Biology Contribution
Rad51 Protein Expression and Survival in Patients with Glioblastoma Multiforme

https://doi.org/10.1016/j.ijrobp.2009.03.018Get rights and content

Purpose

Treatment of glioblastoma multiforme (GBM) continues to pose a significant therapeutic challenge, with most tumors recurring within the previously irradiated tumor bed. To improve outcomes, we must be able to identify and treat resistant cell populations. Rad51, an enzyme involved in homologous recombinational repair, leads to increased resistance of tumor cells to cytotoxic treatments such as radiotherapy. We hypothesized that Rad51 might contribute to GBM's apparent radioresistance and consequently influence survival.

Methods and Materials

A total of 68 patients with an initial diagnosis of GBM were retrospectively evaluated; for 10 of these patients, recurrent tumor specimens were used to construct a tissue microarray. Rad51 protein expression was then correlated with the actual and predicted survival using recursive partitioning analysis.

Results

Rad51 protein was elevated in 53% of the GBM specimens at surgery. The Rad51 levels correlated directly with survival, with a median survival of 15 months for patients with elevated Rad51 compared with 9 months for patients with low or absent levels of Rad51 (p = .05). At disease recurrence, 70% of patients had additional increases in Rad51 protein. Increased Rad51 levels at disease recurrence similarly predicted for improved overall survival, with a mean survival of 16 months from the second craniotomy compared with only 4 months for patients with low Rad51 levels (p = .13).

Conclusion

Elevated levels of the double-stranded DNA repair protein Rad51 predicted for an increase survival duration in patients with GBM, at both initial tumor presentation and disease recurrence.

Introduction

Glioblastoma multiforme (GBM), a devastating malignancy, has a median overall survival of approximately 12 months. Although surgical resection plays an integral role in the treatment of this disease, the infiltration of tumor cells into the surrounding normal brain tissue prohibits a cure by resection alone. Temozolomide chemotherapy plus radiotherapy (RT) has been shown to improve survival. However, despite radiation doses of 60 Gy, most GBM recurrences develop within 2 cm of the tumor bed—the same area targeted by RT 1, 2. Clearly, a better understanding of what mediates this tumor resistance is needed. In response to RT, several stress-response proteins are upregulated, promoting tumor cell survival by leading to their resistance to future insults. Other researchers have shown that the Rad51 protein, in a similar manner, increases in GBM cell lines hours after their exposure to ionizing radiation, presumably also in a defensive response to repair the newly formed DNA damage (3).

The repair of double-strand breaks in somatic cells is mediated through homologous recombinational repair. In eukaryotes, Rad51 plays a crucial role in the homologous recombinational repair pathway by facilitating strand transfer between broken sequences and their undamaged homologues (4). The Rad51 family includes Rad51 and five genes that share strong homology with it: XRCC2, XRCC3, Rad51L1, Rad51L2, and Rad51L3. Cells with mutations in homologous recombinational repair genes exhibit high levels of genetic instability and sensitivity to cross-linking agents and RT (5). A number of reports have demonstrated that Rad51 is not only implicated in the progression of carcinogenesis, but also that Rad51 affects the resistance to anticancer treatments (6). We have a long-standing interest in the treatment of GBM and the causes of radiation resistance seen in this disease. Therefore, we tested GBM tumor specimens specifically for Rad51 activity.

Given Rad51's role in double-strand break repair and ability to promote radiation resistance, we hypothesized that tumor cells with elevated Rad51 at diagnosis might have an increased probability of initially surviving treatment. Furthermore, the cells with elevated Rad51 would become the surviving clonogens responsible for repopulation, subsequently leading to local recurrence and reduced survival. To test this hypothesis, we retrospectively analyzed patient tissue collected at the original diagnostic craniotomy and stained it for Rad51. We then correlated the levels of Rad51 with overall survival. We also analyzed the change in Rad51 level in patients with matched specimens (from both the initial GBM diagnosis and after disease recurrence) to evaluate their overall survival and predicted survival using recursive partitioning analysis (RPA).

Section snippets

Methods and Materials

The data from 68 patients diagnosed with GBM between 1980 and 2002 were retrospectively collected for this intuitional review board-approved study. The inclusion criteria consisted of a pathologic diagnosis of World Health Organization Grade IV astrocytoma and tumors specimens with enough remaining tissue for analysis. From these patients, we collected age, Karnofsky performance status (KPS), gender, date of diagnosis, date of death, date of recurrence, extent of surgery, radiation dose, and

Results

A total of 68 patients with resected GBM were evaluated for their baseline (at diagnosis) Rad51 protein expression levels. The patient characteristics are summarized in Table 1. Of the 68 patients, 46 were men and 22 were women, with a median age of 59 years (range, 22–80). The median KPS was 90 (range, 50–100), and 75% of the patients had a KPS of ≥70. We were able to calculate the RPA score for 43 of the 68 patients. The RPA class was 3 in 10 (25%), 4 in 6 (14%), 5 in 20 (45%), and 6 in 7

Discussion

The purpose of the present study was to determine whether elevated Rad51 at the initial resection would correlate with increased resistance and predict the survival of patients with GBM who had undergone RT. We found that compared with the normal surrounding brain tissue, 53% of the brain tumors had elevated levels of Rad51 protein. Patients with elevated expression of Rad51 at diagnosis had a greater chance of survival than patients with low or absent expression of Rad51. To our knowledge,

Acknowledgments

We thank all members of our laboratory and the shared experimental radiobiology core; we also thank Mike Berens at TGen, who provided us with samples from their multiple-tissue microarrays.

References (12)

There are more references available in the full text version of this article.

Cited by (0)

Supported by the Department of Radiation Oncology, University of Arizona College of Medicine, immunohistochemical and histologic data generated by the Tissue Acquisition and Cellular/Molecular Analysis Shared Service were supported by the Arizona Cancer Center Support Grant (National Institutes of Health Grant CA02307).

Conflict of interest: none.

View full text