Long-term performance of interstial fluid pressure and hypoxia as prognostic factors in cervix cancer

Objectives: Hypoxia and high interstitial fluid pressure (IFP) have been shown to independently predict for nodal and distant metastases, as well as survival, in patients with cervix cancer. Using data from our prospective trial, we updated a cohort of patients treated with definitive radiation alone without chemotherapy, to assess the long-term prognostic impact of these microenvironmental features. Methods: Between April 1994 and January 1999, 107 eligible patients with cervix cancer were entered into a prospective study of tumor oxygenation and IFP prior to primary radiation therapy. Oxygenation data are presented as the hypoxic proportion, defined as the percentage of pO2 readings <5 mm Hg (abbreviated as HP5). Patients with HP5 values >50% were considered to have hypoxic tumors. IFP is presented in mm Hg, divided into high and low IFP groups by the median value. Patients ranged in age from 23 to 78 years with a mean of 53 years. The maximum tumor size ranged from 2 to 10 cm, with a median diameter of 5 cm. FIGO stage was IB in 28 patients, IIA in 4, IIB in 42 and IIIB in 33 patients. Twenty-two patients (21%) had evidence of pelvic lymph node involvement on staging CT abdomen/pelvis or MR pelvis. HP5 ranged from 0% to 99% with a median of 48%. IFP ranged from -3 to 48 mm Hg (median 19 mm Hg). Median follow-up was 6.7 years (range 0.9-10.6). Results: Disease-free survival (DFS) at 5 years was 50%. Five year DFS was 42% for patients with hypoxic tumors (HP5 > 50%), and 58% in patients with oxygenated tumors (HR 1.01 per %, p = 0.05). DFS at 5 years was 42% for patients with interstitial hypertension (IFP >19 mm Hg), and 63% in patients with IFP ≤19 mm Hg (HR 1.05 per mm Hg, p = 0.001). In a multivariate analysis only tumor size (HR 1.2, p = 0.009) pelvic nodal metastases (HR 3.3, p = 0.0004) and IFP (HR 1.06, p = 0.0005) were predictive of DFS. Because an interaction between nodal status and oxygenation was observed (p = 0.03), further analysis indicated a borderline significant impact of HP5 in addition to tumor size in node negative patients (HR 1.01, p = 0.06). These results were similar when local or distant relapse was used as an endpoint. Conclusions: These results confirm our initial finding of the strong independent prognostic impact of IFP for relapse and survival in patients with cervix cancer. In contrast, the independent prognostic impact of HP5 is of borderline significance and is limited to patients without imaging evidence of nodal metastases. However, these findings do not diminish the biologic significance of hypoxia, or the role of hypoxia and IFP as biomarkers of treatment response and as therapeutic targets