Improving Selection for Sentinel Lymph Node Biopsy Among Patients With Melanoma

Area of Science:

• Oncology
• Medical Decision Making
• Biostatistics

Background:

• Refining eligibility criteria for medical procedures like sentinel lymph node biopsy (SLNB) is crucial for patient selection.
• Improving the cost-effectiveness of SLNB in melanoma patients is a significant clinical and economic challenge.

Purpose of the Study:

• To enhance the cost-effectiveness of selecting melanoma patients for sentinel lymph node biopsy (SLNB).
• To compare the accuracy of a patient-centered methodology (PCM) against conventional logistic regression for predicting SLNB positivity.

Main Methods:

• A hybrid prognostic study and decision analytical model was employed, analyzing data from melanoma patients eligible for SLNB across two international centers (2000-2014).
• Individualized SLNB positivity probabilities generated by PCM were compared with those from conventional multiple logistic regression using Area Under the Receiver Operating Characteristic Curve (AUROC) and matched-pair analyses.

Main Results:

• The PCM approach demonstrated superior predictive accuracy, with AUROCs of 0.803 (Australia) and 0.826 (US), compared to conventional methods.
• Simulations indicated that adopting PCM-based minimum cutoff probabilities could reduce the number of SLNBs performed while increasing or maintaining the number of positive results.
• A 23.7% PCM cutoff probability reduced SLNB procedures by 49.9% with a 42.7% positivity rate, indicating improved efficiency.

Conclusions:

• The patient-centered methodology (PCM) significantly outperforms conventional logistic regression in predicting SLNB positivity for melanoma patients.
• Implementing PCM-based, context-tailored minimum cutoff probabilities can optimize patient selection for SLNB, enhancing cost-effectiveness over current guidelines.