Multitask Learning with Convolutional Neural Networks and Vision Transformers Can Improve Outcome Prediction for Head and Neck Cancer Patients.

Neural-network-based outcome predictions may enable further treatment personalization of patients with head and neck cancer. The development of neural networks can prove challenging when a limited number of cases is available. Therefore, we investigated whether multitask learning strategies, implemented through the simultaneous optimization of two distinct outcome objectives (multi-outcome) and combined with a tumor segmentation task, can lead to improved performance of convolutional neural networks (CNNs) and vision transformers (ViTs). Model training was conducted on two distinct multicenter datasets for the endpoints loco-regional control (LRC) and progression-free survival (PFS), respectively. The first dataset consisted of pre-treatment computed tomography (CT) imaging for 290 patients and the second dataset contained combined positron emission tomography (PET)/CT data of 224 patients. Discriminative performance was assessed by the concordance index (C-index). Risk stratification was evaluated using log-rank tests. Across both datasets, CNN and ViT model ensembles achieved similar results. Multitask approaches showed favorable performance in most investigations. Multi-outcome CNN models trained with segmentation loss were identified as the optimal strategy across cohorts. On the PET/CT dataset, an ensemble of multi-outcome CNNs trained with segmentation loss achieved the best discrimination (C-index: 0.29, 95% confidence interval (CI): 0.22-0.36) and successfully stratified patients into groups with low and high risk of disease progression (p=0.003). On the CT dataset, ensembles of multi-outcome CNNs and of single-outcome ViTs trained with segmentation loss performed best (C-index: 0.26 and 0.26, CI: 0.18-0.34 and 0.18-0.35, respectively), both with significant risk stratification for LRC in independent validation (p=0.002 and p=0.011). Further validation of the developed multitask-learning models is planned based on a prospective validation study, which has recently completed recruitment.

Superior loco-regional control after primary surgery compared to chemo-radiotherapy for advanced stage laryngeal cancer.

Objective: The optimal strategy to treat loco-regionally advanced squamous cell carcinoma of the larynx (LSCC) remains to be defined. The goal of this single institution retrospective study was to report on oncologic outcome of advanced LSCC treated with curative intent. Methods: Patients diagnosed and treated for stage T3-T4a LSCC between 2001 and 2014 were retrospectively analyzed. Time-to-event endpoints were calculated beginning from the date of histologic diagnosis, which were analyzed with log-rank test and Cox proportional hazard models. Results: The cohort was divided into two subgroups: primary radiotherapy with concomitant cisplatin (CRT) (n=30, 38%) and primary surgery (n=48, 62%). Median follow-up was 56 months. Locoregional control (LRC) for the primary surgery and CRT were 95% and 50% in 5 years, respectively (p<0.01). Progression free survival (PFS) for the primary surgery and CRT were 61% and 38% in 5 years, respectively (p=0.23). The overall survival (OS) after primary surgery and CRT in 5 years were 63% vs. 65%, respectively (p=0.93). The 5-years LRC was significantly superior after surgery compared to RT for cT3 primaries (100% vs 50%, p= 0.0022). No significant differences were observed in the remaining subgroups regarding cT stage and PFS or OS. Conclusion: Our series demonstrated superior LRC after primary surgery followed by risk-adapted adjuvant (C)RT compared to primary CRT in cT3 LSCC, but no significant difference in PFS or OS in locally-advanced LSCC. The optimal patient selection criteria for the ideal treatment for loco-regionally advanced LSCC still needs to be defined.

Chemoradiation with Cisplatin vs. Carboplatin for Squamous Cell Carcinoma of the Head and Neck (SCCHN).

Cisplatin is the standard for the chemoradiation of squamous cell carcinoma of the head and neck (HNSCC). Many patients cannot receive cisplatin due to impaired renal function. This study investigated carboplatin as an alternative option. In total, 131 patients assigned to two courses of cisplatin (20 mg/m2/d1--5 or 25 mg/m2/d1-4) were matched to 45 patients not suitable for cisplatin and receiving carboplatin (AUC 1.0/d1-5 or AUC 1.5/d1-4). The endpoints included loco-regional control (LRC), metastases-free survival (MFS), overall survival (OS), toxicities, and the completion of chemotherapy. The patients in the carboplatin group were significantly older and had more G3 tumors. Otherwise, the baseline characteristics were balanced. The LRC rates at 2 and 3 years were 77% and 76% in the cisplatin group vs. 69% and 65% in the carboplatin group (p = 0.21). The MFS rates were 83% and 78% vs. 78% and 74% (p = 0.34) and the OS rates 83% and 79% vs. 83% and 75% (p = 0.64), respectively. The outcomes were not significantly different in the subgroups receiving definitive or adjuvant chemoradiation. No significant differences were found regarding toxicities. Non-significantly more patients in the carboplatin group completed their chemotherapy (78% vs. 66%, p = 0.15). Carboplatin was associated with similar outcomes and toxicities as cisplatin, although these patients had worse renal function, more aggressive tumors, and were older. Given the limitations of this study, carboplatin appears an option for patients not suitable for cisplatin.

Longitudinal and Multimodal Radiomics Models for Head and Neck Cancer Outcome Prediction.

Radiomics analysis provides a promising avenue towards the enabling of personalized radiotherapy. Most frequently, prognostic radiomics models are based on features extracted from medical images that are acquired before treatment. Here, we investigate whether combining data from multiple timepoints during treatment and from multiple imaging modalities can improve the predictive ability of radiomics models. We extracted radiomics features from computed tomography (CT) images acquired before treatment as well as two and three weeks after the start of radiochemotherapy for 55 patients with locally advanced head and neck squamous cell carcinoma (HNSCC). Additionally, we obtained features from FDG-PET images taken before treatment and three weeks after the start of therapy. Cox proportional hazards models were then built based on features of the different image modalities, treatment timepoints, and combinations thereof using two different feature selection methods in a five-fold cross-validation approach. Based on the cross-validation results, feature signatures were derived and their performance was independently validated. Discrimination regarding loco-regional control was assessed by the concordance index (C-index) and log-rank tests were performed to assess risk stratification. The best prognostic performance was obtained for timepoints during treatment for all modalities. Overall, CT was the best discriminating modality with an independent validation C-index of 0.78 for week two and weeks two and three combined. However, none of these models achieved statistically significant patient stratification. Models based on FDG-PET features from week three provided both satisfactory discrimination (C-index = 0.61 and 0.64) and statistically significant stratification (p=0.044 and p<0.001), but produced highly imbalanced risk groups. After independent validation on larger datasets, the value of (multimodal) radiomics models combining several imaging timepoints should be prospectively assessed for personalized treatment strategies.

Comparison of Two Cisplatin Regimens for Chemoradiation in Patients With Squamous-cell Carcinoma of the Head and Neck.

BACKGROUND/AIM: Many patients with squamous-cell carcinoma of the head and neck receive cisplatin-based chemoradiation. This retrospective study compared two chemoradiation programs to help identify the optimal cisplatin-regimen. PATIENTS AND METHODS: Forty-one patients assigned to chemoradiation with two cycles of 20 mg/m2/days(d)1-5 were compared to 78 patients assigned to chemoradiation with two cycles of 25 mg/m2/d1-4. Groups were compared for toxicity, loco-regional control (LRC), and survival. RESULTS: Both treatments were associated with similar rates of oral mucositis, radiation dermatitis, xerostomia, nausea, decreased renal function, and hematotoxicity. The cisplatin-regimen had no significant impact on LRC (p=0.41) or survival (p=0.85). Survival was significantly worse with radiotherapy interruptions (>1 week) or discontinuation (p<0.001) and administration of <80% of the planned cisplatin dose (p<0.001). CONCLUSION: Both cisplatin-regimens did not differ significantly regarding toxicities, LRC, and survival. It is important to avoid interruption or discontinuation of radiotherapy and to administer ≥80% of planned cisplatin.

Definitive radiochemotherapy in esophageal cancer - a single institution experience.

Background Definitive radiochemotherapy is the preferred treatment option in patients with the cancer of the cervical esophagus and a viable treatment option in patients with the cancer of lower two thirds of the esophagus, who decline proposed surgical treatment. The purpose of the study was to evaluate the treatment results with definitive radiochemotherapy of patients with esophageal cancer, treated in a single institution in the period from 2010 to 2017. Patients and methods All available medical data for 55 patients with esophageal cancer, who were treated with definitive radiochemotherapy with curative intent, were analyzed retrospectively. Patients were irradiated to a total dose to the tumor of 70 Gy (2 Gy per fraction) in upper third (cervical) tumors or to the mean total dose of 57.6 Gy (1.8 Gy per fraction) in middle third (intrathoracic) tumors. All but one patient received concomitant chemotherapy, with the majority of them (41 patients; 74.5%) receiving concomitant chemotherapy with 5-fluorouracil in continuous 96 hours infusion and cisplatin. The main endpoints of the study were overall survival (OS; death of any cause), locoregional control (LRC; local and/or regional disease recurrence) and disease-free survival (DFS; recurrence of any kind and/or new primary malignoma). Univariate analysis testing the impact of different parameters on survivals and analysis of treatment related side effects were performed as well. Results The mean age of patients was 62 years (SD 9 years; range: 29-80 years). Majority of them had squamous cell cancer (53 patients; 96.4%) in the stage T3 or T4 (47 patients; 85.5%) and/or N+ disease (35 patients; 63.6%). Median follow-up time for the whole group of patients was 16.8 months (range: 0.3-81.8 months). At the time of analysis 14 (25.5%) patients were still alive. Rates for OS, LRC and DFS at two and five years were as follows: 47% and 19.4%; 43.7% and 41%; 32.1% and 11.5%, respectively. Conclusions The study results of treatment with definitive radiochemotherapy in patients with esophageal cancer are similar to the results of other studies. Majority of patients ended the treatment according to the protocol, which at least in part can be attributed to the adequate and well organized supportive treatment in our institution.

Novel Uses of Radioactive Seeds in Surgical Oncology: A Case Series.

The localization of nonpalpable axillary metastatic lymph nodes has been achieved using several techniques in the past. Amongst these techniques, the use of radioactive iodine seeds is increasingly spread, and was initially reserved to breast-conserving surgery. Many studies have assessed the use of radioactive seed localization for the surgical management of breast cancer patients diagnosed with lymph node metastases. However, few articles have reported their utilization in other cancer subtypes and in complex clinical situations. This case series describes the innovative use of radioactive seeds in the axilla in five patients, including one case of squamous cell carcinoma skin cancer, one case of malignant melanoma, and three cases of invasive breast cancer.

Individualized treatment of head neck squamous cell carcinoma patients aged 70 or older with radiotherapy alone or associated to cisplatin or cetuximab: impact of weekly radiation dose on loco-regional control.

The purpose of this study is to evaluate if, in elderly HNC patients, loco-regional control (LRC) is influenced by average weekly radiation dose (AWD). From 2009 to 2017, 150 consecutive HNC elderly patients were analyzed. AWD was calculated by dividing total dose in Gray by overall treatment time in weeks. Patients were divided in 2 groups: Group 1 (70-75 years) and Group 2 (> 75 years). Primary endpoint was LRC; secondary endpoints were overall survival (OS) and compliance to treatment. The median age was 76 years (range 70-92), the distribution of patients by age was 72 and 78 patients in Group 1 and in Group 2, respectively; overall median follow-up was 23 months. Optimal cut-off of AWD for LRC was 9.236 (p = 0.018). Median OS was 73 months. In univariate survival analysis low PS (p = 0.005), T3-T4 (p = 0.021), Stage III-IV (p = 0.046) and AWDLow (< 9.236) (p = 0.018) were significantly associated with lower LRC; low PS (p < 0.001) and Group 2 (p = 0.006) were also associated with lower OS. Considering patients treated with radiotherapy alone AWDLow was significantly associated with lower LRC (p = 0.04) whereas among patient treated with chemoradiotherapy AWD did not affected LRC (p = 0.18). The multivariate analysis confirmed the significant value of PS for the prediction of LRC and OS (p = 0.035 and p < 0.001, respectively). In elderly patients an AWD of > 9.236 Gy was found to be beneficial for RT alone regimen. When radiotherapy alone is indicated in elderly patients an effort should be made to maintain an increased AWD in order to improve LRC.

Comparison of detection methods for HPV status as a prognostic marker for loco-regional control after radiochemotherapy in patients with HNSCC.

OBJECTIVE: To compare six HPV detection methods in pre-treatment FFPE tumour samples from patients with locally advanced head and neck squamous cell carcinoma (HNSCC) who received postoperative (N = 175) or primary (N = 90) radiochemotherapy. MATERIALS AND METHODS: HPV analyses included detection of (i) HPV16 E6/E7 RNA, (ii) HPV16 DNA (PCR-based arrays, A-PCR), (iii) HPV DNA (GP5+/GP6+ qPCR, (GP-PCR)), (iv) p16 (immunohistochemistry, p16 IHC), (v) combining p16 IHC and the A-PCR result and (vi) combining p16 IHC and the GP-PCR result. Differences between HPV positive and negative subgroups were evaluated for the primary endpoint loco-regional control (LRC) using Cox regression. RESULTS: Correlation between the HPV detection methods was high (chi-squared test, p < 0.001). While p16 IHC analysis resulted in several false positive classifications, A-PCR, GP-PCR and the combination of p16 IHC and A-PCR or GP-PCR led to results comparable to RNA analysis. In both cohorts, Cox regression analyses revealed significantly prolonged LRC for patients with HPV positive tumours irrespective of the detection method. CONCLUSIONS: The most stringent classification was obtained by detection of HPV16 RNA, or combining p16 IHC with A-PCR or GP-PCR. This approach revealed the lowest rate of recurrence in patients with tumours classified as HPV positive and therefore appears most suited for patient stratification in HPV-based clinical studies.

Comparing two lower-dose cisplatin programs for radio-chemotherapy of locally advanced head-and-neck cancers.

Radio-chemotherapy is a common treatment for locally advanced squamous cell head-and-neck cancers (LA-SCCHN). Cisplatin (100 mg/m2) every 3 weeks is very common but associated with considerable toxicity. Therefore, cisplatin programs with lower daily doses were introduced. There is a lack of studies comparing lower-dose programs. In this study, 85 patients receiving radio-chemotherapy with 20 mg/m2 cisplatin on 5 days every 4 weeks (group A) were retrospectively compared to 85 patients receiving radio-chemotherapy with 30-40 mg/m2 cisplatin weekly (group B). Groups were matched for nine factors including age, gender, performance score, tumor site, T-/N-category, surgery, hemoglobin before radio-chemotherapy, and radiation technique. One- and 3-year loco-regional control rates were 83 and 69 % in group A versus 74 and 63 % in group B (p = 0.12). One- and 3-year survival rates were 93 % and 73 % in group A versus 91 and 49 % in group B (p = 0.011). On multivariate analysis, survival was significantly better for group A (HR 1.17; p = 0.002). In groups A and B, 12 and 28 % of patients, respectively, did not receive a cumulative cisplatin dose ≥180 mg/m2 (p = 0.016). Toxicity rates were not significantly different. On subgroup analyses, group A patients had better loco-regional control (p = 0.040) and survival (p = 0.005) than group B patients after definitive radio-chemotherapy. In patients receiving adjuvant radio-chemotherapy, outcomes were not significantly different. Thus, 20 mg/m2 cisplatin on 5 days every 4 weeks resulted in better loco-regional control and survival in patients receiving definitive radio-chemotherapy and may be preferable for these patients. Confirmation of these results in a randomized trial is warranted.

Tumour- and treatment-related colostomy rates following mitomycin C or cisplatin chemoradiation with or without maintenance chemotherapy in squamous cell carcinoma of the anus in the ACT II trial.

BACKGROUND: Squamous cell carcinoma of the anus (SCCA) is highly sensitive to chemoradiation (CRT) which achieves good loco-regional control and preserves anal function. However, some patients require permanent stoma formation either as a result of surgery on relapse, poor anal function or treatment-related symptoms. Our aim was to determine patient, tumour and treatment-related colostomy rates following CRT and maintenance chemotherapy in the ACT II trial. PATIENTS AND METHODS: The ACT II trial recruited 940 patients comparing 5FU-based CRT using cisplatin (CisP) or mitomycin C (MMC) with or without additional maintenance chemotherapy. We investigated the association between colostomy-free survival (CFS) and progression-free survival (PFS) with age, gender, T-stage, N-stage, treatment and baseline haemoglobin. RESULTS: The median follow-up was 5.1 years (n = 884 evaluable/940); tumour site canal (84%), margin (14%); stage T1/T2 (52%), T3/T4 (46%); N+ (32%), N0 (62%). Twenty out of 118 (17%) colostomies fashioned before CRT were reversed within 8 months. One hundred and twelve patients had a post-treatment colostomy due to persistent disease (98) or morbidity (14). Fifty-two per cent (61/118) of all pre-treatment colostomies were never reversed. The 5-year CFS rates were 68% MMC/Maint, 70% CisP/Maint, 68% MMC/No-maint and 65% CisP/No-maint. CRT with CisP did not improve CFS when compared with MMC (hazard ratio: 1.04, 95% confidence interval: 0.82-1.31, P = 0.74). The 5-year CFS rates were higher for T1/T2 (79%) than T3/T4 (54%) tumours and higher for node-negative (72%) than node-positive (60%) patients. Significant predictors of CFS were gender, T-stage and haemoglobin, while treatment factors had no impact on outcome. Similar associations were found between PFS and tumour/treatment-related factors. CONCLUSIONS: The majority (52%) of pre-treatment colostomies were never reversed. Neither CRT with 5FU/CisP nor maintenance chemotherapy impacted on CFS. The low risk of colostomy for late effects (1.7%) is likely to be associated with the modest total radiotherapy dose. The predictive factors for CFS were T-stage, gender and baseline haemoglobin. CLINICAL TRIAL REGISTRATION NUMBER: ISRCTN 26715889.

Efficacy of Varunadi Ghritha (polyherbal compound) in treated head and neck cancer cases as a biological response modifier.

BACKGROUND: Persistent immune suppression is reported in Head and Neck Cancers (HNC) even after treatment and a higher recurrence rate was observed in patients with poor CD3 count. Loco regional recurrences and second primary tumours are the common forms of failure in head and neck cancers. Several agents have been tried to overcome this problem without much benefit. In Ayurveda, several plant based products have been reported to have anti-tumour and immunomodulatory properties. AIM: To test the role of Varunadi Ghritha, as an immunomodulator in apparently healthy, treated and controlled HNC patients and to evaluate its effectiveness in preventing locoregional relapses and development of second primary tumours. MATERIALS AND METHODS: Total 78 patients of treated head and neck cancers were randomly selected for intervention and control group. Patients in the intervention group (n = 38) received Varunadi Ghritha, 5gms twice daily for one year and followed up to two years. Patients in the control group (n = 40) were followed up at regular intervals. Immune parameters were assessed in the peripheral blood at base line and at the end of administration of the study compound. RESULTS: In the intervention group, mean percentage increase in CD3, CD19 and CD16 positive cells were significantly higher after the administration of the study compound compared to the control group indicating an immunomodulatory effect of the study compound. A non-significant improvement in disease control was observed in patients with advanced stage of disease in the intervention group. CONCLUSION: Administration of Varunadi Ghritha resulted in an increase in T cell counts in patients with treated HNC.