U.S. Food and Drug Administration Drug Approval Summary: Conversion of Imatinib Mesylate (STI571; Gleevec) Tablets from Accelerated Approval to Full Approval Free

On December 5, 2003, the U.S. Food and Drug Administration converted the approval of imatinib mesylate tablets (Gleevec, Novartis Pharmaceuticals) for the treatment of adult patients with Philadelphia chromosome positive (Ph+) chronic myeloid leukemia (CML) in (a) chronic phase after failure of IFN-α therapy, (b) accelerated phase, and (c) blast crisis from accelerated to regular approval. Accelerated approval, allowed on May 10, 2001 (1), permits approval of a drug for a serious or life-threatening disease for which treatment is inadequate based on an effect on a surrogate end point (in this case, cytogenetic and hematologic responses) that is reasonably likely to predict clinical benefit. Patient follow-up was relatively short at the time of accelerated approval (3-7 months after recruitment of the last chronic phase, accelerated phase, or blast crisis patient), too short to establish clinical benefit. Therefore, Novartis Pharmaceuticals was required to continue patient follow-up to determine response duration and to obtain additional safety data.

Imatinib mesylate subsequently received accelerated approval for treatment of newly diagnosed CML on December 20, 2002; this indication has not yet been converted to regular approval.

Study Populations. Definitions of CML chronic phase, accelerated phase, and blast crisis are summarized in Table 1. Definitions or hematologic and cytogenetic response criteria for each of the three CML phases are summarized in Table 2.

Chronic Phase CML Study. The primary objective of the study was to determine the rate of complete cytogenetic response (CCyR) and major cytogenetic response (MCyR) to imatinib mesylate. Secondary objectives, many of greater interest in assessing clinical benefit, included determination of the rate and duration of complete hematologic response (CHR) and the duration of CCyR and MCyR. Other secondary objectives included evaluation of the safety profile of imatinib mesylate: measurement of time to accelerated phase disease or blast crisis and determination of overall survival. The first patient was recruited on December 6, 1999, whereas the last patient was recruited on May 30, 2000. Data cutoff for this analysis was July 31, 2002.

Study eligibility included Ph+ CML, ages ≥18 years, and with IFN failure or intolerance defined as any of the following: (a) failure to achieve a CHR, lasting for at least 1 month despite ≥6 months of IFN or an IFN-containing regimen, in which IFN was administered at a dose of at least 25 million IU (MIU)/wk. During this treatment period, the cumulative duration of hydroxyurea therapy may not have exceeded 50% of the treatment period with the IFN-containing regimen (hematologic resistance); (b) bone marrow cytogenetics showing >65% Ph+ after 1 year of IFN-based therapy (cytogenetic resistance); (c) an increase in the Ph+ chromosome in bone marrow cells by at least 30 percentage points (e.g., from 20% to 50% or from 30% to 60%), confirmed by two samples at least 1 month apart, or an absolute increase to >65% (cytogenetic refractoriness); and (d) a rising WBC count (to a level >20 × 10⁹/L confirmed by two samples taken at least 2 weeks apart) for patients achieving a CHR while receiving IFN or an IFN-containing regimen at a dose of at least 25 MIU/wk. During this treatment period, the cumulative duration of hydroxyurea therapy may not have exceeded 50% of the treatment period with the IFN-containing regimen (hematologic refractoriness), or (e) intolerance to IFN therapy defined as a grade >3 nonhematologic toxicity persisting for at least 1 month, for patients receiving IFN alone or in combination when IFN was administered at a dose of at least 25MIU/wk. Patients who were intolerant of IFN were to have been diagnosed ≥6 months before the time of entry into the study.

Patients received imatinib mesylate 400 mg p.o. qd. Dose escalation was permitted to 600 mg p.o. qd or to a maximum of 800 mg daily taken as 400 mg bid.

Accelerated Phase CML Study. The primary objective was to determine the rate of hematologic response lasting >4 weeks. Secondary objectives included determination of the duration of hematologic response, overall survival, cytogenetic response, time to blast crisis, and tolerability and safety of ST1571 treatment. The first patient was recruited on August 9, 1999, whereas the last patient was recruited on May 12, 2000. Data cutoff for this analysis was December 16, 2002.

To be eligible for the study, patients had to be ages >18 years, with performance status <2, and with a histologically confirmed diagnosis of Ph+ leukemia of one of the following types: accelerated phase CML (Table 1), relapsed/refractory acute lymphocytic leukemia, relapsed/refractory acute myelocyticleukemia, and relapsed/refractory CML in lymphoid blast crisis.

Patients received imatinib mesylate 400 or 600 mg p.o. qd. Dose escalation was permitted, to a maximum of 800 mg daily, taken as 400 mg bid.

Blast Crisis Phase CML Study. The primary objective was to determine the rate of hematologic response lasting ≥4 weeks. Secondary objectives included determination of the duration of hematologic response, overall survival, cytogenetic response, and tolerability and safety of imatinib mesylate treatment. The first patient was recruited on June 30, 1999, whereas the last patient was recruited on May 12, 2000. Data cutoff for this analysis was July 31, 2002.

To be eligible for the study, patients had to be ages ≥18 years, with performance status ≤2, and with a histologically confirmed diagnosis of Ph+ CML in myeloid blast crisis (Table 1). Both newly diagnosed patients and patients who had received prior therapies for CML accelerated phase or blast crisis were eligible. Newly diagnosed patients were not to have received prior accelerated phase or blast crisis therapies, with the exception of IFN or hydroxyurea.

Patients received imatinib mesylate 400 or 600 mg p.o. qd. Dose escalation was permitted, to a maximum of 800 mg daily, taken as 400 mg bid.

Chronic Phase CML Study.Table 3 summarizes patient demographics characteristics and Table 4 summarizes pertinent data concerning disease history at baseline. Patients studied were in late chronic phase CML, with a median time since diagnosis of 32 months. As expected, patients with cytogenetic failure had the longest IFN exposure, and IFN-intolerant patients the shortest. Even for IFN intolerant patients, however, 26% received IFN for 6 to 12 months and 30% received it for ≥1 year.

The median duration of imatinib mesylate exposure at the time of the data cutoff was 29 months. The maximum imatinib mesylate exposure was 31.5 months. Eighty-one percent of patients were on treatment for ≥24 months.

Confirmed Cytogenetic Response Rate. Confirmed cytogenetic responses are shown in Table 5; 38% of study patients attained a CCyR and 59% had a MCyR. More than 50% of the MCyRs were noted at the time of the first on-study bone marrow (3 months). The percentage of patients with a MCyR steadily increased with increasing time on treatment. Approximately 9% of MCyRs occurred after 1 year of treatment and ∼7% occurred after 2 years of treatment. The MCyR rate was highest in patients with a recent diagnosis of CML (within 1 year of study enrollment) and was lower in patients with laboratory evidence of more advanced disease (i.e., high WBC count, high platelet count, low hemoglobin, and ≥3% blasts in the peripheral blood).

Thirty-five of the 343 patients who had achieved MCyR during treatment had a confirmed loss of response. Of these 35 cases, 7 had achieved a confirmed CCyR. Only 10 (2.9%) of the MCyR patients progressed to accelerated phase or blast crisis. The estimated rate of patients still in MCyR after 18 months is 90.0% [95% confidence interval (95% CI), 86.6-93.4] and at 24 months is 87.8% (95% CI, 83.8-91.7).

Complete Hematologic Response. A CHR was achieved in 503 of 532 study patients (95% CI, 92.3-96.3) CHR was lost during treatment in 117 (23.2%) patients. Sixty-six (13.1%) of these patients progressed to accelerated phase or blast crisis.

Time to Accelerated Phase or Blast Crisis. Of the 532 patients, 85 (16.0%) patients had values indicating progression to accelerated phase or blast crisis. The estimated probabilities (95% CI) of being free of progression to accelerated or blast crisis are 88.4% (85.7-91.2) at 18 months and 85.4% (82.4-88.5) at 24 months.

Survival. At time of analysis, 65 (12.2%) of the 532 patients had died. Of the 64 deaths (1 death was reported after bone marrow transplant), 8 occurred on study treatment and 56 occurred during follow-up after discontinuation of treatment (mostly due to progression; n = 45). The estimated probabilities (95% CI) of being alive are 94.2% (92.2-96.2) at 18 months and 90.8% (88.3-93.2) at 24 months.

Accelerated Phase CML Study. Demographics and performance status of the study population are summarized in Table 6. As indicated, patients receiving imatinib mesylate 400 or 600 mg/d were comparable.

Disease history is provided in Table 7. The median time from accelerated phase diagnosis to study entry was 1.4 and 0.8 months for the imatinib mesylate 400 and 600 mg treatment groups. Approximately 54% of each treatment group had splenomegaly (spleen palpable below the costal margin) and ∼21% of patients had hepatomegaly. The median WBC count of all treated patients was 21.4 × 10⁹/L and ∼10% of each treatment group had basophils ≥20%.

Efficacy results are displayed in Table 8. The primary efficacy variable was confirmed hematologic response. As indicated, the hematologic response rate was higher in patients receiving imatinib mesylate 600 mg/d compared with patients receiving imatinib mesylate 400 mg/d. In addition to imatinib mesylate administration at 600 mg/d, factors associated with improved overall hematologic response rate included gender (females 78%, males 65%), platelets ≥100 × 10⁹/L (76% versus 62%), blasts <15% in bone marrow (82% versus 69%), and hemoglobin ≥100 g/L (77% versus 49%).

MCyR, duration of hematologic response, time to progression, and overall survival were also greater in the imatinib mesylate 600 mg/d treatment group. The median survival in the advanced leukemia population (acute lymphocytic leukemia, acute myelocytic leukemia, and lymphoid blast crisis) was only 5 months, and only two patients are still on treatment.

Blast Crisis CML. Baseline demographics and performance status of the study population are summarized in Table 9. Table 10 summarizes prior disease history. As indicated, 37 patients received imatinib mesylate 400 mg/d and 223 received imatinib mesylate 600 mg/d. One hundred sixty-five (63.5%) patients were considered previously untreated for blast crisis, as either no accelerated phase or blast crisis treatment was recorded or they had received only IFN, hydroxyurea, or low-dose cytosine arabinoside, which was considered palliative treatment in this setting. The remaining 95 (36.5%) patients had received cytotoxic chemotherapy and/or bone marrow transplantation for advanced disease.

Duration of drug exposure is summarized in Table 11. Duration of exposure as shorter in pretreated patients (median 92 days) than in untreated patients (median 142 days).

Efficacy results are displayed in Table 12. The primary efficacy variable was confirmed hematologic response. As indicated, the hematologic response rate was higher in patients receiving imatinib mesylate 600 mg/d compared with patients receiving imatinib mesylate 400 mg/d. In addition to imatinib mesylate dose, other factors influencing hematologic response rate included favorable hematologic variables including a higher platelet count, higher hemoglobin, and <50% blasts in the peripheral blood.

The estimated median duration of hematologic response was 10.1 months (7.3-13.6 months). The estimated proportion of patients still responding at 24 months is 27.2% (16.8-37.7).

Unconfirmed MCyR (single bone marrow cytogenetic examination) was noted in 40 (16.4%) patients. Confirmed MCyRs are shown in Table 12. It should also be noted that ∼50% of patients had other chromosomal abnormalities in addition to being Ph+. The median duration of cytogenetic response was 4.6 months.

Safety.Table 13 lists systemic adverse events and Table 14 lists laboratory adverse events. The most frequent systemic adverse events were fluid retention nausea and muscle cramps. There was a trend suggesting increased grade 3/4 adverse events with advancing phase of disease.

In older patients (ages ≥65 years), with the exception of edema, there was no evidence of an increase in the incidence or severity of adverse events. In women, there was an increase in the frequency of neutropenia as well as grade 1/2 superficial edema, headache, nausea, rigors, vomiting, rash, and fatigue. No differences were seen related to race, but there were insufficient numbers of non-Caucasians for proper evaluation.

Cytopenias, particularly neutropenia and thrombocytopenia, were a consistent finding and occurred with increasing frequency with advancing disease phase (i.e., blast crisis > accelerated phase < chronic phase CML). The median duration of the neutropenic and thrombocytopenic episodes varied from 2 to 4 weeks. Neutropenia and thrombocytopenia were usually managed with either dose reduction or an interruption of treatment. In rare cases, permanent discontinuation of treatment was required.

Grade 3 elevation of transaminases or bilirubin occurred in3% to 6% (Table 14) and were usually managed with dose reduction or interruption. The median duration of these episodes was ∼1 week. Treatment was discontinued permanently because of liver laboratory abnormalities in <1% of patients. However, one patient, who was taking acetaminophen regularly for fever, died of acute liver failure.

Systemic toxicity did not increase with increased duration of treatment. Cytopenias were not consistently reported as adverse events so that treatment duration effects cannot be precisely determined.

Imatinib mesylate has proven to be a very effective drug for the treatment of CML.

The drug initially received accelerated approval for treatment of CML in blast crisis, accelerated phase, and chronic phase after failure of IFN treatment on May 10, 2001. Because the data were consistent and showed a sizable effect, a very rapid review (2.5 months) was possible.

Under the accelerated procedure, the U.S. Food and Drug Administration may approve a new drug or biologic if adequate and well-controlled trials establish that the product has an effect on a surrogate end point that is reasonably likely to predict clinical benefit. Approvals for imatinib mesylate were based on the results of three nonrandomized, single-arm trials including a total of 1,027 CML patients. The surrogate end points supporting imatinib mesylate efficacy were high hematologic and cytogenetic response rates and these were considered likely to lead to a real benefit (generally improved symptoms or survival). The accelerated approval regulations require the applicant to verify and describe the clinical benefit of the drug in additional adequate and well-controlled studies or by continued follow-up of the original study population. Such studies must be conducted with due diligence.

At the time of original approval, patient follow-up was short (3-7 months from the time of accrual of the last study patient). The current submission is based on a median drug exposure of ∼29 months. The maximum duration of study treatment is 32 (chronic phase) to 35 (blast crisis and accelerated phase) months, with 21% of patients treated >1 year and 10% treated >2 years (blast crisis), 61% patients treated >1 year and 45% treated >2 years (accelerated phase), and 91% patients treated >1 year and 81% treated >2 years (chronic phase).

The U.S. Food and Drug Administration converted accelerated approval to regular approval because favorable treatment responses were sustained and clearly represented clinical benefit. In the chronic phase CML study, an estimated 87.8% of patients who achieved MCyR maintained their response 2 years after achieving their initial response. After 2 years of treatment, an estimated 85.4% of patients were free of progression to accelerated phase or blast crisis, and estimated overall survival was 90.8% (95% CI, 88.3-93.2). These results are superior to historical IFN-α or chemotherapy treatment results (2–4). Median survival from randomized trials comparing IFN-α with hydroxyurea or busulfan treatment reports median survival of 61 to 72 months for IFN treatment and 41 to 56 months for hydroxyurea or busulfan.

In accelerated phase CML treatment, results were also favorable. Hematologic response occurred in 72% and MCyR occurred in 27% of treated patients (either imatinib mesylate 400 or 600 mg/d, nonrandomized). These results are superior to historical IFN-α or chemotherapy treatment results (5–7). Overall median survival for patients receiving imatinib mesylate 600 mg/d has not yet been reached and the estimated 24-month survival is 65.8%.

In blast crisis CML, the terminal phase of CML, imatinib mesylate treatment produced a hematologic response rate of 31% and a MCyR rate of 15%. These responses are considerably higher than those described in other blast crisis studies (8). Whereas blast crisis CML is usually fatal in 3 to 6 months (9), imatinib mesylate treatment produced a median survival of 6.9months, with estimated 12- and 24-month survival rates of 32% and 18%, respectively.

In conclusion, imatinib mesylate produced sustained responses in all three phases of CML. Toxicity is acceptable. Imatinib mesylate represents an important addition to the armamentarium of drugs active in the treatment of chronic myelocytic leukemia.