Poster Display session 2 Poster Display session

280P - Defining a SUV decrease cut-off in PET/CT response monitoring after one cycle of preoperative breast cancer chemotherapy (ID 3911)

Presentation Number
280P
Lecture Time
12:00 - 12:00
Speakers
  • Marcin Kubeczko (Gliwice, Poland)
Session Name
Poster Display session 2
Location
Poster Area (Hall 4), Fira Gran Via, Barcelona, Spain
Date
29.09.2019
Time
12:00 - 13:00

Abstract

Background

A decrease in tumor metabolic activity have been shown to be an early indicator of treatment effectiveness for breast cancer.

Methods

We reviewed the records of patients enrolled to the prospective study of prediction of preoperative chemotherapy response (MILESTONE-BreastPred). In the whole group PET/CT was used as an initial method of staging and as response assessment after 1st cycle of chemotherapy. We analysed data of 425 patients, 19.8% TNBC, 12.2% nonluminal HER2-positive and 68% luminal cancers. 88% of patients were stage II or III. pCR rate was 24.8% (ypT0/is ypN0). 31.1% of patients showed progressive disease during follow-up. The aim of the study was to assess the clinical utility of various cut-offs of SUVmax decrease to predict both pCR and disease relapse.

Results

We analysed SUVmax decrease cut-off of 20%, 30%, 40%, 50 and 60%. For pCR prediction, the sensititivity was above 50% for cut-offs 20%-40%, while specificity exceeded 50% for cut-off values 40-60%. In the whole range of cut-offs, there was highly statistically significant difference in pCR rate between low and high SUVmax decrease (p < 0.001). Cut-off of 40% was chosen, providing optimal trade-off between sensitivity and specificity with positive likelihood ratio 1.89, negative likelihood ratio 0.49, PPV 38.4% and PPV 86.0%. Prediction of disease relapse based on 1st cycle SUVmax decrease was more challenging, with sensitivity above 50% for cut-offs 20-30% and specificity above 50% for cut-offs 40-60%. For cut-off 40% the difference in disease relapse between groups was closest to the limit of statistical significance (p = 0.085) and provided PLR 0.81, NLR 1.17, PPV 26.6% and NPV 65.6%.

Conclusions

Prediction of pCR by 1st cycle SUVmax decrease was feasible, with good balance of specificity and sensitivity for cut-off 40%. This cut-off provides the most prominent, although not significant in this analysis, association with the risk of disease relapse. The study was supported by the Polish National Center of Research and Development MILESTONE project–Molecular diagnostics and imaging in individualized therapy for breast, thyroid and prostate cancer, grant no.STRATEGMED2/267398/4/NCBR/2015.

Legal entity responsible for the study

The authors.

Funding

The study was supported by the Polish National Center of Research and Development MILESTONE project – Molecular diagnostics and imaging in individualized therapy for breast, thyroid and prostate cancer, grant no. STRATEGMED 2/267398/4/NCBR/2015.

Disclosure

All authors have declared no conflicts of interest.

Collapse