The use of artificial intelligence in reconstructive surgery for head and neck cancer: a systematic review

OBJECTIVES

The popularity of artificial intelligence (AI) in head and neck cancer (HNC) management is increasing, but postoperative complications remain prevalent and are the main factor that impact prognosis after surgery. Hence, recent studies aim to assess new AI models to evaluate their ability to predict free flap complications more effectively than traditional algorithms. This systematic review aims to summarize current evidence on the utilization of AI models to predict complications following reconstructive surgery for HNC.

METHODS

A combination of MeSH terms and keywords was used to cover the following three subjects: "HNC," "artificial intelligence," and "free flap or reconstructive surgery." The electronic literature search was performed in three relevant databases: Medline (Ovid), Embase (Ovid), and Cochrane. Quality appraisal of the included study was conducted using the TRIPOD Statement.

RESULTS

The review included a total of 5 manuscripts (n = 5) for a total of 7524 patients. Across studies, the highest area under the receiver operating characteristic (AUROC) value achieved was 0.824 by the Auto-WEKA model. However, only 20% of reported AUROCs exceeded 0.70. One study concluded that most AI models were comparable or inferior in performance to conventional logistic regression. The highest predictors of complications were flap type, smoking status, tumour location, and age.

DISCUSSION

Some models showed promising results. Predictors identified across studies were different than those found in existing literature, showing the added value of AI models. However, the algorithms showed inconsistent results, underlying the need for better-powered studies with larger databases before clinical implementation.

Stereotactic Radiation for Ultra-Central Non-Small Cell Lung Cancer: A Safety and Efficacy Trial (SUNSET)

INTRODUCTION

The use of stereotactic body radiotherapy (SBRT) for tumors in close proximity to the central mediastinal structures has been associated with a high risk of toxicity. This study (BLINDED FOR REVIEW) aimed to determine the maximally tolerated dose (MTD) of SBRT for ultra-central (UC) non-small cell lung carcinoma (NSCLC), using a time-to-event continual reassessment methodology (TITE-CRM).

METHODS

Patients with T1-3N0M0 (≤ 6 cm) NSCLC were eligible. The MTD was defined as the dose of radiotherapy associated with a ≤ 30% rate of grade (G) 3-5 pre-specified treatment-related toxicity occurring within 2 years of treatment. The starting dose level was 60 Gy in 8 daily fractions. The dose-maximum hotspot was limited to 120% and within the planning tumor volume (PTV); tumors with endobronchial invasion were excluded. This primary analysis occurred two years after completion of accrual.

RESULTS

Between March 2018 and April 2021, 30 patients were enrolled at 5 institutions. The median age was 73 years (range: 65-87) and 17 (57%) were female. PTV was abutting proximal bronchial tree in 19 (63%), esophagus 5 (17%), pulmonary vein 1 (3.3%) and pulmonary artery 14 (47%). All patients received 60 Gy in 8 fractions. The median follow-up was 37 months (range: 8.9-51). Two patients (6.7%) experienced G3-5 adverse events related to treatment: 1 patient with G3 dyspnea and 1 G5 pneumonia; the latter had CT findings consistent with a background of interstitial lung disease. Three-year overall survival was 72.5% (95% confidence interval [CI]: 52.3-85.3%), progression-free survival 66.1% (95% CI: 46.1-80.2%), local control 89.6% (95% CI: 71.2-96.5%), regional control 96.4% (95% CI: 77.2-99.5%) and distant control 85.9% (95% CI: 66.7-94.5%). Quality of life scores declined numerically over time, but the decreases were not clinically or statistically significant.

CONCLUSIONS

60 Gy in 8 fractions, planned and delivered with only a moderate hotspot, has a favorable adverse event rate within the pre-specified acceptability criteria, and results in excellent control for UC tumors.

Stereotactic Radiation Therapy in Early Non-Small Cell Lung Cancer and Interstitial Lung Disease: A Nonrandomized Clinical Trial

Importance

Patients with interstitial lung disease (ILD) and early-stage non-small cell lung cancer (NSCLC) have been reported to be at high risk of toxic effects after stereotactic ablative radiotherapy (SABR), but for many patients, there are limited alternative treatment options.

Objective

To prospectively assess the benefits and toxic effects of SABR in this patient population.

Design, Setting, and Participants

This prospective cohort study was conducted at 6 academic radiation oncology institutions, 5 in Canada and 1 in Scotland, with accrual between March 7, 2019, and January 12, 2022. Patients aged 18 years or older with fibrotic ILD and a diagnosis of T1-2N0 NSCLC who were not candidates for surgical resection were enrolled.

Intervention

Patients were treated with SABR to a dose of 50 Gy in 5 fractions every other day.

Main Outcomes and Measures

The study prespecified that SABR would be considered worthwhile if median overall survival-the primary end point-was longer than 1 year, with a grade 3 to 4 risk of toxic effects less than 35% and a grade 5 risk of toxic effects less than 15%. Secondary end points included toxic effects, progression-free survival (PFS), local control (LC), quality-of-life outcomes, and changes in pulmonary function. Intention-to-treat analysis was conducted.

Results

Thirty-nine patients enrolled and received SABR. Median age was 78 (IQR, 67-83) years and 59% (n = 23) were male. At baseline, 70% (26 of 37) of patients reported dyspnea, median forced expiratory volume in first second of expiration was 80% (IQR, 66%-90%) predicted, median forced vital capacity was 84% (IQR, 69%-94%) predicted, and median diffusion capacity of the lung for carbon monoxide was 49% (IQR, 38%-61%) predicted. Median follow-up was 19 (IQR, 14-25) months. Overall survival at 1 year was 79% (95%, CI 62%-89%; P < .001 vs the unacceptable rate), and median overall survival was 25 months (95% CI, 14 months to not reached). Median PFS was 19 months (95% CI, 13-28 months), and 2-year LC was 92% (95% CI, 69%-98%). Adverse event rates (highest grade per patient) were grade 1 to 2: n = 12 (31%), grade 3: n = 4 (10%), grade 4: n = 0, and grade 5: n = 3 (7.7%, all due to respiratory deterioration).

Conclusions and Relevance

In this trial, use of SABR in patients with fibrotic ILD met the prespecified acceptability thresholds for both toxicity and efficacy, supporting the use of SABR for curative-intent treatment after a careful discussion of risks and benefits.

Trial Registration

ClinicalTrials.gov Identifier: NCT03485378.

Single vs. multiple fraction non-inferiority trial of stereotactic ablative radiotherapy for the comprehensive treatment of oligo-metastases/progression: SIMPLIFY-SABR-COMET

BACKGROUND

Radiotherapy delivery regimens can vary between a single fraction (SF) and multiple fractions (MF) given daily for up to several weeks depending on the location of the cancer or metastases. With limited evidence comparing fractionation regimens for oligometastases, there is support to explore toxicity levels to nearby organs at risk as a primary outcome while using SF and MF stereotactic ablative radiotherapy (SABR) as well as explore differences in patient-reported quality of life and experience.

METHODS

This study will randomize 598 patients in a 1:1 ratio between the standard arm (MF SABR) and the experimental arm (SF SABR). This trial is designed as two randomized controlled trials within one patient population for resource efficiency. The primary objective of the first randomization is to determine if SF SABR is non-inferior to MF SABR, with respect to healthcare provider (HCP)-reported grade 3-5 adverse events (AEs) that are related to SABR. Primary endpoint is toxicity while secondary endpoints include lesional control rate (LCR), and progression-free survival (PFS). The second randomization (BC Cancer sites only) will allocate participants to either complete quality of life (QoL) questionnaires only; or QoL questionnaires and a symptom-specific survey with symptom-guided HCP intervention. The primary objective of the second randomization is to determine if radiation-related symptom questionnaire-guided HCP intervention results in improved reported QoL as measured by the EuroQoL-5-dimensions-5levels (EQ-5D-5L) instrument. The primary endpoint is patient-reported QoL and secondary endpoints include: persistence/resolution of symptom reporting, QoL, intervention cost effectiveness, resource utilization, and overall survival.

DISCUSSION

This study will compare SF and MF SABR in the treatment of oligometastases and oligoprogression to determine if there is non-inferior toxicity for SF SABR in selected participants with 1-5 oligometastatic lesions. This study will also compare patient-reported QoL between participants who receive radiation-related symptom-guided HCP intervention and those who complete questionnaires alone.

TRIAL REGISTRATION

Clinicaltrials.gov identifier: NCT05784428. Date of Registration: 23 March 2023.

Gut microbiome predicts gastrointestinal toxicity outcomes from chemoradiation therapy in patients with head and neck squamous cell carcinoma

OBJECTIVES

Chemoradiation (CRT) in patients with locally advanced head and neck squamous cell cancer (HNSCC) is associated with significant toxicities, including mucositis. The gut microbiome represents an emerging hallmark of cancer and a potentially important biomarker for CRT-related adverse events. This prospective study investigated the association between the gut microbiome composition and CRT-related toxicities in patients with HNSCC, including mucositis.

MATERIALS AND METHODS

Stool samples from patients diagnosed with locally advanced HNSCC were prospectively collected prior to CRT initiation and analyzed using shotgun metagenomic sequencing to evaluate gut microbiome composition at baseline. Concurrently, clinicopathologic data, survival outcomes and the incidence and grading of CRT-emergent adverse events were documented in all patients.

RESULTS

A total of 52 patients were included, of whom 47 had baseline stool samples available for metagenomic analysis. Median age was 62, 83 % patients were men and 54 % had stage III-IV disease. All patients developed CRT-induced mucositis, including 42 % with severe events (i.e. CTCAE v5.0 grade ≥ 3) and 25 % who required enteral feeding. With a median follow-up of 26.5 months, patients with severe mucositis had shorter overall survival (HR = 3.3, 95 %CI 1.0-10.6; p = 0.02) and numerically shorter progression-free survival (HR = 2.8, 95 %CI, 0.8-9.6; p = 0.09). The gut microbiome beta-diversity of patients with severe mucositis differed from patients with grades 1-2 mucositis (p = 0.04), with enrichment in Mediterraneibacter (Ruminococcus gnavus) and Clostridiaceae family members, including Hungatella hathewayi. Grade 1-2 mucositis was associated with enrichment in Eubacterium rectale, Alistipes putredinis and Ruminococcaceae family members. Similar bacterial profiles were observed in patients who required enteral feeding.

CONCLUSION

Patients who developed severe mucositis had decreased survival and enrichment in specific bacteria associated with mucosal inflammation. Interestingly, these same bacteria have been linked to immune checkpoint inhibitor resistance.

Canadian Consensus Recommendations for the Management of Operable Stage II/III Non-Small-Cell Lung Cancer: Results of a Modified Delphi Process

The treatment paradigm for patients with stage II/III non-small-cell lung cancer (NSCLC) is rapidly evolving. We performed a modified Delphi process culminating at the Early-stage Lung cancer International eXpert Retreat (ELIXR23) meeting held in Montreal, Canada, in June 2023. Participants included medical and radiation oncologists, thoracic surgeons and pathologists from across Quebec. Statements relating to diagnosis and treatment paradigms in the preoperative, operative and postoperative time periods were generated and modified until all held a high level of consensus. These statements are aimed to help guide clinicians involved in the treatment of patients with stage II/III NSCLC.

Predictors of prolonged treatment time intervals in oral cavity cancer

OBJECTIVES

Delays in treatment time intervals have been associated with overall survival in oral cavity squamous cell carcinoma (OCSCC). The aim of this study was to identify bottlenecks leading to prolonged treatment intervals.

MATERIAL AND METHODS

A retrospective analysis was conducted using a cohort of OCSCC patients who underwent surgery and adjuvant radiation therapy. The endpoints of interest were prolonged treatment intervals. Multivariable logistic regression was used to adjust for patient and tumour characteristics.

RESULTS

Median diagnosis-to-treatment interval (DTI) and surgery to initiation of postoperative radiation therapy interval (S-PORT) were 39 days (IQR 30-54) and 64 days (IQR 54-66), respectively. Prolonged DTI was associated with older age, worse Charlson Comorbidity index scores and worse T stages. Patients with prolonged DTI had longer times to preoperative imaging reports (25 vs 9 days; P < 0.01). Time to preoperative pathology did not differ. Prolonged S-PORT was associated with longer times to pathology report (28 vs 18 days; P < 0.01), to maxillofacial consult (38 vs 15 days; P < 0.01) and to maxillofacial approval of radiation (50 vs 28 days; P < 0.01). In patients requiring medical oncology consults, those with prolonged S-PORT had longer waiting times until consultation (58 vs 38 days; P = 0.02). Multivariate analysis showed independent predictors of prolonged DTI: time to preoperative imaging; and prolonged S-PORT: time to pathology report, time to maxillofacial consult, and time to medical oncology consult.

CONCLUSIONS

Strategies targeting these organizational bottlenecks may be effective for shortening treatment time intervals, hence representing potential opportunities for improving oncological outcomes in OCSCC patients.

Oncologic Significance of Therapeutic Delays in Patients With Oral Cavity Cancer

Importance

Oral cavity cancer often requires multidisciplinary management, subjecting patients to complex therapeutic trajectories. Prolonged treatment intervals in oral cavity cancer have been associated with poor oncological outcomes, but there has yet to be a study investigating treatment times in Canada.

Objective

To report treatment delays for patients with oral cavity cancer in Canada and evaluate the outcomes of treatment delays on overall survival.

Design, Setting, and Participants

This multicenter cohort study was performed at 8 Canadian academic centers from 2005 to 2019. Participants were patients with oral cavity cancer who underwent surgery and adjuvant radiation therapy. Analysis was performed in January 2023.

Main Outcomes and Measures

Treatment intervals evaluated were surgery to initiation of postoperative radiation therapy interval (S-PORT) and radiation therapy interval (RTI). The exposure variables were prolonged intervals, respectively defined as index S-PORT greater than 42 days and RTI greater than 46 days. Patient demographics, Charlson Comorbidity Index, smoking status, alcohol status, and cancer staging were also considered. Univariate (log rank and Kaplan-Meier) and multivariate (Cox regression) analyses were performed to determine associations with overall survival (OS).

Results

Overall, 1368 patients were included; median (IQR) age at diagnosis was 61 (54-70) years, and 896 (65%) were men. Median (IQR) S-PORT was 56 (46-68) days, with 1093 (80%) patients waiting greater than 42 days, and median (IQR) RTI was 43 (41-47) days, with 353 (26%) patients having treatment time interval greater than 46 days. There were variations in treatment time intervals between institutions for S-PORT (institution with longest vs shortest median S-PORT, 64 days vs 48 days; η2 = 0.023) and RTI (institution with longest vs shortest median RTI, 44 days vs 40 days; η2 = 0.022). Median follow-up was 34 months. The 3-year OS was 68%. In univariate analysis, patients with prolonged S-PORT had worse survival at 3 years (66% vs 77%; odds ratio 1.75; 95% CI, 1.27-2.42), whereas prolonged RTI (67% vs 69%; odds ratio 1.06; 95% CI, 0.81-1.38) was not associated with OS. Other factors associated with OS were age, Charlson Comorbidity Index, alcohol status, T category, N category, and institution. In the multivariate model, prolonged S-PORT remained independently associated with OS (hazard ratio, 1.39; 95% CI, 1.07-1.80).

Conclusions and Relevance

In this multicenter cohort study of patients with oral cavity cancer requiring multimodal therapy, initiation of radiation therapy within 42 days from surgery was associated with improved survival. However, in Canada, only a minority completed S-PORT within the recommended time, whereas most had an appropriate RTI. An interinstitution variation existed in terms of treatment time intervals. Institutions should aim to identify reasons for delays in their respective centers, and efforts and resources should be directed toward achieving timely completion of S-PORT.

The protective role of postoperative radiation therapy in low and intermediate grade major salivary gland malignancies: A study of the Canadian Head and Neck Collaborative Research Initiative

BACKGROUND

The objective of this study was to examine the utility of postoperative radiation for low and intermediate grade cancers of the parotid and submandibular glands.

METHODS

The authors conducted a retrospective, Canadian-led, international, multi-institutional analysis of a patient cohort with low or intermediate grade salivary gland cancer of the parotid or submandibular gland who were treated from 2010 until 2020 with or without postoperative radiation therapy. A multivariable, marginal Cox proportional hazards regression analysis was performed to quantify the association between locoregional recurrence (LRR) and receipt of postoperative radiation therapy while accounting for patient-level factors and the clustering of patients by institution.

RESULTS

In total, 621 patients across 14 tertiary care centers were included in the study; of these, 309 patients (49.8%) received postoperative radiation therapy. Tumor histologies included 182 (29.3%) acinic cell carcinomas, 312 (50.2%) mucoepidermoid carcinomas, and 137 (20.5%) other low or intermediate grade primary salivary gland carcinomas. Kaplan-Meier LRR-free survival at 10 years was 89.0% (95% confidence interval [CI], 84.9%-93.3%). In multivariable Cox regression analysis, postoperative radiation therapy was independently associated with a lower hazard of LRR (adjusted hazard ratio, 0.53; 95% CI, 0.29-0.97). The multivariable model estimated that the marginal probability of LRR within 10 years was 15.4% without radiation and 8.8% with radiation. The number needed to treat was 16 patients (95% CI, 14-18 patients). Radiation therapy had no benefit in patients who had early stage, low-grade salivary gland cancer without evidence of nodal disease and negative margins.

CONCLUSIONS

Postoperative radiation therapy may reduce LLR in some low and intermediate grade salivary gland cancers with adverse features, but it had no benefit in patients who had early stage, low-grade salivary gland cancer with negative margins.

Agreement among Radiation Oncologists and Radiologist in Contouring GTV on Planning and Mid-Treatment MRI for Locally Advanced Oropharynx Cancer

PURPOSE/OBJECTIVE(S)

Over the 7-week radiotherapy course, patients with head and neck cancers undergo significant anatomical changes, including weight loss and tumor shrinkage (with complete response at mid-treatment in as high as 50% of patients). The current standard of care is to maintain the same plan for the entire treatment duration, unless major dosimetric deviations are suspected. The use of MRI for treatment adaptation has the advantage of increased soft tissue contrast and is being integrated into several clinical practices with the recent implementation of the MR-Linac technology. In this interim secondary analysis of a phase 2, randomized trial, we assessed inter-observer variability (IOV) in the delineation of GTV at planning and mid-treatment in locally advanced oropharynx cancer (OPC).

MATERIALS/METHODS

Three expert radiation oncologists (RO) and 1 expert radiologist, with a minimum of 5 years' experience in MRI contouring, blinded to each other's volumes, were provided with planning and mid-treatment MRI (T1 gadolinium, T2-weighted and DWI) of 9 patients with locally advanced OPC. Primary gross tumor volume (GTV) was independently contoured by all 3 ROs and the radiologist. Contouring agreement was analyzed using the dice similarity coefficient (DSC) index and the average Hausdorff distance.

RESULTS

Nine cases of locally advanced OPC were included. Fifty percent of the patients were stage 3 (AJCC 8th edition) and all were p16+ OPC. There was moderate IOV in the delineation of GTV on planning MRI as evidenced by mean DSC index and Hausdorff distance of 0.75 (range 0.53 - 0.87) and 0.30 cm (max 2.14 cm). At mid-treatment, the GTV IOV had a mean DSC index and Hausdorff distance of 0.30 (range 0.19 - 0.76) and 0.55 cm (max 2.83), respectively. The contours of the radiologist compared to RO shown a mean DSC index and Hausdorff distance on planning MRI 0.63 (range 0.53 - 0.69) and 0.41 cm (max 2.14 cm), and on mid-treatment MRI of 0.36 (range 0.19 - 0.59) and 0.68 cm (max 1.08 cm).

CONCLUSION

Although this study demonstrates the presence of moderate IOV between 3 experienced head and neck ROs and one radiologist in an academic institution for the delineation of GTV on planning MRI, we observed a higher IOV between the 4 experts on mid-treatment MRI. We also experienced a higher IOV between the radiologist and RO on planning and mid-treatment images. These results show high degree of GTV volume variability on mid-treatment MRI. Therefore, peer review of contours is important and guidelines for MR adaptive tumor delineation are needed.

Stereotactic Boost and Short-Course Radiotherapy for p16-Associated Oropharynx Cancer (SHORT-OPC): First Planned Interim Safety Analysis from a Randomized Phase II Trial

PURPOSE/OBJECTIVE(S)

There is a need for safe treatment de-intensification in p16+ oropharynx cancer (OPC). The standard of care (SOC) radiotherapy (RT) regimen is cumbersome and associated with high toxicity. Stereotactic radiotherapy (SBRT) and multimodality image guidance is an opportunity to precisely target the gross tumor while safely reducing elective irradiation dose. We aim to assess the safety and efficacy of a short course RT for p16+ OPC, consisting of an SBRT boost to the gross tumor volume (GTV) followed by de-escalated elective irradiation.

MATERIALS/METHODS

In this randomized phase II trial, patients with p16-positive, stage I-II OPSCC with primary tumor <30 cc (8th Ed AJCC) are planned with combined CT, MRI and FDG-PET, and randomized to 1) SBRT boost (14 Gy in 2 fractions) to the GTV followed with de-escalated RT (+/- Cisplatin) to a dose of 40 Gy in 20 fractions, or 2) SOC RT (+/- Cisplatin) to a dose of 70 Gy in 33 fractions to the GTV and 59.4-54Gy (or equivalent) to the intermediate-to-low dose elective region. Patients are stratified by stage (I vs. II) and use of chemotherapy. The primary endpoint of the trial is locoregional control at 2 years, powered for a sample size of 100 patients. A Bayesian adaptive design includes 2 planned safety interim analysis using grade ≥ 3 subacute toxicities >40% as a stopping criterion, and 1 planned futility analysis. Acute adverse events (AE) are defined as those occurring ≤ 60 days from RT, subacute AE between 60-180 days after RT, and late AE >180 days from RT. This is the first planned toxicity analysis.

RESULTS

Twenty-one patients were randomly assigned and eligible (11 in SOC and 10 in experimental arm). Median age was 69 years (range 49-84); 29% and 71% had stage T1 and T2, while 10%, 85% and 1 patient had N0, N1 and N2 disease, respectively. RT alone and chemoradiation was administered in 67% and 33% of patients, respectively. At a median follow-up of 11 months (range 1.7-17.6), there was 1 local recurrence at the primary tumor site in the SOC arm (at 10 month) and no recurrence in the experimental arm. All enrolled patients remain alive at the time of analysis. There was a 54.5% rate of grade 3 acute AE in the SOC arm and 30.0% rate of grade 3 acute AE in the experimental arm. More specifically, 1, 5 (45%), 2 (18%), and 2 (18%) versus 0, 1, 1 and 1 patient developed acute grade 3 dysphagia, mucositis, pain and dermatitis in the SOC and experimental arm, respectively. There was no acute grade 4 or 5 toxicity. There was no grade ≥ 3 subacute toxicity or late toxicity in both arms.

CONCLUSION

This primary safety analysis showed that SBRT boost followed by a short course of de-escalated elective irradiation in p16+ OPC has limited early toxicity and meets criteria for study continuation.

Dynamic Prediction of Toxicities in Head and Neck Cancer Radiotherapy by 3D Convolutional Neural Network Using Daily Cone-Beam CTs

PURPOSE/OBJECTIVE(S)

Radiotherapy (RT) is essential in head and neck cancer (HNC) treatments, but often causes significant toxicity. Different machine learning models have shown promise in predicting RT-induced toxicity, but none have yet integrated the fluctuating anatomical changes. By integrating daily cone-beam CTs (CBCT) allowing sequential anatomical views, our aim is to build a dynamic predictive model for three major HNC RT toxicities: reactive feeding tube placement, hospitalization and radionecrosis (RN).

MATERIALS/METHODS

292 HNC cases treated with curative RT between 2017 and 2019 at our institution were retrospectively analyzed for clinical and radiological data. VoxelMorph, a deep deformable registration model, integrated the daily anatomical deformations between each CBCT and the planning CT, then converted them to Jacobian determinant matrix (Jf). Resnet, a convolutional neural network with multiple layers was trained using a 5-fold cross validation to integrate both radiological and clinical data. Each toxicity was classified as a binary decision using the cross-entropy loss to account for a class imbalance. Its predictive performance was compared to the baseline model using only clinical data.

RESULTS

The cohort included 78% men and 22% women, with a median age of 63 years (range 35-84). Primary cancer sites were 46% oropharynx, 19% larynx, 14% oral cavity, 7.5% nasopharynx, 5% hypopharynx, 4% unknown primary and 5% others; and stage ranged between Tx-4b N0 and 3b M0 (AJCC 8th Ed). Induction chemotherapy, concurrent chemotherapy, and adjuvant RT was used in 9%, 57% and 20% of patients, respectively. The incidence of feeding tube, hospitalization and RN was 19.9%, 7.2%, and 3.8%, respectively. Integrating Jf from the 10th RT CBCT showed better accuracy for each toxicity prediction: feeding tube (69.1% > 57.2%), hospitalization (75.3% > 63.1%) and RN (85.8% > 75.7%). Integrating both the raw CBCT and Jf improved hospitalization prediction (79.0% > 73.6%). Substituting Jf for the raw CBCT improved the prediction for RN (79.7% > 74.7%) and hospitalization (73.6% > 64.4%). For feeding tube, predictive performance of the Jf model trained against deformations showed a positive correlation between its performance and the RT received (r2 > 0.9) with increasing RT fractions, with a maximum accuracy of 83.1% at the 25th fraction. No such correlation was found for RN or hospitalization prediction.

CONCLUSION

To our knowledge, this is the first study showing promising results to predict HNC RT toxicities using daily per-treatment CBCT. Next steps involve integrating both the radiomic and the dosimetric inputs to build a more powerful model. This could expand to predict therapeutic outcomes and, ultimately, could guide decisions in individualized RT.

Assessment of Precision Irradiation in Early Non-Small Cell Lung Cancer and Interstitial Lung Disease (ASPIRE-ILD): Primary Analysis of a Phase II Trial

PURPOSE/OBJECTIVE(S)

The use of stereotactic ablative radiotherapy (SABR) in patients with fibrotic interstitial lung disease (ILD) has been associated with an increased risk of toxicity, but patients with ILD and lung cancer may have no other options for curative-intent treatment. The goal of the ASPIRE-ILD trial was to assess the benefits and toxicities of SABR in patients with fibrotic ILD.

MATERIALS/METHODS

We enrolled patients with fibrotic ILD and a diagnosis of T1-2N0 NSCLC who were not candidates for surgery. All patients were centrally reviewed prior to enrollment to confirm the presence and subtype of ILD. After stratification by the ILD-GAP score (a measure of ILD severity and prognosis), patients were treated with SABR to a dose of 50 Gy in 5 fractions EOD (BED = 100 Gy10), with a built-in de-escalation protocol in case of unacceptable toxicity. The primary endpoint was overall survival (OS), powered to distinguish 1-year OS >70% vs. an unacceptable rate of ≤50%. Secondary endpoints included toxicity (CTC-AE version 4.0), progression-free survival (PFS), local control (LC), patient-reported outcomes (FACT-L quality of life and cough severity), and changes in pulmonary function tests (PFTs). The study pre-specified that SABR would be considered worthwhile if median OS was >1 year, with a grade 3-4 toxicity risk <35% and a grade 5 toxicity risk <15%. Target accrual was 39 treated patients.

RESULTS

Thirty-nine patients were enrolled and treated with SABR between March 2019 and January 2022, all to a dose of 50 Gy in 5 fractions, at 5 institutions in Canada and 1 in Scotland. Median age was 78 years (interquartile range: 67-83), 59% were male, and 92% had a history of smoking (median 43 pack-years). At baseline, 70% reported dyspnea, median FEV1 was 80% predicted and median DLCO was 49% predicted. ILD-GAP scores were as follows: ≤2 (i.e., best ILD status): n = 14; 3-5: n = 23; ≥6 (i.e., worst ILD status): n = 2. Median follow-up was 19 months. OS at 1-year was 78.9% (p<0.001 by binomial test vs. the unacceptable rate). Median OS was 25 months, median PFS was 19 months, and 2-year LC was 92%. AE rates (possibly, probably or definitely related) were as follows (highest grade per patient): grade 1-2: n = 12 (31%); grade 3: n = 4 (10%); grade 4; n = 0; grade 5 n = 3 (7.7%, all due to respiratory deterioration). AE rates did not differ by ILD-GAP category or ILD subtype. FACT-L scores trended downward over time (p = 0.07), and cough severity scale scores worsened over time (p = 0.02). Comparing last-available PFTs with baseline, DLCO declined (median: -4%; p = 0.046), FVC trended downward (median: -2.5%; p = 0.11), and FEV1 remained stable (median change: 0%).

CONCLUSION

The use of SABR in patients with ILD met the pre-specified acceptability thresholds for both toxicity and efficacy, supporting the use of SABR for curative-intent treatment after a careful discussion of risks and benefits. Further studies exploring pharmacologic options to reduce toxicity may be beneficial in this population. ().

A competing risk analysis of the patterns and risk factors of recurrence in early-stage non-small cell lung cancer treated with stereotactic ablative radiotherapy

INTRODUCTION

To assess patterns of recurrence after stereotactic ablative radiotherapy (SABR) in patient ineligible to surgery with early-stage non-small cell lung cancer (ES-NSCLC), report survival and treatment after first recurrence.

METHODS

We performed a retrospective analysis on 1068 patients with ES-NSCLC and 1143 lesions. Between group differences were estimated using competing risk analysis and cause-specific hazard ratios were calculated. Overall survival (OS) after first recurrence was calculated.

RESULTS

Median follow-up was 37.6 months. Univariate analysis demonstrated that ultra-central location was associated with higher risk of regional recurrence (RR) and distant metastasis (DM) (p = 0.004 and 0.01). Central lesions were associated with higher risk of local recurrence (LR) and RR (p < 0.001). Ultra-central lesions were associated with shorter OS (p = 0.002) compared to peripheral lesions. In multivariate analysis, central location was the only factor associated with increased LR and RR risks (p = 0.016 and 0.005). Median OS after first recurrence was 14.8 months. There was no difference in OS after first recurrence between ultra-central, central, and peripheral lesions (p = 0.83). Patients who received a second SABR course had an OS of 51.3 months, compared to 19.5 months with systemic therapy and 8.1 months with supportive care (p < 0.0001).

DISCUSSION

The main prognostic factor for LR and RR risks was central location. Ultra-central and central tumors might benefit from treatment intensification strategies such as dose escalation and/or addition of systemic therapy to improve radiotherapy outcomes. After a first recurrence post SABR, patients with contralateral lung recurrences and those who were eligible to receive a second course of SABR had improved OS.

Long-term outcomes of modern multidisciplinary management of sinonasal cancers: The M. D. Anderson experience

PURPOSE

To report long-term outcomes of modern radiotherapy for sinonasal cancers.

METHODS AND MATERIALS

A retrospective analysis of patients with sinonasal tumors treated with intensity-modulated radiotherapy or proton therapy. Multivariate analysis was used to determine predictive variables of progression free survival (PFS) and overall survival (OS).

RESULTS

Three hundred and eleven patients were included, with median follow-up of 75 months. The most common histologies were squamous cell (42%), adenoid cystic (15%), and sinonasal undifferentiated carcinoma (15%). Induction chemotherapy was administered to 47% of patients; 68% had adjuvant radiotherapy. Ten-year local control, regional control, distant metastasis free survival, PFS, and overall survival rates were 73%, 88%, 47%, 32%, and 51%, respectively. Age, non-nasal cavity tumor site, T3-4 stage, neck dissection, and radiation dose were predictive of PFS, while age, non-nasal cavity tumor site, T3-4 stage, positive margins, neck dissection, and use of neoadjuvant chemotherapy were predictive of OS. There was a 13% rate of late grade ≥3 toxicities.

CONCLUSION

This cohort of patients with sinonasal cancer treated with modern radiotherapy demonstrates favorable disease control rate and acceptable toxicity profile.

Longitudinal diffusion and volumetric kinetics of head and neck cancer magnetic resonance on a 1.5T MR-Linear accelerator hybrid system: A prospective R-IDEAL Stage 2a imaging biomarker characterization/ pre-qualification study

Objectives

We aim to characterize the serial quantitative apparent diffusion coefficient (ADC) changes of the target disease volume using diffusion-weighted imaging (DWI) acquired weekly during radiation therapy (RT) on a 1.5T MR-Linac and correlate these changes with tumor response and oncologic outcomes for head and neck squamous cell carcinoma (HNSCC) patients as part of a programmatic R-IDEAL biomarker characterization effort.

Methods

Thirty patients with pathologically confirmed HNSCC who received curative-intent RT at the University of Texas MD Anderson Cancer Center, were included in this prospective study. Baseline and weekly Magnetic resonance imaging (MRI) (weeks 1-6) were obtained, and various ADC parameters (mean, 5 th , 10 th , 20 th , 30 th , 40 th , 50 th , 60 th , 70 th , 80 th , 90 th and 95 th percentile) were extracted from the target regions of interest (ROIs). Baseline and weekly ADC parameters were correlated with response during RT, loco-regional control, and the development of recurrence using the Mann-Whitney U test. The Wilcoxon signed-rank test was used to compare the weekly ADC versus baseline values. Weekly volumetric changes (Δvolume) for each ROI were correlated with ΔADC using Spearman's Rho test. Recursive partitioning analysis (RPA) was performed to identify the optimal ΔADC threshold associated with different oncologic outcomes.

Results

There was an overall significant rise in all ADC parameters during different time points of RT compared to baseline values for both gross primary disease volume (GTV-P) and gross nodal disease volumes (GTV-N). The increased ADC values for GTV-P were statistically significant only for primary tumors achieving complete remission (CR) during RT. RPA identified GTV-P ΔADC 5 th percentile >13% at the 3 rd week of RT as the most significant parameter associated with CR for primary tumor during RT (p <0.001). Baseline ADC parameters for GTV-P and GTV-N didn't significantly correlate with response to RT or other oncologic outcomes. There was a significant decrease in residual volume of both GTV-P & GTV-N throughout the course of RT. Additionally, a significant negative correlation between mean ΔADC and Δvolume for GTV-P at the 3 rd and 4 th week of RT was detected (r = -0.39, p = 0.044 & r = -0.45, p = 0.019, respectively).

Conclusion

Assessment of ADC kinetics at regular intervals throughout RT seems to be correlated with RT response. Further studies with larger cohorts and multi-institutional data are needed for validation of ΔADC as a model for prediction of response to RT.

Incidence and predictive factors of retropharyngeal lymph node metastases in patients with oropharyngeal cancer undergoing multimodality treatment planning imaging

BACKGROUND

We investigated the incidence and predictive factors of retropharyngeal lymph node (RPLN) metastases in patients with oropharyngeal cancer (OPC) undergoing multimodality treatment planning imaging before radiotherapy.

METHODS

Consecutive patients with OPC treated with curative-intent radiotherapy from 2017 to 2019 were retrospectively analyzed. Treatment planning comprised contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and fluorodeoxyglucose-positron emission tomography (FDG-PET) unless contraindicated.

RESULTS

Of 300 patients, 66 (22%) had radiological evidence of RPLN involvement on planning images, compared to 17 (6%) on diagnostic CT alone. On multivariate analysis, RPLN involvement was statistically (p < 0.05) associated with tonsil, soft palate, and posterior pharyngeal wall primaries, and with disease extension to the soft palate or vallecula.

CONCLUSIONS

Multimodality treatment planning imaging reveals a high rate of RPLN metastases from OPC compared to diagnostic CT alone. Patients with tonsil, soft palate, or posterior pharyngeal wall primaries or disease extending to the soft palate or vallecula appear at higher risk.

Artificial intelligence to predict outcomes of head and neck radiotherapy

Head and neck radiotherapy induces important toxicity, and its efficacy and tolerance vary widely across patients. Advancements in radiotherapy delivery techniques, along with the increased quality and frequency of image guidance, offer a unique opportunity to individualize radiotherapy based on imaging biomarkers, with the aim of improving radiation efficacy while reducing its toxicity. Various artificial intelligence models integrating clinical data and radiomics have shown encouraging results for toxicity and cancer control outcomes prediction in head and neck cancer radiotherapy. Clinical implementation of these models could lead to individualized risk-based therapeutic decision making, but the reliability of the current studies is limited. Understanding, validating and expanding these models to larger multi-institutional data sets and testing them in the context of clinical trials is needed to ensure safe clinical implementation. This review summarizes the current state of the art of machine learning models for prediction of head and neck cancer radiotherapy outcomes.

University of Montreal's Clinician-Investigator Program: A 10-Year Descriptive Evaluation

PURPOSE

Clinician-investigators have an important role in the development and implantation of new therapies and treatment modalities; however, there have been several reports highlighting a pending shortage in the clinician-investigators' workforce. In Canada, the Royal College has promoted the development of clinician-investigators programs (CIP) to facilitate the training of these individuals. There is currently a paucity of data regarding the outcomes of such programs. This study aims to identify the strengths and areas of improvement of the Montreal University CIP.  Methods: An internet-based 51-question survey was distributed to all the alumni from the University of Montreal CIP. Participation was voluntary and no incentives were provided. The response rate was 64%.  Results: Among respondents, 50% (n=16) had completed their clinical residency and all CIP requirements. The majority of these individuals (63%) had become independent investigators and had secured provincial and national funding. Satisfaction of the respondents was high regarding the overall program (85%), the research skills developed during the CIP (84%) and the financial support obtained during the program (72%). The satisfaction rate regarding career planning was lower (63%).  Conclusion: This survey demonstrates that, while indicators are favorable, some areas still require improvement. Several steps to improve the CIP have been identified; notably, the transition from the CIP to early independent career has been identified as critical in the development of clinician-investigators and steps have been taken to improve this progression.

What is the future of treatment de-escalation for HPV-positive oropharyngeal cancer? A review of ongoing clinical trials

Background

Human papillomavirus (HPV)-associated oropharyngeal squamous cell carcinoma (OPSCC) has increased in incidence in recent decades. With higher cure rates in younger populations, long-term survivors may live with acute- and long-term toxicity, leading to increased interest in de-escalation treatment strategies for HPV-related OPSCC. Herein, we have examined the current landscape of clinical trials in this context.

Methods

A review of active clinical trials related to de-escalation of HPV-associated OPSCC treatment was performed using the clinicaltrials.gov database from inception to January 2022. A search using the key words "oropharyngeal cancer" and "HPV" was completed. Three investigators independently reviewed each trial, with any discrepancies settled by a fourth. Data collected from each study included study phase, study design, primary, and secondary endpoints, and de-escalation treatment strategies. A final 24 articles were selected for full text review.

Results

Many trials (n=19, 79%) were non-randomized, and most studies employed a phase II design (n=14, 58%). Only 13% (n=3) were randomized trials, and 8% (n=2) included a phase III component. The most frequent primary endpoint was progression-free survival (PFS) (n=9, 37.5%). With regards to the identified de-escalation strategies, all the studies (n=24) had at least one component assessing changes in RT dose/fractionation and/or a reduction in RT volumes. A smaller percentage of trials assessed surgical interventions (n=9, 37.5%) and/or changes in systemic therapy (n=8, 33.3%).

Conclusion

A small number of randomized trials are underway, and a transition to more randomized phase III trials in the future will be important to change clinical practice.

Oligometastases: Emerging Evidence

The hypothesis that ablative therapies (such as surgery or radiation) can be used to cure patients with a limited number of metastases was influential in changing practice. Early assertions of efficacy were based on observational studies, often without control groups, showing better-than-expected outcomes. However, in the past decade, new evidence from randomized trials has emerged, which in some cases have affirmed old hypotheses, but in other cases have raised new questions. In this review, we discuss the challenges in defining oligometastatic disease, summarize the randomized evidence evaluating metastasis-directed therapy in patients with oligometastatic disease, provide context for the difficulty in generating randomized evidence, and examine ongoing phase III studies.

Automated Detection of Brain Metastases on T1-Weighted MRI Using a Convolutional Neural Network: Impact of Volume Aware Loss and Sampling Strategy

BACKGROUND

Detection of brain metastases (BM) and segmentation for treatment planning could be optimized with machine learning methods. Convolutional neural networks (CNNs) are promising, but their trade-offs between sensitivity and precision frequently lead to missing small lesions.

HYPOTHESIS

Combining volume aware (VA) loss function and sampling strategy could improve BM detection sensitivity.

STUDY TYPE

Retrospective.

POPULATION

A total of 530 radiation oncology patients (55% women) were split into a training/validation set (433 patients/1460 BM) and an independent test set (97 patients/296 BM).

FIELD STRENGTH/SEQUENCE

1.5 T and 3 T, contrast-enhanced three-dimensional (3D) T1-weighted fast gradient echo sequences.

ASSESSMENT

Ground truth masks were based on radiotherapy treatment planning contours reviewed by experts. A U-Net inspired model was trained. Three loss functions (Dice, Dice + boundary, and VA) and two sampling methods (label and VA) were compared. Results were reported with Dice scores, volumetric error, lesion detection sensitivity, and precision. A detected voxel within the ground truth constituted a true positive.

STATISTICAL TESTS

McNemar's exact test to compare detected lesions between models. Pearson's correlation coefficient and Bland-Altman analysis to compare volume agreement between predicted and ground truth volumes. Statistical significance was set at P ≤ 0.05.

RESULTS

Combining VA loss and VA sampling performed best with an overall sensitivity of 91% and precision of 81%. For BM in the 2.5-6 mm estimated sphere diameter range, VA loss reduced false negatives by 58% and VA sampling reduced it further by 30%. In the same range, the boundary loss achieved the highest precision at 81%, but a low sensitivity (24%) and a 31% Dice loss.

DATA CONCLUSION

Considering BM size in the loss and sampling function of CNN may increase the detection sensitivity regarding small BM. Our pipeline relying on a single contrast-enhanced T1-weighted MRI sequence could reach a detection sensitivity of 91%, with an average of only 0.66 false positives per scan.

EVIDENCE LEVEL

3 TECHNICAL EFFICACY: Stage 2.

RADT-10. THE LOST METASTASES: DEEP LEARNING’S POTENTIAL IN RADIOSURGERY QUALITY ASSURANCE

Introduction

Identifying, segmenting, measuring, and following multiple brain metastases treated with radiosurgery can be time consuming and error prone. Machine learning has shown promise for automated detection and segmentation. Recently, a U-Net inspired model combining volume aware loss functions and volume aware sampling methods was trained in an industrial-academic partnership. A total of 530 clinically annotated T1 gadolinium MRIs were used. Initial validation showed a high sensitivity (91%) with an average of 0.66 false positives per MRI. The goal of the present work was to characterize those “false positives” which may represent clinically undetected metastases.

METHODS

The images used for model development were clinically annotated for radiosurgery planning. Lesions had first been identified by a radiologist, second by clinicians during tumor board review, third by the treating radiation oncologist and the treating neurosurgeon (potentially after segmentation by a trainee) and finally by fellow radiation oncologists during quality assurance rounds. Despite these multiple checks, 10 patients (2%) had brain lesions considered potential clinical misses when all “false positives” were manually reviewed by a single investigator. Further detailed review including prior and subsequent imaging was used to arbitrate the nature of these lesions.

RESULTS

Among the 10 cases, four were confirmed as undetected metastases: two lesions required subsequent radiosurgery and 2 patients died prior to further imaging. The six other lesions were adjudicated as true “false positives” (typically vascular).

CONCLUSION

The multi-tier radiosurgery workflow at our institution left very few unidentified brain metastases (0.8%). Despite this low error rate, our AI algorithm still detected two lesions that required further treatment. Future investigations will focus on potential roles of AI in simplifying and accelerating our workflow. It also remains to be established if more undetected metastases would be seen in community settings where workflows include fewer sequential imaging reviews.

Magnetic Resonance-Guided Radiation Therapy for Head and Neck Cancers

Despite the significant evolution of radiation therapy (RT) techniques in recent years, many patients with head and neck cancer still experience significant toxicities during and after treatments. The increased soft tissue contrast and functional sequences of magnetic resonance imaging (MRI) are particularly attractive in head and neck cancer and have led to the increasing development of magnetic resonance-guided RT (MRgRT). This approach refers to the inclusion of the additional information acquired from a diagnostic or planning MRI in radiation treatment planning, and now extends to online high-quality daily imaging generated by the recently developed MR-Linac. MRgRT holds numerous potentials, including enhanced baseline and planning evaluations, anatomical and functional treatment adaptation, potential for hypofractionation, and multiparametric assessment of response. This article offers a structured review of the current literature on these established and upcoming roles of MRI for patients with head and neck cancer undergoing RT.

Long-term Locoregional Control With Unilateral Radiation for AJCC-7 T1-2N2b Tonsillar Cancer

OBJECTIVES

Unilateral radiation to cervical nodes has been used as a de-escalation strategy in well-lateralized tonsil cancers. The efficacy of this approach with multiple ipsilateral nodes is not established. The study hypothesis was that unilateral radiation for American Joint Committee on Cancer (AJCC)-7 T1-2N2b tonsillar cancer results in a low rate of contralateral nodal failure.

MATERIALS AND METHODS

This study was a retrospective chart review of patients with AJCC-7 T1-2N2b tonsillar cancer from 2 academic institutions who were treated with unilateral radiation. The primary endpoint was the contralateral nodal failure rate. Locoregional control, overall survival, and the need for gastrostomy tube placement were additional endpoints.

RESULTS

The study cohort included 66 patients treated between 2005 and 2016. The median follow-up time was 80.9 months; contralateral nodal failure occurred in 2/66 (3.0%) patients at 4.1 and 20.9 months, respectively. Both patients underwent salvage treatment with long-term subsequent survival. Overall locoregional control at both 2 and 5 years was 93.9% and the median duration of control was not reached. Overall survival at 5 years was 92.4%.

CONCLUSIONS

The use of unilateral radiation for AJCC-7 T1-2N2b tonsillar cancer resulted in low rates of contralateral nodal failure. This outcome demonstrates the safety of considering unilateral radiation treatment in patients with a relatively high ipsilateral nodal burden.

Changes in Apparent Diffusion Coefficient (ADC) in Serial Weekly MRI during Radiotherapy in Patients with Head and Neck Cancer: Results from the PREDICT-HN Study

Background: The PREDICT-HN study aimed to systematically assess the kinetics of imaging MR biomarkers during head and neck radiotherapy. Methods: Patients with intact squamous cell carcinoma of the head and neck were enrolled. Pre-, during, and post-treatment MRI were obtained. Serial GTV and ADC measurements were recorded. The correlation between each feature and the GTV was calculated using Spearman’s correlation coefficient. The linear mixed model was used to evaluate the change in GTV over time. Results: A total of 41 patients completed the study. The majority (76%) had oropharyngeal cancer. A total of 36 patients had intact primary tumours that can be assessed on MRI, and 31 patients had nodal disease with 46 nodes assessed. Median primary GTV (GTVp) size was 14.1cc. The rate of GTVp shrinkage was highest between pre-treatment and week 4. Patients with T3-T4 tumours had a 3.8-fold decrease in GTVp compared to T1-T2 tumours. The ADC values correlated with residual GTVp. The median nodal volume (GTVn) was 12.4cc. No clinical features were found to correlate with GTVn reduction. The overall change in ADC for GTVn from pre-treatment was significant for 35th−95th percentiles in weeks 1−4 (p < 0.001). Conclusion: A discrepancy in the trajectory of ADC between primary and nodal sites suggested that they exhibit different treatment responses and should be analysed separately in future studies.

Oligometastatic Head and Neck Cancer: Challenges and Perspectives

Simple Summary

Oligometastasis represents a disease state and an opportunity for cure when metastases emerge. Emerging evidence supports that most head and neck cancer patients with oligometastatic disease are likely to benefit from curative intent local ablative therapy if appropriate selection criteria are applied. Biomarkers to predict development of oligometastasis, as well as to identify which patients could benefit from a radical intent approach, are under investigation. This review summarizes recent knowledge about the characteristics, investigational efforts, and evidence for local ablation regarding oligometastasis in head and neck cancer. We also describe the challenges and opportunities in patient selection and discuss the role of radiotherapy and immunotherapy combinations to enhance anti-tumor immunity.

Abstract

A minority of patients with metastatic head and neck squamous cell carcinoma (HNSCC) present with oligometastatic disease. Oligometastasis not only reflects a disease state, but might also present an opportunity for cure in the metastatic setting. Radical ablation of all oligometastatic sites may confer prolonged survival and possibly achieve cure in some patients. However, substantial debate remains about whether patients with oligometastatic disease could benefit from curative intent therapy or whether aggressive treatments expose some patients to futile toxicity. Optimal selection of patients, carefully balancing the currently known prognostic factors against the risks of toxicity is critical. Emerging evidence suggests that patients with a limited burden of disease, viral-related pharyngeal cancer, metachronous metastasis and lung-only metastasis may benefit most from this approach. Efforts are underway to identify biomarkers that can detect oligometastasis and better select patients who would derive the maximum benefit from an aggressive radical approach. The combination of radiotherapy and immunotherapy promises to enhance the anti-tumoral immune response and help overcome resistance. However, optimization of management algorithms, including patient selection, radiation dose and sequencing, will be critical in upcoming clinical trials. This review summarizes recent knowledge about the characteristics and investigational efforts regarding oligometastasis in HNSCC.

Planning Trade-offs for Stereotactic Ablative Radiotherapy in Patients with 4-10 Metastases: A Sub-study of the SABR-COMET-10 randomized trial

BACKGROUND

The use of stereotactic ablative radiotherapy (SABR) for metastases often involves trade-offs, balancing adequate target coverage with the need to maintain a safe dose to organs at risk (OARs). Very few studies have evaluated this complexity in patients with >3 oligometastases. To explore these trade-offs, we analyzed planning outcomes from the first 60 patients enrolled on a randomized trial of SABR in patients with 4-10 metastases.

METHODS

Radiation plans for patients enrolled on the BLINDED-FOR-REVIEW randomized trial were analyzed. Data abstracted included target locations and sizes, dose prescriptions and target coverage, and OAR doses, and these parameters were evaluated based on target location to identify locations where compromise of planning constraints was required.

RESULTS

SABR was planned for 332 lesions in 296 separate PTVs. The median PTV size was 9.0 cc per lesion (interquartile range [IQR]: 5.3-24.3 cc) and 89.1 cc per patient (IQR: 37.0-177.0). The most common prescriptions were 30-35 Gy in 5 fractions (43% of lesions) and 20-24 Gy in 1 fraction (33% of lesions). The maximum dose within the PTV was a median of 124% of prescription (IQR: 117-130%). Only 3.3% of lesions had a dose to 95% of the PTV (D95) <95% of prescription; this was more common for spinal (25%) and lymph node (8.4%) targets (p<0.001), translating to median D95 values of 100.5% (IQR: 99.6%-101.5%) and 100.0% (IQR: 84.6%-100.5%), respectively. All OARs constraints were met in 95.2% of patients, with the only exceptions being lymph node targets (n=3).

CONCLUSIONS

SABR planning for patients with 4-10 metastases was achievable without dosimetric compromise in the large majority of patients. Nodal and spinal targets were most likely to lead to compromise of target coverage or OAR constraints. Further research is needed to determine how to best balance these trade-offs.

Current Therapy for Metastatic Head and Neck Cancer: Evidence, Opportunities, and Challenges

Management of metastatic head and neck squamous cell carcinoma is evolving as new systemic therapies have led to improvements in survival, and as advances in locoregional therapy and the increased numbers of patients with HPV-associated cancers who develop oligometastases raise the possibility of ablation of limited numbers of metastases. We review the data regarding first-line immunotherapy in PD-L1-expressing metastatic head and neck squamous cell carcinoma, the experience with aggressive local management of oligometastases, and promising novel immunotherapies, targeted therapies, and HPV-specific treatments. For patients with metastatic head and neck squamous cell carcinoma that is PD-L1 expressing, first-line systemic therapy is pembrolizumab or pembrolizumab with chemotherapy. Inclusion of chemotherapy is associated with higher objective response proportion in all biomarker subgroups and may have a greater impact on survival in HPV-associated cancers. For patients with oligometastatic disease, particularly when metastases are metachronous, current evidence supporting the role of local ablation is limited to a small number of retrospective studies. Based on retrospective data, patients with a smaller number of metastases, lung metastases, and/or virally associated head and neck squamous cell carcinoma are most likely to benefit from an aggressive ablative approach. Additionally, we review emerging evidence for targeted therapy in metastatic head and neck squamous cell carcinoma, including with agents that inhibit mutant HRAS or NOTCH1, or overexpressed EGFR. Studies of antiangiogenic agents in combination with immune checkpoint blockade, and combination immunotherapy, are also under study.

Cross-institutional outcome prediction for head and neck cancer patients using self-attention neural networks

In radiation oncology, predicting patient risk stratification allows specialization of therapy intensification as well as selecting between systemic and regional treatments, all of which helps to improve patient outcome and quality of life. Deep learning offers an advantage over traditional radiomics for medical image processing by learning salient features from training data originating from multiple datasets. However, while their large capacity allows to combine high-level medical imaging data for outcome prediction, they lack generalization to be used across institutions. In this work, a pseudo-volumetric convolutional neural network with a deep preprocessor module and self-attention (PreSANet) is proposed for the prediction of distant metastasis, locoregional recurrence, and overall survival occurrence probabilities within the 10 year follow-up time frame for head and neck cancer patients with squamous cell carcinoma. The model is capable of processing multi-modal inputs of variable scan length, as well as integrating patient data in the prediction model. These proposed architectural features and additional modalities all serve to extract additional information from the available data when availability to additional samples is limited. This model was trained on the public Cancer Imaging Archive Head–Neck-PET–CT dataset consisting of 298 patients undergoing curative radio/chemo-radiotherapy and acquired from 4 different institutions. The model was further validated on an internal retrospective dataset with 371 patients acquired from one of the institutions in the training dataset. An extensive set of ablation experiments were performed to test the utility of the proposed model characteristics, achieving an AUROC of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$80\%$$\end{document}80%, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$80\%$$\end{document}80% and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$82\%$$\end{document}82% for DM, LR and OS respectively on the public TCIA Head–Neck-PET–CT dataset. External validation was performed on a retrospective dataset with 371 patients, achieving \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$69\%$$\end{document}69% AUROC in all outcomes. To test for model generalization across sites, a validation scheme consisting of single site-holdout and cross-validation combining both datasets was used. The mean accuracy across 4 institutions obtained was \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$72\%$$\end{document}72%, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$70\%$$\end{document}70% and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$71\%$$\end{document}71% for DM, LR and OS respectively. The proposed model demonstrates an effective method for tumor outcome prediction for multi-site, multi-modal combining both volumetric data and structured patient clinical data.

Comprehensive Quantitative Evaluation of Variability in MR-guided Delineation of Oropharyngeal Gross Tumor Volumes and High-risk Clinical Target Volumes: An R-IDEAL Stage 0 Prospective Study

PURPOSE

Tumor and target volume manual delineation remains a challenging task in head-and-neck cancer radiotherapy. The purpose of this study was to conduct a multi-institutional evaluation of manual delineations of gross tumor volume (GTV), high-risk clinical target volume (CTV), parotids, and submandibular glands on treatment simulation MR scans of oropharyngeal cancer (OPC) patients.

METHODS

Pre-treatment T1-weighted (T1w), T1-weighted with gadolinium contrast (T1w+C) and T2-weighted (T2w) MRI scans were retrospectively collected for 4 OPC patients under an IRB-approved protocol. The scans were provided to twenty-six radiation oncologists from seven international cancer centers who participated in this delineation study. In addition, patients' clinical history and physical examination findings, along with a medical photographic image and radiological results, were provided. The contours were compared using overlap/distance metrics using both STAPLE and pair-wise comparisons. Lastly, participants completed a brief questionnaire to assess participants' experience and CTV delineation institutional practices.

RESULTS

Large variability was measured between observers' delineations for GTVs and CTVs. The mean Dice Similarity Coefficient values across all physicians' delineations for GTVp, GTVn, CTVp, and CTVn were 0.77, 0.67, 0.77, and 0.69, respectively, for STAPLE comparison and 0.67, 0.60, 0.67, and 0.58, respectively, for pair-wise analysis. Normal tissue contours were defined more consistently when considering overlap/distance metrics. The median radiation oncology clinical experience was 7 years. The median experience delineating on MRI was 3.5 years. The GTV-to-CTV margin used was 10 mm for six of seven participant institutions. One institution used 8 mm and three participants (from three different institutions) used a margin of 5 mm.

CONCLUSION

The data from this study suggests that appropriate guidelines, contouring quality assurance sessions, and training are still needed for the adoption of MR-based treatment planning for head-and-neck cancers. Such efforts should play a critical role in reducing delineation variation and ensure standardization of target design across clinical practices.

Upfront DPYD Genotyping and Toxicity Associated with Fluoropyrimidine-Based Concurrent Chemoradiotherapy for Oropharyngeal Carcinomas: A Work in Progress

Background: 5-FU-based chemoradiotherapy (CRT) could be associated with severe treatment-related toxicities in patients harboring at-risk DPYD polymorphisms. Methods: The studied population included consecutive patients with locoregionally advanced oropharyngeal carcinoma treated with carboplatin and 5-FU-based CRT one year before and after the implementation of upfront DPYD*2A genotyping. We aimed to determine the effect of DPYD genotyping on grade ≥3 toxicities. Results: 181 patients were analyzed (87 patients before and 94 patients following DPYD*2A screening). Of the patients, 91% (n = 86) were prospectively genotyped for the DPYD*2A allele. Of those screened, 2% (n = 2/87) demonstrated a heterozygous DPYD*2A mutation. Extended genotyping of DPYD*2A-negative patients later allowed for the retrospective identification of six additional patients with alternative DPYD variants (two c.2846A>T and four c.1236G>A mutations). Grade ≥3 toxicities occurred in 71% of the patients before DPYD*2A screening versus 62% following upfront genotyping (p = 0.18). When retrospectively analyzing additional non-DPYD*2A variants, the relative risks for mucositis (RR 2.36 [1.39–2.13], p = 0.0063), dysphagia (RR 2.89 [1.20–5.11], p = 0.019), and aspiration pneumonia (RR 13 [2.42–61.5)], p = 0.00065) were all significantly increased. Conclusion: The DPYD*2A, c.2846A>T, and c.1236G>A polymorphisms are associated with an increased risk of grade ≥3 toxicity to 5-FU. Upfront DPYD genotyping can identify patients in whom 5-FU-related toxicity should be avoided.

Orofacial Pain and Snoring/Obstructive Sleep Apnea in Individuals with Head and Neck Cancer: A Critical Review

AIMS

(1) To summarize current knowledge on the prevalence, intensity, and descriptors of orofacial pain and snoring/obstructive sleep apnea (OSA) before and after head and neck cancer (HNC) treatment; and (2) to propose future directions for research.

METHODS

The median prevalence for each condition was estimated from the most recent systematic reviews (SRs) and updated with new findings retrieved from the PubMed, Web of Science, Embase, and Cochrane databases up to December 2021.

RESULTS

The prevalence of HNC pain seems relatively stable over time, with a median of 31% before treatment in three studies to a median of 39% at 1 month to 16 years after treatment in six studies. HNC pain intensity remains mild to moderate. There was a threefold increase in temporomandibular pain prevalence after surgery (median 7.25% before to 21.3% after). The data for snoring prevalence are unreliable. The OSA/HNC prevalence seems relatively stable over time, with a median of 72% before treatment in three studies to 77% after treatment in 14 studies.

CONCLUSION

With the exception of temporomandibular pain, the prevalence of HNC pain and OSA seems to be stable over time. Future studies should: (1) compare the trajectory of change over time according to each treatment; (2) compare individuals with HNC to healthy subjects; (3) use a standardized and comparable method of data collection; and (4) assess tolerance to oral or breathing devices, since HNC individuals may have mucosal sensitivity or pain.

Risk stratification after recurrence of human papillomavirus (HPV)-related and non-HPV-related oropharyngeal cancer: A secondary analysis of NRG Oncology RTOG 0129 and 0522

BACKGROUND

No risk-stratification strategies exist for patients with recurrent oropharyngeal cancer (OPC).

METHODS

Retrospective analysis using data from prospective NRG Oncology clinical trials RTOG 0129 and 0522. Eligibility criteria included known p16 status and smoking history, and locoregional/distant recurrence. Overall survival (OS) was measured from date of recurrence. Recursive partitioning analysis was performed to produce mutually exclusive risk groups.

RESULTS

Hundred and fifty-four patients were included with median follow-up after recurrence of 3.9 years (range 0.04-9.0). The most important factors influencing survival were p16 status and type of recurrence, followed by surgical salvage and smoking history (≤20 vs. >20 pack-years). Three significantly different risk groups were identified. Patients in the low-, intermediate-, and high-risk groups had 2-year OS after recurrence of 81.1% (95%CI 68.5-93.7), 50.2% (95%CI 36.0-64.5), and 20.8% (95%CI 10.5-31.1), respectively.

CONCLUSION

Patient and tumor characteristics may be used to stratify patients into risk groups at the time of OPC recurrence.

Radiomics feature stability of open-source software evaluated on apparent diffusion coefficient maps in head and neck cancer

Radiomics is a promising technique for discovering image based biomarkers of therapy response in cancer. Reproducibility of radiomics features is a known issue that is addressed by the image biomarker standardisation initiative (IBSI), but it remains challenging to interpret previously published radiomics signatures. This study investigates the reproducibility of radiomics features calculated with two widely used radiomics software packages (IBEX, MaZda) in comparison to an IBSI compliant software package (PyRadiomics). Intensity histogram, shape and textural features were extracted from 334 diffusion weighted magnetic resonance images of 59 head and neck cancer (HNC) patients from the PREDICT-HN observational radiotherapy study. Based on name and linear correlation, PyRadiomics shares 83 features with IBEX and 49 features with MaZda, a sub-set of well correlated features are considered reproducible (IBEX: 15 features, MaZda: 18 features). We explore the impact of including non-reproducible radiomics features in a HNC radiotherapy response model. It is possible to classify equivalent patient groups using radiomic features from either software, but only when restricting the model to reliable features using a correlation threshold method. This is relevant for clinical biomarker validation trials as it provides a framework to assess the reproducibility of reported radiomic signatures from existing trials.

Patient-Reported Outcomes after Intensity-Modulated Proton Therapy for Oropharynx Cancer

PURPOSE

To report patient-reported outcomes (PROs) derived from the Functional Assessment of Cancer Therapy-Head and Neck (FACT-HN) tool, in patients with oropharynx cancer (OPC) treated with intensity-modulated proton therapy (IMPT) in the context of first-course irradiation.

MATERIALS AND METHODS

Patients with locally advanced OPC treated with radical IMPT between 2011 and 2018 were included in a prospective registry. FACT-HN scores were measured serially during and 24 months following IMPT. PRO changes in the FACT-HN scores over time were assessed with mixed-model analysis.

RESULTS

Fifty-seven patients met inclusion criteria. Median age was 60 years (range, 41-84), and 91% had human papillomavirus-associated disease. In total, 28% received induction chemotherapy and 68% had concurrent chemotherapy. Compliance to FACT-HN questionnaire completion was 59%, 48%, and 42% at 6, 12, and 24 months after treatment, respectively. The mean FACT-General (G), FACT-Total, and FACT-Trial Outcome Index (TOI) score changes were statistically and clinically significant relative to baseline from week 3 of treatment up to week 2 after treatment. Nadir was reached at week 6 of treatment for all scores, with maximum scores dropping by 15%, 20%, and 39% compared to baseline for FACT-G, FACT-Total, and FACT-TOI, respectively. Subdomain scores of physical well-being, functional well-being, and head and neck additional concerns decreased from baseline during treatment and returned to baseline at week 4 after treatment.

CONCLUSIONS

IMPT was associated with a favorable PRO trajectory, characterized by an acute decline followed by rapid recovery to baseline. This study establishes the expected acute, subacute, and chronic trajectory of PROs for patients undergoing IMPT for OPC.

Vocal-cord Only vs. Complete Laryngeal radiation (VOCAL): a randomized multicentric Bayesian phase II trial

BACKGROUND

Radiotherapy, along with laser surgery, is considered a standard treatment option for patients with early glottic squamous cell cancer (SCC). Historically, patients have received complete larynx radiotherapy (CL-RT) due to fear of swallowing and respiratory laryngeal motion and this remains the standard approach in many academic institutions. Local control (LC) rates with CL-RT have been excellent, however this treatment can carry significant toxicities include adverse voice and swallowing outcomes, along with increased long-term risk of cerebrovascular morbidity. A recent retrospective study reported improved voice quality and similar local control outcomes with focused vocal cord radiotherapy (VC-RT) compared to CL-RT. There is currently no prospective evidence on the safety of VC-RT. The primary objective of this Bayesian Phase II trial is to compare the LC of VC-RT to that of CL-RT in patients with T1N0 glottic SCC.

METHODS

One hundred and fifty-five patients with T1a-b N0 SCC of the true vocal cords that are n ot candidate or declined laser surgery, will be randomized in a 1:3 ratio the control arm (CL-RT) and the experimental arm (VC-RT). Randomisation will be stratified by tumor stage (T1a/T1b) and by site (each site will be allowed to select one preferred radiation dose regimen, to be used in both arms). CL-RT volumes will correspond to the conventional RT volumes, with the planning target volume extending from the top of thyroid cartilage lamina superiorly to the bottom of the cricoid inferiorly. VC-RT volumes will include the involved vocal cord(s) and a margin accounting for respiration and set-up uncertainty. The primary endpoint will be LC at 2-years, while secondary endpoints will include patient-reported outcomes (voice impairment, dysphagia and symptom burden), acute and late toxicity radiation-induced toxicity, overall survival, progression free survival, as well as an optional component of acoustic and objective measures of voice analysis using the Consensus Auditory-Perceptual Evaluation of Voice.

DISCUSSION

This study would constitute the first prospective evidence on the efficacy and safety of VC-RT in early glottic cancer. If positive, this study would result in the adoption of VC-RT as standard approach in early glottic cancer.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03759431 Registration date: November 30, 2018.

Longitudinal characterization of the tumoral microbiome during radiotherapy in HPV-associated oropharynx cancer

PURPOSE

To describe the baseline and serial tumor microbiome in HPV-associated oropharynx cancer (OPC) over the course of radiotherapy (RT).

METHODS

Patients with newly diagnosed HPV-associated OPC treated with definitive radiotherapy +/- concurrent chemotherapy were enrolled in this prospective study. Using 16S rRNA gene sequencing, dynamic changes in the tumor site microbiome during RT were investigated. Surface tumor samples were obtained before RT and at week 1, 3 and 5 of RT. Radiological primary tumor response at mid-treatment was categorized as complete (CR) or partial (PR).

RESULTS

Ten patients were enrolled, but 9 patients were included in the final analysis. Mean age was 62 years (range: 51-71). As per AJCC 8th Ed, 56%, 22% and 22% of patients had stage I, II and III, respectively. At 4-weeks, 6 patients had CR and 3 patients had PR; at follow-up imaging post treatment, all patients had CR. The baseline diversity of the tumoral versus buccal microbiome was not statistically different. For the entire cohort, alpha diversity was significantly decreased over the course of treatment (p = 0.04). There was a significant alteration in the bacterial community within the first week of radiation. Baseline tumor alpha diversity of patients with CR was significantly higher than those with PR (p = 0.03). While patients with CR had significant reduction in diversity over the course of radiation (p = 0.01), the diversity remained unchanged in patients with PR. Patients with history of smoking had significantly increased abundance of  Kingella (0.05) and lower abundance of Stomatobaculum (p = 0.03) compared to never smokers.

CONCLUSIONS

The tumor microbiome of HPV-associated OPC exhibits reduced alpha diversity and altered taxa abundance over the course of radiotherapy. The baseline bacterial profiles of smokers vs. non-smokers were inherently different. Baseline tumor alpha diversity of patients with CR was higher than patients with PR, suggesting that the microbiome deserves further investigation as a biomarker of radiation response.

Durvalumab therapy following chemoradiation compared with a historical cohort treated with chemoradiation alone in patients with stage III non-small cell lung cancer: A real-world multicentre study

BACKGROUND

The PACIFIC trial demonstrated that durvalumab therapy following chemoradiation (CRT) was associated with improved overall survival (OS) in patients with stage III non-small cell lung cancer (NSCLC). It is unclear whether the results obtained as part of randomised controlled trials are a reflection of real-world (RW) data. Several questions remain unanswered with regard to RW durvalumab use, such as optimal time to durvalumab initiation, incidence of pneumonitis and response in PD-L1 subgroups.

METHODS

In this multicentre retrospective analysis, 147 patients with stage III NSCLC treated with CRT followed by durvalumab were compared with a historical cohort of 121 patients treated with CRT alone. Survival curves were estimated using the Kaplan-Meier method and compared with the log-rank test in univariate analysis. Multivariate analysis was performed to evaluate the effect of standard prognostic factors for durvalumab use.

RESULTS

Median OS was not reached in the durvalumab group, compared with 26.9 months in the historical group (hazard ratio [HR]: 0.56, 95% confidence interval [CI]: 0.37-0.85, p = 0.001). In the durvalumab group, our data suggest improved 12-month OS in patients with PD-L1 expression ≥50% (100% vs 86%, HR: 0.25, 95% CI: 0.11-0.58, p = 0.007). There was no difference in OS between patients with a PD-L1 expression of 1-49% and patients with PD-L1 expression <1%. Delay in durvalumab initiation beyond 42 days did not impact OS. Incidence of pneumonitis was similar in the durvalumab and historical groups. In the durvalumab group, patients who experienced any-grade pneumonitis had a lower 12-month OS than patients without pneumonitis (85% vs 95%, respectively; HR: 3.3, 95% CI: 1.2-9.0, p = 0.006).

CONCLUSIONS

This multicentre analysis suggests that PD-L1 expression ≥50% was associated with favourable OS in patients with stage III NSCLC treated with durvalumab after CRT, whereas the presence of pneumonitis represented a negative prognostic factor.