P04.08 Biopsy-only glioblastoma (BO-GBM) as a heterogeneous group of patients

BACKGROUND

“Biopsy-only” glioblastoma is associated with a heterogeneous functional and survival outcome. BO-GBM patients is an understudied group of patients associated to a poor outcome, which has been reported to represent 21% of histologically confirmed GBM in the US National Cancer Data Base. Pattern of care included radiotherapy-temozolomide (RT-TMZ) standard regimen completed in 15% of patients, any other form of oncologic treatment in 60%, and supportive care alone in 25% of patients. Our objective was to explore pattern of care and prognosis associated to BO-GBM in our center.

MATERIAL AND METHODS

Patients with BO-GBM included in a prospective regional glioma SIRIC cohort initiated in 2014 and closed in 2017 were retrospectively reviewed for patients characteristics, MRI finding, treatment allocation and delivery. PFS and OS were analyzed.

RESULTS

Of 535 patients included in the cohort, 86 patients were referred > 3 months post-surgery and were excluded from this analysis while 449 patients were included at initial surgery, of which 158 patients (35%) underwent biopsy only. Of 158 patients, 18 patients were excluded for missing data leaving 139 patients for the present analysis. Fifty-four (39%) were referred to RT-TMZ (50 patients completed concomitant treatment), 68 (49%) considered unfitted for RT received chemotherapy upfront (CT-UF) (of which 4 were subsequently referred to RT), 17 (12%) were referred to palliative care only (PC). Groups differed at baseline for age (mean 60, 68, and 69 years, for RT-TMZ, CT-UF, and PC respectively); for KPS (70, 60, and 50 for RT-TMZ, CT-UF, and PC respectively); for mean tumor surface measured on gadolinium-enhanced T1-weighted (793, 1420, 1412 mm2 for RT-TMZ, CT-UF, PC); for tumor extension (bilateral in 6.4% and 29.3% for RT-TMZ and CT-UF respectively); for mean steroid intake (45, 60, 100 mg daily respectively). Median OS was 14 months (95% CI, 9.65–18.71), 8 months (95% CI, 4.62–7.67), and 2 months (95% CI, 0.67–3.33) for RT-TMZ, CT-UF, and PC respectively.

CONCLUSION

Inoperable GBM constitute a large and heterogeneous population in which one third of patients are amenable to standard of care, with survival outcome close to the one of patients who underwent surgery. Patients considered unfit for RT-TMZ at diagnosis fail to be referred subsequently to RT after CT and exhibit a poor survival outcome. Thus, reliable criteria are needed to help selecting patients for adequate treatment while new strategies are warranted for BO-GBM unfit for RT.

Limited impact of prognostic factors in patients with recurrent glioblastoma multiforme treated with a bevacizumab-based regimen

Bevacizumab has demonstrated activity in patients with recurrent glioblastoma. However, the impact of prognostic factors associated with recurrent glioblastoma treated with cytotoxic agents has not been determined in patients treated with bevacizumab. To analyze the prognostic factors and clinical benefits of bevacizumab and irinotecan treatment in patients with recurrent glioblastoma. This monocentric study retrospectively analyzed all patients with recurrent glioblastoma who were treated with at least one cycle of bevacizumab and irinotecan at our institution from April 2007 to May 2010. Multivariate analysis was used to analyze prognostic factors for overall survival (OS) from the initiation of bevacizumab administration. Among the 100 patients that were identified (M/F: 65/35), the median age was 57.9 years (range: 18-76). Karnofsky Performance Status (KPS) was <70 in 44 patients and ≥ 70 in 56 patients; 83 % of the patients were on steroids. The median tumor area was 2012 mm². The median progression free survival was 3.9 months (CI 95 %: 3.4-4.3). The median OS was 6.5 months (CI 95 %: 5.6-7.4). Multivariate analysis revealed that OS was affected by KPS (p = 0.024), but not by gender, age, steroid treatment, number of previous lines of treatment, tumor size, or time from initial diagnosis. KPS was improved in 30 patients, including 14/44 patients with an initial KPS <70. The median duration of maintained functional independence (KPS ≥ 70) was 3.75 months (CI 95 %: 2.9-4.6). The median OS from initial diagnosis was 18.9 months (CI 95 %: 17.5-20.3). In patients with recurrent glioblastoma treated with bevacizumab, KPS was revealed as the only factor to impact OS. The clinical benefits associated with this regimen appear valuable. A positive impact of bevacizumab administration on OS of patients with glioblastoma multiforme is suggested.

[Cardiac tamponade with anterior interventricular vein thrombosis complicating central venous catheter insertion in a neonate]

Tamponade is a rare but particularly serious complication of central venous catheters in the newborn. Tamponade can be due to the endocardic aggression caused by the continuous flow of a hyperosmotic solution or by a mechanical injury that can result in perforation of the atrial wall. The risk of tamponade is present whatever is the position of the tip of the catheter, although it has been shown that this risk is increased when this tip is in the right auricle. The originality of our observation is the discovery at the post-mortem examination of an anterior interventricular vein thrombosis, without any lesion of the atrial wall. In the event of the diagnosis of tamponade in living newborn, this etiology must be required because of its therapeutic implications.

Impact of Karnovsky performance status (KPS) on outcome of elderly patients (pts) with glioblastoma (GBM) and activity of temozolomide (TMZ) as first line therapy: Retrospective analysis of a cohort

1515

Background: Due to increasing incidence of GBM in the elderly, prognostic factors and therapeutic strategies need to be considered in this population. Recently radiotherapy has shown survival improvement in patients with KPS ≥ 70 (ANOCEF, EANO 2005), while chemotherapy with TMZ may be considered as a therapeutic option (Chinot at al, Cancer 2003). Methods: We analyzed retrospectively all pts above 70 years old with GBM, who were referred to our institution from May 1998 to October 2004; all responses to TMZ were reviewed. Results: We identified 136 pts registered. Median age was 74 (range 70–87), and 43% had a KPS < 70. Surgery consisted of stereotactic biopsy (SB), 29% ; partial surgery (PS), 12%; and gross total removal (GTR), 29% of pts. Diagnosis was strongly suggested by neuroradiology in 30% of cases. Treatment consisted of TMZ (5 days standard schedule) (group A) as first-line treatment in 89 pts (65%), radiotherapy and/or nitroso-urea based regimens in 40 pts (30%) (group B) and best supportive care in 7 (5%) pts. For the all cohort, median of overall survival (OS) was 7 months (m), strongly impacted by KPS (3.9 vs 8.7 m, p<0.0001 for KPS < 70 and ≥70 respectively) and age (8.2 vs 6.0 m, p<0.007 for age < 75 years vs. ≥ 75). OS was 6.6, 7.3, and 8.4 mos in the case of SB, PS, and GTR respectively and was 5.2 for neuroradiologic diagnosis. In group A, median time to tumour progression (TTP) and OS were 4.7 & 7.3 m. KPS impacted TTP (2.9 vs. 5.1 m, p=0.0002) and OS (4.9 vs. 8.7 m, p<0.0001) for KPS < and ≥ 70 respectively. Response rates (RR) were for 71 pts evaluable 28%; SD 35%; PD 37% associated with an OS of 11.7, 7 and 3.2 m respectively. RR was 34% for histologically proven GBM, vs 22% in cases of neuroradiologic diagnosis. In group B, TTP and OS were 4.3 & 6.7 m respectively. Conclusions: KPS appeared to have a major impact on outcomes in elderly patients with GBM. Future trials designs should take this impact into consideration. TMZ appeared to be effective in elderly patients with newly diagnosed GBM. This alternative approach is currently being tested against RT alone in international trials. Impact of MGMT status in the TMZ population will be presented.

No significant financial relationships to disclose.

MGMT expression correlates with response rate and survival in patients with inoperable glioblastoma (GBM) treated with neoadjuvant temozolomide (TMZ)

1574

Background: Methylation of the promotor of O6-alkylguanine alkyltransferase (MGMT), a DNA repair gene, may enhance chemosensitivity to alkylating agents. In GBM, this methylation has been correlated to survival as well as to the benefit of adding TMZ concomitant and adjuvant to radiotherapy (RT) (Hegi, NEJM, 2005). We examine the relationship between MGMT expression and objective response rate to dose intense TMZ schedule administered as neoadjuvant treatment before RT in inoperable GBM, as previously presented (Chinot, ASCO, 2005). Methods: Thirty patients were included in this phase II trial that tested TMZ (150 mg/m2/day) on days 1 to 7 and 15 to 21 of each 28 days cycle for up to 4 cycles prior to RT. We analysed retrospectively MGMT expression by immunochemistry (streptavidin-peroxydase) after antigen retrieval using anti-MGMT antibody (Abcys, 1/100) in 25 formalin-fixed paraffin embedded samples from the study population. Results: In the eligible population (n = 28) response rates (RR) were of 25% (95% CI, 8.63% to 41.37%); SD 32%; PD 43%. Median progression free survival (PFS) and overall survival (OS) were 3.8 and 5.8 months, respectively. MGMT expression was analysed in 25 pts while material was considered as inadequate in 3 pts because of insufficient tumor material. The median percentage of cells that expressed MGMT in tumor nuclei was 35% and so was chosen as cut-off. Low MGMT expression was significantly associated with a high RR (55%) while tumor that exhibit high MGMT expression was associated to a RR of 9% (chi-2 p=0.004). MGMT was also strongly correlated to PFS (log rank p=0.009) and OS (log rank p=0.003). Conclusion: Despite limited number of patients, our study strongly supports the predictive value of MGMT expression for objective response to TMZ in addition to its prognostic value for PFS and OS in GBM. If confirmed in prospective study, MGMT expression may help to guide therapeutic decisions as well as more targeted trial design.

[Table: see text]

No significant financial relationships to disclose.