Current cutaneous squamous cell carcinoma staging systems may not adequately capture patient's potential morbidity.
Clinical Pearls
This was a nested case-control study using population data obtained from 2013-2015 from the National Disease Registration Survey on patients with cutaneous squamous cell carcinoma with (n=887) and without (n=887) metastasis to validate 4 different staging systems (American Joint Committee on Cancer 8th Edition (AJCC8), Brigham and Women’s Hospital (BWH), Tübingen, and Salamanca T3 refinement).
The authors found the BWH had the highest specificity (92.8%, 95% confidence interval 90.8-94.3%), and positive predictive value (PPV: 13.2%, 95%CI 10.6-16.2%). AJCC8 had a greater negative predictive value (NPV: 99.2%, 95%CI 99.2-99.3%), and defined worsening of outcomes with increasing stage (monotonicity) compared to BWH and Tübingen (p<.001).
AJCC8 staging performed worse across measured metrics in a subset analysis of only head and neck cSCCs (Case: n=422; Control: n=422).
Salamanca T3 refinement staging (Case: n=37; Control, n=37) performed worse than all other staging systems across most measured metrics.
Discussion:
Cutaneous squamous cell carcinoma (cSCC) are non-melanoma skin cancers (NMSCs) that are the second most common type of skin cancer, with an estimated incidence of 1 million cases annually. While surgery can be curative and despite many cases of cSCC having good prognosis, up to 4% of cSCCs may be of a more aggressive variant leading to metastasis and, in 2% of cases, death. Several staging systems have been developed to assist in risk-stratifying patients to assist in appropriate healthcare resource allocation.
The authors of this study performed a nested case-control study using population data to validate four existing staging criteria: the AJCC8, BWH, Tübingen, Salamanca. In addition to assessing sensitivity, specificity, NPV and PPV, the authors determined distinctiveness (how different stages within a single system differed from each other), homogeneity (how much similar a single stage is between systems), monotonicity (how much outcomes worsened with increasing stage within a system), and Harrell’s concordance index (c-index: probability a patient with metastasis will be assigned a higher stage). After adjusting the PPV and NPV for an expected 2% metastasis prevalence within the general population, the authors found the BWH had the greatest specificity and PPV while the AJCC8 had higher NPV, homogeneity, and monotonicity than either the BWH or Tübingen. Notably the AJCC8 fared worse in a subset analysis of head and neck cSCCs despite being initially designed for them.
Limitations of the study include limitations to available pathology reports and imputations made to recategorize available cases and controls if parameters were not available in pathology reports and lack of uniform reporting of perineural invasion. Data also excluded genital and oral mucosa cSCCs and therefore cannot be generalized to these tumor types.
Recently, when compared with AJCC8 and BWH staging systems, improved accuracy metrics, including PPV, have been demonstrated using the clinically validated 40-gene expression profile (40-GEP, DecisionDx-SCC, Castle Biosciences Inc) test for predicting individualized risk of metastasis in patients with cSCC. This molecular test classifies patients as low (Class 1), moderate (Class 2A), or high (Class 2B) risk for metastasis within three years of diagnosis based on primary tumor gene expression. The 40-GEP provides independent prognostic value for individualized metastatic risk assessment in cSCC; this ancillary tool can be used by clinicians to complement their current methods for risk assessment.
Venables et al. rightfully indicate the importance of accurately identifying patients at high risk for metastasis so they may be considered for more intense follow-up and surveillance (e.g., imaging, sentinel lymph node biopsy, adjuvant therapy), as clinical trials and development of therapies have been rapidly advancing, giving patients efficacious treatment options. On the other hand, accurate identification of patients at low risk for metastasis is also essential, as this can help avoid excessive surveillance or overtreatment and guide allocation of healthcare resources to the appropriate patients. Overall, this article highlights the shortcomings of existing staging systems, as they are rooted in subjective measurements by pathologists, impacted by treatment choice, and over- or under-estimate risk for many patients.Individualized biologic risk assessment, such as that provided by the 40-GEP test, can contribute additional, objective information about the primary tumor that is independent of clinical presentation to improve risk assessment in patients with cSCC.
Коментари