Introduction

Lymphoedema develops after surgery in 30% patients. In a prospective, multi-centre UK study, we used a validated perometer arm measurement to determine 1) the factors predicting lymphoedema development and 2) the effect of a lymphoedema diagnosis on QoL and survival.

Methods

Participants (n = 1100) undergoing axillary clearance at 9 centres underwent arm volume measurements pre and post-surgery by perometry, and completed FACT-B+4 and Lymphoedema checklist questionnaires. Relative arm volume increase (RAVI) of >10% diagnosed lymphoedema. Predictors of lymphoedema development were determined using logistic regression, and changes in QoL were assessed using Generalised Estimating Equation (GEE) analyses.

Results

Median patient age was 56 (range 22 to 90) years; 78% received radiotherapy and 65% chemotherapy. Lymphoedema was detected in 21.4% of women by perometry and 24.4% underwent sleeve application by 24 months.

Initial decreases in QoL scores post-surgery were greater in patients with lymphoedema and took longer to return to baseline values (FACT-B p=0.014, TOI p=0.008, ARM subscale p<0.001).

RAVI at 1 month (p<0.001), BMI in three categories (≤25, >25-≤30 and >30, p=0.05), ER status (p=0.05) and number of positive nodes (p<0.001) were used to develop a novel scoring model (AUROC 0.80) to predict lymphoedema. Out of 826 patients used for the model, 75% of patients had low scores (≤1) at 1 month and 11.6% developed lymphoedema by 24 months, whereas 20% who scored moderate risk (1.5-2) had a 31.3% risk and 5% who scored high risk (2.5-4) and had a 66.7% risk. Using the model scores, 75% of patients could be reassured regarding their lymphoedema risk.

Local recurrence was 1.7%. One hundred and twenty-nine patients had died of breast cancer (n=88) or had distant recurrence (n=41) across the study.

Lymphoedema (RAVI≥10%) by 9 months was an independent predictor of post 9 months distant disease-free cancer survival (Table).

Distant Disease Free Survival (DDFS) after Lymphoedema diagnosis

  Single variable Multivariable 
Variable (between 3 and 9 months) Hazard ratio (HR) (95% CI) p-value HR (95% CI) p-value 
RAVI ≥ 10% Yes: 105 vs No 955 2.30 (1.39-3.81) 0.001 1.73 (1.01-2.96) 0.047 
ER Status Negative: 171 953 2.94 (1.92-4.49) <0.001 1.734 (1.06-2.83) 0.028 
No. positive nodes* 969 1.07 (1.05-1.09) <0.001 1.05 (1.03-1.08) <0.001 
Adjuvant CT Yes: 654 vs No 963 0.83 (0.54-1.27) 0.39 0.60 (0.37-0.964) 0.035 
Hormone treatment Yes: 808 vs No 964 0.45 (0.28-0.72) 0.001 
Tumour size* 959 1.02 (1.01-1.024) <0.001 1.02 (1.01-1.03) <0.001 
Grade           
0/1: 67 961 1 (-) <0.001 1 (-) <0.001 
2: 428   3.96 (0.54-29.324)   2.69 (0.36-20.33)   
3: 436   11.59 (1.61-83.54)   6.91 (0.94-50.90)   
Undiff: 30   17.58 (2.12-146.184)   12.54 (1.43-109.96)   
  Single variable Multivariable 
Variable (between 3 and 9 months) Hazard ratio (HR) (95% CI) p-value HR (95% CI) p-value 
RAVI ≥ 10% Yes: 105 vs No 955 2.30 (1.39-3.81) 0.001 1.73 (1.01-2.96) 0.047 
ER Status Negative: 171 953 2.94 (1.92-4.49) <0.001 1.734 (1.06-2.83) 0.028 
No. positive nodes* 969 1.07 (1.05-1.09) <0.001 1.05 (1.03-1.08) <0.001 
Adjuvant CT Yes: 654 vs No 963 0.83 (0.54-1.27) 0.39 0.60 (0.37-0.964) 0.035 
Hormone treatment Yes: 808 vs No 964 0.45 (0.28-0.72) 0.001 
Tumour size* 959 1.02 (1.01-1.024) <0.001 1.02 (1.01-1.03) <0.001 
Grade           
0/1: 67 961 1 (-) <0.001 1 (-) <0.001 
2: 428   3.96 (0.54-29.324)   2.69 (0.36-20.33)   
3: 436   11.59 (1.61-83.54)   6.91 (0.94-50.90)   
Undiff: 30   17.58 (2.12-146.184)   12.54 (1.43-109.96)   

*Per unit increase

Conclusions

Lymphoedema is associated with lasting quality of life deficits and early distant relapse.

Women at low risk of lymphoedema (75%) can be reassured using the scoring model. Early Arm measurements at 1 month post-surgery a useful measurement time to determine lymphoedema risk and enable patients to be reassured (75%) or plan for lymphoedema monitoring.

Citation Format: Bundred NJ, Foden P, Riches K, Morris J, Evans A, Todd C, Bramley M, Skene A, Purushotham A, Keeley V. Prediction model for lymphoedema, and effect of Lymphoedema diagnosis on quality of life (QoL) and distant recurrence from breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-07-01.