Loss of Heterozygosity (LOH) Test: Tumor Genetics and Results

Loss of heterozygosity, or LOH, describes a tumor region that has lost the normal difference between the two chromosome copies inherited from each parent. The change can occur because one copy is deleted, because part of one parental chromosome is replaced by a duplicate of the other, or through a more complex rearrangement. LOH is important in cancer because it can remove the remaining working copy of a tumor-suppressor gene, expose a harmful inherited or acquired variant, and create a genomic pattern associated with defective DNA repair. A report may describe LOH at one gene, a chromosome arm, or across a percentage of the tumor genome. Those results are not interchangeable. A locus-specific finding may support biallelic inactivation of a gene such as BRCA1, BRCA2, TP53, or RB1. A high genomic LOH score may be one component of a homologous recombination deficiency assessment. Interpretation depends on tumor purity, copy number, assay design, cutoff, cancer type, and the treatment context for which the test was validated.

• LOH means loss of allelic diversity, not always physical loss of DNA.
• Deletion-associated LOH and copy-neutral LOH have different copy-number patterns.
• Locus-specific LOH can provide evidence that both copies of a tumor-suppressor gene are affected.
• Genome-wide LOH is a genomic-instability measure and may contribute to an HRD result.
• LOH assays and cutoffs are platform specific and cannot be substituted without validation.
• A tumor LOH finding does not by itself establish an inherited cancer syndrome.

Table of Contents

• What loss of heterozygosity means
• Why tumors develop LOH
• How LOH is measured
• Interpreting locus-specific LOH
• Genomic LOH and homologous recombination deficiency
• Negative and indeterminate results
• Technical and biological pitfalls
• Clinical next steps

What Loss of Heterozygosity Means

At many positions in the genome, the DNA copy inherited from one parent differs slightly from the copy inherited from the other. A person is heterozygous at such a position because two different alleles are present. In tumor cells, one allele may disappear or become undetectable, creating loss of heterozygosity.

The simplest mechanism is a deletion. If one chromosome segment is lost, only the allele on the remaining copy is observed. This is sometimes called copy-loss LOH. However, LOH can occur without a change in total copy number. In copy-neutral LOH, the tumor loses one parental segment and duplicates the other, leaving two copies that are genetically the same. Mitotic recombination, chromosome loss followed by duplication, or other repair errors can produce this pattern.

LOH is therefore an allelic state, not a synonym for deletion. A copy-number-only test may identify a one-copy loss but cannot necessarily determine whether heterozygosity was lost. Conversely, a SNP-based assay can detect copy-neutral LOH even when two copies remain. This distinction is one reason SNP chromosomal microarray and appropriately designed sequencing assays provide more information than methods that measure copy number alone.

LOH can involve a small region around one gene, a whole chromosome arm, or most of a chromosome. Tumors often contain multiple LOH segments. The meaning depends on what lies within the region, how the event arose, and whether the assay is using LOH as a gene-level “second hit” or as a genome-wide instability score.

In a mixed tumor specimen, normal cells retain both parental alleles. Their DNA dilutes the tumor signal, so LOH may appear as allelic imbalance rather than complete disappearance of one allele. Bioinformatic models estimate tumor purity, ploidy, and local copy number to decide whether the observed ratio is consistent with LOH. The output is consequently a model-based interpretation rather than a direct visual count of chromosome copies.

Why Tumors Develop LOH

Many tumor-suppressor genes follow a two-hit pattern. One copy is inactivated first by a germline or somatic variant, and the second functional copy is later lost or disabled. LOH is a common second-hit mechanism. When the remaining normal allele is removed, the cell may lose an important brake on proliferation, DNA repair, or cell-cycle control.

For example, a person can inherit a pathogenic variant in one copy of a cancer-predisposition gene. Normal cells retain the other working copy. If a tumor develops LOH that removes the working allele, both copies become functionally compromised. Tumors can follow the same sequence without an inherited variant: one allele acquires a somatic mutation and the other is lost.

LOH is not proof of complete gene inactivation. The assay must determine which allele was retained. If a tumor retains the harmful variant and loses the normal allele, the finding supports biallelic loss. If it loses the harmful allele and retains the normal one, LOH may not disable the gene. Phasing—the process of linking the mutation to a particular parental or chromosome copy—can be difficult, especially in low-purity specimens.

Some LOH regions are large and contain many genes. The selective advantage may arise from loss of one critical gene, several cooperating genes, or general chromosome instability. A recurrent chromosome-arm LOH can have prognostic value even when the exact target is uncertain. In blood cancers, copy-neutral LOH can duplicate an oncogenic mutation or make a previously heterozygous mutation effectively homozygous, affecting diagnosis or risk assessment.

A tumor may also accumulate widespread LOH because a DNA repair pathway is defective. Homologous recombination repair normally fixes double-strand breaks accurately. When this pathway fails, characteristic genomic scars—including LOH, telomeric allelic imbalance, and large-scale state transitions—can accumulate. These scars are historical evidence of instability; they do not directly measure whether the pathway is currently defective [1].

LOH is thus both a mechanism and a biomarker. At one locus it can explain gene inactivation. Across the genome it can summarize a pattern. Confusing those roles leads to overinterpretation.

How LOH Is Measured

LOH analysis requires information about alleles. Historically, laboratories compared polymorphic microsatellite markers in tumor and matched normal DNA. If the normal sample showed two alleles but the tumor showed loss or marked reduction of one, LOH was called. Modern assays commonly use thousands of single-nucleotide polymorphisms, or SNPs, distributed across the genome.

SNP microarray measures both total signal intensity and the relative contribution of the two alleles. This combination can distinguish one-copy deletion, normal heterozygosity, copy-neutral LOH, and some higher-copy states. It is useful for genome-wide analysis but requires adequate tumor content and careful modeling of complex aneuploid genomes.

Next-generation sequencing can infer LOH from allelic fractions at germline SNPs, read depth, and copy-number patterns. Tumor-only panels may estimate the state using population-frequency data and computational models. Matched tumor-normal sequencing provides a clearer baseline because it directly identifies which positions were heterozygous in the person’s normal DNA.

Whole-exome and whole-genome sequencing offer broader coverage and may define breakpoints more precisely. Whole-genome methods can assess copy number, allelic imbalance, structural variation, and mutational signatures together. Clinical implementation depends on validation, computational approach, and turnaround time.

Targeted gene testing may report whether a pathogenic variant has accompanying LOH at that locus. This is narrower than a genomic LOH score. The laboratory may use nearby SNPs, mutation VAF relative to purity and copy number, or direct phasing. A report should explain whether biallelic status is confirmed, inferred, or unknown.

Result type	What is measured	Typical use
Locus-specific LOH	Allelic loss around one gene or variant	Evidence for a second hit or biallelic inactivation
Chromosome-arm LOH	Allelic loss across a large arm	Tumor classification, prognosis, or clonal characterization
Copy-neutral LOH	Loss of allelic diversity with no net copy loss	Detection of duplicated mutations or acquired uniparental disomy
Genomic LOH percentage	Fraction of evaluable genome affected by qualifying LOH segments	One form of genomic-scarring or HRD assessment
Composite HRD score	LOH plus other genomic-scar measures, or a separate algorithm	Prediction in specific validated treatment settings

The specimen is usually formalin-fixed tumor tissue, although fresh tumor, bone marrow, or other material may be used. A pathologist estimates tumor content and selects the region. Necrosis, low tumor fraction, degraded DNA, and highly complex copy number can make the result indeterminate.

Interpreting Locus-Specific LOH

A locus-specific report asks whether the tumor has lost one allele around a particular gene. This is most informative when a pathogenic or likely pathogenic variant has already been detected. The interpretation should describe the variant, local copy number, LOH state, and whether the combined evidence supports biallelic inactivation.

Consider a tumor with a pathogenic BRCA2 variant. If the tumor loses the chromosome copy carrying the normal BRCA2 allele and retains the mutated copy, the result supports loss of both functional copies. That can strengthen the biological evidence for homologous recombination repair deficiency. If no LOH is found, the second allele could still be inactivated by another mechanism such as a second sequence variant, promoter methylation, or structural rearrangement. Absence of LOH is therefore not proof that the gene remains functional.

Variant allele fraction must be interpreted with purity and copy number. A high VAF may suggest LOH, but amplification of the mutant allele, deletion of the normal allele, copy-neutral duplication, or contamination by normal cells can produce different ratios. VAF alone should not be labeled LOH without an assay validated to resolve these possibilities.

The phrase “LOH at the gene” can also be misleading when the event spans a large chromosome region. The biological target may be the named gene, but the tumor has lost allelic diversity across many neighboring genes. The report should avoid implying a precise gene-specific deletion if the platform only defines a broad segment.

In hereditary cancer evaluation, tumor LOH can provide supporting evidence but cannot classify a germline variant by itself. A pathogenic variant detected in tumor may be inherited or acquired. Confirmation in blood, saliva, or another appropriate non-tumor sample is needed. Conversely, a person with a known germline variant may have a tumor that did not lose the normal allele; this does not negate the inherited diagnosis.

Locus-specific LOH can also matter in hematologic malignancies. Copy-neutral LOH may duplicate mutations in genes such as JAK2 or other myeloid drivers, increasing mutant dosage. The finding is interpreted with morphology, blood counts, cytogenetics, and the complete mutation profile rather than as a stand-alone cancer diagnosis [2].

Genomic LOH and Homologous Recombination Deficiency

Genomic LOH is usually reported as the proportion of the assessable tumor genome contained in qualifying LOH segments. Algorithms may exclude whole-chromosome events, very small regions, or regions that cannot be modeled. Each platform defines its own segment rules and threshold for “LOH high.”

Genomic LOH is not identical to HRD. It is one possible genomic scar associated with homologous recombination deficiency. Some assays combine LOH with telomeric allelic imbalance and large-scale state transitions. Others integrate tumor BRCA1/2 status, mutational signatures, or additional repair genes. The Jackson Laboratory’s clinical education guidance explicitly distinguishes LOH from an HRD score because the components vary by laboratory [3].

A high score indicates that the tumor has accumulated a pattern consistent with prior repair deficiency. It does not identify the cause. The cause may be biallelic BRCA1 or BRCA2 loss, alteration of another homologous recombination gene, epigenetic silencing, or an unknown mechanism. A low score does not exclude every functionally important repair defect, particularly in a young tumor that has not had time to accumulate scars or in an assay with limited coverage.

Clinical utility is assay and disease specific. HRD testing has been studied most extensively in high-grade ovarian cancer, where validated companion-diagnostic or genomic-instability assays can contribute to decisions about PARP-inhibitor strategies in defined settings. Evidence, drug labels, and recommended cutoffs can change. A “LOH-high” result from one platform should not be assumed equivalent to “HRD-positive” from another.

Several additional cautions apply:

• Genomic scars can persist after the tumor restores homologous recombination through a resistance mechanism.
• Prior treatment and clonal evolution may alter current drug sensitivity without erasing the historical scar.
• Tumor purity and ploidy estimation affect the score.
• Borderline values near the cutoff are vulnerable to analytic variation.
• A positive HRD-related result is not a guarantee of response, and a negative result is not a guarantee of no benefit.

Recent reviews emphasize that available HRD methods differ in analytes, algorithms, thresholds, and evidence base [1,4]. The exact commercial or laboratory assay named on the report is therefore essential.

Negative and Indeterminate Results

A negative locus-specific LOH result means the assay did not find qualifying allelic loss in the evaluated region. It may indicate that both parental alleles remain, but it does not establish that both gene copies function normally. A second hit could be a sequence change outside the panel, a small exon-level deletion, promoter methylation, or another regulatory alteration.

A “LOH low” genomic result means the percentage or score did not meet that assay’s threshold. It should not be translated into “DNA repair normal.” Homologous recombination function is not measured directly, and different algorithms may classify the same tumor differently. A low result also cannot exclude a treatment response that occurs through mechanisms unrelated to HRD.

An indeterminate result can arise when:

• tumor content is below the validated minimum;
• DNA quantity or quality is inadequate;
• extensive necrosis or normal-cell contamination dilutes allelic signals;
• ploidy and complex copy number prevent reliable modeling;
• the genome has too few informative heterozygous SNPs;
• sequencing coverage is uneven;
• the specimen fails assay-specific quality controls.

A test failure should not be reported or interpreted as LOH low. The report may recommend another tumor block, a repeat biopsy, a different platform, or paired normal analysis. The clinical value of repeating depends on whether the result would change management.

Tumor heterogeneity also affects negative results. A biopsy may sample a clone without LOH while another metastatic site contains the event. Conversely, a subclonal LOH event can fall below the assay’s detection threshold. Broad plasma profiling may sometimes capture multiple sites, but ctDNA-based copy-number and LOH analysis has its own tumor-fraction limitations.

When a treatment decision rests on the result, ask whether the assay was the validated companion or complementary diagnostic used in the relevant evidence. A generic “genomic instability” test may not be interchangeable with the test specified in a trial, label, or guideline.

Technical and Biological Pitfalls

The first major pitfall is confusing LOH with copy loss. A region with one copy usually has LOH, but a region with two copies can also have LOH if both copies derive from the same parental chromosome. Conversely, some copy-number gains retain both alleles. Accurate interpretation requires combined allelic and copy-number information.

The second is assuming that LOH identifies which allele was lost. Without phasing or matched normal data, the assay may show allelic imbalance but not whether the tumor retained the pathogenic or normal allele. Statements about biallelic inactivation should reflect the actual evidence.

Third, tumor purity can distort every measurement. Normal cells contribute heterozygous DNA and can mask LOH. Very high ploidy and subclonal copy-number changes produce overlapping allele fractions that challenge computational models. Different algorithms may choose different purity and ploidy solutions.

Fourth, formalin-fixed tissue introduces fragmentation and chemical damage. Although modern assays are designed for this material, low-quality DNA can reduce informative coverage. Decalcified bone specimens may be particularly difficult.

Fifth, genomic LOH cutoffs are not universal. The percentage of the “genome” included in the denominator, minimum segment size, handling of whole-chromosome loss, and exclusion of low-quality regions can differ. Comparing raw numbers across laboratories can be misleading.

Finally, LOH is a feature of the tested tumor at a particular time. It does not necessarily describe normal tissues, every metastatic site, or future resistant clones. A tumor LOH result can suggest the need for germline testing but cannot determine inheritance. Similarly, a germline pathogenic variant does not guarantee that a specific tumor has LOH or is sensitive to a particular drug.

Good reports state the specimen, tumor percentage, method, tested regions, definition of LOH, threshold, result, limitations, and intended clinical interpretation. When one of these elements is missing, the laboratory can often provide clarification.

Clinical Next Steps

Start by identifying which type of LOH result was issued. Is it a local finding at one gene, a copy-neutral chromosome-arm event, a genomic LOH percentage, or one component of a composite HRD score? The same word can represent very different analyses.

For a locus-specific finding, ask:

1. Which gene and chromosome region are involved?
2. Is there a pathogenic variant in the retained allele?
3. Was biallelic inactivation demonstrated or only suspected?
4. Was matched normal DNA tested?
5. Does the result suggest separate germline confirmation?

For a genomic LOH or HRD-related result, ask what assay and cutoff were used, whether the specimen met tumor-content requirements, and what clinical evidence links that exact result to the proposed treatment. Review current disease-specific guidelines and drug labeling. A molecular tumor board can help when the result is borderline, discordant with gene findings, or generated by a nonstandard platform.

If the result is indeterminate, the next step may be another tissue block, repeat extraction, a fresh biopsy, or a different validated test. Retesting is most appropriate when the result will influence a meaningful decision and adequate material can be obtained safely.

Genetic counseling is indicated when a pathogenic variant in a hereditary cancer gene is detected or when personal and family history suggests inherited risk. Tumor LOH can support the biological relevance of a variant, but germline testing must be performed and interpreted separately. Relatives should not be tested based only on a tumor LOH report.

Keep the full molecular report with the pathology record. Treatment indications and interpretations evolve, and future review may require the exact platform, score, segment definition, and variant context. The practical goal is not to label a tumor simply “LOH positive.” It is to determine what kind of allelic loss occurred, what biological claim the assay supports, and whether that claim has validated clinical value for this patient’s cancer.

References

• Homologous recombination deficiency (HRD) testing landscape. 2025. Review article.
• Comparison of the copy-neutral loss of heterozygosity identified by SNP array and next-generation sequencing. 2022. Comparative study.
• Homologous Recombination Deficiency (HRD) Testing: FAQ. 2025. Clinical education resource.
• Recent Advances in Genomic Approaches for the Detection of Homologous Recombination Deficiency in Cancer. 2024. Review article.
• Diagnosis of Ovarian Carcinoma Homologous Recombination Deficiency: Review of Current Assays and Development of the OncoScan Genomic Instability Metric. 2023. Assay study and review.

Disclaimer

This article offers general education about tumor LOH testing and cannot classify an individual result or recommend cancer treatment. LOH, genomic-instability, and HRD assays are not interchangeable, and decisions should use the complete laboratory report, current disease-specific guidance, and specialist review. Tumor results that suggest inherited risk require separate germline testing and genetic counseling.