“…3). Though these variables turned out to be statistically insignificant in CPH analysis, they are still considered important in decision making and prognosis in a clinical setting 18,19 .Figure 4Univariate & Multivariable CPH analyses. Variables are sorted in descending order of hazard ratio.…”
The Cox proportional hazards model commonly used to evaluate prognostic variables in survival of cancer patients may be too simplistic to properly predict a cancer patient’s outcome since it assumes that the outcome is a linear combination of covariates. In this retrospective study including 255 patients suitable for analysis who underwent surgical treatment in our department from 2000 to 2017, we applied a deep learning-based survival prediction method in oral squamous cell carcinoma (SCC) patients and validated its performance. Survival prediction using DeepSurv, a deep learning based-survival prediction algorithm, was compared with random survival forest (RSF) and the Cox proportional hazard model (CPH). DeepSurv showed the best performance among the three models, the c-index of the training and testing sets reaching 0.810 and 0.781, respectively, followed by RSF (0.770/0.764), and CPH (0.756/0.694). The performance of DeepSurv steadily improved with added features. Thus, deep learning-based survival prediction may improve prediction accuracy and guide clinicians both in choosing treatment options for better survival and in avoiding unnecessary treatments.
“…3). Though these variables turned out to be statistically insignificant in CPH analysis, they are still considered important in decision making and prognosis in a clinical setting 18,19 .Figure 4Univariate & Multivariable CPH analyses. Variables are sorted in descending order of hazard ratio.…”
The Cox proportional hazards model commonly used to evaluate prognostic variables in survival of cancer patients may be too simplistic to properly predict a cancer patient’s outcome since it assumes that the outcome is a linear combination of covariates. In this retrospective study including 255 patients suitable for analysis who underwent surgical treatment in our department from 2000 to 2017, we applied a deep learning-based survival prediction method in oral squamous cell carcinoma (SCC) patients and validated its performance. Survival prediction using DeepSurv, a deep learning based-survival prediction algorithm, was compared with random survival forest (RSF) and the Cox proportional hazard model (CPH). DeepSurv showed the best performance among the three models, the c-index of the training and testing sets reaching 0.810 and 0.781, respectively, followed by RSF (0.770/0.764), and CPH (0.756/0.694). The performance of DeepSurv steadily improved with added features. Thus, deep learning-based survival prediction may improve prediction accuracy and guide clinicians both in choosing treatment options for better survival and in avoiding unnecessary treatments.
“…6 Studies addressing this topic are scarce and exclusively of retrospective nature. [9][10][11][12][13][14] Additionally, the described disease constellation is relatively rare. Out of 807 patients who received primary TLM for OSCC, OPSCC, or HPSCC, and whose outcome was reviewed previously, 1,2,15 only 65 (8.1%) met these criteria.…”
Section: Discussionmentioning
confidence: 99%
“…11,12 In line with the present study, for pT1 carcinomas, Kadletz et al described significantly higher RFS estimates for patients who received PORT as well. 14 Most of the previous studies focused on patients with OSCC [10][11][12][13] and/or OPSCC. 12,14 In the present study, carcinomas localized in the oral cavity, oropharynx, and hypopharynx were included and analyzed together as well as separately.…”
Section: Discussionmentioning
confidence: 99%
“…6,8 One reason is that no results of prospective randomized trials addressing this topic are available. Only a few retrospective studies [9][10][11][12][13][14] provide data upon which therapeutic recommendations and rational clinical decision making can be based.…”
Objectives/Hypothesis
Indication for postoperative radiotherapy in patients with locally circumscribed tumors (pT1–pT2) and a single ipsilateral lymph node metastasis (pN1) is debatable. The aim of this study was to evaluate the oncological long‐term outcome of patients with pT1‐pT2 pN1 squamous cell carcinoma (SCC) of the oral cavity, the oropharynx, and the hypopharynx without extracapsular spread (ECS) after a margin‐negative surgical resection, who either received or did not receive postoperative (chemo)radiotherapy.
Study Design
Retrospective case series.
Methods
The oncological outcome of patients with pT1‐pT2 pN1 SCC without ECS was evaluated retrospectively. All patients underwent primary tumor resection that included transoral laser microsurgery and neck dissection at an academic tertiary referral center.
Results
Of 65 identified patients treated between 1986 and 2015 (18 oral cavity, 30 oropharynx, 17 hypopharynx), 21 (32%) received postoperative radiotherapy, and 44 (68%) were treated by surgery alone. The group of patients receiving postoperative treatment showed a significantly superior 5‐year disease‐specific (94.4% vs. 73.2%, P = .029) and recurrence‐free survival (85.2% vs. 43.2%, P = .002), as well as a higher local control rate (90.2% vs. 64.9%, P = .042). The overall survival was 71.4% vs. 62.6% (P = .53). The mean follow‐up was 80.7 months.
Conclusions
Patients with locally circumscribed carcinomas and a single ipsilateral ECS‐negative lymph node metastasis seem to benefit from postoperative radiotherapy.
Level of Evidence
4 Laryngoscope, 130:E530–E538, 2020
“…However, locally advanced disease of this risk status is indicated by tumour‐free surgical margins and/or NM without ECS . Approximately 9%‐42% of people with HNC belong to the IR class . The benefit of postoperative radiotherapy (PORT) may not be uniform across various patient subgroups of IR‐HNC.…”
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
