<h3>Purpose/Objective(s)</h3> Following standard-of-care chemoradiation therapy (CRT), 30-50% of locally advanced cervical cancer (LACC) patients experience recurrence. Thus, more sophisticated prognostic markers and targeted therapeutic options are needed. The Cancer Genome Atlas (TCGA) molecular classification of cervical cancer did not identify clinically significant patient strata for overall prognosis or risk of recurrence. Notably, the TCGA study did not collect data on clinical factors such as tumor size and nodal status and thus did not incorporate these into their clustering strategies. We hypothesized that at different stages of initial presentation (no nodal disease, locoregional nodal disease, distant nodal disease), different gene programs may be important for progression and resistance to standard of care therapy. <h3>Materials/Methods</h3> We utilized institutional PET and RNA-seq data (<i>N</i>=99) to separate the gene expression profiles of patients with LACC into three groups (Negative nodal (NN) disease, pelvic/locoregional nodal (LRN) disease, aortic/distant (DN) nodal disease) as assessed by pre-treatment PET imaging. Next, to globally assess gene contribution to treatment resistance, we constructed an Rscript to perform univariate Kaplan Meier (KM) analysis for disease recurrence, for all genes in each pre-treatment PET nodal group. FDR multiple hypothesis correction was employed. Lastly, the lists of genes significantly associated with treatment resistance in each PET nodal group were compared. Pathway Enrichment Analysis (PAE) was performed using EnrichR. <h3>Results</h3> When combining all genes statistically associated with recurrence following CRT in at least one nodal cohort, only 7% were shared between two groups (3% LRN-NN, 2% DN-NN, 2% LRN-DN), and <1% was shared by all three groups. Thus, the genes shown to be significant predictors of recurrence following standard of care chemoradiation therapy are largely different depending on the degree of spread of cervical cancer at the time of initiation of therapy (21% NN, 24% DN, 49% LRN). Concordantly, PAE shows striking differences in pathways associated with recurrence. The Immune Response, (p=3.9*10<sup>−6</sup>), Epithelial to Mesenchymal Transition (p=1.6*10<sup>−10</sup>), and MYC/E2F targets (p=5.3*10<sup>−8</sup>) were most enriched in NN, LRN, and DN, respectively. Lastly, expression levels of the top 5 genes associated with recurrence within nodal groups identifies high and low risk clusters of patients' prognostic for recurrence (p=0.00304, p=0.00137, and p=0.019 for NN, LRN, DN, respectively). <h3>Conclusion</h3> Our novel findings suggest that different genes drive cervical cancer progression following CRT depending on the degree of nodal spread at presentation. Utilizing these gene clusters identifies high and low risk groups of patients in each presenting nodal strata.