-blind, randomized study of primary hormonal treatment of Stage D2 prostate carcinoma: flutamide versus diethylstilbestrol. J Clin Oncol 14:2250-2257, 1996.

18. Kuhn JM, Billebaud T, Navratil H, et al: Prevention of the transient adverse effect of a gonadotropin releasing hormone analog (Buserelin) in metastatic carcinoma by administration of an antiandrogen (Nilutamide). N Engl J Med 321:413-418, 1989.

19. Fleshner NE, Trachenberg J: Treatment of advanced prostate cancer with the combination of finasteride plus flutamide: Early results. Eur Urol 24(Suppl 2):106-112, 1993.

20. Shamdas GJ, Ahman FR, Matzer MB, Ritchie JM: Leukoerythroblastic anemia in metastatic prostate cancer. Clinical and prognostic significance in patients with hormone-refractory disease. Cancer 71:3594-3600, 1993.
21. Modder B, Foley JE, Fisher JW: The in vitro and in vivo effects of testosterone and steroid metabolites on erythroid colony forming cells (CFU-E). J Pharmacol Exp Ther 207:1004:1012, 1978.

© The Prostate Cancer Research Institute (PCRI), 1997. All rights reserved