Chronic Lymphocytic Leukaemia: Census of Patients Treated in Italian Haematology Units

This study was conducted by contacting the population of the Italian haematology units and collecting from 68% of them data concerning the number of patients with chronic lymphocytic leukaemia visited over the previous 12 months, with the aim of obtaining an overview of the treatment of this disease and comparing the results with the prevalence estimates found in literature. The projection obtained (about 17,000 patients visited in the previous 12 months) is probably overestimated because of double-counting of patients who may have been treated at two different facilities during the year, although it is also underestimated since the internal medicine units were not involved. The balance of these two opposite factors is not known. It is important to bear in mind the approximation with which the count was performed in facilities for which no official data were available. Albeit with these limits, the results obtained are in line with some existing prevalence data and make it possible to determine the portion of patients at different Binet stages and in the various age ranges, identifying the corresponding therapeutic treatments. Use of the CIRS scale to classify patients as FIT and UNFIT was seen to be still somewhat limited.


Introduction. Chronic Lymphocytic Leukaemia (CLL)
is the most common form of leukaemia in the Western world, with an incidence that in Italy is estimated at between 5.0 and 5.5 cases per 100,000 men and between 3.5 and 4.0 cases every 100,000 women; 1,2,4 only the populations of Australia, United States and Ireland are characterised by a higher incidence. 7,9,10 The risk of a diagnosis of CLL increases significantly with age: the estimated incidence for the population over 70 years of age is approximately 20 cases every 100,000. 3,5 The prevalence data available for Italy, calculated on the basis of mean survival, are discordant: some sources talk about 20,000-22,000 cases, 3,8 others about approximately 12,000 cases. 6 The idea of performing a census of patients with CLL originates, on the one hand, from the need to estimate the number of individuals with the disease using an alternative approach and, on the other, from the need to collect information of use for understanding the way this disease is treated in the different parts of our country.
The census is in itself an ambitious aim that by definition has to tackle the many difficulties implicit to data collection. The method proposed, which is based on the quantification of the number of patients visited in each facility, represents a compromise that does not aim to provide exact results, rather to offer realistic estimates of the phenomena that may influence the choice of therapy concerning the treatment of CLL in Italy. Since the survey was conducted in almost 70% of the facilities defined originally, the authors believe that the data obtained and corresponding projections, albeit with the limits that will be discussed, constitute a very useful contribution to improving current knowledge in this area.
Materials and Methods. The data was collected using a questionnaire containing questions of interest (Annex 1). These questions concerned the overall numbers and percentages of patients visited/treated, rather than specific patients. In short, the questionnaire included the following items: -The characteristics of the hospital facility -The number of specialists working in the unit -The number of specialists in the unit treating CLL -The number of patients with CLL visited in the unit over the previous 12 months -The number of patients with CLL visited in the unit over the previous month -The percentage of patients diagnosed over the previous 12 months -The percentages of patients in the various Binet stages A, B and C -The mean visit frequency for the different Binet stages -The percentages of patients in the different age ranges -The percentages of patients from other regions (migration) -The percentages of patients in the different treatment lines -The distribution of the treatments administered over the different treatment lines outside clinical studies -The therapeutic objectives in the different treatment lines -Use of the CIRS scale and calculation of the score obtained The choice of the population of facilities to be contacted focused on Haematology units (initial list of 204 centres), excluding internal medicine and general medicine facilities. This choice, which was due primarily to the limited concentration of patients with CLL in these units, most likely makes the final projections an underestimate.
In order to collect reliable data concerning the whole unit, it was essential to identify an appropriate contact person to be interviewed. Consequently, the initial screening of the questionnaire included a specific question asking whether the respondent was able to provide data for the facility as a whole. In addition, for the haematology units with a presumably higher number of patients, the authors used a list of specialists who are known to be a reference point for the treatment of CLL in their units.
The information on the number of patients with CLL visited at the facility was sought in two different ways: for the previous month and for the previous 12 months, specifying which of the two values should be considered more reliable and whether this value was based on the data available or on a personal estimate. This made it possible to correct the values provided according to the following criteria, in order to obtain a better estimate of the number of patients visited over the previous 12 months: The facilities that provided effectively available data (cases A+C) rather than personal estimates were a minority; however, these facilities are concentrated amongst those treating a higher number of patients (on average, 111 patients over the previous 12 months vs. 79 in cases B+D).
The correction was performed using an algorithm that, starting from the number of patients visited in the previous month (considered more simple to quantify in the absence of other information), projects the annual value taking into consideration the mean frequency of patient visits in each of the Binet stages (Annex 2).
It is important to point out that the correction, in most cases, did not significantly influence the 12month estimates declared during the interview and that in all produced a 7% reduction in patients.
The B cases were examined individually because they were more uncertain; when values provided and corrected values were discordant the centres were recontacted and verification interviews performed with another contact person. More generally, the same approach was taken for structures that had provided inconsistent or unrealistic values. Overall, a doublecheck was performed on approximately 30% of facilities.
The interview was conducted by telephone and only in a few cases in person and always subject to appointment; during the first contact the respondent was told about the aims of the study, emphasising the need to collect reliable data on the treatment of CLL at the facility. The interviews were conducted by specialised personnel with consolidated experience in the medical field. The survey lasted approximately 4 months, from November 2014 to February 2015.
The final projections for the universe of haematology units were performed by considering the size of the 68 centres that refused the interview by dividing the facilities into four ranges according to the number of beds.

Results.
Estimate of the number of patients. Overall, in 136 out of a total of 204 facilities identified (of which 5 did not administer treatment), the number of patients with CLL visited in the previous 12 months was 11,526 units; if this number is projected over the entire population of haematology units, we obtain an estimate of 17,044 patients (28 cases every 100,000 inhabitants).
In order to evaluate this value, it is necessary to consider the possibility that the same patient may, during the year, have been treated at more than one centre. From this point of view the information collected concerning healthcare mobility can be of help: on average, 15% of patients visited in a facility lives in a different region to that in which the unit is found. These patients are more likely to be visited also in a centre in their home region.
The projection of the diagnoses of CLL made in the previous 12 months was 2,966 patients (4.9 cases every 100,000 inhabitants), equal to approximately 17% of the whole. Newly diagnosed patients are particularly prone to double-counting, as the diagnosing facility in many cases is different to that in which the patient is subsequently treated. We therefore believe that the effective percentage and number of new diagnoses is in actual fact significantly lower.
At hub centres, which see a higher number of patients, the percentage of new diagnoses was higher.
The units contacted that treat CLL, employee an average of 8-9 specialists and residents (median =7), of whom approximately half deal actively with the condition (median = 3).
Binet stages. The specialists interviewed were asked to break down the patients visited over the previous year at the facility according to Binet stage: in all (figure 1), approximately half (53%) of the patients were in stage A (seen on average every 5-6 months), 29% were in stage B (seen on average every 3-4 months) and 18% were in stage C (seen on average every 1-2 months).
It is likely that, compared to the total population of patients with CLL, the weight of patients in stage A is underestimated because these patients, apart from the diagnosis phase, are seen even less frequently than once a year, in which case they were not counted. For the same reason, the visit frequency of patients in stage A is probably overestimated compared to the total number of patients in stage A.
Age ranges. Age is one of the most influential variables in patient evaluation and the choice of treatment: of the three age ranges (up to 65 years, 66-75 years and over 75 years) used in the questionnaire, the most common was the intermediate range (41%, approximately 7000 patients). The patients who for the treatment of CLL are usually defined "young" (up to 65 years) account for approximately one quarter (figure 2).  Test. It was estimated that approximately half of the patients visited over the previous 12 months had each of the different cytogenetic tests (figure 3), with a slightly higher frequency only for 17p deletion, which is more selective for therapy. GENE IGVH, with FISH +12, was the cytogenetic test less frequently performed.

10
Of the various flow cytometry tests, the conventional CD38 test was used, whereas use of the CD49d test was still somewhat limited.
Treatment lines. 40% of patients with CLL seen during the previous 12 months was awaiting initial treatment  The doctors interviewed estimated that, on average, approximately 20% of W&W patients will start treatment in the next 12 months.
Treatments administered. We then collected the distributions of the pharmacological treatment administered in the unit to patients visited in the previous year, making a distinction between the three different patient age ranges: up to 65 years, 66-75 years, over 75 years. First-line therapy. By excluding the 8% of patients involved in clinical studies, the treatments administered as first-line therapy (figure 5) are greatly concentrated on RFC (Rituximab -Fludarabine -Cyclophosphamide) in young patients (73%) and on R-Benda (Rituximab -Bendamustine) in patients aged between 66 and 75 years (63%). For older patients, with poorer general conditions, treatments with Chlorambucil (in combination with Rituximab, but also in monotherapy and in combination with other drugs) are very common; the percentage of patients treated with R-Benda drops in this case to 29% (as shown in the rituximab + bendamustine chart in figure 5).
The therapeutic objective of first-line treatment is, with a similar frequency (41%), increase in OS or increase in PFS / CR, whereas symptom control/palliative care is already defined as a treatment objective in almost 20% of cases. Second-and subsequent line therapy. Again excluding the 8% of patients involved in clinical studies, in treatment lines subsequent to the first (figure 6) the most commonly used treatment overall is R-Benda, especially in patients up to 75 years.
Chlorambucil is very commonly used in patients over 65, in combination with Rituximab or often in other regimens in elderly patients.  In these treatment lines (figure 7), the most common therapeutic objective is an increase in PFS/ CR (53% of cases), which is higher than in first-line treatment (26%). As expected, OS is the primary endpoint for a mere 21% of patients. In this subset of patients, the category of patients for whom symptom control/ palliative care is the therapeutic objective is obviously higher (26%) than was reported for patients receiving first-line therapy. Cirs Scale and Score Calculation. The CIRS score classifying patients as FIT and UNFIT (where a score of ≤ 6 is classified as fit and > 6 as unfit) is partially used in the clinical practice of the units: 40% of respondents replied that this happens rarely, another 38% replied "quite often". Only a minority, primarily in the centres treating more patients, make frequent (13%) or continuous (8%) use of the scoring system ( figure 8).
This result may have a close relationship with the choice of treatment, as all the most recent treatment indications refer to the CIRS score.

Discussion and Conclusion.
The vast coverage obtained amongst haematology facilities allows us to make comparisons with existing epidemiological data on CLL. Considering that the estimates obtained from this study present two limits of an opposite nature, the double counting of certain patients (overestimation) and the exclusion of medical units (underestimation), the number of patients obtained was seen to be in line with the incidence and prevalence of the disease most commonly reported in literature for this country. The data obtained is in agreement with that found in literature in terms of disease stage, age, lines of therapy and treatment options. However, the data concerning the biological tests used are surprising. Currently, the most important parameter, also in the light of the novel treatments available, is the study of 17p/p53 deletion, followed by 11q mutation and testing. CD38 and Zap-70 tests would now appear to be rather obsolete, given their unclear clinical value, whereas CD49d has shown a greater prognostic value. These data reflect the methodology and technology used, highlighting that in some centres flow cytometry (which is now essential for diagnosing CLL) is more commonly used, whilst FISH and PCR testing is less common. The same can be said for CIRS score calculation: it is probably used more by centres conducting a greater number of clinical trials or with a higher number of patients. It is essential for use of the CIRS scale to become part of clinical practice precisely in the light of the new medicinal products now available.
This study is the first large-scale survey of its kind to be conducted in Italy and provides us with incidence data and a picture of the clinical and biological characteristics of patients with CLL. In addition, its added value is the possibility to obtain an evaluation that reflects today's therapeutic orientation and allows us to hypothesise how the current treatment scenario will change with the advent of new medicinal products.

S5
From now on we will be talking about patients with Chronic Lymphocytic Leukaemia. The questions that follow concern the work of the whole department/ unit in which you work and not just your personal work. Do you think you have adequate information about the treatment of Chronic Lymphocytic Leukaemia in the unit as a whole?