TP53 Genotype but Not p53 Immunohistochemical Result Predicts Response to Preoperative Short-Term Radiotherapy in Rectal Cancer

Abstract

Objective

To evaluate and compare the predictive power of p53 gene analysis versus p53 immunohistochemical staining in terms of response to preoperative short-term radiotherapy using 25 Gy in operable rectal cancer.

Summary Background Data

Recent studies show that p53 may be a determinant of radiosensitivity being required for induction of apoptosis in case of radiation-induced DNA damage.

Methods

Preirradiation biopsy samples of 64 patients with rectal carcinoma were analyzed. Genetic alterations of the p53 gene were detected by complete direct sequencing of exons 2 to 10. Expression of the nuclear phosphoprotein p53 was assessed by immunohistochemical staining. Results were correlated with histopathology of resected specimens and follow-up data, respectively.

Results

Mutations of the p53 gene were present in 45% of tumors. Patients with a normal p53 gene had a significant survival advantage. Comparing pre- and postradiotherapy T category, a reduction was seen in patients with normal p53 genotype only. A mutant p53 genotype was highly specific in indicating stable disease concerning T category after irradiation. Protein overexpression was detected in 61%. Overexpression of the p53 protein was not related to survival or response. The concordance between immunohistochemistry and sequencing was only 0.51.

Conclusions

The authors show that downstaging after short-term radiation may occur but is seen in tumors with normal p53 gene only. Moreover, p53 genotype but not p53 immunohistochemistry is predictive for response to preoperative short-term radiotherapy and patient survival.

In view of local recurrence rates of up to 50% after “curative” resection of rectal cancer, preoperative pelvic radiotherapy was introduced. ^1–3 Short-term preoperative radiotherapy using 25 Gy resulted in significant improvement in overall survival, disease-free survival, and local control compared with surgery only and with postoperative radiotherapy. ⁴ In addition, the implementation of the technique of total mesorectal excision has been shown to be associated with a significant reduction in local recurrence rates. ⁵ Today the combination of both strategies is recommended as a standard procedure for treatment of locally advanced rectal cancer, although there is still no agreement as to the optimal intensity and sequencing of adjuvant treatment. ⁶

The effect of radiotherapy is based on induction of DNA damage. DNA damage is a well-known trigger for p53-dependent apoptosis induction. ^7,8 The p53 gene is the most commonly mutated gene in cancer, with a prevalence of about 50% in rectal cancer. ⁹ Functional p53 has been shown to be crucial in determining the radiosensitivity of tumor cells. ¹⁰

The value of p53 to predict outcome of various anticancer treatments has been investigated in several studies. Data suggesting a correlation between response to (neo)adjuvant treatment and p53 status are inconsistent and differ among different cancers. Most data have been accrued from studies using immunohistochemical analysis to detect p53 protein overexpression, which is a possible but not exclusive determinant of mutation in the p53 gene. ^11–14

Data on p53 gene analysis generated from sufficient numbers of patients to be correlated with clinical results are rare but appear to be more consistent regarding their relation to treatment response. ^15–18

For colorectal cancer, p53 gene mutations have been reported to be associated with decreased survival. ¹⁹ Some studies report p53 protein overexpression to be a useful parameter for predicting tumor response to radiotherapy. ^9,20

The aim of our study was to evaluate whether p53 genotype or p53 protein overexpression, or both, are predictive factors of tumor response to radiotherapy and outcome of patients with operable distal rectal cancer.

METHODS

Patients

Sixty-four consecutive patients (21 women, 43 men; mean age 64.9) with resectable rectal carcinoma were studied retrospectively. These patients had been recruited from 1994 to 1998 for a multicenter pilot study implementing preoperative short-term radiotherapy. Patients received 25 Gy in 10 fractions of 2.5 Gy each over 5 consecutive days in 1 week, with surgery to be performed the following week (range 3–6 days). Patient and disease characteristics are shown in Table 1. Tumors were staged according to the TNM system of the UICC. ²¹ Response to radiotherapy was assessed by comparison of pre- and posttreatment T category as assessed by transrectal ultrasound and histopathology, respectively.

Table 1. PATIENT CHARACTERISTICS

Adjuvant chemotherapy (typically 5-fluorouracil and leucovorin) was offered to patients with node-positive tumors.

Tumor Material

Histologic sections from paraffin-embedded preoperative biopsy samples and surgical specimens were used for p53 immunohistochemistry and for DNA extraction for p53 sequencing analysis.

Immunohistochemical Staining

Immunohistochemistry with p53 (DO-1 Ab, 1:20; Immunotech, Marseille, France) was performed on sections of formalin-fixed (7.5% buffered formalin) and paraffin-embedded representative tumor tissues from all 64 carcinomas. Sections were cut to 4 μm, then mounted on silicone-coated slides, deparaffinized, and heated in citrate buffer (pH 6.0) using microwave radiation. Antibodies were incubated overnight at 4°C. Positive reactions were visualized using the avidin-biotin method.

A specimen was scored positive when more than 5% of the tumor cells showed positive nuclear staining reactions (5% cutoff point).

DNA Sequencing

DNA extraction from paraffin-embedded tissue and sequencing analysis of polymerase chain reaction products corresponding to exons 2 to 10 of the p53 gene was performed in our surgical research laboratory as described in detail previously. ¹⁸

Statistical Methods

Cancer-specific survival time was calculated according to the Kaplan-Meier method for TP53 wild-type and TP53 mutant genotypes as well as for positive and negative immunohistochemistry. The log-rank test was used to compare the groups. The concordance between both methods was calculated.

Response to radiotherapy (downstaging) was defined as reduction in depth of tumor invasion (T category) when comparing preoperative endoluminal ultrasound and histopathology of the resected specimen. Associations between response to radiotherapy and TP53 genotype as well as immunohistochemistry were calculated using the Fisher exact test.

For the p53 genotype, sensitivity (the probability that normal genotype indicates response) and specificity (the probability that mutant genotype indicates no change) were calculated, and corresponding 95% confidence intervals are given. P < .05 was considered significant.

RESULTS

Preoperative, histopathologic, and postoperative patient characteristics are given in Table 1. In terms of age, gender, type of resection, and histopathologic features such as grading and blood and lymph vessel invasion, no difference was found between patients with p53 normal tumors (wild-type) and mutant tumors.

There were no postoperative deaths. In all patients, longitudinal and circumferential resection margins were clear. With a median follow-up of more than 3 years (37.37 ± 2.4 months; range 32.7–62.7) in this series of 64 patients, 5 isolated local failures (8%) and 13 distant failures (20%) were observed (the latter all with local control, except one) (see Table 1). Disease-specific survival rates were 77% at 2 years and 67% at 5 years (Fig. 1).

Figure 1. Overall survival of 64 patients with preoperative short-term radiation for rectal cancer. Mean follow-up was 35.4 months (median 37.4).

Overall survival was significantly diminished in patients with abnormal p53 genotype (P = .05) (Fig. 2). In contrast, no difference in relapse-free survival could be found between patients with or without p53 protein overexpression (P = .17).

Figure 2. Kaplan-Meier analysis comparing overall survival of patients with and without TP53 mutations as assessed by complete direct sequencing showing a significant survival advantage for those exhibiting a normal p53 gene (P = .049).

After radiotherapy, a response (defined as reduction in depth of tumor invasion resulting in a reduction of at least one T category) was found in 14 of the 64 patients (22%) compared with the pretreatment level (Fig. 3). In terms of the p53 genotype, T reduction was observed in 14 of 35 tumors with a normal p53 gene (40%); in the group with mutant TP53 gene, no such effect was detected. The mutant p53 genotype was significantly associated with no response to radiotherapy (P = .000). TP53 gene analysis was highly specific in indicating stable disease with respect to changes in T category (100%) (Table 2).

Figure 3. Dynamic of T category in patients without (wild-type) and with (mutant) TP53 mutations. T category was assessed at time of diagnosis clinically (cT category) using transrectal ultrasound and after radiotherapy and surgery by means of pathohistology (pT category). Downstaging occurred exclusively in patients with a normal p53 gene.

Table 2. CORRELATION BETWEEN p53 GENOTYPE AND RESPONSE TO PREOPERATIVE SHORT-TERM RADIATION

P = 0.000.

Specificity: 100% (95% confidence interval 100–88).

Sensitivity: 40% (95% confidence interval 24–58).

Response was determined in changes of T-stage (depth of tumor invasion). Partial remission (PR) was defined as lesser invasion resulting in a reduction of T stage for at least one category. Stable disease (SD) was defined as no change in T category. Complete remission (CR) indicates disappearance of tumor, which was seen in one patient.

No relationship was found between response and result of p53 immunohistochemistry.

The proportion of stage 1 tumors after radiotherapy was found to be almost double in the group with a normal p53 gene (see Table 1), although this difference did not reach statistical significance.

Prevalence of TP53 mutation as assessed by complete direct sequencing was 45% (29/64). Mutations are characterized in Table 3. Overexpression of the p53 protein as assessed by immunohistochemistry was seen in 61% (39 patients). The concordance between TP53 mutation and immunoreactivity was only 0.51.

Table 3. TP53 MUTATIONS CHARACTERIZED IN PATIENTS WITH RECTAL CARCINOMA

DISCUSSION

Genetic testing of the p53 gene and immunohistochemical analysis of p53 protein overexpression were used to determine their respective power to predict response to preoperative short-term radiotherapy in patients with operable distal rectal cancer.

For patients with negative p53 immunohistochemical staining, Adel et al ²² found a significant reduction of local failure when comparing patients with and without preoperative radiotherapy. When they analyzed the 76 patients with radiotherapy, they found a significant reduction of local failure in the group with negative p53 immunohistochemical staining. However, this result was based on only five events of local recurrence, four of which occurred in the group with positive staining. Their local failure rate of 7% in patients after radiotherapy corresponds to ours of 8%, but considering the small number of events we did not include freedom from local relapse in our statistical analysis. In terms of overall survival, we found that a normal p53 genotype was significantly associated with a survival advantage. Adel et al ²² did not report on survival data; Fu et al ²³ showed a trend toward a greater 5-year survival rate in patients staining negative for p53 and positive for p21. However, neither series included genetic analysis of p53. In our analysis we did not observe any predictive power of p53 immunohistochemical staining in terms of overall survival, nor did we find reduction in T category to be related to the immunohistochemical result, whereas p53 gene analysis was predictive for both. In a large study Kressner et al ¹⁶ correlated cancer-specific survival with p53 status, comparing p53 sequencing and immunohistochemical analysis. In accordance to our results, they found that sequencing provided prognostic information but immunohistochemical analysis did not.

In our institution we preformed comparative analyses for three different cancers, looking at the concordance of p53 gene analyses and p53 immunohistochemistry and its relation to response to neoadjuvant treatments. Results of p53 gene analysis were found to be consistent for patients with rectal, lung, and breast cancer in that response to neoadjuvant treatment was significantly related to p53 genotype. Results of immunohistochemical analysis were found to be adequately related to treatment response in patients with breast cancer only. ^17,18

Reviewing the analyses dealing with p53 immunohistochemical and treatment response, inconsistent results are reported, which might be attributed to various problems affecting this technique. ^11,24,25 It has been shown that false-negative immunohistochemical results can be explained by the presence of distinct mutations (e.g., deletions, insertions) that prevent transcription. However, the clinical value of positive immunohistochemistry in the absence of TP53 mutation remains unclear. Drawbacks of the immunohistochemistry technique include interobserver variability and different scoring systems, which may lead to conflicting results. Currently we have no clinically approved scoring standard for judging positivity and/or intensity of p53 immunohistochemical staining. It is not yet clear whether a distinct scoring system can reach the specificity of genetic analysis or even surpass it.

Short-term preoperative radiotherapy, with its advantage of low toxicity and minimal delay until surgery, has been criticized for not providing enough time or even dose to allow tumor shrinkage. In our study, we compared pre- and posttreatment T categories as assessed by transrectal ultrasound and histopathology, respectively. Transrectal ultrasound has been shown to be an accurate modality for staging rectal cancer, in particular for the T category. ^26–28

We observed a reduction of T category in 22% of all patients but in 40% of patients when considering only tumors with a normal p53 gene. In addition, this reduction was exclusive to the group with a normal p53 gene. Whether this downstaging will also translate into fewer local failures in the group with a normal p53 gene will become apparent only in larger studies because the number of events is small. However, we found a higher-than-average proportion of stage 1 tumors in the group with wild-type p53. In the Swedish Rectal Cancer Trial, significantly more patients had stage 1 tumors in the radiotherapy plus surgery group versus the surgery-alone group. One might speculate that this difference may have been caused by response to radiotherapy in TP53 normal tumors.

The 100% association (specificity) between mutant TP53 and stable disease in T category found in our series makes a deficient p53 gene likely to be a crucial factor in treatment failure. Whether this effect is specific for this particular type of radiotherapy protocol merits further study. Whether long-term radiotherapy will result in an even higher response rate considering the p53 genotype remains to be determined.

The p53 genotype was specific in providing prognostic as well as predictive information in patients with preoperative short-term radiotherapy for rectal cancer. In contrast, p53 immunohistochemistry did not contribute to either prognostic or predictive information.

FIGURE

Figure 4. Autoradiography showing sequencing result of a part of exon 7 of the p53 gene. Normal control sequence (left) is compared with mutated sequence in pretreatment biopsy specimen of rectal cancer of patient #R39 (right). This tumor carries a G-A transition (point mutation, arrow) giving rise to a change of the amino acid sequence of the p53 protein by incorporation of serine instead of glycine.

Acknowledgments

The authors thank Wolfgang Pflanzl, Angelika Rainer, and Martin Klimpfinger, heads of the departments of pathology, for their cooperation and support in providing tumor samples in this multicenter study.