ESTRO 2022

Session Item

Sunday

May 08

10:30 - 11:30

Room D3

Head & neck

Co-Chair: Jørgen Johansen, Denmark;

Chair: , Germany

Session Code: 2220

Session Type: Proffered Papers

Track: Clinical

10:50 - 11:00

Positive predictive value of post radiotherapy FDG PET-CT is affected by treatment and HPV-status

Suyun Zhou, United Kingdom

Presentation Number: OC-0433

Abstract

Abstract Title:

Positive predictive value of post radiotherapy FDG PET-CT is affected by treatment and HPV-status

Authors:

Suyun Zhou¹, Christopher Chan², Hesham Dyab³, Robert Rulach¹, Fraser Hendry³, Carole Maxfield³, Mary Frances Hendry³, Allan James¹, Carolynn Lamb¹, Derek Grose¹, Christina Wilson¹, Stefano Schipani¹, Yee Cheng Lau⁴, Claire Paterson¹

¹Beatson West of Scotland Cancer Centre, Clinical Oncology, Glasgow, United Kingdom; ²University Hospital Ayr, General Medicine, Glasgow, United Kingdom; ³NHS Greater Glasgow and Clyde, Radiology, Glasgow, United Kingdom; ⁴Royal Alexandra Hospital, Medicine, Glasgow, United Kingdom

Show Affiliations

Purpose or Objective

The high negative predictive value (NPV) of post-chemoradiation (CRT) PET-CT to guide the need for neck dissection (ND) in HNSCC was shown by the PET-Neck study. The positive predictive value (PPV) of this investigation remains under scrutiny with increasing evidence that it is affected by several factors.

The aim of this study was to assess the PPV of PET-CT stratified by treatment modality and tumour HPV-status.

Material and Methods

Patients with node-positive HNSCC treated with radical ((C)RT) between January 2013 and July 2019 were identified from the PET-CT database. PET-CT responses were classified as complete (CR), incomplete (ICR) or equivocal (EQR) in keeping with definitions from the PET-Neck study, by radionucleotide radiologists. Patient demographics and clinical outcomes were obtained from electronic patient records.

A true positive PET-CT was defined as a pathologically positive ND, or relapsed disease in the neck/distally during follow-up.

Results

434 HNSCC patients were identified, patient and tumour characteristics are detailed in Table 1. Median time from end of RT to PET scan was 93 (IQR 97-101) days. Median follow-up was 42.7 (IQR 30.8-60.0) months.

343 (79%) patients received CRT and 91 (21%) had RT alone.

In the CRT group (n=343), 212 (61.8%) achieved a CR; 131 (38.2%) less than CR with 39 cases of recurrence, thus the PPV of PET-CT for CRT was 29.8% (95% CI 22.1-38.4), see Table 2.

In the RT-only group (n=91), 51 (56%) achieved a CR; 40 (44%) a less than CR with 22 cases of recurrence. Hence the PPV of PET-CT was 55.0 (95% CI 38.5-70.7) -significantly higher than the PPV of the CRT group, p=0.005.

When stratifying for HPV status, 310 (71.4%) patients had HPV-positive and 124 (28.6%) HPV-unrelated HNSCC.

Of the 310 HPV-positive group, 190 (61.3%) patients achieved a CR; 120 (38.7%) less than CR with 30 cases of recurrence. Thus the PPV of PET-CT was 25% (95CI% 17.4-33.7).

In the HPV-unrelated HNSCC group (n=124), 73 (58.9%) achieved a CR; 51 (41.1%) less than CR with 31 cases of recurrence. The PPV of PET-CT was 60.8% (95%CI 46.1-74.2). This is significantly higher than the PPV of the HPV-positive group, p<0.001.

Of the 120 HPV-positive patients who achieved a less than CR; 76 (24.5%) achieved EQR with 10 cases of relapse; and 44 (14.2%) ICRs, of which 20 patients had recurrence. PPV for PET-CT was 13.2% (95%CI 6.5-22.9) and 45.5% (95%CI 30.4-61.2) for EQR and ICR groups, respectively. Within the HPV-positive cohort, the PPV was significantly lower for the EQR compared to the ICR group, p<0.001.

PPV for HPV-positive group was significantly lower than that of the HPV-unrelated group for both EQR and ICR groups, p=0.013 and p=0.038, respectively.

Conclusion

The PPV for PET-CT at 12 weeks is lower for patients who have HPV-positive disease and for those who received CRT, especially amongst those who achieved an EQR. Additional biomarkers are needed to help select patients for further surveillance or ND.