The Prognostic Impact of Comorbidity in Peripheral T-Cell Lymphomas: a Swedish Lymphoma Registry Study

Introduction

Comorbidity has repeatedly been established to have a negative impact on the survival of patients with cancer, but it has not been investigated in peripheral T-cell lymphomas (PTCLs). The impact of comorbidity depends on the prognosis of the malignancy, and because PTCLs are aggressive diseases that generally respond less well to treatment compared to aggressive B-cell lymphomas, we aimed to investigate how comorbidities affect outcome in a large population-based cohort of PTCL patients.

Materials and methods

Through the Swedish Lymphoma Registry all patients diagnosed with T-cell lymphomas between 2000 and 2009 were identified. After informed consent, data on comorbidity at the time of diagnosis was collected from the patient records and classified according to the Charlson Comorbidity Index (CCI). All cases of diabetes mellitus were awarded 1 CCI point because complications to diabetes mellitus could not reliably be evaluated from available medical records. Non-melanoma skin tumors were not recorded as malignancies.

Results

In total 755 patients with non-cutaneous, non-leukemic PTCL were identified. Comorbidity data was available for 676 of the 755 patients (90%). Four-hundred twenty-five (56%) of the patients had no co-morbidity, while 267 (35%) had CCI scores between 1 and 7 points. The 5-year overall survival (OS) rates for patients with CCI 0, 1 and ≥2 were 41%, 23% and 12 % respectively. There was no significant difference in survival of patients with CCI 2 or higher. Individual CCI factors with prognostic impact on OS in univariable analysis were previous malignancy (Hazard ratio [HR] 1.68, p=0.001), active malignancy (HR 2.04, p<0.001), renal failure (HR 5.56, p<0.001), liver disease (HR 2.66, p=0.018), peptic ulcer (HR 1.82, p=0.020), diabetes mellitus (HR 1.85, p<0.001), myocardial infarction (HR 1.62, p=0.001), congestive heart failure (HR 2.17, p<0.001), stroke (HR 1.74, p=0.005) and autoimmune disease (HR 1.50, p=0.034). The cause of death was known in 451 patients and lymphoma related death was more common among patients with CCI 0 (82%) than with CCI 1 (69%) or CCI ≥2 (65%) (p=0.008 and p=0.001 respectively).

In multivariable analysis, comorbidity was an independent negative factor for both OS and progression-free survival (PFS). Among patients receiving chemotherapy (N=551) the presence of comorbidity together with age, male gender, stage III-IV, LDH>ULN, extranodal involvement >1, WHO PS >1, gastrointestinal involvement and NK/T-cell, nasal type lymphoma were independent risk factors for OS with HR 1.46 for CCI 1 (p=0.004) and HR 1.76 for CCI ≥2 (p<0.001). The impact of comorbidity was age dependent as patients above the median age of the cohort (>67 years) had inferior OS (HR 1.64, p=0.006) and PFS (HR1.51, p=0.027) only if they presented with CCI ≥2.

Up-front autologous stem cell transplantation (auto-SCT) was planned in 154 out of 755 patients. Comorbidity data was available in 149, and auto-SCT as part of first-line therapy was performed in 101 (68%) of these patients. In this group comorbidities were uncommon but diabetes mellitus was most frequent (n=10) with an impact on OS (HR 3.31, p=0.01, n=10) and PFS (HR 2.86, p=0.003) and was an independent risk factor in multivariable analysis (data not shown). Using the CCI, instead of individual comorbidities, independent prognostic factors for OS were age (HR 1.028, p=0.017), skin involvement (HR 2.29, p=0.016) and CCI (CCI 1 HR 1.94, p=0.019 and CCI ≥2 HR 3.67, p=0.004). For PFS, skin involvement (HR 3.70, p<0.001) and CCI (CCI 1 HR 1.80, p=0.039 and CCI ≥2 HR 3.56, p=0.012) but not age, were independent adverse prognostic factors, while ALKneg ALCL histologic subtype was associated with favorable PFS (HR 0.49, p=0.028). There was no difference in the proportion of patients with planned auto-SCT actually performed between the CCI groups, although there were only 5 patients with CCI ≥2 in this group.

Conclusions

Our data demonstrates that even though PTCLs are aggressive diseases with a substantial proportion of cases responding poorly to chemotherapy, comorbidity has an important impact on outcome, not the least in patients planned for up-front auto-SCT. Despite medical advances in the treatment of several of the conditions included in the score, the CCI still seems to be a useful tool to capture comorbidity.