Intended for healthcare professionals

Clinical Review Regular review

Tumour markers in malignancies

BMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7232.424 (Published 12 February 2000) Cite this as: BMJ 2000;320:424

This article has a correction. Please see:

  1. Annika Lindblom, clinical geneticist (annika.lindblom{at}cmm.ki.se)a,
  2. Annelie Liljegren, oncologistb
  1. a Department of Clinical Genetics, Karolinska Hospital, S171 76 Stockholm, Sweden
  2. b Department of South Stockholm Oncology, Huddinge University Hospital, Sweden
  1. Correspondence to: A Lindblom
  • Accepted 24 September 1999

In western Europe today, a third of all people develop a malignant disease at least once in their lifetime. Treatment has improved, and for many diseases, such as leukaemia and testicular cancer, the prognosis is much better today than it was 20 years ago. Screening for early diagnosis has also led to lower mortality for diseases such as breast cancer and cervical cancer; many malignancies, however, are still diagnosed after the metastatic process has already started, indicating a poor prognosis.

Tumour markers are usually proteins associated with a malignancy and might be clinically usable in patients with cancer. A tumour marker can be detected in a solid tumour, in circulating tumour cells in peripheral blood, in lymph nodes, in bone marrow, or in other body fluids (ascites, urine, and stool). A tumour marker may be used to define a particular disease entity, in which case it may be used for diagnosis, staging, or population screening. Markers may also be used to detect the presence of occult metastatic disease, to monitor response to treatment, or to detect recurrent disease (table). Recently they have even been used as targets for therapeutic intervention in clinical trials.

Summary points

Tumour markers are commonly proteins associated with malignancy, offering a putative clinical use in cancer

A tumour marker can be detected in a solid tumour, in circulating tumour cells in peripheral blood, in lymph nodes, in bone marrow, or in other body fluids (urine or stool)

A tumour marker can be used for population screening and for detection, diagnosis, staging, prognosis, or follow up of malignant diseases

A specific tumour marker is a fusion protein associated with a malignant process in which an oncogene is translocated and fused to an active promoter of another gene

Unspecific markers include the oncofetal proteins (such as the carcinoembryonic antigen or α fetoprotein) expressed by many different types of cancer

Methods

This article is based largely on our experience, discussions with colleagues, reviews, and original articles on the subject, as well as on a textbook on chromosomal rearrangement in tumour cells. We used the following key words for Medline searches for 1966-98: cancer, tumour markers, prognosis, various malignant diseases, and RT-PCR [reversed transcriptase and polymerase chain reaction].

Specificity in tumour markers

Tumour specific proteins

A specific tumour marker is expressed only in tumour cells. The best example is the so called fusion proteins associated with malignant processes in which an oncogene is translocated and fused to an active promoter of another gene. The result is a constantly active production of the fusion protein, leading to the development of a malignant clone. The Philadelphia chromosome in chronic myeloid leukaemia is the best known example (figure).1 DNA sequences can be recombined not only through translocations but also through inversions and insertions. By recombining DNA in this manner, fusion genes may be created or destroyed, or the regulatory control of genes may be interfered with. These mechanisms frequently occur in haematological malignancies but also in some solid tumours of mesodermal origin (table).2

Non-specific proteins or markers related to malignant cells

Oncofetal antigens are another kind of marker, less stringent but still very useful. These are expressed in cells during embryological development and in cancer cells. The most commonly used oncofetal antigen, carcinoembryonic antigen, is expressed in all gastrointestinal tumours as well as in many other tumours.3 α Fetoprotein is used to diagnose hepatocellular cancer but is also expressed in testicular and ovarian cancer.4

Cell specific proteins overexpressed in malignant cells

Some proteins are expressed normally by differentiated cells but are expressed at higher rates in the corresponding tumour cells, which is why a relative increase in serum concentrations can be used as a tumour marker; this is the case with prostate specific antigen concentrations in prostate cancer.5 Cell specific proteins are used for diagnostic purposes —for example, the tyrosinase protein expressed in melanocytes.6

Figure1

Philadelphia chromosome in chronic myeloid leukaemia (fluorescent in situ hybridisation)

Common methods used to identify tumour proteins

Immunohistochemistry

Traditionally, most methods have used monoclonal antibodies and immunohistochemistry to identify tumour marker proteins.7 These methods can be used directly on the tumours for diagnostic purposes, as is done in melanoma, or to identify the protein in serum, as is done in thyroid cancer. Efforts have also been made to identify tumour marker proteins that are detectable in peripheral blood, bone marrow, or lymph nodes to improve the staging of disease and to obtain prognostic information. This identification of micrometastasis used first histology and later immunohistology; none of these methods, however, has yet been proved to give enough relevant information for them to be incorporated into staging protocols.7

Reversed transcriptase and polymerase chain reaction

Recently, a technique for simple and rapid amplifying of DNA—the polymerase chain reaction—has been available. This technique can also be applied to RNA, after RNA is first altered into DNA by using an enzymatic reaction with reversed transcriptase. This method—“reversed transcriptase and polymerase chain reaction” (RT-PCR)—makes it possible to study very small amounts of gene expression and has been shown to be a much more sensitive method for detecting micrometastasis.8 Amplification by polymerase chain reaction allows detection of transcripts from a single tumour cell among 10 to 100 million normal cells. The success of the marker used depends on its specificity and sensitivity. In many solid tumours the use of specific markers is often limited because the heterogeneity of the disease leads to most markers being expressed in only a small proportion of the tumours. Reversed transcriptase and polymerase chain reaction was first used to show the bcr/abl translocation in patients with chronic myeloid leukaemia in 19889 but has now been used experimentally for detecting micrometastases in a wide variety of malignant diseases.8

Common methods for identifying tumour proteins

  • Immunohistochemistry

  • Fluorescent in situ hybridisation (FISH)

  • Reversed transcriptase and polymerase chain reaction (RT-PCR)

Tumour markers in malignant disease

View this table:

Tumour markers in malignant disease

Tumour markers and other prognostic factors in the most common malignant diseases are reviewed below. Hereditary cancer and the genes involved will not be discussed.

The cancer specific marker carcinoembryonic antigen has been used for a long time in follow up to detect early relapse in colorectal cancer. Its use is still controversial, however, and the benefit of the use of this marker in the clinic is therefore unclear.10 It was recently shown that detecting carcinoembryonic antigen by using reversed transcriptase and polymerase chain reaction on lymph nodes at surgery for colorectal cancer is a prognostic factor in stage II of the disease, and it is hoped that this method can be used to identify a subgroup of patients who will benefit from adjuvant treatment.11 In breast cancer a new cancer antigen, CA15.3, is used to monitor the treatment of patients with the disease, as well as to detect recurrent disease, and another new marker, CA27.29, may predict relapse of disease and can be used to detect whether treatment has failed.12 The only tumour marker routinely used clinically is the oestrogen receptor, the status of which is useful when deciding on adjuvant hormone treatment.12 Many other markers are currently being studied experimentally, but no consensus has been reached for clinical use. To optimise treatment in breast cancer, a marker is still needed to identify the 30% of patients in the lymph node negative group who will have a relapse. Overexpression of the HER-2/neu oncogene has been discussed as being related to poor prognosis in breast cancer,13 and recently antibodies directed against this protein have been used in clinical trials.14

In ovarian cancer the cancer antigen marker CA125 is used for follow up, and an increase in serum concentrations may predict recurrent disease or serous adenocarcinoma. The CA19.9 (or CA72.4) marker has been used experimentally in a similar way to follow patients with ovarian mucinous adenocarcinoma. The data are still not sufficient for recommendation of routine use of any of the breast or ovarian cancer markers.12 Prostate specific antigen and prostate specific membrane antigen, diagnosed with immunohistochemistry, have long since been used as serum tumour markers of relapse in prostate cancer15; they have sometimes even been used for population screening and diagnosis.16 Recently, the detection of circulating prostate cancer cells in the bloodstream by using a reversed transcriptase and polymerase chain reaction assay for prostate specific antigen has been investigated as a predictor of surgical failure, and further development and use of this technique could give clinically relevant information.17

Tyrosinase is used in the diagnosis of melanoma.6 The use of reversed transcriptase and polymerase chain reaction to detect circulating melanoma cells was described in 1991 as the first example of detecting haematogenous spread of melanoma cells from a solid tumour in peripheral blood.18 The tissue specific expression of thyroglobulin and parathyroid hormone is commonly used for diagnosis and follow up of patients with thyroid cancer.19 Many types of sarcoma have recently been associated with chromosomal translocations and fusion proteins (table).2028 The identification of several specific fusion products for the very rare sarcomas has meant a diagnostic improvement. An exact clinical pathological diagnosis is often difficult, while the demonstration of the specific fusion protein is diagnostic for these rare sarcomas. This advance also opens the way to developing drugs directed at these proteins, in the same way as is already done against the HER-2/neu in breast cancer. The specific embryological markers human chorionic gonadotrophin and α fetoprotein are used for diagnosis and follow up of patients with testicular cancer.29

Translocations creating fusion proteins are commonly involved in haematological diseases.2 The Philadelphia chromosome in chronic lymphocytic leukaemia is the best known example. Half of all childhood cases of acute lymphatic leukaemia are hyperdiploid, which indicates a highly favourable prognosis.30 The Philadelphia chromosome in acute lymphatic leukaemia is indicative of poor prognosis.31 Mixed lineage leukaemia (involving a rearrangement of 11q23) in very young children predicts a poor outcome.32 Beside these chromosomal rearrangements, more than 100 other cytogenetic changes are used for additional prognostic information in different haematological malignancies. Also in lymphoma many chromosomal rearrangements are known about, even if they are still not commonly used clinically yet.2 The t(8:14) is characteristic of Burkitt's-type lymphomas and leukaemias. The result is a juxtaposition and activation of the c-myc gene.2 In experimental studies in lymphoma, soluble CD25 has been shown to be the most sensitive serum marker for tumour burden,33 and a high concentration of soluble CD44 indicates a poor prognosis.34

Conclusion

Tumour markers are mainly used to diagnose specific malignancies. The methods commonly involve immunohistochemistry and cytogenetics, including fluorescent in situ hybridisation (FISH), and reversed transcriptase and polymerase chain reaction. Markers to be used in population based screening for early diagnosis—such as screening for early colorectal cancer in stool—are needed. The only marker that is sometimes used for screening is prostate specific antigen. Markers used for staging are also needed to optimise treatment; the oestrogen receptor is an important marker for this purpose in breast cancer, and the carcinoembryonic antigen marker looks similarly promising for improved staging of colorectal cancer. The sentinel node technique can improve staging,35 but more and better markers and techniques are needed in screening, staging, and follow up of malignant disease.

Acknowledgments

We thank Drs Ulf Bergerheim, Elisabeth Blennow, Bo Frankendal, Henrik Grönberg, Johan Hanson, Bertil Johansson, Olle Larsson, Per-Uno Malmström, Hans Strander, Christina Wedelin, and Jan Zedenius for valuable contributions.

Footnotes

  • Competing interests None declared.

  • Funding Swedish Cancer Society gave financial support.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract