AMSTERDAM – Clinicians can use a simple tool to assess comorbidity in patients who are being considered for treatment of chronic lymphocytic leukemia in a clinical trial or routine practice, according to a retrospective analysis of data from a randomized controlled trial.

The cumulative illness rating scale (CIRS) proved to be helpful in identifying thresholds at which comorbidity might have an impact on mortality or the tolerability of chemotherapy with or without rituximab (Rituxan) in the CLL8 trial, Dr. Valentin Goede of the German Chronic Lymphocytic Leukemia Study Group (GCLLSG) reported. Patients with a high number of comorbidities or a high score in any category of comorbidity had worse outcomes in the trial.

"Chronic lymphocytic leukemia is a disease that primarily occurs in elderly patients," Dr. Goede said in an interview at the annual congress of the European Hematology Association. A relatively high proportion of patients is thus likely to have cardiovascular, metabolic, respiratory, gastrointestinal, genitourinary, neurologic, and other health problems "that might impact on the tolerability of the treatments."

How to assess the impact is not standardized, said Dr. Goede of the University of Cologne, Germany, and this compounds an existing predicament in that elderly patients and those with comorbid diseases tend to be excluded from clinical trials. A key problem is how to assess a patient’s "fitness" for treatment objectively, rather than on the basis of a physician’s clinical experience or opinion, he said.

"What the GCLLSG has done over the past 10 years is integrate a comorbidity score into our trials, [so that we can] try to have a more objective measurement of the fitness of the patients," Dr. Goede said.

For the current study, the GCLLSG looked at data from the phase III CLL8 trial to see whether the CIRS could be used to stratify patients based on their levels of comorbidity. The CLL8 trial compared fludarabine, cyclophosphamide, and rituximab (FCR) chemotherapy with FC. A total of 817 patients were involved, and while the trial excluded "unfit" patients, "it was not clear if this was the correct cutoff," said Dr. Goede.

Although the comorbidity burden was low in the population of patients studied, CIRS results correlated with overall survival and the frequency of grade 3-4 side effects, he reported. Items graded in the CIRS include cardiovascular, respiratory, gastrointestinal, genitourinary, and other categories. Each was scored individually and a total score obtained.

Patients with four or more comorbidities or with higher CIRS scores in any category (2-4) were more likely to have worse survival than were those with fewer comorbidities or lower CIRS category scores (hazard ratio, 1.9; P less than .01). If they met both of these criteria, their hazard ratio rose to 7.2 (P less than .001), when compared with patients who had fewer comorbidities or scores of 0-1 in the categories rated.

Though the number of comorbidities did not predict grade 3 or 4 toxicity, a higher score of 2-4 in any category did (P less than .001).

The CIRS might thus be a useful tool for future clinical trials "to stratify between patients who are more susceptible and less susceptible to toxicity, and [those who are susceptible] to early mortality," Dr. Goede suggested. Refinements and further validation are of course required.

"For normal practitioners, I would say that if you do this assessment in a patient, ... based on these data, you are able to anticipate the outcome of the patient a little bit," he added. This may also help identify comorbidities that need to be addressed in order to improve patients’ overall outcomes.

Compared with the complete geriatric assessment (CGA), the CIRS might be an easier tool to use in the clinic, he added. As a geriatrician and hematologist, Dr. Goede has experience with using both in clinical practice.

"The complete geriatric assessment is quite complicated, it’s very resource intensive, and it takes a very long time to do [it], and you really need people who are skilled to do the assessment," he noted. While the CGA takes about 1 hour, the CIRS takes around 1-3 minutes.

"The main problem with the CGA, as with other simple tools, is that they do not really tell you what the cutoff is," Dr. Goede added. The CGA is also not validated specifically in the hematologic setting.

The CIRS could potentially be useful in assessing comorbidity in other hematologic malignancies or tumor types, but there are no data. The tool itself is, however, validated in hospitalized patients (J. Am. Geriatr. Soc. 2008;56:1926-31) and in the family practice setting (J. Clin. Epidemiol. 2005;58:603-8).

The CLL8 study was sponsored by the GCLLSG with funding from Roche. Dr. Goede has received honoraria from Roche, Mundi-Pharma, and Gilead.

Simple Comorbidity Scale Judged Useful in Leukemia