Originally published as JCO Early Release 10.1200/JCO.2005.04.907 on July 11 2005

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ahlgren, J. D. Search for Related Content PubMed PubMed Citation Articles by Ahlgren, J. D. Related Articles Related Article Social Bookmarking                            What's this?

Thymidylate Synthase in Nodal Metastases: Prognostic for Chemosensitivity But Not for Adjuvant Therapy?

The George Washington University Medical Center, Washington, DC

Every year, more than 150,000 patients are diagnosed in the United States with colorectal cancer. Of these patients, approximately 60,000 will present with stage III disease, with nodal spread but no distant metastases. It is well established that, as a whole, this group will obtain a survival benefit from adjuvant chemotherapy.^1-5

Colorectal cancer patients with lower thymidylate synthase (TS) levels seem to have improved survival. This was evident in the recent meta-analysis by Popat et al,⁶ which showed significant reduction of survival in patients with high TS, including both patients with advanced disease and patients with local disease. In patients with advanced disease, the hazard ratio (HR) for high TS was greater (HR = 2.39; 95% CI, 1.43 to 4.01) when TS was measured in metastases than when it was measured in the primary tumor (HR = 1.33; 95% CI, 1.07 to 1.66). For localized disease, patients with high TS also had poorer overall survival (pooled HR = 1.35; 95% CI, 1.07 to 1.80).

In this issue, Öhrling et al⁷ have addressed the prognostic value of TS level in nodal metastases from colorectal cancer by means of a retrospective analysis of a large database from several Scandinavian adjuvant trials. This study extends the earlier work of these authors in which they examined the prognostic significance of TS level in primary colorectal cancer.⁸ The level of TS expression in lymph nodes (LNs) did provide additional prognostic information beyond that provided by TS expression in the primary tumor. In the population of 348 patients with stage III colorectal cancer from whom archived tumor was available, higher TS expression in LNs predicted poorer prognosis both for disease-free survival (DFS) (P = .04) and overall survival (OS) (P = .02). Even when the TS level in the primary tumor was low, higher TS levels in the LNs provided additional prognostic information. Expression of TS was different between the primary tumor and LNs in 28% of patients. Among the 92 patients with low TS expression in the primary tumor, those with a high TS level in the LNs had significantly poorer cancer-specific survival (P = .02).

It is only in the subgroup analysis of patients receiving or not receiving adjuvant fluorouracil (FU) -based chemotherapy that anomalies arise. Here, LN TS level continued to predict survival in the patients who did not receive adjuvant therapy (DFS, P = .03; OS, P = .04). However, among the patients who received FU-based chemotherapy, the prognostic value of LN TS level failed to reach statistical significance (DFS, P = .2; OS, P = .5). This result is not in agreement with what might be expected based on our current understanding of the influence of TS level on chemosensitivity to FU. Elevated levels of TS have been demonstrated in colon cancer cells with both de novo⁹ and acquired¹⁰ resistance to FU, and lower intratumoral TS level predicted both improved response (P = .001) and survival (P = .02) in patients in one study in metastatic colorectal cancer who were treated with FU-based chemotherapy.¹¹ If lower TS levels predict improved response to FU, one would expect greater, rather than less, benefit from adjuvant FU-based chemotherapy among patients with lower TS levels. Thus, if anything, the survival difference between the low and high TS cohorts would be expected to be greater in patients receiving chemotherapy, rather than smaller. However, similar anomalous results have been previously reported. In fact, four studies^8,12-14 have suggested a greater chemotherapy benefit in patients with high TS levels, and the authors' previous study demonstrated a deleterious effect (significantly reduced survival) for FU-based chemotherapy in patients whose tumors had low TS expression.⁸

Could the unexpected effect of chemotherapy be explained by some bias inherent in the methodology of this study? The authors have looked for such factors. Scoring of the TS level (0 to 1 v 2 to 3) was based on the highest staining area found, even if that area was small. Thus, examining a larger number of samples could increase the opportunity to find a higher scoring region and, thus, lead to a higher TS score. Heterogeneous TS expression was observed among LNs in 37% of patients. The median number of nodes containing tumor was two (range, one to 20 nodes), so patients with higher numbers of nodes might have a higher probability of finding small areas with higher TS scores. The authors analyzed the relationship between TS expression and number of LNs examined and found such a correlation (P = .03). As expected, there was a correlation between survival and number of involved LNs (DFS, P = .05; OS, P = .04). However, in multivariate analysis, only TS expression, not the number of nodes analyzed, retained prognostic significance. Thus, sampling artifact seems unlikely to have been a major source of bias.

Could the finding of small areas of high TS expression be a marker for tumor heterogeneity, suggesting genomic instability as a contributor to prognosis? The authors have examined this possibility as well. In the primary tumor, TS heterogeneity seemed to have no influence on outcome. In the LNs, heterogeneity between nodes showed a trend toward better, rather than poorer, prognosis.

Is it possible that TS expression (at least as assessed by immunohistochemistry) may be a marker for some other feature of tumor biology that might be even more powerful prognostically than the influence of TS level on chemosensitivity? A study by Allegra et al¹⁵ provides food for thought. In that study, low TS activity was associated with higher Ki-67 staining (P = .006) and more frequent overexpresion of p53. In the same study, patients with p53 overexpression had improved survival when administered FU-based adjuvant therapy, whereas patients with normal p53 actually had reduced survival with adjuvant FU (P = .01), which was another unexpected result. However, in that study, overall clinical outcome was not convincingly influenced by TS, Ki-67, or p53 expression, and neither TS nor Ki-67 expression seemed to influence the effectiveness of FU-based adjuvant therapy.

In their earlier study of TS level in the primary tumor, Edler et al⁸ had reported that patients with low TS who were treated with FU-based adjuvant chemotherapy actually had a significantly poorer survival than patients treated with surgery alone (DFS, P = .010; OS, P = .008), whereas patients with high TS expression had a small trend toward improved survival with adjuvant therapy (DFS, P = .2; OS, P = .3), and only patients with the highest TS expression had a significant benefit (DFS, P = .04). This is the opposite trend of what would be predicted on the basis of reduced chemosensitivity with higher TS expression. In fact, there is no currently understood theoretical basis by which to explain how adjuvant chemotherapy would lead to a worsened survival in any subset of patients, as seen in the patients with low TS levels in the previous study⁸ and also as reported by Allegra et al¹⁵ in patients with low p53 expression.

In the present study by Öhrling et al,⁷ this apparent deleterious effect of chemotherapy in patients with low TS expression is no longer seen, but neither patients with low TS nor patients with high TS were observed to have any benefit from FU-based chemotherapy. This remains at odds with most prospective studies. The authors point out that their study is a retrospective study in which their patients received a heterogeneous mix of chemotherapy regimens, without control for dose, duration, or other parameters. Nevertheless, most of the regimens used with their patients have previously shown clear benefit in controlled trials. It is possible that the lack of a chemotherapy effect may represent inadequate statistical power in this retrospective study, but the lack of an overall chemotherapy benefit remains an unexplained result, which must be considered in the context of the unexpected subset results.

Although it is known that increased TS level reduces the efficacy of FU-based chemotherapy in advanced colorectal cancer, data on the utility of TS level (and of molecular markers in general) in predicting the efficacy of adjuvant chemotherapy are sparse and, in some cases, conflicting. It may be that the influence of TS level on the efficacy of adjuvant chemotherapy has yet to be isolated from a host of interrelated confounding variables. The fact that some of the results reported here cannot be adequately explained on the basis of current knowledge of the effect of TS level on chemosensitivity to FU in no way detracts from their importance. The negative influence of elevated LN TS level on survival in the group as a whole has been clearly demonstrated. The anomalous results with chemotherapy should be viewed as a new and different window through which it may be possible to learn more about the biology of tumor metastasis, progression, and chemosensitivity.

As the authors clearly point out in their closing discussion, further prospective randomized trials can offer the opportunity to evaluate the role of TS as a predictor of benefit from adjuvant chemotherapy. With FU-based chemotherapy likely to be the control arm in most new trials of adjuvant therapy in colon cancer, the opportunity now exists to study the effect of TS level within the framework of such trials at minimal additional expense. An appropriate structure would be to prospectively stratify patients for such a trial into low- and high-TS groups. Patients would then be randomly assigned to either an FU-based or non–FU-based regimen. This would allow prospective data on the effect of TS level to be obtained for only the additional cost of obtaining prerandomization TS levels and would not compromise the results of any underlying FU-based versus non–FU-based question in any way. In fact, given what we now know (or don't know) about the effect of TS level on the efficacy of adjuvant therapy in colorectal cancer, the results of the primary question would be strengthened by balancing the arms for one more potential source of bias.

Author's Disclosures of Potential Conflicts of Interest

The author indicated no potential conflicts of interest.

REFERENCES

1. Gill S, Loprinzi CL, Sargent DJ, et al: Pooled analysis of fluorouracil-based adjuvant therapy for stage II and III colon cancer: Who benefits and by how much? J Clin Oncol 22:1797-1806, 2004[Abstract/Free Full Text]

2. Moertel CG, Fleming TR, Macdonald JS, et al: Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma: A final report. Ann Intern Med 122:321-326, 1995[Abstract/Free Full Text]

3. International Multicentre Pooled Analysis of Colon Cancer Trials Investigators: Efficacy of adjuvant fluorouracil and folinic acid in colon cancer: International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) Investigators. Lancet 345:939-944, 1995[CrossRef][Medline]

4. O'Connell MJ, Mailliard JA, Kahn MJ, et al: Controlled trial of fluorouracil and low-dose leucovorin given for 6 months as postoperative adjuvant therapy for colon cancer. J Clin Oncol 15:246-250, 1997[Abstract/Free Full Text]

5. Francini G, Petrioli R, Lorenzini L, et al: Folinic acid and 5-fluorouracil as adjuvant chemotherapy in colon cancer. Gastroenterology 106:899-906, 1994[Medline]

6. Popat S, Matakidou A, Houlston RS: Thymidylate synthase expression and prognosis in colorectal cancer: A systematic review and meta-analysis. J Clin Oncol 22:529-536, 2004[Abstract/Free Full Text]

7. Öhrling K, Edler D, Hallström M, et al: Detection of thymidylate synthase expression in lymph node metastases of colorectal cancer can improve the prognostic information. J Clin Oncol 23:5628-5634, 2005[Abstract/Free Full Text]

8. Edler D, Glimelius B, Hallström M, et al: Thymidylate synthase expression in colorectal cancer: A prognostic and predictive marker of benefit from adjuvant fluorouracil-based chemotherapy. J Clin Oncol 20:1721-1728, 2002[Abstract/Free Full Text]

9. Aschele C, Debernardis D, Casazza S, et al: Immunohistochemical quantitation of thymidylate synthase expression in colorectal cancer metastases predicts for clinical outcome to fluorouracil-based chemotherapy. J Clin Oncol 17:1760-1770, 1999[Abstract/Free Full Text]

10. Clark JL, Berger SH, Mittelman A, et al: Thymidylate synthase gene amplification in a colon tumor resistant to fluoropyrimidine chemotherapy. Cancer Treat Rep 71:261-265, 1987[Medline]

11. Leichman CG, Lenz H-J, Leichman L, et al: Quantitation of intratumoral thymidylate synthase expression predicts for disseminated colorectal cancer response and resistance to protracted-infusion fluorouracil with weekly leucovorin. J Clin Oncol 15:3223-3229, 1997[Abstract]

12. Johnston PG, Fisher ER, Rockette HE, et al: The role of thymidylate synthase expression in prognosis and outcome of adjuvant chemotherapy in patients with rectal cancer. J Clin Oncol 12:2640-2647, 1994[Abstract/Free Full Text]

13. Yamachika T, Nakanishi H, Inada K, et al: A new prognostic factor for colorectal carcinoma, thymidylate synthase, and its therapeutic significance. Cancer 82:70-77, 1998[CrossRef][Medline]

14. Takenoue T, Hirokazu N, Matsuda K, et al: Relation between thymidylate synthase expression and survival in colon carcinoma, and determination of appropriate application of 5-fluorouracil by immunohistochemical method. Ann Surg Oncol 7:193-198, 2000[CrossRef][Medline]

15. Allegra CJ, Parr AL, Wold LE, et al: Investigation of the prognostic and predictive value of thymidylate synthase, p53, and Ki-67 in patients with locally advanced colon cancer. J Clin Oncol 20:1735-1743, 2002[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Article

• Detection of Thymidylate Synthase Expression in Lymph Node Metastases of Colorectal Cancer Can Improve the Prognostic Information
  Katarina Öhrling, David Edler, Marja Hallström, Peter Ragnhammar, and Henric Blomgren
  JCO 2005 23: 5628-5634 [Abstract] [Full Text]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ahlgren, J. D. Search for Related Content PubMed PubMed Citation Articles by Ahlgren, J. D. Related Articles Related Article Social Bookmarking                            What's this?