Purpose: Loss of function or expression of the mismatch repair gene MLH1 has been implicated in experimentally acquired resistance to cisplatin (CDDP) and other anticancer agents. The clinical significance of MLH1 expression was evaluated in advanced thoracic squamous cell carcinoma of the esophagus (ESCC) treated by neoadjuvant chemotherapy.

Experimental Design: We investigated MLH1 and P53 expression by immunohistochemistry in the surgical specimens of 107 patients who had undergone preoperative chemotherapy using CDDP along with 5-FU and ADM. These findings were correlated with the clinical outcome for this treatment. Biopsy samples before chemotherapy in 20 of these patients, and another 43 surgical specimens without chemotherapy, were also examined as control samples.

Results: In surgical specimens of ESCC, low MLH1 expression was not frequent without chemotherapy, whereas it was commonly observed after chemotherapy (14 versus 37%, P = 0.0057). Comparison between samples before and after chemotherapy revealed that MLH1 expression was unchanged during chemotherapy in 12 of 20 patients (60%) but was from high to low in 8 of 20 patients (40%). In the surgical specimen after neoadjuvant chemotherapy, MLH1 expression was not correlated with any clinicopathological factors, including the response to chemotherapy. However, low MLH1 showed poorer prognosis than high MLH1 (5-year survival 40.6 versus 19.3%, P = 0.0393), and in multivariate analysis, MLH1 was an independent prognostic factor for this multimodal treatment, following lymph node metastasis and clinical response to chemotherapy. Positive p53 expression, which was not affected by chemotherapy, was weakly associated with a poor response and clinical outcome, although this trend was not significant.

Conclusions: In advanced ESCC, expression of MLH1 is reduced during CDDP-based chemotherapy, and this may partly account for poor postoperative survival.

ESCC2 shows rather high clinical efficacy for chemotherapy among the digestive tract cancers. Thus, we have used preoperative (neoadjuvant) chemotherapy for advanced ESCCs since 1989 to improve operative curability and prognosis. The principle of neoadjuvant chemotherapy generally postulates two discordant properties of cancer cells: (a) major chemosensitivity and (b) minor chemoresistance. Surgery is required for the latter. Chemosensitive tumors can be eradicated by repeating chemotherapy alone, but chemoresistant tumors lead to poor prognosis even if they were resected by surgery. The mechanism for acquisition of chemotherapy resistance is not well understood. Clonal selection of chemo-resistant cells or induction chemo-resistance by epigenetic or genetic event are considered to occur during chemotherapy.

The MMR system detects and repairs damaged DNA, including mismatches, small insertions and deletions, and drug-induced adducts (1). Furthermore, MMR works as a DNA checkpoint system, which generates apoptotic signals when DNA damage is not repaired appropriately (2). Disorder of the MMR system is frequently implicated in carcinogenesis of various cancers, such as hereditary nonpolyposis colorectal carcinoma (3). One possible mechanism is that loss of MMR causes mutations of the promoter lesion and induces transcription in several oncogenes, including transforming growth factor-βIIR (4) and insulin-like growth factor receptor (5). Recently, another role of the MMR system has been revealed to be directly involved in the drug sensitivity of cancer cells. Chemo-resistance by MMR deficiency is strongly related to the checkpoint function, rather than the MMR function. Loss of the checkpoint function permits cancer cells to survive with DNA damaged by chemotherapy; furthermore, damaged DNA is inherited as a gene mutation and may cause further malignant transformation. The MMR system involves several molecules, including MLH1, MSH2, MSH3, MSH6, and PMS2 (6). Among them, loss of MLH1 is associated with resistance to various anticancer agents, including CDDP, ADM, and 5-FU (7, 8, 9, 10). In the cultured cell lines, loss of MLH1 appears after treatment with these drugs, and MLH1-deficient cells exhibit resistance to them compared with MLH1-proficient counterparts (7). In clinical samples, after CDDP-based chemotherapy, expression of MLH1 was decreased in ovarian cancer (11) and breast cancer (12). Moreover, MSI, which results from disorder of the MMR system and loss of MLH1 protein, is frequently induced during CDDP-based chemotherapy (13). It is of clinical importance that the loss of MLH1 or MSI is significantly associated with the clinical failure of chemotherapy.

To understand the implications of MLH1 in acquired chemo-resistance of ESCC, we investigated MLH1 expression in cancer tissue samples before and after and also with or without chemotherapy using CDDP, ADM, and 5-FU. We also examined the influence on the clinical outcome.

Patients and Specimens.

A group of 156 patients with advanced thoracic esophageal squamous cell carcinomas, who were diagnosed as T3 and T4 (UICC TNM), underwent surgical treatment at Osaka University Medical School and Osaka Medical Center for Cancer and Cardiovascular Diseases from 1994 to 2000. A group of 113 patients underwent neoadjuvant chemotherapy followed by esophagectomy, whereas 43 patients, who did not consent to chemotherapy, underwent only surgery. The treatment regimen of preoperative chemotherapy in our department has been described elsewhere (14). In brief, chemotherapy consisted of administration of 70 mg/m2 CDDP, 35 mg/m2 ADM as a drip i.v. infusion for 2 h on day 1, and 700 mg/m2 5-FU as a continuous i.v. infusion for 24 h on day 1–7 for a month. Four to 6 weeks after completing chemotherapy, surgical treatment, including subtotal esophagectomy and regional lymph nodes resection, was performed. A group of 103 patients underwent curative operation without residual tumor (R0), whereas in the remaining 10 cases, microscopic and/or macroscopic tumors remained (R1–2). Most patients received two courses of chemotherapy regimen, except for five patients who received three courses. R0 operation had been done for 94 patients (92%) of the former group and three patients (60%) of the latter. Six cases with no residual tumor in the surgical specimens after chemotherapy (grade 3) were excluded; thus, 150 patients comprising 107 with and 43 without chemotherapy were enrolled in this study.

The removed esophagus was fixed in 10% formaldehyde and sliced into 5-mm serial longitudinal sections. A representative section from each patient with the deepest tumor infiltration was subjected to immunohistochemistry examination. Small pieces (5 mm) of tumor and noncancerous mucosa were removed and kept at −80°C for Western blot analysis. In 20 cases with neoadjuvant chemotherapy, biopsy samples before chemotherapy were also available for immunohistochemistry.

Immunohistological Staining Procedure and Evaluation of Staining.

Immunostaining was performed using streptavidin-preoxidase complex methods for MLH1 and P53. Primary antibodies were incubated overnight at room temperature with for MLH1 [G168-15, dilution 1:100 (PharMingen)] and at 4°C for p53 [DO-7, dilution 1:100 (Novocastra Laboratories, Newcastle, United Kingdom)]. For MLH1, intact nuclear staining of adjacent non-neoplastic epithelium served as an internal positive control. The immunoreactivity of MLH1 was evaluated according to the intensity and frequency of the positive nuclear stained cells, as reported previously (12). The staining intensity of each cancer cell was scored as no staining = 0; weaker intensity than normal epithelium = 1; same intensity as normal epithelium = 2; and stronger intensity than normal epithelium = 3. Scores 2 and 3 were classified as positive cancer cells. Tumors containing >50% positive cancer cells were classified as high MLH1 and those with ≤50% positive cancer cells as low MLH1. Immunohistochemical evaluation of MLH1 was repeated twice by two pathologists (K. K. and Y. D.). Among several criteria for MLH1 evaluation, this cutoff line yielded the most reproducible results and the best clinical significance for postoperative survival. For P53 staining, when >10% of the cancer cells showed positive staining, the tumors were classified as positive, following other P53 immunohistochemical studies (15).

Immunoblot Analysis.

The tissue samples were minced and homogenized with lysis buffer [10 mm Tris/HCl (pH 7.4), 150 mm NaCl, 1% Triton X-100, 5 mm DTT, 0.1 mm phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, and 2 μg/ml leupeptin], and the extract was centrifuged at 14,000 rpm for 25 min at 4°C. Aliquots (60 μg) of supernatant proteins were separated by 8% PAGE, followed by electroblotting onto a PVDF membrane. Antihuman MLH1 mouse monoclonal antibody [G168-15, dilution 1:250 (PharMingen, San Diego, CA)] was used. Detection of the protein bands was performed using the Amersham enhanced chemiluminescence detection system (Amersham, Arlington Heights, IL) according to the manufacturer’s instructions. An equal amount of protein from each tissue extract was confirmed by immunoblot by β-actin [A2066, dilution 1:200 (Sigma, St. Louis, MO)].

Evaluation of the Effect of Chemotherapy.

The effect of neoadjuvant chemotherapy was evaluated by two different modalities. The clinical response of chemotherapy was evaluated by tumor size reduction on the computed tomography scan before and after chemotherapy, according to RECIST criteria (16), as CR, 100% regression of the disease; PR, with >30% decrease in the sum of the longest diameter of the targets; progressive disease, with >20% increase; and stable disease, for the remainders. The histological response of chemotherapy was evaluated by the proportion of viable cancer cells against whole cancer tissue on the H&E-stained section of the surgical specimens according to the Japanese Society for Esophageal Diseases criteria (17): grade 0, no histological effect; grade 1a, viable cancer cells accounted for more than two-thirds of the tumor tissue; grade 1b, viable cancer cells accounted for between one-third and two-thirds of the tumor tissue; grade 2, viable cancer cells account for less than one-third of the tumor tissue; and grade 3, no residual viable cancer cells. Samples of the last grade were excluded from this study. CR and PR in clinical response and grade 1b and 2 in histological response were regarded as effective.

Statistical Analysis.

Associations between expression and parameters in clinicopathological factors were analyzed by the Mann-Whitney U, Spearman’s rank correlation, or Fisher’s exact tests. Changes of expression during chemotherapy were analyzed by Wilcoxon signed-rank test. Ten patients who underwent noncurative operation (R1–2) were excluded from the survival analysis. Cause-specific postoperative survival was analyzed by the Kaplan-Meier method for the remaining 97 patients and was statistically assessed by a Log-rank test (median follow-up period is 38.3 months in high MLH1 and 42.5 months in low MLH1). A Cox proportional hazard model was used to assess the risk ratio with simultaneous contribution from several covariates. In all analyses, Ps < 0.05 were considered statistically significant. All statistical analyses were performed using the software package Stat View ver. 5.0 (Abacus Concepts, Inc., Berkeley, CA).

MLH1 Expression in Esophageal Cancers with or Without Chemotherapy.

The noncancerous squamous epithelium of the esophagus expressed MLH1 in the basal and parabasal layers. In the interstitial tissue, lymphocytes also expressed MLH1 weakly, but that in the germinal centers expressed it strongly. Most of the ESCC expressed more intense MLH1 staining in their nuclei, relative to the normal squamous epithelium. However, cancer cells sometimes lost MLH1 expression homogeneously or heterogeneously or showed weaker MLH1 expression in the center of the colonies (Fig. 1). Immunohistochemical findings for MLH1 were confirmed by immunoblot analysis using representative surgical specimens with distinct MLH1 expression. A band with strong intensity was located at Mr 80–85,000, which corresponds to the molecular weight of MLH1. A strong band appeared in high MLH1 tumor samples and normal tissue, but the bands of the samples of low MLH1 were faint or totally absent from the blot (Fig. 2).

When MLH1 expression was classified as high or low, the low level was not common, observed in 14% (6 of 43) advanced ESCC patients without chemotherapy. Low MLH1 was more frequently detected in 37.4% (40 of 107) patients after chemotherapy, and the difference of the frequency was statistically significant (P = 0.0057; Table 1). There was no significant difference in the clinicopathological background between ESCC patients with or without chemotherapy. The staining pattern of MLH1 reduction with or without chemotherapy was basically identical. In 20 patients, biopsy samples before chemotherapy were available for immunohistochemistry and compared with their surgical specimens after chemotherapy. Among 19 high MLH1 tumors before chemotherapy, 11 (57.9%) had been equally evaluated as high MLH1 after chemotherapy, whereas 8 (42.1%) turned out to be of low MLH1 after chemotherapy. Only 1 was of low MLH1 in the biopsy specimen before chemotherapy with no change after chemotherapy (Table 2).

Positive expression of P53 was observed in 70.1% (75 of 107) of tumors after neoadjuvant chemotherapy, and this frequency was similar to 60.5% (26 of 43) from surgical specimens without chemotherapy. In paired samples before and after chemotherapy, expression of P53 did not change in the majority of cases (18 of 20; Tables 1 and 2).

Clinical Significance of MLH1 Expression in Patients with Neoadjuvant Chemotherapy and Surgery.

Because loss of MLH1 was induced during chemotherapy, we tried to reveal its clinical significance among 107 patients with neoadjuvant chemotherapy and surgery. MLH1 expression in 107 and 97 (RO) surgical specimens after chemotherapy did not show any correlation with clinicopathological features, including sex, age, tumor location, histological grade, tumor length, tumor depth, and lymph node metastasis (Table 3). The effect of chemotherapy was evaluated by both tumor size reduction in radiographical examination and histological examination of the removed esophagus. There was no significant correlation between MLH1 expression after chemotherapy and the effect of chemotherapy, showing CR + PR in 64.2% (43 of 67) and grade1b-2 29.9% (20 of 67) of high MLH1 patients but 57.5% (23 of 40) and 22.5% (9 of 40) of low MLH1 patients (P = 0.541 and 0.5024, respectively) (Table 4a). In contrast, P53 expression in surgical specimen after chemotherapy tended to be correlated with tumor size reduction by chemotherapy, showing CR + PR in 56% (42 of 75) of P53 (+) patients but 75% (24 of 32) of P53 (−) patients (P = 0.083). However, very little difference was seen in the histological examination of the removed esophagus, showing grade1b-2 28% (21 of 75) of P53 (+) patients but 25% (8 of 32) of P53 (−) patients (P = 0.8161) (Table 4b).

The cause-specific survival rate of 97 ESCC patients treated by neoadjuvant chemotherapy and surgery, excluding grade 3 and noncurative surgery, was shown to be 69.2, 39.5, and 33.5% at 1, 3, and 5 years. Among the clinicopathological factors, lymph node metastasis (hazard ratio 2.6, P = 0.0224) and the effect of chemotherapy (clinical response: hazard ratio 2.657, P = 0.0019; histological response: hazard ratio 2.351, P = 0.0303) were revealed to be significant prognostic factors, whereas tumor depth was not (hazard ratio 1.318, P = 0.4204; Table 5). MLH1 was a significant prognostic factor, showing 5-year survival rates of 40.6% for high MLH1 and 19.3% for low MLH1 (P = 0.0393). Patients with P53(−) tended to show better prognosis than P53(+) patients, although the difference was not statistically significant (5-year survival rate 47.2% for P53(−) and 27.8% for P53(+), P = 0.184; Fig. 3, A and B). Next, multivariate analysis was performed with lymph node metastasis, the effect of chemotherapy (clinical and histological response) and MLH1 expression as variables. Low MLH1 expression as well as lymph node metastasis and poor clinical response were revealed to be independent significant prognostic factors for separating the poor prognosis group (Table 5). Because interaction of the apoptotic signal from MMR and p53 has been reported, postoperative survival was evaluated according to the coexpression pattern of MLH1 and P53. As shown in Fig. 3 C, ESCC patients with disorder of both molecules, i.e., low MLH1/P53(+), exhibited extremely poor prognosis (5-year survival 11.6%), and the remaining three patterns exhibited similar survival curves (5-year survival 36.3–50%). Among 64 patients who died of esophageal cancer, the first tumor recurrence was observed in the lymph nodes (28, 43.8%), as hematogenic metastasis (19, 29.7%), both in lymph nodes and as hematogenic metastasis (15, 23.4%), and in others sites (2, 3.1%). No special trend was present in the recurrence organ between patients with or without chemotherapy and with or without MLH1 expression.

Novel insight of the MLH1 function is focused on its implication for drug resistance of cancer cells. In the present study, we found that MLH1 disappeared during neoadjuvant chemotherapy with CDDP, ADM, and 5-FU, and low MLH1 was associated with poor survival of ESCC patients treated by neoadjuvant chemotherapy and surgery. The decline of MLH1 expression has been shown by comparison of biopsy samples before chemotherapy with surgical specimens after chemotherapy, although the number of patients was limited. Because biopsy samples may not always represent the MLH1 expression of whole tumor attributable to heterogeneity of cancer cells, we supported this observation by comparing surgical specimens with or without preoperative chemotherapy in the larger cohort.

To date, MLH1 has not been investigated well in human ESCC, probably because of the observation that MSI is not frequent in ESCC, and MLH1 is rarely disturbed in squamous cell carcinoma of other organs. In contrast, MSI and loss of MLH1 expression is frequently observed and should play a central role in the carcinogenesis of other digestive cancers, including colorectal cancers (18), gastric cancers (19), and adenocarcinoma of the esophagus (20). Interestingly, studies of these cancers have revealed that although disorder of the MMR system is strongly associated in the carcinogenesis, it does not contribute to the malignant potentiality of the cancer cells. Thus, tumors with MSI or loss MLH1 show similar or sometimes better prognosis compared with those without these disorders. Concerning ESCC in this series, although the number is small, there was no difference in the prognosis between low and high MLH1 tumors when chemotherapy was not performed (data not shown).

Reduction of MLH1 expression was frequently observed after chemotherapy and was associated with poor prognosis, in ESCC as well as breast cancers (12, 21) and T-cell leukemia (22). What needs to be explained is that MLH1 expression did not correlate with the effect of chemotherapy, and both MLH1 expression and chemotherapy effect were independently involved in the prognosis in this study. Because chemotherapy was repeated only twice as a neoadjuvant therapy, its effect might reflect the survival of high MLH1 cells, which account for most of the cancer tissue before chemotherapy and whose drug sensitivity is regulated by unknown genes other than MLH1. Poor prognosis is attributable to tumor recurrence mostly as metastases. It would be of interest to investigate MLH1 expression in metastatic or recurrent tumors. Some patients were treated three times by neoadjuvant chemotherapy because the effect was not enough for curative operation. Although curative surgery was done, all were low for MLH1 and died of tumor relapse.

The function of MLH1 in repairing DNA mismatches or DNA adducts by anticancer drug is still being elucidated, and it remains uncertain how MLH1 generates proapoptotic signals, when detecting unrepairable DNA damage. Several experiments have reported that MLH1 is associated with the G2-M checkpoint (23); it requires the rad gene to induce apoptosis (24), and loss of p53 enhances the drug resistance by MLH1 deletion (25, 26). In the present study, we observed this synergy of p53 and MLH1 dysfunction related to the poorest prognosis for this treatment. A novel mechanism of drug resistance is demonstrated by MLH1 deficiency. When loss of DNA repair (i.e., MLH1 deficiency) and mutagenic agents (i.e., anticancer drugs) are combined, genomic mutations should rapidly accumulate and finally affect other drug-resistant genes. Point mutation of O6-methylguanine-DNA-methyl transferase gene, a well-known drug resistance gene, was induced by BCNU and O6-benzylguanine treatment in a very short time in MLH1-deficient cells (27). Likewise, amplification of the carbamyl-P synthetase/asparate transcarbamylase/dihydroorotase gene has been done by N-phosphonacetyl-L-aspartase treatment in MLH1-deficient cells (28). We are concerned that such mutations involve not only drug-resistant genes but also other oncogenes related to tumor malignancy potentiality. Continuous chemotherapy which is not effective may result in further malignant potentiality and poorer prognosis than no treatment.

This MLH1 study leads us to a novel strategy for ESCC, especially for its relapse after multimodal treatment. For the patients with high MLH1, the same chemotherapy, including CDDP, ADM, and 5-FU, might still be effective; however, for those with low MLH1, this chemotherapy might not be effective even it had been in preoperative treatment. The reduction of MLH1 during chemotherapy may be attributable to DNA methylation, and therefore, de-methylating agents, such as 5-aza-2′-deoxycytidine, might lead to recovery of MLH1 expression and chemo-sensitivity (29). Such combination therapy is now undergoing clinical trial. It is of interest that, contrary to our findings, MLH1-deficient cells are more sensitive than MLH1-proficient cells to some anticancer agents, including camptothecin and etoposide (30). This is probably because the DNA repair function of MLH1 is more implicated than its checkpoint function in its effectiveness as an antitumor agent. With respect to the use of radiation, some studies show that radiation sensitivity is not affected by MSI and MLH1 status (31), whereas others have shown that MLH1-deficient tumor might be more sensitive because of G2-M abrogation (32). The possibility of the usefulness of radiotherapy still remains in low MLH1 tumors. Treatment of ESCC patients has become more complicated with improvements in chemotherapy and radiotherapy. Biological markers, which are now extensively investigated by molecular techniques, such as cDNA microarray, would play a central role in the determination of therapeutic modality. The relationship between MLH1 and drug sensitivity is not only of interest in basic science but also should provide unique and novel insights for the clinical treatment of ESCC patients in the future.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This work was supported by Grants-in-Aid for Scientific Research (C; No. 09671303 and 10671185) and a Grant-in-Aid for Encouragement of Young Scientist (No. 09770936) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

2

The abbreviations used are: ESCC, esophageal squamous cell cancer; CDDP, cisplatin; ADM, doxorubicine; 5-FU, 5-fluorouracil; MMR, mismatch repair; MSI, microsatellite instability; PR, partial response; CR, complete response.

Fig. 1.

Immunohistochemical expression of MLH1 in ESCCs immunohistochemical staining of MLH1 in normal squamous epithelium (A). Esophageal cancers accompanied by nuclear MLH1 expression were classified as high (B) or low (C and D), according to the frequency and intensity of the stained cells. In some cases, colonies of positive and negative cells coexisted (E). Original magnification: ×200 (A–D), ×100 (E).

Fig. 1.

Immunohistochemical expression of MLH1 in ESCCs immunohistochemical staining of MLH1 in normal squamous epithelium (A). Esophageal cancers accompanied by nuclear MLH1 expression were classified as high (B) or low (C and D), according to the frequency and intensity of the stained cells. In some cases, colonies of positive and negative cells coexisted (E). Original magnification: ×200 (A–D), ×100 (E).

Close modal
Fig. 2.

Comparison of MLH1 expression in normal esophageal tissues and esophageal cancers by Western blot analysis and immunostainning. For a high MLH1 tumor in immunohistochemistry, a strong band appeared in cancer tissue in the Western blot analysis (Case 2), whereas for low MLH1 tumors, the bands for MLH1 were absent (Case 1) or weaker (Case 3) than the adjacent noncancerous epithelium. The band of MLH1 was located at Mr 80–85,000. The cell extract from TE2 was used as a positive control for MLH1 expression, and the bands for actin showed the equality of the loaded samples. N, normal epithelium; T, esophageal cancer; IHC, immunohistochemical expression of MLH1.

Fig. 2.

Comparison of MLH1 expression in normal esophageal tissues and esophageal cancers by Western blot analysis and immunostainning. For a high MLH1 tumor in immunohistochemistry, a strong band appeared in cancer tissue in the Western blot analysis (Case 2), whereas for low MLH1 tumors, the bands for MLH1 were absent (Case 1) or weaker (Case 3) than the adjacent noncancerous epithelium. The band of MLH1 was located at Mr 80–85,000. The cell extract from TE2 was used as a positive control for MLH1 expression, and the bands for actin showed the equality of the loaded samples. N, normal epithelium; T, esophageal cancer; IHC, immunohistochemical expression of MLH1.

Close modal
Fig. 3.

Survival curves for ESCC patients treated by neoadjuvant chemotherapy and surgery, according to the status of MLH1 and p53. Cause-specific survival curves were plotted by the Kaplan-Meier method for 97 patients with neoadjuvant chemotherapy and surgery. Patients undergoing noncurative surgery (R1 and R2) were excluded. Patients at grade 3 were also excluded because expression of MLH1 and p53 was evaluated with surgical specimens. Survival curves were classified by MLH1 expression (A), p53 expression (B), and coexpression of MLH1 and p53 (C). Five-year survival rates are indicated for each curve. Ps were calculated using Log-rank tests.

Fig. 3.

Survival curves for ESCC patients treated by neoadjuvant chemotherapy and surgery, according to the status of MLH1 and p53. Cause-specific survival curves were plotted by the Kaplan-Meier method for 97 patients with neoadjuvant chemotherapy and surgery. Patients undergoing noncurative surgery (R1 and R2) were excluded. Patients at grade 3 were also excluded because expression of MLH1 and p53 was evaluated with surgical specimens. Survival curves were classified by MLH1 expression (A), p53 expression (B), and coexpression of MLH1 and p53 (C). Five-year survival rates are indicated for each curve. Ps were calculated using Log-rank tests.

Close modal
Table 1

Expression of MLH1 in surgical specimens with or without neoadjuvant chemotherapy in ESCC patients

A. MLH1 expressiona
HighLowTotal
Neoadjuvant chemotherapy Treated 67 (62.6%) 40 (37.4%) 107 (100%) 
 Untreated 37 (86.0%) 6 (14.0%) 43 (100%) 
B. P53 expressionb     
  Positive Negative Total 
Neoadjuvant chemotherapy Treated 75 (70.1%) 32 (29.9%) 107 (100%) 
 Untreated 26 (60.5%) 17 (39.5%) 43 (100%) 
A. MLH1 expressiona
HighLowTotal
Neoadjuvant chemotherapy Treated 67 (62.6%) 40 (37.4%) 107 (100%) 
 Untreated 37 (86.0%) 6 (14.0%) 43 (100%) 
B. P53 expressionb     
  Positive Negative Total 
Neoadjuvant chemotherapy Treated 75 (70.1%) 32 (29.9%) 107 (100%) 
 Untreated 26 (60.5%) 17 (39.5%) 43 (100%) 
a

Fisher’s exact test; P = 0.0057.

b

Fisher’s exact test; P = 0.3357.

Table 2

Comparison of MLH1 expression before and after chemotherapy in ESCC patients

A. MLH1 expressionPostaTotal
HighLow
Preb High 11 19 
 Low 
 Total 11 20 
B. P53 expression  Posta  Total 
  Positive Negative  
Preb Positive 12 12 
 Negative 
 Total 14 20 
A. MLH1 expressionPostaTotal
HighLow
Preb High 11 19 
 Low 
 Total 11 20 
B. P53 expression  Posta  Total 
  Positive Negative  
Preb Positive 12 12 
 Negative 
 Total 14 20 
a

Post, surgery specimen after chemotherapy.

b

Pre, prechemotherapy biopsy sample.

Table 3

Relationship between expression of MLH1 and clinicopathological parametersa for the patients with neoadjuvant chemotherapy and surgery (RO)

MLH1 expressionTotal n = 97 (%)
High n = 61 (%)Low n = 36 (%)
Age 58.9 ± 8.3 59.0 ± 6.8 59.0 ± 7.8 
Gender    
 Male 49 (80.3) 33 (91.7) 82 (84.5) 
 Female 12 (19.7) 3 (8.3) 15 (15.5) 
Locationb    
 Ut 12 (19.7) 5 (13.9) 17 (17.5) 
 Mt 28 (45.9) 18 (50.0) 46 (47.4) 
 Lt 21 (34.4) 13 (36.1) 34 (35.1) 
Histological type    
 G1 17 (27.9) 4 (11.1) 21 (21.6) 
 G2 30 (49.1) 22 (61.1) 52 (54.6) 
 G3 14 (23.0) 10 (27.8) 24 (24.8) 
Depth of invasion    
 T3 21 (34.4) 13 (36.1) 34 (35.1) 
 T4 40 (65.6) 23 (63.9) 63 (64.9) 
LN metastasis    
 NO 18 (29.5) 8 (22.2) 26 (26.8) 
N1 and/or M1 lymph 43 (70.5) 28 (77.8) 71 (73.2) 
MLH1 expressionTotal n = 97 (%)
High n = 61 (%)Low n = 36 (%)
Age 58.9 ± 8.3 59.0 ± 6.8 59.0 ± 7.8 
Gender    
 Male 49 (80.3) 33 (91.7) 82 (84.5) 
 Female 12 (19.7) 3 (8.3) 15 (15.5) 
Locationb    
 Ut 12 (19.7) 5 (13.9) 17 (17.5) 
 Mt 28 (45.9) 18 (50.0) 46 (47.4) 
 Lt 21 (34.4) 13 (36.1) 34 (35.1) 
Histological type    
 G1 17 (27.9) 4 (11.1) 21 (21.6) 
 G2 30 (49.1) 22 (61.1) 52 (54.6) 
 G3 14 (23.0) 10 (27.8) 24 (24.8) 
Depth of invasion    
 T3 21 (34.4) 13 (36.1) 34 (35.1) 
 T4 40 (65.6) 23 (63.9) 63 (64.9) 
LN metastasis    
 NO 18 (29.5) 8 (22.2) 26 (26.8) 
N1 and/or M1 lymph 43 (70.5) 28 (77.8) 71 (73.2) 
a

Based on Tumor-Node-Metastasis classification diagnosed by computed tomography scan and esophagograph before chemotherapy.

b

Ut, upper thoracic esophagus; Mt, middle thoracic esophagus; Lt, lower thoracic esophagus.

Table 4A

Relationship between expression of MLH1 and chemotherapy effect

MLH1Clinical responseHistological responseTotal
CR + PRaSD + PDaGrade 1b–2aGrade 0–1aa
High 43 (64.2%) 24 (35.8%) 20 (29.9%) 47 (70.1%) 67 (100%) 
Low 23 (57.5%) 17 (42.5%) 9 (22.5%) 31 (77.5%) 40 (100%) 
 66 41 29 78 107 
MLH1Clinical responseHistological responseTotal
CR + PRaSD + PDaGrade 1b–2aGrade 0–1aa
High 43 (64.2%) 24 (35.8%) 20 (29.9%) 47 (70.1%) 67 (100%) 
Low 23 (57.5%) 17 (42.5%) 9 (22.5%) 31 (77.5%) 40 (100%) 
 66 41 29 78 107 
Table 4B

Relationship between expression of P53 and chemotherapy effect

P53Clinical responseHistological responseTotal
CR + PRaSD + PDaGrade 1b–2aGrade 0–1aa
Positive 42 (56.0%) 33 (44.0%) 21 (28.0%) 54 (72.0%) 75 (100%) 
Negative 24 (75.0%) 8 (25.0%) 8 (25.0%) 24 (75.0%) 32 (100%) 
 66 41 29 78 107 
P53Clinical responseHistological responseTotal
CR + PRaSD + PDaGrade 1b–2aGrade 0–1aa
Positive 42 (56.0%) 33 (44.0%) 21 (28.0%) 54 (72.0%) 75 (100%) 
Negative 24 (75.0%) 8 (25.0%) 8 (25.0%) 24 (75.0%) 32 (100%) 
 66 41 29 78 107 
a

CR, PR, SD, PD, Grade 0–1a, and Grade 1b–2 were described in “Materials and Methods.”

Table 5

Prognostic factors in ESCC patients treated by neoadjuvant chemotherapy and surgery by Cox proportional hazard model

Univariate analysisMultivariate analysis
Risk ratio95% CIPRisk ratio95% CIP
Depth of invasion       
 T3      
 T4 1.318 0.673–2.579 0.4204 N.D.   
LN status       
 NO     
N1 and/or M1 lymph 2.600 1.145–5.904 0.0224 2.815 1.185–6.691 0.0191 
Clinical response       
 CR + PR     
 SD + PD 2.657 1.435–4.923 0.0019 2.281 1.199–4.340 0.0119 
Histological response       
 Grade 1b–2     
 Grade 0–1a 2.351 1.085–5.096 0.0303 1.482 0.651–3.373 0.3489 
MLH1 expression       
 High     
 Low 1.888 1.022–3.488 0.0425 2.020 1.146–4.231 0.0178 
P53 expression       
 Negative      
 Positive 1.617 0.790–3.309 0.1881 N.D.   
Univariate analysisMultivariate analysis
Risk ratio95% CIPRisk ratio95% CIP
Depth of invasion       
 T3      
 T4 1.318 0.673–2.579 0.4204 N.D.   
LN status       
 NO     
N1 and/or M1 lymph 2.600 1.145–5.904 0.0224 2.815 1.185–6.691 0.0191 
Clinical response       
 CR + PR     
 SD + PD 2.657 1.435–4.923 0.0019 2.281 1.199–4.340 0.0119 
Histological response       
 Grade 1b–2     
 Grade 0–1a 2.351 1.085–5.096 0.0303 1.482 0.651–3.373 0.3489 
MLH1 expression       
 High     
 Low 1.888 1.022–3.488 0.0425 2.020 1.146–4.231 0.0178 
P53 expression       
 Negative      
 Positive 1.617 0.790–3.309 0.1881 N.D.   

N.D., not done; CI, confidence interval.

1
Fink D., Aebi S., Howell S. B. The role of DNA mismatch repair in drug resistance.
Clin. Cancer Res.
,
4
:
1
-6,  
1998
.
2
Gong J. G., Costanzo A., Yang H. Q., Melino G., Kaelin W. G., Jr., Levrero M., Wang J. Y. The tyrosine kinase c-Abl regulates p73 in apoptotic response to cisplatin-induced DNA damage.
Nature (Lond.)
,
399
:
806
-809,  
1999
.
3
Bronner C. E., Baker S. M., Morrison P. T., Warren G., Smith L. G., Lescoe M. K., Kane M., Earabino C., Lipford J., Lindblom A., et al Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer.
Nature (Lond.)
,
368
:
258
-261,  
1994
.
4
Lu S. L., Kawabata M., Imamura T., Akiyama Y., Nomizu T., Miyazono K., Yuasa Y. HNPCC associated with germline mutation in the TGF-beta type II receptor gene.
Nat. Genet.
,
19
:
17
-18,  
1998
.
5
Cui H., Horon I. L., Ohlsson R., Hamilton S. R., Feinberg A. P. Loss of imprinting in normal tissue of colorectal cancer patients with microsatellite instability.
Nat. Med.
,
4
:
1276
-1280,  
1998
.
6
Fishel R. The selection for mismatch repair defects in hereditary nonpolyposis colorectal cancer: revising the mutator hypothesis.
Cancer Res.
,
61
:
7369
-7374,  
2001
.
7
Aebi S., Kurdi-Haidar B., Gordon R., Cenni B., Zheng H., Fink D., Christen R. D., Boland C. R., Koi M., Fishel R., Howell S. B. Loss of DNA mismatch repair in acquired resistance to cisplatin.
Cancer Res.
,
56
:
3087
-3090,  
1996
.
8
Drummond J. T., Anthoney A., Brown R., Modrich P. Cisplatin and adriamycin resistance are associated with MutLalpha and mismatch repair deficiency in an ovarian tumor cell line.
J. Biol. Chem.
,
271
:
19645
-19648,  
1996
.
9
Fink D., Nebel S., Aebi S., Zheng H., Cenni B., Nehme A., Christen R. D., Howell S. B. The role of DNA mismatch repair in platinum drug resistance.
Cancer Res.
,
56
:
4881
-4886,  
1996
.
10
Meyers M., Wagner M. W., Hwang H. S., Kinsella T. J., Boothman D. A. Role of the hMLH1 DNA mismatch repair protein in fluoropyrimidine-mediated cell death and cell cycle responses.
Cancer Res.
,
61
:
5193
-5201,  
2001
.
11
Samimi G., Fink D., Varki N. M., Husain A., Hoskins W. J., Alberts D. S., Howell S. B. Analysis of MLH1 and MSH2 expression in ovarian cancer before and after platinum drug-based chemotherapy.
Clin. Cancer Res.
,
6
:
1415
-1421,  
2000
.
12
Mackay H. J., Cameron D., Rahilly M., Mackean M. J., Paul J., Kaye S. B., Brown R. Reduced MLH1 expression in breast tumors after primary chemotherapy predicts disease-free survival.
J. Clin. Oncol.
,
18
:
87
-93,  
2000
.
13
Watanabe Y., Koi M., Hemmi H., Hoshai H., Noda K. A change in microsatellite instability caused by cisplatin-based chemotherapy of ovarian cancer.
Br. J. Cancer
,
85
:
1064
-1069,  
2001
.
14
Doki Y., Ishikawa O., Mano M., Hiratsuka M., Sasaki Y., Kameyama M., Ohigashi H., Murata K., Yamada T., Miyashiro I., Yokoyama S., Ishiguro S., Imaoka S. Cytokeratin deposits in lymph nodes show distinct clinical significance from lymph node micrometastasis in human esophageal cancers.
J. Surg. Res.
,
107
:
75
-81,  
2002
.
15
Miyata H., Doki Y., Shiozaki H., Inoue M., Yano M., Fujiwara Y., Yamamoto H., Nishioka K., Kishi K., Monden M. CDC25B and p53 are independently implicated in radiation sensitivity for human esophageal cancers.
Clin. Cancer Res.
,
6
:
4859
-4865,  
2000
.
16
Therasse P., Arbuck S. G., Eisenhauer E. A., Wanders J., Kaplan R. S., Rubinstein L., Verweij J., Van Glabbeke M., van Oosterom A. T., Christian M. C., Gwyther S. G. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada.
J. Natl. Cancer Inst. (Bethesda)
,
92
:
205
-216,  
2000
.
17
Japanese Society for Esophageal Diseases .
Guidelines for the Clinical and Pathologic Studies on Carcinoma of the Esophagus
, Ed. 9
55
-58, Kanehara Tokyo  
1999
.
18
Miyakura Y., Sugano K., Konishi F., Ichikawa A., Maekawa M., Shitoh K., Igarashi S., Kotake K., Koyama Y., Nagai H. Extensive methylation of hMLH1 promoter region predominates in proximal colon cancer with microsatellite instability.
Gastroenterology
,
121
:
1300
-1309,  
2001
.
19
Fleisher A. S., Esteller M., Tamura G., Rashid A., Stine O. C., Yin J., Zou T. T., Abraham J. M., Kong D., Nishizuka S., James S. P., Wilson K. T., Herman J. G., Meltzer S. J. Hypermethylation of the hMLH1 gene promoter is associated with microsatellite instability in early human gastric neoplasia.
Oncogene
,
20
:
329
-335,  
2001
.
20
Eads C. A., Lord R. V., Wickramasinghe K., Long T. I., Kurumboor S. K., Bernstein L., Peters J. H., DeMeester S. R., DeMeester T. R., Skinner K. A., Laird P. W. Epigenetic patterns in the progression of esophageal adenocarcinoma.
Cancer Res.
,
61
:
3410
-3418,  
2001
.
21
Tomita S., Deguchi S., Miyaguni T., Muto Y., Tamamoto T., Toda T. Analyses of microsatellite instability and the transforming growth factor-beta receptor type II gene mutation in sporadic breast cancer and their correlation with clinicopathological features.
Breast Cancer Res. Treat.
,
53
:
33
-39,  
1999
.
22
Hayami Y., Komatsu H., Iida S., Utsunomiya A., Hanada S., Hua X. J., Huiping N., Harada S., Tsuboi K., Banno S., Wakita A., Kato T., Ueda R. Microsatellite instability as a potential marker for poor prognosis in adult T cell leukemia/lymphoma.
Leuk. Lymphoma
,
32
:
345
-349,  
1999
.
23
Brown R., Hirst G. L., Gallagher W. M., McIlwrath A. J., Margison G. P., van der Zee A. G., Anthoney D. A. hMLH1 expression and cellular responses of ovarian tumour cells to treatment with cytotoxic anticancer agents.
Oncogene
,
15
:
45
-52,  
1997
.
24
Paulovich A. G., Margulies R. U., Garvik B. M., Hartwell L. H. RAD9, RAD17, and RAD24 are required for S phase regulation in Saccharomyces cerevisiae in response to DNA damage.
Genetics
,
145
:
45
-62,  
1997
.
25
Vikhanskaya F., Colella G., Valenti M., Parodi S., D’Incalci M., Broggini M. Cooperation between p53 and hMLH1 in a human colocarcinoma cell line in response to DNA damage.
Clin. Cancer Res.
,
5
:
937
-941,  
1999
.
26
Branch P., Masson M., Aquilina G., Bignami M., Karran P. Spontaneous development of drug resistance: mismatch repair and p53 defects in resistance to cisplatin in human tumor cells.
Oncogene
,
19
:
3138
-3145,  
2000
.
27
Liu L., Schwartz S., Davis B. M., Gerson S. L. Chemotherapy-induced O(6)-benzylguanine-resistant alkyltransferase mutations in mismatch-deficient colon cancer.
Cancer Res.
,
62
:
3070
-3076,  
2002
.
28
Chen S., Bigner S. H., Modrich P. High rate of CAD gene amplification in human cells deficient in MLH1 or MSH6.
Proc. Natl. Acad. Sci. USA
,
98
:
13802
-13807,  
2001
.
29
Strathdee G., Sansom O. J., Sim A., Clarke A. R., Brown R. A role for mismatch repair in control of DNA ploidy following DNA damage.
Oncogene
,
20
:
1923
-1927,  
2001
.
30
Jacob S., Aguado M., Fallik D., Praz F. The role of the DNA mismatch repair system in the cytotoxicity of the topoisomerase inhibitors camptothecin and etoposide to human colorectal cancer cells.
Cancer Res.
,
61
:
6555
-6562,  
2001
.
31
Yan T., Schupp J. E., Hwang H. S., Wagner M. W., Berry S. E., Strickfaden S., Veigl M. L., Sedwick W. D., Boothman D. A., Kinsella T. J. Loss of DNA mismatch repair imparts defective cdc2 signaling and G(2) arrest responses without altering survival after ionizing radiation.
Cancer Res.
,
61
:
8290
-8297,  
2001
.
32
Davis T. W., Wilson-Van Patten C., Meyers M., Kunugi K. A., Cuthill S., Reznikoff C., Garces C., Boland C. R., Kinsella T. J., Fishel R., Boothman D. A. Defective expression of the DNA mismatch repair protein, MLH1, alters G2-M cell cycle check point arrest following ionizing radiation.
Cancer Res.
,
58
:
767
-778,  
1998
.