Treatment-Free Remission and TKI Discontinuation in CML Labeling: FDA and EMA Approaches for Nilotinib

Treatment-free remission in CML labeling: how FDA and EMA handled TKI discontinuation for nilotinib

Among the BCR-ABL tyrosine kinase inhibitors (TKIs) for chronic myeloid leukemia (CML), nilotinib (Tasigna) is the one with a regulator-reviewed treatment-free remission (TFR) pathway written into both the US prescribing information and the EU Summary of Product Characteristics (SmPC). Both agencies based the language on the same two single-arm studies, ENESTfreedom and ENESTop. Both also tied discontinuation to strict eligibility criteria, sensitive molecular monitoring and rapid re-treatment. The regulatory instruments differed. FDA approved a labeling supplement, cleared a 510(k) BCR-ABL test for identifying discontinuation candidates and monitoring treatment-free remission the same day, and imposed postmarketing requirements. CHMP declined the proposed extension of indication but accepted changes to SmPC sections 4.2, 4.4, 4.8 and 5.1.

Key takeaways

• FDA (US): Tasigna NDA 022068/S-026, approved December 22, 2017, updated the package insert so that patients with Ph+ CML in chronic phase who had received Tasigna for at least 3 years and achieved a sustained MR4.5 response may be considered for treatment discontinuation. This covers both newly diagnosed patients and patients resistant or intolerant to imatinib ³²³.
• EMA (EU): The Tasigna variation (H-798-II-84-G) assessment report was first published 19 July 2017 ¹⁸⁰²⁰⁶. CHMP considered discontinuation possible for eligible patients but did not find the results sufficient for an extension of the indication. It accepted SmPC changes to dosing, warnings, adverse reactions and clinical efficacy instead ^197198199.
• Evidence base: Two open-label, multicenter, single-arm studies: ENESTfreedom (first-line nilotinib) and ENESTop (nilotinib after prior imatinib) ³⁰⁸³⁰⁹.
• Diagnostic linkage (US): The label requires an FDA-authorized BCR-ABL test with sensitivity of at least MR4.5 ³⁶⁰. The MolecularMD MRDx BCR-ABL Test (K173492) was cleared the same day, December 22, 2017, with an intended use covering identification of nilotinib-treated patients who may be candidates for discontinuation and monitoring of TFR ^235236237.
• Other TKIs: The Sprycel (dasatinib) US label says that the effect of stopping treatment after MMR or MR4.5 on long-term disease outcome "has not been established" ²¹². CHMP stated that the nilotinib stopping approach cannot be extrapolated directly to other TKIs ²¹.

US labeling: Tasigna prescribing information

Where the TFR content sits

The US label spreads TFR content across several sections:

The 2021 and 2024 Tasigna labels keep sections 2.2 and 2.3 and continue to report the 96-week results in sections 14.3 and 14.4 ^{243244245246247249}.

Eligibility (section 2.2)

All patients must have Ph+ chronic-phase CML with typical BCR-ABL transcripts (e13a2/b2a2 or e14a2/b3a2) and at least 3 years of Tasigna. MR4.5 is defined as BCR-ABL/ABL ≤0.0032% on the International Scale (IS). Eligibility must be determined with an FDA-authorized BCR-ABL test, and the same FDA-authorized test must be used to monitor response on and off treatment ⁴⁷.

• Newly diagnosed patients: at least MR4.0 (≤0.01% IS) sustained for one year before discontinuation, with MR4.5 at the last assessment. They must have no history of accelerated phase or blast crisis and no prior discontinuation attempt that ended in relapse ⁴⁸.
• Patients resistant or intolerant to imatinib: imatinib must have been the only therapy before Tasigna, and MR4.5 must have been sustained for at least the year immediately before discontinuation. The same exclusions apply ⁴⁹.

Monitoring and reinitiation (sections 2.2, 2.3 and 5.16)

• BCR-ABL transcripts and a CBC with differential are measured monthly in year 1, every 6 weeks in year 2 and every 12 weeks after that. After loss of MR4.0, testing moves to every 2 weeks until levels stay below MMR for four consecutive measurements ⁵⁰.
• Newly diagnosed patients must restart Tasigna within 4 weeks of losing MMR (MR3.0, ≤0.1% IS) ³⁶¹. They restart at the dose used before discontinuation, with monthly BCR-ABL monitoring until MMR is regained and every 12 weeks after that ³⁶⁵.
• Post-imatinib patients must restart within 4 weeks of losing MMR or of confirmed loss of MR4.0, meaning two consecutive measurements at least 4 weeks apart ³⁶²³⁶⁶.
• If MMR is not reached after 3 months of re-treatment, BCR-ABL kinase domain mutation testing should be performed ³⁶³.

Safety signal specific to discontinuation

The label states that musculoskeletal symptoms (myalgia, pain in extremity, arthralgia, bone pain, spinal pain, musculoskeletal pain) were reported more often in the first year after discontinuation than in the year before. The rate of new symptoms generally fell in the second year ²⁹⁶. Table 10 gives all-grade incidences for patients who stayed in TFR for more than 48 weeks:

Among patients with musculoskeletal symptoms during TFR, the symptoms had not resolved by the cutoff in 23/53 (43%) of newly diagnosed patients and 32/57 (56%) of post-imatinib patients ²⁹⁹.

EU labeling: Tasigna SmPC

Eligibility and monitoring (sections 4.2 and 4.4)

The SmPC allows a physician experienced in CML to consider TFR in adults with Ph+ CML in chronic phase. Two pathways apply:

• First-line nilotinib: 300 mg twice daily for at least 3 years, with a deep molecular response (MR4.5) sustained for at least the year immediately before stopping ¹⁰⁸.
• Nilotinib after prior imatinib: nilotinib for at least 3 years, with MR4.5 sustained for at least the preceding year ¹⁰⁹.

Section 4.4 requires confirmed typical transcripts (e13a2/b2a2 or e14a2/b3a2), because these allow response depth and loss of remission to be quantified ¹¹². The monitoring schedule matches the US schedule (monthly, then every 6 weeks, then every 12 weeks, including CBC with differential). The SmPC does not require an agency-authorized test. It requires a quantitative test validated on the IS and sensitive to at least MR4.5 ^113114346.

Re-treatment rules are close to the US rules, with a few product-information details of their own:

• First-line: loss of MR4 without loss of MMR triggers BCR-ABL testing every 2 weeks. Loss of MMR requires restarting within 4 weeks at 300 mg twice daily, or 400 mg once daily if the dose had been reduced before stopping ¹¹⁵.
• After imatinib: confirmed loss of MR4 (two measurements at least 4 weeks apart) or loss of MMR requires restarting within 4 weeks at 300 mg or 400 mg twice daily ¹¹⁶.

Section 4.8

Text attributed to SmPC section 4.8 states that patients may have musculoskeletal symptoms more often after stopping Tasigna than before. In ENESTfreedom (N=190), these symptoms occurred in 24.7% of patients within a year after discontinuation, compared with 16.3% in the year before on treatment ³⁴⁷³⁴⁸. These figures use a different denominator and time frame from the US Table 10 subset (patients in TFR more than 48 weeks), so the two sets of numbers should not be compared directly.

Supporting evidence: ENESTfreedom and ENESTop

The trials measured MR4.5 with the MolecularMD MRDx BCR-ABL Test ³¹⁷³⁰⁹. The 510(k) decision summary for that test cites the Tasigna discontinuation studies as clinical performance evidence ²³⁸.

The two primary endpoints are not the same. ENESTfreedom measured remaining in MMR. ENESTop also required no confirmed loss of MR4 and no re-initiation ¹²⁹¹³⁰.

How each regulator weighed the evidence

CHMP

According to the assessment report, ENESTfreedom did not meet its 48-week primary endpoint, because the lower confidence limit fell below the prespecified 50% threshold. ENESTop met its primary objective. Most relapses occurred within the first 24 weeks, and response generally returned after nilotinib was restarted ¹⁹⁶. CHMP noted these limitations:

• single-arm design ¹⁹³
• relatively young, fit and mostly Caucasian study populations, which limits generalizability ²⁰⁰²⁰¹
• immature overall and progression-free survival data ¹⁹³
• the risk that monitoring in routine practice may be less intensive than in the trials, plus uncertainty about late relapse and resistance emerging off treatment ^{118119202203204}
• musculoskeletal pain after stopping, which was generally manageable but has no clear mechanism ²⁰⁵

CHMP treated strict eligibility, frequent molecular monitoring and prompt re-initiation as essential conditions. It concluded that the data did not justify a new indication but did justify product-information changes ^197198199. CHMP also noted that academic TFR studies exist for imatinib and, to a lesser extent, dasatinib. It stated that the nilotinib approach cannot be extrapolated directly to other TKIs ²¹.

FDA

An FDA multidisciplinary review for S-026 could not be located in Drugs@FDA. The supplement's documents there are the labeling and the approval letter ²⁸²²⁸³, so FDA's reasoning has to be inferred from those two documents. The approval letter imposed three postmarketing requirements under section 505(o):

• PMR 3323-1 (relapse risk): final reports with at least 60 months of follow-up from both trials. The final 5-year report was due October 2020 ³²⁸³²⁹.
• PMR 3323-2 (resistance): annual tracking of late relapse, loss of response and mutations, plus gene-expression comparisons. The final 10-year report was due February 2026 ³²⁶³²¹.
• PMR 3323-3 (long-term safety): mortality, late adverse events, pancreatitis, diabetes, arterial vascular disease and other conditions, in patients still in remission and in those re-treated. The final report was due February 2026 ³²⁷³²⁰.

These PMRs target the same uncertainties CHMP raised: durability of remission, emergent resistance and long-term safety. On the same day, FDA cleared the MRDx test with a discontinuation and TFR-monitoring intended use ²³⁶²³⁷. Together with the label's instruction to use "the same FDA-authorized test" before and after stopping ⁴⁷³⁶⁰, this makes the US pathway more closely tied to a specific diagnostic than the EU pathway, which requires only a validated test with MR4.5 sensitivity ¹¹⁴.

Status across the CML TKI class

• Nilotinib follow-on products (US): The Danziten (nilotinib tartrate), Cavhanza and Cipla nilotinib D-tartrate labels include a "Treatment Discontinuation in Patients With Ph+ CML-CP Who Have Achieved a Sustained Molecular Response (MR4.5)" section describing musculoskeletal symptoms after discontinuation ^{141142143144145146}. The excerpts reviewed did not show a complete stopping and monitoring protocol in these labels ^147148149.
• Dasatinib (US): Sprycel and Phyrago state in section 2.6 that the effect of stopping treatment after MMR or MR4.5 on long-term disease outcome has not been established ²¹²²¹³.
• Imatinib (US): No TFR guidance was found in the Gleevec label excerpts reviewed ²⁷²²⁷³.
• Asciminib (EU): The Scemblix EPAR mentions TFR only as a treatment goal in the benefit-risk context. It contains no asciminib stopping recommendation ¹⁸¹⁹²⁰.
• Ponatinib (EU): The Iclusig EPAR discusses stopping for lack of haematologic response by 3 months, which is a non-response rule and not TFR ²⁵.

Practical implications for labeling and regulatory strategy

1. The same dataset can lead to different label architecture. FDA put TFR into Dosage and Administration, Warnings and Precautions and Clinical Studies. CHMP refused the indication extension and placed the content in SmPC sections 4.2, 4.4, 4.8 and 5.1 ^197198199.
2. Molecular monitoring is a central labeling element. Both labels specify transcript type, MR4.5 sensitivity, the monitoring schedule and 4-week re-initiation windows ^{47112114361362}. In the US this came with a cleared test ²³⁷.
3. Expect long follow-up obligations. FDA's PMRs extend to 10-year reporting on resistance and safety ³²⁶³²⁷. The EU SmPC has since been updated with 264-week data ³³²²⁷⁴.
4. Off-treatment adverse events go in the label. Both regulators required disclosure of musculoskeletal symptoms after discontinuation ²⁹⁶³⁴⁷.
5. Discontinuation claims are product-specific. Neither agency extended nilotinib's TFR language to other TKIs, and the dasatinib US label explicitly says the effect of stopping has not been established ²¹²¹².