1. The EUROCARE-6 CML survival estimates revealed that the prognostic improvement highlighted by RCTs was confirmed in the European real-world setting.

2. There are still large differences in CML survival throughout Europe: the prompt introduction of TKIs in daily clinical practice is undelayable.

The European incidence of chronic myeloid leukemia (CML) was about 1.1/100,000 inhabitants (1), increasing to about 4.0/100,000 in patients aged 75–99 at the time of diagnosis. The disease is characterized by the presence of the BCR-ABL1 fusion gene located in the Philadelphia (Ph) chromosome and is classified as being in a chronic (CP), accelerated (AP), or blastic phase (BP), with the last two phases accounting for about 4% and 3% of cases, respectively (2, 3) and being associated with a worse prognosis (4).

For many years, CML was associated with a poor life expectancy (5), but the 2001 introduction of imatinib mesylate, the first tyrosine kinase inhibitor (TKI) and, more recently, of second- and third-generation TKIs (dasatinib, nilotinib, bosutinib, and ponatinib) has profoundly changed the CML curative-intent treatment, previously based on hematopoietic stem cell transplantation. TKIs have greatly improved CML survival rates and now make it possible to consider CML a chronic disease (6–11). Imatinib was approved as first-line treatment for all CML phases and is now available as a generic drug, as its patent has expired. Dasatinib and nilotinib were approved in 2006–2007 as second-line treatments for patients resistant to, or intolerant of, previous treatments (including imatinib): dasatinib in all CML phases and nilotinib only in the CP or AP. Since 2010–2011, both have been authorized for the first-line treatment of newly diagnosed Ph-positive adult cases of CP CML. Bosutinib was licensed in the United States in 2012 (and in Europe in 2013) for the treatment of adults with CP, AP, or BP CML who are resistant to, or intolerant of, previous treatments with one or more TKIs. In December 2017, the recommendation was extended in the United States to include newly diagnosed adult patients with CP CML. Ponatinib was approved in the United States in 2012 (and in Europe in 2013) for the treatment of adults with CP, AP, or BP CML who are resistant to, or intolerant to, other TKIs and also for the treatment of those with CP, AP, or BP CML who have the T315I mutation, which is known to be involved in resistance to all previous TKIs.

The 5-year survival estimates for patients with CML increased from 1997 to 2008 throughout Europe (particularly after 2000), although with large differences among European countries (10, 12): they increased slightly in Southern Europe, more in the United Kingdom, and considerably more in Northern, Central, and Eastern Europe, although in the latter region, survival remained lower than elsewhere (10). These improvements were plausibly linked to the widespread introduction of targeted and other new treatments (10). There was only a small increase in survival estimates among the elderly, possibly because of the under-use of imatinib (90% of patients aged 20–59 received imatinib, 75% of those aged 60–79, and 46% of those aged ≥80) and the newer TKIs (13). Furthermore, the cancer registry (CR) of Girona showed that the 5-year survival rate in patients with CML treated with TKIs in 1994–2008 was about 80%, compared with 44% among those who were not (14).

Population-based studies including all cases occurring in the region covered by a CR reflect the effectiveness of healthcare services in controlling the disease and are more likely to highlight socioeconomic disparities potentially associated with cancer survival. People who live in more affluent areas have better access to optimal care than those living in deprived areas, and this leads to discrepancies in overall survival (OS) figures (15). Moreover, access to optimal treatment is related to per capita income and healthcare investments (16).

Clinical practice, particularly in oncological settings, often relies on randomized controlled trials (RCTs) because they provide more detailed information than population-based studies. However, the amount of data may be overwhelming (17), and it can be difficult to determine the health systems’ sustainability, in terms of finance and uptake of new practices. As a consequence, oncological organizations have developed frameworks to help clinicians and policymakers quantify the real value of new therapies (17–21). Generalizing trial results to everyday clinical practice is not straightforward because of low overall trial accrual (<5% of all newly diagnosed patients with cancer) and under-representation by age, gender, disease stage, co-morbidities, and socioeconomic status. However, despite these limitations, approved treatments are frequently offered to patients who would have been ineligible for the related trials, but they rarely show the benefits detected in RCTs; furthermore, a survival advantage detected by RCTs is not always subsequently confirmed in real-life setting.

This raises questions as to how the results of population-based studies using real-world data can add to the results of RCTs involving patients with CML and to the findings on the extent of RCTs’ effectiveness for the patient population as a whole. In an attempt to answer these questions, we compared the survival of patients with CP CML participating in RCTs with the data from EUROCARE, the widest collaborative population-based study on cancer survival in Europe (22).

The comparison of EUROCARE-6 CML survival estimated in patients diagnosed in 2000–2006 vs. 2007–2013 confirmed that the prognostic improvement highlighted by RCTs was verifiable in real-world settings. In particular, the EUROCARE-6 OS values in many countries ( Table 3 ) were very similar to CP CML OS reported in RCTs ( Table 1 ) (58, 59). Moreover, the same brilliant achievement was observed comparing the estimated values of 5-year cause-specific survival in CP CML RCTs (58, 59) with 5-/3-year conditional crude RS estimated in almost all European countries in 2007–2013 ( Supplementary Table 4 ). This means that the optimal outcome figures obtained in controlled settings are achievable (and, indeed, are achieved) in real-world settings, too. The high concordance between CRs and RCTs survival results could be explained by the fact that TKIs are responsible of the quite complete disappearance of AP and BP worse prognosis CML phases. Almost all patients are diagnosed in CP (or have been quickly brought back to CP), so survival results reported in the whole population are close to those of RCTs. Moreover, the high concordance between CRs and RCTs survival results could be related to the fact that we compared quite homogeneous groups of patients with CML aged lower than 65 years with probably few comorbidities.

Previous population studies reported similar or inferior survival results but estimated only on national or small pooled samples.

Swedish CML Registry (779 CMLs, from 2002 to 2010; median age, follow-up: 60 years, 61 months) showed 5-year RS close to 1.0 for those younger than 60 years, 0.9 for those aged 60 to 80 years, and 0.6 for those older than 80 years (70). Swedish Cancer Registry (2,662 CMLs, from 1973 to 2013; median age: 69 years) reported clear improvements in life expectancy over the study period (71). Swedish Cancer Registry and Swedish Cause of Death Registry (CMLs, from 1970 to 2012) showed 5-year OS increasing from 0.18 to 0.82, during the study period; between 2006 and 2012, 5-year RS was close to normal for 40-year-old but considerably lower for 80-year-old patients (72). UK’s Haematological Malignancy Research Network (242 CMLs, from 2004 to 2011; median age: 59 years) showed 5-year OS of 78.9% (72.3% to 84.0%) and 5-year RS of 88.6% (81.0% to 93.3%) (73). Other national studies are aligned with our survival results (74–80).

European Treatment and Outcome Study (EUTOS) (2,904 CMLs, from 2008 to 2013; median age, follow-up: 55 years, 29 months) showed a 30 months OS of 92% (81). US Surveillance, Epidemiology, and End Results (SEER) (13,869 CMLs, from 1975 to 2009) reported lower survival values: 5-year RS ratios increased from 0.26 in 1975–1989 to 0.36 in 1990–2000 and 0.56 in 2001–2009 (82). Moreover, SEER (5,138 CMLs, from 2000 to 2005) showed 5-year OS improvement for all patients during the study period (83, 84). Compared with patients diagnosed in 2000, 5-year OS improved among 15–44 years (from 71.6% to 86.4%), 45–64 years (from 67.5% to 76.3%), 65–74 years (from 38.1% to 51.2%), and 75–84 years patients (from 19.2% to 36.4%) (83).

Population-based studies using real-world survival data reveal differences from the values observed in RCTs that are often related to treatment disparities and largely due to different socioeconomic conditions. They also provide information concerning treatment effectiveness in everyday clinical practice without any patient or outcome selection: they are therefore more representative of what happens in real-life, despite lacking in clinical details offered by RCTs, particularly in relation to disease stage at the time of diagnosis and first-line treatments. The findings of RCTs are often used to guide clinical practice (particularly in oncology), but patient selection can reduce their applicability to the general population (17, 18, 20, 21). Conversely, results of population-based CR studies that fully cover the target population are less affected by patient selection biases, and they provide useful data complementing RCTs outcomes.

However, these two information sources need to be integrated and require the use of new study designs and methods of analysis. High-resolution population-based studies, which include representative patients, present more detailed clinical information than that which is routinely collected by population-based CRs: this approach may help to reduce the gap between RCTs and real-world studies (hrstudies.it; https://www.ipaac.eu/en/work-packages/wp7/).

In an attempt to quantify the difference between RCTs and population-based studies using tangible data, we compared OS and cause-specific survival observed in the RCTs included in our previous systematic review (58), and OS and RS values estimated using EUROCARE-6 (22) cases diagnosed up to age 64 over a comparable period of time. It was the first time that this was done for CML, considering all European regions and pooling survival results. Our study shows that CML survival values tend to become very similar between RCTs and population-based settings, regardless of the survival analysis methods used. However, some differences still persist, in particular in Eastern European countries, where OS values were lower than elsewhere, especially in the first period of time being considered: this is probably due to a delayed introduction of TKIs in daily clinical practice. To underline that the date of the introduction of TKIs reimbursement varied greatly between Europe: this could be useful to interpret the different survival outcomes observed by countries ( Supplementary Table 5 ). Also to notice that the allogeneic bone marrow transplantations medium rate was 0.62 per million for Eastern European countries in comparison with 0.81 per million for other European countries [ Supplementary Table 6 , by calendar year from 2000 to 2022 and by country; data provided by the European Society for Blood and Marrow Transplantation (EBMT), Chronic Malignancies Working Party (CMWP)].

Residual discrepancies can be attributed to different case selection criteria: RCTs select patients on the basis of well-defined inclusion and exclusion criteria, and the results cannot be readily extended to the general population, whereas population-based studies involve unselected patients but often lack detail and, in the case of CML, the morphology code might be not very precise. Moreover, RCTs almost always record cancer-specific mortality, with off-study survival being reported by the investigator after study discontinuation, whereas population studies systematically update life status of all registered patients and use RS to make adjustments for general mortality by age, gender, and geographical area.

RCTs also generally include patients without comorbidities who are younger than those encountered in real-life populations: for example, it has been found that the elderly, women, and members of racial and ethnic minorities are less likely to be enrolled in American cooperative group cancer trials than patients who are younger, male, and Caucasian (85, 86).

Our previous meta-analysis did not reveal any difference in the OS of patients treated with the first- or the new-generation TKIs (58, 59). In the only two RCTs for which 5-year OS data are available [DASISION (23, 32) and ENESTnd (38, 45)], the 60-month OS value was similar in the patients treated with imatinib and those treated with dasatinib or nilotinib, and similar to EUROCARE-6 OS data for patients diagnosed in 2007–2013. To underline that second-generation TKIs introduction time in clinical practice (2006–2007) limits a strict comparison with survival data of previous years, but imatinib can be considered an historical arm because it has been introduced in 2001. Moreover, CML survival values under imatinib or second-generation TKIs are fairly superimposable (60 months RCTs OS ≥ 90%, Table 1 ).

We compared the first-line treatment of RCT patients with newly diagnosed CP CML with all treatment lines administered to patients with CML from the general population (including a small percentage of patients with AP and BP CML who have a different prognosis). Unfortunately, CRs do not routinely collect information on CML phase and treatment line; thus, it was not possible to select CP CML cases receiving first-line treatment. To overcome this drawback, we analyzed 5-/3-year conditional OS and RS to remove the contribution of BP and AP CML and improve estimates comparability. Considering conditional OS and RS for patients diagnosed in 2007–2013, population-based CRs survival values were very similar to those observed in the RCTs.

Code 9876 (Ph− atypical CML or aCML) was not included but, as most CRs do not distinguish Ph+ CML and Ph− aCML, and as 78.0% of cases are classified as CML NOS ( Table 2 ), some aCML cases were inevitably included. This has little impact on our analysis as 90%–95% of CML diagnoses have the characteristic t(9;22)(q34;q11.2) reciprocal translocation, leading to the Ph chromosome and to the BCR-ABL1 fusion gene that is the target for specific TKIs (4). However, this partly explains why OS values for ICD-O-3 code 9863, including CML NOS and (probably) patients with poorer prognosis (such as aCML cases not targeted by TKIs), were, at all evaluable times and in all evaluable regions, lower compared to the values for Ph+ CML for which TKIs are indicated.

Code 9875 (Ph+ CML) was hardly used in Northern Europe or the United Kingdom and Ireland, and the implausibly small number of cases in the other regions/countries considered is attributable to differences in registration criteria or inaccurate pathological description. It is also likely that the underuse of code 9875 for Ph+ CML is due to a bad translation of the ICD-O-3 classification: code 9863 refers to “chronic myeloid leukemia, NOS” and code 9875 refers to “chronic myelogenous leukemia, BCR/ABL positive” (Ph+ CML) and, although hematologists normally correctly diagnose cases of code 9875 as Ph+ CML, the use of the word “myelogenous” is ambiguous for non-hematologists. This may also explain the considerable difference in the use of code 9875 between specialized hematological registries and general CRs. CRs should correctly code CML morphology by specifying ICD-O-3 9875 (Ph+ CML) or 9876 (Ph− aCML), the phase of the disease at the time of diagnosis, first-line therapy, and the occurrence of transformation into AP or BP to make a more precise analysis possible: one that is potentially comparable with other types of studies. Some strategies should be adopted to avoid CML code misuse and to reduce the number of CML NOS cases, such as to plan specific training courses to increase the precision of coding or to link CML population-based data with other available data sources, for example, national health insurance databases, to discover patients really treated with TKIs (87). Unfortunately, 9875 (Ph+ CML) code is so underused in CRs in the studied period 2000–2013 not to permit to design a population-based study excluding 9863 code (CML NOS).

A clear improvement in real-world CML survival was observed in European regions and countries comparing EUROCARE-6 with RCTs OS data. However, some discrepancies with RCTs still remain. Our results suggest an insufficient adoption of adequate protocols in daily clinical practice in countries where CML survival values still remain lower in real-life than those obtained in RCTs. In future works, it will be of interest to focus on populations usually excluded from RCTs, such as older patients, or with comorbidities and other cancers.

Austria: M. Hackl (National CR); Belgium: E. Van Eycken (National CR); Bulgaria: Z. Valerianova (National CR); Croatia: M. Sekerija (National CR); Cyprus: P. Pavlou (National CR); Czech Republic: L. Dušek (National CR); Denmark: H. Storm (National CR); Estonia: M. Mägi; K. Innos* (National CR); Finland: N. Malila; J. Pitkäniemi (National CR); France: M. Velten (Bas Rhin CR); X. Troussard (Basse Normandie, Haematological Malignancies CR); A.M. Bouvier; V. Jooste* (Burgundy, Digestive CR); A.V. Guizard (Calvados, General CR); G. Launoy (Calvados, Digestive CR); S. Dabakuyo Yonli (Cote dOr, Gynaecologic (Breast) CR); M. Maynadié (Cote dOr, Haematological Malignancies CR); A.S. Woronoff (Doubs CR); J.B. Nousbaum (Finistere, Digestive CR); G. Coureau (Gironde, General CR); A. Monnereau* (Gironde, Haematological Malignancies CR); I. Baldi (Gironde, Central Nervous System CR); K. Hammas (Haut-Rhin CR); B. Tretarre (Herault CR); M. Colonna (Isere CR); S. Plouvier (Lille Area CR); T. D’Almeida (Limousin CR); F. Molinié; A. Cowppli-Bony (Loire-Atlantique/Vendée CR); S. Bara (Manche CR); C. Schvartz (Marne-Ardennes, Thyroid CR); G. Defossez (Poitou-Charentes CR); B. Lapôtre-Ledoux (Somme CR); P. Grosclaude (Tarn CR); Germany: S. Luttmann (Bremen CR); R. Stabenow [Common CR of 4 Federal States (Brandenburg, Mecklenburg-West Pomerania, Saxony-Anhalt, Thüringen)]; A. Nennecke (Hamburg CR); J. Kieschke (Lower Saxony CR); S. Zeissig (Rhineland-Palatinate CR); B. Holleczek (Saarland CR); A. Katalinic* (Schleswig-Holstein CR); Iceland: H. Birgisson (National CR); Ireland: D. Murray; P.M. Walsh (National CR); Italy: G. Mazzoleni; F. Vittadello (Alto Adige CR); F. Cuccaro (Barletta-Andria-Trani CR); R. Galasso (Basilicata CR); G. Sampietro (Bergamo CR); S. Rosso (Biella CR); M. Magoni (Brescia CR); M. Ferrante (Catania-Messina-Enna CR); A. Sutera Sardo (Catanzaro CR); M.L. Gambino (Como CR); P. Ballotari; E. Giacomazzi (Cremona and Mantova CR); S. Ferretti (Ferrara CR); A. Caldarella; G. Manneschi (Firenze-Prato CR); G. Gatta*; M. Sant*; P. Baili*; F. Berrino*; L. Botta; A. Trama; R. Lillini; A. Bernasconi; S. Bonfarnuzzo; C. Vener; F. Didonè; P. Lasalvia; G. Del Monego; M.C. Magri; L. Buratti (Fondazione IRCCS Istituto Nazionale dei Tumori, Milan); D. Serraino; L. Dal Maso (Friuli Venezia Giulia CR); R. Capocaccia* (Epidemiologia e Prevenzione Board); R. De Angelis*; E. Demuru; C. Di Benedetto; S. Rossi*; M. Santaquilani; S. Venanzi (Istituto Superiore di Sanità, Rome); R.A. Filiberti (Genova CR); S. Iacovacci (Latina CR); V. Gennaro (Liguria, mesotheliomas CR); A.G. Russo (Lodi CR); G. Spagnoli (Modena CR); L. Cavalieri d’Oro (Monza and Brianza CR); M. Fusco; M.F. Vitale (Napoli CR); M. Usala (Nuoro CR); F. Vitale (Palermo CR); M. Michiara (Parma CR); G. Chiranda (Piacenza CR); G. Cascone; E. Spata (Ragusa CR); L. Mangone (Reggio Emilia CR); F. Falcini (Romagna CR); R. Cavallo (Salerno CR); D. Piras (Sassari CR); A. Madeddu; F. Bella (Siracusa CR); A.C. Fanetti (Sondrio CR); S. Minerba (Taranto CR); G. Candela; T. Scuderi (Trapani CR); R.V. Rizzello (Trento CR); F. Stracci (Umbria CR); G. Tagliabue (Varese CR); M. Rugge (Veneto CR); A. Brustolin (Viterbo CR); Latvia: S. Pildava (National CR); Lithuania: G. Smailyte (National CR); Malta: M. Azzopardi (National CR); Norway: T.B. Johannesen* (National CR); Poland: J. Didkowska; U. Wojciechowska (National CR); M. Bielska-Lasota* (National Institute of Public Health-National Institute of Hygiene-National Research Institute, Warsaw); Portugal: A. Pais (Central Portugal CR); J.L. Pontes (Northern Portugal CR); A. Miranda (Southern Portugal CR); Slovakia: C. Safaei Diba (National CR); Slovenia: V. Zadnik; T. Zagar (National CR); Spain: C. Sánchez-Contador Escudero; P. Franch Sureda (Balearic Islands, Mallorca CR); A. Lopez de Munain; M. De-La-Cruz (Basque Country CR); M.D. Rojas, A. Aleman (Canary Islands CR); A. Vizcaino (Castellon CR); R. Marcos-Gragera (Girona CR); M.J. Sanchez (Granada CR); M.D. Chirlaque (Murcia CR); M. Guevara Eslava*; E. Ardanaz (Navarra CR); J. Galceran; M. Carulla (Tarragona CR); Switzerland: Y. Bergeron (Fribourg CR); C. Bouchardy (Geneva CR); S. Mohsen Mousavi (Graubünden and Glarus CR); S. Mohsen Mousavi (Eastern Switzerland CR); A. Bordoni (Ticino CR); The Netherlands: O. Visser* (National CR); UK-England: J. Rashbass (National CR); UK-Northern Ireland: A. Gavin* (National CR); UK-Scotland: D. Morrison (National CR); UK-Wales: D. W. Huws* (National CR).*EUROCARE Steering Committee

The datasets presented in this article are not readily available. See EUROCARE-6 Collaborative Group Rules (http://www.eurocare.it). Requests to access the datasets should be directed to http://www.eurocare.it.

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent from the participants’ legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

CV designed and carried out the study and analyzed the data; SR and PM did quality controls and analyzed the data; RA and MS designed the study and data quality checks; RM-G, HP, MM, XT, and GP provided advice and revised the results. EUROCARE-6 Working Group collected, prepared, and transmitted raw data for the study database; corrected data after quality controls; and checked the results of the analyses. All authors contributed to the article and approved the submitted version.

This study was funded by the Compagnia di San Paolo, the Cariplo Foundation and the European Commission (grant number 801520 HP-JA-2017, Innovative Partnership for Action Against Cancer, iPAAC Joint Action). The sources of the funding played no role in designing the study, collecting, analyzing, or interpreting the data, writing the report, or deciding whether or not to submit the article for publication. This research was (partially) funded by Italian Ministry of Health “Ricerca Corrente” funds.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.