Correspondence

Alleles Lost and Gained in Malignant Cells

N Engl J Med 1994; 331:1591-1592December 8, 1994

Article

To the Editor:

In their study of the correlation between the allelic loss of chromosome 18q and survival among patients with colorectal cancer (July 28 issue),1 Jen et al. used microsatellite polymorphisms to distinguish between two alleles in tumor DNA. A “relative intensity of the two alleles in the tumor DNA [that] differed from the relative intensity in the non-neoplastic tissue DNA by a factor of at least 1.5” was used as the criterion for loss of heterozygosity.2 However, this criterion is not stringent enough to support the loss of heterozygosity in some cases of liver cancer.

We used 12 microsatellite polymorphism markers to amplify chromosome 1p in 30 paired samples of hepatocellular carcinoma tissue and non-neoplastic liver tissue. In nine pairs, the difference in the allelic ratio was more than 6 for several consecutive markers, and the loss of heterozygosity was subsequently confirmed by Southern blot analysis. A difference of 2 or 3 in the ratio was found for several consecutive markers in six other pairs of samples. However, Southern blot analysis of DNA from these tissues disclosed a twofold to threefold increase in the copy number of one allele, instead of loss of the other allele.3 Duplication or multiplication in chromosome 1 was possible in these six hepatocellular carcinomas, as shown for chromosome 8 in liver tumors in another study.4 The tumor samples in the remaining pairs had the same allelic ratios.

Although the criterion adopted by Jen et al. is reliable for detecting an imbalance in the gene dosage between two alleles in tumor DNA, it fails to distinguish the loss of one allele from an increased copy number of the other allele. The determination of loss of heterozygosity by microsatellite polymorphisms should therefore meet a more stringent criterion, such as a difference in the allelic ratio by a factor of at least 3. Unless there is corroboration by independent methods, a difference in the allelic ratio that is not larger than 3 should be classified as an indeterminate finding rather than definite evidence of loss of heterozygosity, and such cases should be excluded from a prognostic analysis.

Pei-Jer Chen, M.D., Ph.D.
Shiou-Hwei Yeh, M.S.
Ding-Shinn Chen, M.D.
National Taiwan University Hospital, Taipei, Taiwan 10016

4 References

1
Jen J, Kim H, Piantadosi S, et al. Allelic loss of chromosome 18q and prognosis in colorectal cancer. N Engl J Med 1994;331:213-221
Full Text | Web of Science | Medline

2
Gruis NA, Abeln EC, Bardoel AF, Devilee P, Frants RR, Cornelisse CJ. PCR-based microsatellite polymorphisms in the detection of loss of heterozygosity in fresh and archival tumour tissue. Br J Cancer 1993;68:308-313
CrossRef | Web of Science | Medline

3
Yeh SH, Chen PJ, Chen HL, Lai MY, Wang CC, Chen DS. Frequent genetic alterations at the distal region of chromosome 1p in human hepatocellular carcinomas. Cancer Res 1994;54:4188-4192
Web of Science | Medline

4
Fujiwara Y, Monden M, Mori T, Nakamura Y, Emi M. Frequent multiplication of the long arm of chromosome 8 in hepatocellular carcinoma. Cancer Res 1993;53:857-860
Web of Science | Medline

To the Editor:

Jen et al. use the allelic loss of chromosome 18q to predict metastasis in colorectal cancer. Allelic loss is an excellent predictor of which patients will have a distant metastasis (sensitivity of 95 percent) but a poor predictor of which patients will not have a distant metastasis (specificity of 52 percent). Thus, the loss of chromosome 18q can be used to rule in the potential for metastasis but not to rule it out. In terms of therapy, this means that people who do not have allelic loss of chromosome 18q will still have to be treated as if they were at risk for metastasis.

Harry B. Burke, M.D., Ph.D.
University of Nevada School of Medicine, Reno, NV 89520

Author/Editor Response

The authors reply:

To the Editor: We agree with Chen and his colleagues that an allelic imbalance determined by the polymerase chain reaction (PCR) cannot distinguish between a loss of one allele and a gain of the other. However, when found with chromosome 18q markers in colorectal cancer, the imbalance almost always reflects a loss. We base this interpretation on an extensive set of Southern blot analyses with probes from every other chromosome arm as controls1; on cytogenetic analyses2,3; and on evaluation of xenografts, in which allelic losses are unambiguous because of the absence of contaminating non-neoplastic components (unpublished data).

In our opinion, PCR-based assays can never reliably distinguish between loss and gain of a chromosome, no matter what the allelic ratio is. For other chromosomal regions (such as those described by Chen et al.), independent methods should be used to determine whether losses or gains are characteristically present, so that the results of PCR-based assays can be interpreted appropriately. It should also be noted that allelic imbalances in PCR-based assays, whether they represent losses or gains of the relevant chromosomal regions, may still be useful prognostic indicators, because the extent of the imbalance may correlate with the degree of aneuploidy.

In response to Dr. Burke's comments, the potential clinical usefulness of chromosome 18q allelic loss must be considered in the context of the stage of disease. We found that among patients with stage II cancer, the absence of chromosome 18q allelic loss in the tumor was associated with a five-year survival rate of nearly 95 percent, whereas patients with tumors that had chromosome 18q allelic loss had a five-year survival of about 50 percent. Patients with a high expected survival rate are not likely to benefit from further therapy, whereas those with a poorer prognosis may benefit from further therapy. Studies are being conducted by the cooperative oncology groups to address the use of adjuvant therapy in relation to prognostic markers, including the status of chromosome 18q.

Stanley R. Hamilton, M.D.
Jin Jen, Ph.D.
Bert Vogelstein, M.D.
Johns Hopkins University School of Medicine, Baltimore, MD 21205-2196

3 References

1
Vogelstein B, Fearon ER, Kern SE, et al. Allelotype of colorectal carcinomas. Science 1989;244:207-211
CrossRef | Web of Science | Medline

2
Muleris M, Salmon RJ, Zafrani B, Girodet J, Dutrillaux B. Consistent deficiencies of chromosome 18 and of the short arm of chromosome 17 in eleven cases of human large bowel cancer: a possible recessive determinism. Ann Genet 1985;28:206-213
Medline

3
Bardi G, Johansson B, Pandis N, et al. Cytogenetic aberrations in colorectal adenocarcinomas and their correlation with clinicopathologic features. Cancer 1993;71:306-314
CrossRef | Web of Science | Medline

Citing Articles (1)