Case 287 -- Medicolegal case of 56 year old female with one lymph node metastasis and two oral squamous cell tumors

Contributed by James R. Davie M.D. Ph.D., Patricia A. Swalsky, Sydney Finkelstein, MD
Published on line in November, 2001

PATIENT HISTORY:

A previously healthly 56 year-old female with no smoking or alcohol history was diagnosed in 1991 with leukoplakia and lichen planus of the tongue, oropharynx, and gingiva.  Over the next six years, the patient was closely followed, with administration of numerous laser ablative treatments.

In 1998, the patient was noted to have a new 0.6 cm friable ulceration of the right anterior lower gingival mucosa believed consistent with granulomatous ulceration vs. possible squamous cell carcinoma.  Repeated requests for referral to perform definitive biopsy were refused by the patient's health maintenance organization (HMO).  Five months later, the patient was biopsied under a different health plan, which confirmed the presence of moderately differentiated squamous cell carcinoma.  Surgical resection and regional radiation treatment were performed.

In 1999, the patient had a new lesion of the right posterior base of the tongue, and a mass of the right upper neck, confirmed by biopsy and fine needle aspirate to be squamous cell carcinoma, well differentiated.  A rebiopsy of the area of original right anterior lower gingival mucosa in 1999 demonstrated the presence of residual squamous cell carcinoma.  Surgical resection of the tongue and right cervical lymph node chain was performed; histologic exam revealed well-differentiated squamous cell carcinoma in the tongue lesion and in one of 14 lymph nodes.  The patient succumbed to metastatic disease within the year.

Legal proceedings were initiated by the patient's estate against the first HMO on the basis that the four-month delay in treatment allowed the gingival squamous cell carcinoma to develop metastases to the cervical lymph node and base-of-tongue, contributing to the patient's demise.  The HMO denied liability, claiming that there was no proof that the early 1998 gingival squamous cell carcinoma had any relation to either the later 1999 base-of-tongue tumor or the 1999 cervical lymph node metastasis.

Microdissection genotyping was requested to elucidate the genetic relationship between the two oral squamous cell tumors and the lymph node metastasis.

MICRODISSECTION GENOTYPING PROCEDURE: 

Microdissection genotyping was performed with the specific objective of defining the relationship between three sites of squamous cell carcinoma:

1) gingival tumor (first biopsy in 1998 and repeat biopsy in 1999)

2) base of tongue tumor (first biopsy in 1999)

3) cervical lymph node (first biopsy in 1999)

Each tumor site was microdissected using a dissecting microscope from unstained four-micron thick formalin-fixed, paraffin-embedded histologic recut sections.  Specific sites were selected for microdissection according to histologic features, ensuring that tumor was accurately sampled and that necrotic areas were avoided (the latter have degraded DNA less likely to be PCRed successfully).  "Normal" non-neoplastic tissue was also analyzed to verify that each microsatellite locus tested had heterozygous (different-sized) alleles and would therefore "informative" for determination of a potential loss of heterozygosity. To provide even greater assurance in the significance of the results and the conclusions drawn, each tumor was sampled at two or more sites to enable a comparison not only between each site of tumor formation but also within each tumor.

Microdissected tissue was purified for DNA, and subjected to PCR reactions using microsatellite-specific, fluorescent-tagged primers.  The fluorescent tagged primers allow the PCR products to be analyzed by product size (which indicates allele identity) and product quantity using an ABI 3100 capillary gel electropherometer.   The ratio between the product quantity for each of the two alleles at each microsatellite locus was used to render a loss of heterozygosity call of  No LOH, LOH L (low level loss of heterozygosity), or LOH H (high level loss of heterozygosity). 

MICRODISSECTION GENOTYPING THEORY: Microdissection genotyping is based the knowledge that each individual primary carcinoma will, over time, acquire a unique profile of acquired mutational damage (a "mutational fingerprint") that in turn may be expected to be present in its metastases.   This profile of mutational change may differ between a tumor and its metastases by the acquisition of additional mutational alterations with time, but early events will be common between a primary tumor and its progeny. Therefore, the early 'fingerprint' of shared mutational damage to link primary with secondary sites of tumor formation, and the presence of additional mutations may provide an indication of temporal sequence. Microsatellites One method of tracking mutational damage is through microsatellite DNA sequences. Microsatellites are short segments of DNA (up to 200 bp) containing a variable number of tandem repeats of an identical 1-6 base pair motif;  microsatellites are therefore also known as STRs (short tandem repeats).  Variations in the number of repeats give rise to multiple alleles for a given microsatellite, which can be differentiated by performing flanking PCR of the alleles and simply looking at the size of the PCR product, which corresponds to the number of repeat units within the microsatellite.   Given the ease in allele haplotyping, the variety of alleles for microsatellites (up to several dozen), and the frequency of microsatellites within the genome (roughly one microsatellite every 10 kilobases), microsatellites make useful markers for high-resolution 'tagging' of specific chromosomal regions for linkage analysis. Loss of Heterozygosity Figure 1: LOH in Knudson's "Double-Hit" Tumorigenesis Model   Loss of heterozygosity (LOH) for important oncogene/anti-oncogenes can be detected by determining LOH in nearby microsatellite markers.   In Figure 1 , microsatellite loci alleles(also known as short tandem repeats, or STR, loci) are shown as similarly-colored circles along the 17p arms.  If no deletions occur, the ratio of the individual quantity of each of the two alleles at each microsatellite locus is approximately equal (Fig. 1, left and middle panels).  According to Knudson's 'double-hit' hypothesis, deactivation of anti-oncogenes such as TP53 (shown above) requires both genes to be inactivated.   Point mutations/small deletions in anti-oncogenes have no effect on nearby microsatellites, and cannot be detected by microsatellite LOH studies (Fig 1, middle panel).   If, however, one of the gene-inactivating events involves a gross deletion of a chromosome portion, then nearby microsatellite markers will also be deleted (Fig. 1, right panel), which will cause a gross change in the allelic ratio for the affected microsatellite markers that can be detected by microdissection genotyping as a loss-of-heterozygosity (LOH).    It should be noted that the system used here is capable of detecting two different allelic loss events:  loss of heterozygosity via loss of the smaller microsatellite allele (designated as LOH T), or loss of heterozygosity via loss of the larger microsatellite allele (designated as LOH B).  These two mutually exclusive LOH events are color-coded differently as either blue or red in Table 1 of the RESULTS section (below) to reflect that these two events do not represent compatibility in mutational profile. Figure 2:  Microsatellite PCR analysis:  Allele Identification & Quantitation for determination of Loss of Heterozygosity (LOH).

"Informative" vs. "Non-informative" microsatellite loci Normal microdissected tissue samples are first evaluated to determine informative status with respect to individual microsatellite loci. When a particular microsatellite marker in a normal tissue sample manifests only a single peak then the patient is designated as non-informative for that marker. (If a patient is homozygous for a given microsatellite locus, then the microsatellite alleles are the same size following PCR, generating a single peak, and a loss of one of the two alleles cannot be detected.  This is why a homozygous microsatellite locus is considered "non-informative.").    Assessment of relative level of Loss of Heterozygosity For "informative" (heterozygous) microsatellite loci, the alleles are assessed as being in balance when the ratio of the individual allele peak heights (which correspond to allele dosage) falls within the range of 0.66 to 1.50. Values beyond this range are classified as being loss-of-heterozygosity (LOH) with two categories:  Low level allelic imbalance (LOH L) is said to exist when the microsatellite allelic peak height ratios falls into the range from 0.50 to 0.66 or from 1.50 to 2.00. High level allelic imbalance (LOH H) is present when the allele ratios fell below 0.50 or above 2.00. To afford a conservative assessment of the presence of mutation, in this study only high level allelic loss is classified as indicative of mutation. Using these criteria, it would be necessary for no less than 50% of the cells in any given microdissection target to be mutated in order for the microdissection genotyping approach used here to detect this alteration. These stringent conditions are justified given the ability of tissue microdissection to optimally sample tumor at a particular site coupled with the desire not to overcall mutational change that may, in fact, not be present.   Microsatellite Instability Microsatellite instability is seen more commonly in certain tumor types, such as gastrointestinal tumors.   Mutations in DNA mismatch repair enzymes (MMR) cause the number of repeats in microsatellites to contract or expand at an accelerated pace, which is detected as the appearance of novel alleles in addition to the germline alleles upon microdissection genotyping of neoplastic tissue.

RESULTS AND DISCUSSION:

Table 1

RESULTS OF MICRODISSECTION GENOTYPING OF SEVEN SITES

COMPARATIVE LOSS OF HETEROZYGOSITY (LOH) AT 11 MICROSATELLITE LOCI

  Non-informative microsatellite markers not shown in table:  D17S1289 (at 17p13, near site of p53 gene).

KEY:
NO LOH:  Normal two alleles are noted for the microsatellite locus.  Relative ratio between the two alleles does not exceed 2:1.

LOH T:  Loss of heterozygosity of the "Top" (larger sized) allele.   Relative ratio between the two alleles exceeds 2:1.
LOH B: Loss of heterozygosity of the "Bottom" (smaller sized) allele.  Relative ratio between the two alleles exceeds 2:1.

MSI:  Microsatellite instability noted at the microsatellite locus (generation of novel alleles due to DNA mismatch repair defects).
 

Cumulative mutational damage is shown in Table 1 for all the informative markers determined using the non-neoplastic tissue sample. Only a single microsatellite marker, Locus 1, was without allelic loss in all tumor samples. Three markers, Locus 3, Locus 7 and Locus 8 were equivalently mutated in all tumor samples. These changes may be proposed as part of the field effect for cancer formation and progression that is noted for head and neck cancer. All the remaining markers showed discordant patterns of involvement between the different tumors and may be used to decide the origin for the metastatic squamous cell carcinoma in the cervical lymph node.

In seven instances (Locus 2, Locus 4, Locus 6, Locus 9, Locus 10, Locus 11) the presence or absence of mutational change between the gingival squamous cell carcinoma and the cervical lymph node deposit is identical. In contrast, only a single microsatellite marker is concordant between the base of tongue tumor and the cervical lymph node. The temporal accumulation of allelic loss (LOH) can be formulated into a model using either the gingival tumor or the base of tongue tumor as the primary cancer for cervical lymph node metastatic seeding:

  When the early 1998 gingival squamous cell carcinoma is evaluated as the primary tumor for the lymph node metastasis, all lymph node deposit LOH is accounted for by the primary tumor. The gingival tumor requires only the acquisition of a single allelic loss event after the point of metastatic seeding.

In contrast, when the base of tongue squamous cell carcinoma is invoked to account for metastatic seeding to the cervical lymph node, tumor in the lymph node must acquire 2 additional alterations while the primary tumor must acquire 4 additional alterations after the point of metastatic seeding.

It is clear that the gingival primary squamous cell carcinoma generates a much better fit with the data and therefore can be affirmed as representing the source of metastatic seeding to the cervical lymph node. By the same token, the base of tongue tumor can be affirmed as not accounting for cervical lymph node metastasis.

The data in Table 1 also points out the consistency of mutational change within each tumor based on the sampling of two or three sites within each neoplasm. For each tumor there are fewer differences than coincidences, affirming the reliability of the microdissection genotyping process. The gingival and base of tongue squamous cell carcinomas are seen to be independent primaries, and the metastasis in the cervical lymph node traced to the gingival squamous cell carcinoma.

The degree of concordance or discordance in the genotypic profile of mutational damage will enable the discrimination between de novo tumor formation versus metastasis. It should be noted that head and neck cancer, such as the lesions present here, are felt to arise in the context of a "field effect" for cancer formation, in which cells in early stages of dysplasia (and therefore expected to have a small number of microsatellites with LOH) have a slight proliferative advantage over normal cells, and over a period of years, replace normal cells over extensive areas.  (Given the patient's multi-year history of extensive oropharyngeal leukoplakia, this would not be unexpected.)  Therefore, identical mutational changes between independent primary tumors arising in the head and neck may be considered to be part of the field effect. Notwithstanding field effect changes, each independent squamous cell carcinoma arising in the head and neck may be expected to exhibit its own unique profile of additional acquired mutational changes.

REFERENCES:

1. Bennett, P.  "Demystified...: Microsatellites.[Review]", Molecular Pathology vol 53(4):p 177-183, Aug 2000.
2. Williams, HK "Molecular pathogenesis of oral squamous carcinoma.[Review]" Molecular Pathology. 53(4):165-172, August 2000.
3.  John  M. Butler, Forensic DNA Typing.  Academic Press, San Diego, CA.  ©2001
4. Mao, L. et al.  "Microsatellite alterations as clonal markers for the detection of human cancer." Proc Natl Acad Sci U S A 91, 9871-9875, 1994.
5. Paul D.P. Pharoah and Carlos Caldas, "Molecular genetics and the assessment of human cancers [Review]", 
6. Expert Reviews in Molecular Medicine, 1999. Link:  http://www-ermm.cbcu.cam.ac.uk/99000526a.pdf