Thrombosis as a Treatment Complication in Hodgkin Lymphoma Patients

A Comprehensive Analysis of Three Prospective Randomized German Hodgkin Study Group (GHSG) Trials

S. Borchmann; H. Müller; I. Hude; M. Fuchs; P. Borchmann; A. Engert


Ann Oncol. 2019;30(8):1329-1334. 

In This Article


Here, we studied the incidence, type, risk factors and timing of thrombotic events in prospectively treated HL patients. The total incidence of 3.3% is at the lower end of other, comparable studies.[13–20] The reason for this might be the prospective nature of our cohort, not over-representing patients with symptoms or complications which might be a bias of retrospective study designs. Other single-center, hospital-record based studies might have estimated a higher incidence due to a larger proportion of high-risk patients.[20,23] Additionally, we included a slightly younger and healthier trial population of HL patients compared with the general HL patient population. However, considering all of these aspects, the incidence of any thrombotic event of 7.3% in advanced-stage patients is high compared with the general cancer population and in line with the high-risk group of cancer patients as defined by Khorana et al..[12] The incidence of any thrombotic events in early-stage patients was low, with 0.7% in early-favorable stage and 1.3% in early-unfavorable stage HL and independent of the treatment regimen used.

DVT of the arm was the most common venous event followed by DVT of the leg and PE. The high frequency of DVT of the arm likely reflects the high frequency of local axillary and periclavicular lymph node affection in HL and the high number of thrombotic events directly linked to tumor compression and indwelling central venous catheters or port-a-caths observed in the present analysis.[2,21] Additionally, arm veins are almost universally used for chemotherapy administration, which is associated with venous inflammation and irritation. This could potentially result in thrombotic events.[24]

The major risk factor for thrombotic events in our study was advanced stage disease. Additionally, most events occurred in the earlier stages of chemotherapy. Taken together, this is in line with previous studies that found high-tumor burden and more aggressive lymphoma to be associated with a higher risk of thrombotic events.[20,25] Additionally, patients who received BEACOPP chemotherapy every 14 days were at higher risk of thrombotic events compared with patients receiving BEACOPP at longer intervals, albeit at a higher dose (BEACOPPesc), suggesting dose-density as a potential risk factor. This is in line with the fact that chemotherapy itself is a major risk factor for thrombotic events in cancer patients.[5,6]

In advanced-stage HL patients, only age and smoking were associated with a higher risk of thrombotic events in our cohort. Our analysis of age as a risk factor for thrombosis is somewhat hampered by the fact that patients older than 60 years were excluded from the HD14 and HD15 trial due to the high toxicity of BEACOPPesc in this age group. However, in addition to age being a risk factor for thrombosis when modeled continuously, we observed a tendency towards increased risk already in patients between the ages of 50 and 60 years. Taken together, it is likely that patients older than 60 years and diagnosed with advanced stage HL are at particularly high risk for thrombosis and warrant increased attention and potential prophylaxis. Of note, the widely used Khorana score[12] was not prognostic in our study, reflecting marked differences between HL patients and the general cancer population with regard to certain patient characteristics, such as age.

In a subset analysis of female patients receiving oral contraception, we observed a trend towards an increased risk of thrombosis in those patients. Low patient and event numbers in this sub-cohort did not permit conclusive statistical testing. However, the observed trend is in line with the well-recognized increased risk of thrombotic events in women using oral contraception in the general, non-cancer population.[26]

One of the main strengths of our study is the large, prospectively treated cohort of HL patients comprising three randomized, multi-center clinical trials.

The conclusions concerning the treatment types in advanced stage HL rely on a randomized comparison, giving us fairly unbiased estimates. Additionally, the incidences of thrombotic events in early- and advanced stages of HL are based on high-quality data.

An important limitation of our study is the lack of detailed information on central venous catheters and prophylactic anticoagulation in all patients. While we were able to retrieve these data in patients with reported thrombotic events from the available documentation, this was not possible for all patients.

To conclude, we present the largest study of thrombotic events in prospectively treated HL patients to date. We give a high-quality estimate of the incidence of thrombotic events in HL patients in early- and advanced stages and highlight particularly vulnerable patient groups, such as older patients, patients with advanced stage disease or smokers. Interestingly, in advanced stage patients alone, the risk of any thrombotic event in our cohort was comparable to the high-risk group in the general ambulatory cancer population, as defined by the Khorana score.[12] However, considering the surprisingly low prognostic value of established risk factors in our HL cohort, further collaborative efforts and interventional trials are needed to define a patient population in which prophylactic anticoagulation can be clearly recommended. Currently, prophylactic anticoagulation should be considered in selected high-risk advanced stage HL patients after discussing the risks and benefits with the patient.