Article Text

Download PDFPDF

Mutant K-ras2 in serum
  1. H J N Andreyev1,
  2. R Benamouzig2,
  3. M Beranek3,
  4. P Clarke4,
  5. D Cunningham5,
  6. A R Norman5,
  7. W Giaretti6,
  8. A F P M de Goeij7,
  9. B J Iacopetta8,
  10. E Jullian9,
  11. K Krtolica10,
  12. J Q Lee11,
  13. S T Wang11,
  14. N Lees12,
  15. F Al-Mulla13,
  16. O Muller14,
  17. M Pauly15,
  18. V Pricolo16,
  19. A Russo17,
  20. C Troungos18,
  21. N Urosevic19,
  22. R Ward20
  1. 1Imperial College, London UK
  2. 2Hospital Avicenne, Bobigny, France
  3. 3Charles University Hospital, Hradec Kralove, Czech Republic
  4. 4Institute of Cancer Research, Sutton, UK
  5. 5Royal Marsden Hospital, London UK
  6. 6Istituto Nazionale per la Ricerca sul Cancro, Genova, Italy
  7. 7Maastricht University, the Netherlands
  8. 8University of Western Australia, Nedlands, Australia
  9. 9Groupe Hospitalier Cochin-Saint Vincent de Paul, Paris, France
  10. 10Institute of Nuclear Sciences “Vinca” Belgrade, Yugoslavia
  11. 11National Cheng Kung University, Taiwan
  12. 12Wythenshawe Hospital, Manchester, UK
  13. 13Kuwait University, Kuwait
  14. 14Max-Planck-Institut fur Molekulare Physiologie, Dortmund, Germany
  15. 15Centre Universitaire de Luxembourg, Luxembourg
  16. 16Rhode Island Hospital, Brown University, USA
  17. 17University of Palermo, Italy
  18. 18University of Athens, Greece
  19. 19Military Medical Academy, Belgrade, and University of Western Australia, Nedlands, Australia
  20. 20St Vincent’s Hospital, Darlinghurst, Australia
  1. Correspondence to:
    Dr H J N Andreyev, Department of Medicine and Therapeutics, Chelsea and Westminster Hospital, 369 Fulham Rd, London SW10 9NH, UK;

Statistics from

Ryan and colleagues’ careful and well conducted study (Gut 2003;52:101–8) raises once again the interesting issue of whether molecular analysis and knowledge of mutations of the Kirsten ras gene in particular, have a role in the management of patients with colorectal cancer.

The two RASCAL (Kirsten Ras in Colorectal Cancer Collaborative group) studies1,2 which eventually enrolled data from 4266 patients from 42 centres in 21 countries showed that although the frequency of Kirsten ras mutations at codons 12 and 13 may vary a little between populations, overall they are only present in just over one third of patients. This is significantly less frequent than quoted by Ryan et al. In addition, the RASCAL study also showed that Kirsten ras mutations are not associated with sex, age, tumour site, or Dukes’ stage.

Much more importantly, however, they also showed that of the 12 different possible specific point mutations at codons 12 or 13, only a guanine to thymidine mutation, which are found in less than 10% of all patients, is an independent variable for relapse and death. Indeed, the most common mutation, guanine to adenine, exerted an effect on survival barely different from wild-type ras. Others have shown there is a reasonable biological basis for these findings.3,4

As a collaborative group, therefore, we felt it was generally unhelpful to consider Kirsten ras mutations collectively, rather than the effect of each individual mutation separately when considering the prognosis of patients with colorectal cancer. With this in mind, we believe that two crucial questions are left unanswered by the study of Ryan et al. Without definitive answers to these, it is unlikely that serum detection of mutated Kirsten ras gene will ever be clinically useful.

Firstly, detection of any new mutation in the serum of a patient who has had colorectal cancer raises the possibility of tumour recurrence, although the presence of a Kirsten ras mutation is not colon specific. However, more than one third of patients in Ryan’s study did not clinically relapse despite the presence of a serum mutation for very long periods. Others have also detected tumour DNA in patients for long periods without evidence of overt clinical recurrence.5 Therefore, screening for the presence of Kirsten ras mutations per se is unlikely to be clinically helpful very often. However, it would have been very interesting to know whether the patients who did not relapse in this study were those who had what the RASCAL group have defined as “benign” rather than “aggressive” mutations.

Secondly, the authors do not say whether each postoperative serum mutation which they detected corresponded with the genotype of the primary tumour. While a number of studies have suggested that occasional metastases may carry a different mutation to the primary, this is very uncommon. If the mutation is not the same as the primary tumour, this raises many issues, not least that despite the very careful use of controls, nested polymerase chain reaction techniques may be too sensitive for clinical practice.

For an individual with colorectal cancer, a multitude of factors may shape the clinical course. To improve our knowledge, we must seek to study the most subtle variation in molecular profiles possible, as small differences may be very important. While rigorous small prognostic clinical studies may be intriguing, they may have significant limitations. Some of us have argued previously that there is only one way forward.6 All bodies funding prospective therapeutic studies in colorectal cancer should insist that a panel of molecular markers are measured and recorded in all patients using a standardised technique, even if they are not included as an end point, and make these results readily available for subsequent analysis. Studies which are performed prospectively carry the least risk of methodological flaws.


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.