Contributions
Abstract: PB1996
Type: Publication Only
Background
The International Prognostic Score (IPS), which was introduced and validated 2 decades ago by Hasenclever and Diehl for advanced stages, is the most widely used risk stratification tool in HL. However, it has been criticised for its low validity in early bulky disease and modest predictive power in unfavourable disease. In the present era, with more than 80% of patients (pts) cured by first-line treatment, we need an easy-to-apply and accurate index to help deciding risk-adapted approaches. The German Hodgkin Study Group (GHSG) proposed an alternative score, which has not yet gained broad acceptance.
Aims
We retrospectively compared the influence of the IPS and the GHSG score on overall survival (OS), event-free survival (EFS) and relapse rate (RR) in an uniformly treated cohort in our center.
Methods
From Jan/97 to Dec/16, 292 consecutive pts with classical HL (80% nodular sclerosis) were treated with Stanford V. Median age was 32 (15-72), 35% had advanced stage disease and 39% had bulky mediastinal disease (all of whom received radiotherapy). The IPS was low-risk in 56%, intermediate-risk in 33% and high-risk in 11%. By the GHSG classification, 18% were early favourable, 33% early unfavourable and 49% advanced; the differences in distribution were significant (p<0.01).
Results
Overall response rate was 95%. With a median follow up of 9.5 years, median OS is not yet reached. OS at 12 years varied from 89 to 47% in IPS groups, and from 87 to 54% in GHSG groups (p<0.05 for all comparisons). EFS varied from 74 to 34% for IPS and from 75 to 58% for GHSG (p<0.05). RR was similar in low and intermediate-risk IPS (23 vs 26%) and was 35% in high-risk IPS; in GHSG groups it was 15%, 22% and 31% respectively. The IPS was a better discriminator than the GHSG score for OS (AUC 0.69 vs 0.63, p=0.005) and EFS (AUC 0.61 vs 0.60, p=0.006).
Conclusion
The distribution of pts by IPS and GHSG score showed significant differences in our cohort, with an inversion of low and high-risk groups. This discrepancy is attributable to the high proportion (39%) of bulky mediastinal disease in our series, which is also the probable cause of a paradoxically high RR in the low-risk IPS group. The prognostic performance of the GHSG score was not superior to the IPS, reinforcing the need to evolve beyond clinical scores and to incorporate novel biomarkers into clinical practice, in order to accurately identify pts wirh the poorest risk.
Session topic: 17. Hodgkin lymphoma – Clinical
Keyword(s): Hodgkin's Lymphoma, lymphoma, prognosis
Abstract: PB1996
Type: Publication Only
Background
The International Prognostic Score (IPS), which was introduced and validated 2 decades ago by Hasenclever and Diehl for advanced stages, is the most widely used risk stratification tool in HL. However, it has been criticised for its low validity in early bulky disease and modest predictive power in unfavourable disease. In the present era, with more than 80% of patients (pts) cured by first-line treatment, we need an easy-to-apply and accurate index to help deciding risk-adapted approaches. The German Hodgkin Study Group (GHSG) proposed an alternative score, which has not yet gained broad acceptance.
Aims
We retrospectively compared the influence of the IPS and the GHSG score on overall survival (OS), event-free survival (EFS) and relapse rate (RR) in an uniformly treated cohort in our center.
Methods
From Jan/97 to Dec/16, 292 consecutive pts with classical HL (80% nodular sclerosis) were treated with Stanford V. Median age was 32 (15-72), 35% had advanced stage disease and 39% had bulky mediastinal disease (all of whom received radiotherapy). The IPS was low-risk in 56%, intermediate-risk in 33% and high-risk in 11%. By the GHSG classification, 18% were early favourable, 33% early unfavourable and 49% advanced; the differences in distribution were significant (p<0.01).
Results
Overall response rate was 95%. With a median follow up of 9.5 years, median OS is not yet reached. OS at 12 years varied from 89 to 47% in IPS groups, and from 87 to 54% in GHSG groups (p<0.05 for all comparisons). EFS varied from 74 to 34% for IPS and from 75 to 58% for GHSG (p<0.05). RR was similar in low and intermediate-risk IPS (23 vs 26%) and was 35% in high-risk IPS; in GHSG groups it was 15%, 22% and 31% respectively. The IPS was a better discriminator than the GHSG score for OS (AUC 0.69 vs 0.63, p=0.005) and EFS (AUC 0.61 vs 0.60, p=0.006).
Conclusion
The distribution of pts by IPS and GHSG score showed significant differences in our cohort, with an inversion of low and high-risk groups. This discrepancy is attributable to the high proportion (39%) of bulky mediastinal disease in our series, which is also the probable cause of a paradoxically high RR in the low-risk IPS group. The prognostic performance of the GHSG score was not superior to the IPS, reinforcing the need to evolve beyond clinical scores and to incorporate novel biomarkers into clinical practice, in order to accurately identify pts wirh the poorest risk.
Session topic: 17. Hodgkin lymphoma – Clinical
Keyword(s): Hodgkin's Lymphoma, lymphoma, prognosis