June 21, 2010

AFFYMAX AND TAKEDA ANNOUNCE PHASE 3 TRIALS MEET PRIMARY ENDPOINTS FOR INVESTIGATIONAL DRUG, HEMATIDE™/PEGINESATIDE, TO TREAT ANEMIA IN CHRONIC RENAL FAILURE WITH SOME DIFFERENCES NOTED IN SECONDARY ANALYSES

PALO ALTO, Calif., DEERFIELD, Ill., and OSAKA, Japan, June 21, 2010 – Affymax, Inc. (Nasdaq:AFFY) and Takeda Global Research & Development Center, Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited (TSE, 4502), today announced preliminary top-line results from the Phase 3 clinical program for the investigational drug, Hematide™/peginesatide, for the treatment of anemia in chronic renal failure patients. The primary efficacy endpoint, the mean change in hemoglobin (Hb) from baseline, in each of the four Phase 3 studies (EMERALD 1, EMERALD 2, PEARL 1 and PEARL 2) met the statistical criteria for non-inferiority, when Hematide was compared to epoetin and darbepoetin, in correcting and/or maintaining Hb in the target range. Hematide also met the statistical criterion for non-inferiority in the combined four studies for the adjudicated cardiovascular composite safety endpoint (CSE), which was composed of death, stroke, myocardial infarction, congestive heart failure, unstable angina, and arrhythmia (hazard ratio (HR) 1.06, 90 percent confidence interval (CI) 0.91 – 1.22). The median duration of follow-up for patients on study drug in the four trials was 1.3 years.

In a subgroup analysis of CSE events in the EMERALD studies in maintenance treatment of anemia in dialysis patients, the frequency of CSE events was balanced between Hematide and the comparator (HR 0.95, 90 percent CI 0.79 – 1.13).

A difference in CSE events was noted, however, when a subgroup analysis was conducted in non-dialysis patients. In the PEARL trials, which evaluated correction and maintenance treatment of anemia in non-dialysis patients, the frequency of CSE events was higher in the Hematide group (21.6 percent) versus the comparator (17.1 percent) (HR 1.34, 90 percent CI 1.03 – 1.73).

The Hematide Phase 3 program, which involved 2,609 randomized patients, consisted of four open-label, randomized active-controlled clinical trials in the U.S. and Europe, including two studies in non-dialysis patients (PEARL 1 and 2) and two others in dialysis patients (EMERALD 1 and 2). In all studies, Hematide was dosed once every four weeks while comparator drugs were dosed more frequently according to their product labels. In these studies, epoetin was dosed one-to three-times per week and darbepoetin was dosed every two weeks. The Hb target range was 11-12 g/dL for non-dialysis patients and 10-12 g/dL for those on dialysis.

“Completion of these four Phase 3 studies is a key milestone and we look forward to pre- NDA discussions with the FDA,” said Arlene M. Morris, chief executive officer of Affymax, Inc. “We are continuing to evaluate the data, in particular the non-dialysis studies, and the impact on the timing of an NDA submission.”

“The focus of Takeda and Affymax is patients’ needs, which are paramount in drug development. Together, we are working to bring this treatment option to patients with chronic renal failure and to physicians who treat them,” said Azmi Nabulsi, M.D., M.P.H, president of Takeda Global Research & Development Center.

EMERALD 1 & 2 (Dialysis)

In the EMERALD studies, patients were randomized two to one to receive Hematide or epoetin. The starting dose of Hematide was based on a conversion from the prior epoetin dose.

The primary efficacy endpoint of these two studies was a mean change in Hb between baseline and the evaluation period, which was between weeks 29 and 36 following entry into the study. This endpoint was met in both studies. The secondary endpoints of the two studies were the proportion of patients who received red blood cell (RBC) transfusions and the proportion of patients who maintained Hb within the target range during the evaluation period.

EMERALD 1

* Lower limit of the 95 percent two-sided confidence interval for the difference from epoetin (Hematide – epoetin) met the pre-specified criterion for non-inferiority (lower bound of confidence interval > - 1.0 g/dL)

Regarding the secondary endpoints, the proportion of patients who received RBC transfusions was similar between groups (10.3 percent in the Hematide group and 8.6 percent in the epoetin group). The proportion of patients with a mean Hb within the target range during the evaluation period was statistically significantly lower in the Hematide group (63.0 percent) compared to the epoetin group (71.7 percent).

The mean Hb level during the evaluation period was 11.1 g/dL for Hematide and 11.3 g/dL for epoetin. Serious adverse events (SAEs) occurred in 58 percent of Hematide patients and in 63 percent of epoetin patients.

EMERALD 2

* Lower limit of the 95 percent two-sided confidence interval for the difference from epoetin (Hematide – epoetin) met the pre-specified criterion for non-inferiority (lower bound of confidence interval > - 1.0 g/dL)

Regarding the secondary endpoints, the proportion of patients who received RBC transfusions was similar (7.6 percent in the Hematide group and 9.9 percent in the epoetin group). The proportion of patients with a mean Hb within the target range during the evaluation period was also similar (63.5 percent in the Hematide group vs 65.9 percent in the epoetin group).

The mean Hb level during the evaluation period was 11.1 g/dL for Hematide and 11.1 g/dL for epoetin. SAEs occurred in 49 percent of Hematide patients and in 52 percent of epoetin patients.

PEARL 1 & 2 (Non-dialysis)

In the PEARL studies, patients were randomized into one of three treatment groups: Hematide starting dose of 0.025 mg/kg administered once every four weeks, Hematide starting dose of 0.04 mg/kg administered once every four weeks and darbepoetin starting dose of 0.75 mcg/kg administered every two weeks.

The primary endpoint for these two studies was a mean change in Hb between baseline and the evaluation period, which was between weeks 25 and 36 following entry into the study. This endpoint was met in both studies. The secondary endpoints of the two studies were the proportion of patients who received RBC transfusions and the proportion of patients who responded during the correction and evaluation periods.

PEARL 1

* Lower limit of the 97.5 percent two-sided confidence interval for the difference from darbepoetin (Hematide – darbepoetin) met the pre-specified criterion for non-inferiority (lower bound of confidence interval > - 1.0 g/dL)

Regarding the secondary endpoints, the proportion of patients who received RBC transfusions was 6.2 percent in the Hematide 0.025 mg/kg group, 7.3 percent in the Hematide 0.04 mg/kg group and 4.9 percent in the darbepoetin group. These were not statistically different. The proportions of patients with a mean Hb within the target range during the evaluation period were similar (93 percent in the 0.025 mg/kg Hematide group, 94 percent in the 0.04 mg/kg Hematide group compared to 94 percent in the darbepoetin group).

The mean Hb levels during the evaluation period were 11.5 g/dL for 0.025 mg/kg Hematide, 11.6 g/dL for 0.04 mg/kg Hematide and 11.5 g/dL for darbepoetin. SAEs occurred in 48 percent of patients in the 0.025 mg/kg Hematide group, 46 percent in the 0.04 mg/kg Hematide group and 43 percent in darbepoetin patients.

PEARL 2

* Lower limit of the 97.5 percent two-sided confidence interval for the difference from darbepoetin (Hematide – darbepoetin) met the pre-specified criterion for non-inferiority (lower bound of confidence interval > - 1.0 g/dL)

Regarding the secondary endpoints, the proportions of patients who received RBC transfusions was higher in the Hematide groups (11.4 percent in the 0.025 mg/kg, 10.4 percent in the 0.04 mg/kg) than the darbepoetin group (4.9 percent). The difference between the 0.025 mg/kg Hematide group and the darbepoetin group was statistically significant. The proportions of patients with a mean Hb within the target range during the evaluation period were similar (91 percent in the 0.025 mg/kg Hematide group and 93 percent in the 0.04 mg/kg Hematide group compared to 95 percent in the darbepoetin group).

The mean Hb levels during the evaluation period were 11.6 g/dL for 0.025 mg/kg Hematide, 11.7 g/dL for 0.04 mg/kg Hematide and 11.4 g/dL for darbepoetin. SAEs occurred in 52 percent of patients in the 0.025 mg/kg Hematide group, 49 percent in the 0.04 mg/kg Hematide group and 43 percent in darbepoetin patients.

Analysis of Cardiovascular Composite Endpoint

The CSE was evaluated in a pooled analysis of all four Phase 3 pivotal trials. The CSE was defined as time to first event, consisting of death from any cause or a cardiovascular event (myocardial infarction, stroke, or a serious adverse event of congestive heart failure, unstable angina, or arrhythmia). The criterion for non-inferiority required the upper limit of the two-sided 90% confidence interval for the estimated HR to be less than 1.3. CSE events were adjudicated by a prospectively-established, independent, blinded Event Review Committee.

For the CSE from the pooled EMERALD and PEARL studies, 22.4 percent of patients in the Hematide group and 21.6 percent of patients in the control group (epoetin or darbepoetin) experienced at least one CSE event. The pre-specified criterion for non-inferiority was met with a HR of 1.06, 90 percent CI (0.91, 1.22).

In a subgroup analysis of dialysis patients (EMERALD studies), 22.8 percent of patients in the Hematide group and 24.4 percent in the epoetin group experienced at least one CSE event, HR 0.95, 90 percent CI (0.79, 1.13).

Of note, in a subgroup analysis of non-dialysis patients (PEARL studies), 21.6 percent of patients receiving Hematide and 17.1 percent of patients receiving darbepoetin experienced at least one CSE event, HR 1.34, 90 percent CI (1.03, 1.73). The HR in the non-dialysis patients was primarily driven by higher rates of death, unstable angina, and arrhythmia events in the Hematide treated patients. The percentages of patients experiencing stroke, myocardial infarction, and congestive heart failure were similar between treatment groups.

Further investigation into these data is ongoing.

Affymax, Inc. Conference Call Details

Affymax will hold a conference call today, Monday, June 21, 2010 at 8:00 a.m. Eastern Time (5:00 a.m. Pacific Time) to discuss this announcement. To participate in the live call via phone, please dial 877-354-4057 from the United States and Canada or +1-224-357-2391 internationally. The conference ID is 83101977. Please dial in approximately ten minutes prior to the start of the call. Individuals interested in listening to the live call via webcast may do so by visiting www.Affymax.com and clicking on the “Investor Relations” section. A replay of the webcast will be available on Affymax’s website for 30 days.

About Hematide

Hematide is a novel investigational synthetic, PEGylated peptidic compound that binds to and activates the erythropoietin receptor and thus acts as an ESA.

Affymax and Takeda are collaborating on the development of Hematide and plan to co-commercialize the product once approved in the United States. Phase 3 clinical trials investigated the potential for Hematide to treat anemia associated with chronic renal failure. The product, upon approval, will be commercialized outside the United States (in the European Union and Japan) by Takeda.

About Anemia in Chronic Renal Failure (CRF)

Anemia in CRF affects many individuals with Chronic Kidney Disease (CKD). According to the National Kidney Foundation, 26 million Americans - 1 in 9 U.S. adults - have CKD.^i,ii Anemia develops in the early stages of CKD and worsens as patients progress towards total kidney failure and need a dialysis machine to eliminate waste and water from their blood.² In severe or prolonged cases of anemia, the lack of oxygen in the blood can cause serious and sometimes fatal damage to the heart and other organs.ⁱⁱⁱ

About Takeda Pharmaceuticals North America, Inc. and Takeda Global Research & Development Center, Inc.

Based in Deerfield, Ill., Takeda Pharmaceuticals North America, Inc. and Takeda Global Research & Development Center, Inc. are subsidiaries of Takeda Pharmaceutical Company Limited, the largest pharmaceutical company in Japan. The respective companies currently market oral diabetes, insomnia, rheumatology and gastroenterology treatments and seek to bring innovative products to patients through a pipeline that includes compounds in development for diabetes, cardiovascular disease, gastroenterology, neurology and other conditions. To learn more about these Takeda companies, visit www.tpna.com.

About Takeda Pharmaceutical Company Limited

Located in Osaka, Japan, Takeda is a research-based global company with its main focus on pharmaceuticals. As the largest pharmaceutical company in Japan and one of the global leaders of the industry, Takeda is committed to strive towards better health for patients worldwide through leading innovation in medicine. Additional information about Takeda is available through its corporate website, www.takeda.com.

About Affymax, Inc.

Affymax, Inc. is a biopharmaceutical company committed to developing novel drugs to improve the treatment of serious and often life-threatening conditions. For additional information, please visit www.affymax.com.

This release contains forward-looking statements, including statements regarding timing, scope and results of an NDA filing for Hematide and the likelihood and timing of the commercialization of Hematide. The companies’ actual results may differ materially from those indicated in these forward-looking statements due to risks and uncertainties, including the FDA’s interpretation of the data from the Phase 3 studies, in particular with respect to the subgroup analyses in the non-dialysis groups, the risks relating to the continued safety and efficacy of Hematide in clinical development, the potential for once per month dosing and room temperature stability, regulatory requirements and approvals, research and development efforts, industry and competitive environment, intellectual property rights and disputes and other matters that are described in Affymax's quarterly report on Form 10-Q filed with the Securities and Exchange Commission. The scientific information discussed in this release related to Hematide is preliminary and investigative. Hematide has not been approved by the U.S. Food and Drug Administration or any other regulatory body, and no conclusion can be drawn regarding the safety or effectiveness of Hematide to treat anemia associated with chronic renal failure in dialysis or non dialysis patients. Investors are cautioned not to place undue reliance on these forward-looking statements, which speak only as of the date of this release. The companies undertake no obligation to update any forward-looking statement in this press release.

###

ⁱ National Kidney Foundation. http://www.kidney.org/kidneydisease/ckd/index.cfm. Accessed March 17, 2010
ⁱⁱ National Kidney Foundation. “Anemia and Chronic Kidney Disease.” http://www.kidney.org/Atoz/pdf/anemia.pdf. Accessed March 17, 2010
ⁱⁱⁱ National Heart Lung Blood Institute. http://www.nhlbi.nih.gov/health/dci/Diseases/anemia/anemia_whatis.html. Accessed on March 17, 2010

Corporate Contact:

Sylvia Wheeler
Vice President, Corporate Communications
Affymax, Inc.
650-812-8861

Josephine Zammuto
Corporate Communications
Takeda Global Research & Development Center, Inc.
224-532-0716

Seizo Masuda
Corporate Communications
Takeda Pharmaceutical Company Limited
+1 81-3-33278-2037