Dr. Breen's Berner-U Presentation

Shared with permission from The Alpenhorn; published June 2012

Berner-U Presentation May 8th 2012
by Dr. Matthew Breen
(Summarized by Pat Long, and edited by Dr. Breen)

Human and Canine genetic researchers have been operating independently, but the advantages of collaboration are becoming more apparent. The sequencing of the Human Genome took 15 years and cost between 3 and 5 billion dollars. The mouse genome took 5 years and cost $100M. Conversely, the canine genome took about one year and cost $47M, and finally the horse genome took about 6 months and cost $15M. The research increases in speed each year with better tools. The canine genome, however, is most useful in cancer research. Unlike the mouse, canines and humans both get spontaneous cancers.

In any discussion of genetic research, it is necessary to understand the vocabulary. Genome is the sum total of all the genetic material for an organism. Chromosomes are the structures that house the DNA, a human cell has 23 pairs of chromosomes, one set of those 23 chromosomes comes from each parent. The canine cell has 38 pairs of chromosomes. The DNA in those chromosomes are made up of the double helix, a twisted ladder, with the rungs constituting the base pairs, those letters we always see to depict the nucleotides A, C, G, and T. In the canine genome, it is estimated that there are approximately 2.4 billion base pairs, and about 19,000 genes.

Humans in general have more genetic diversity than purebred dogs. When the canine genome project was initiated, the decision was made to use a breed with as little genetic variation as possible. Berners were one of the handful of breeds considered, but it was ultimately decided to use a Boxer for the genetic mapping project, a female Boxer named Tasha. Of all the AKC registered purebred dogs, 75% of them are made up of only 20 breeds. The top five breeds in 2009 were: Labrador Retrievers, Yorkshire Terriers, German Shepherd Dogs, Golden Retrievers, and Beagles. Through selection, many breeds now have a limited level of genetic variation.

Quite simply, the lack of genetic variation within a purebred population helps to simplify the study of genetics. There are more than 300 recognized breeds, and there are more than 360 inherited diseases, more than any mammal except humans. Most of those diseases are recessive, and about 50% of those genetic diseases that we know about occur predominately in one or a few closely related breeds. A pure bred population is strongly enriched for 'risk' alleles. Because there are large families with multiple generations, it helps increase the ability to map genes. A small number of founders, popular sires, genetic bottlenecks - the very things that help set breed type and consistency reduces genetic variation.

Dogs are plagued by the greatest number of documented, naturally occurring genetic disorders of any non-human species. Current estimates suggest that up to 25% of the purebred dogs living in US households my have or carry a serious genetic disease. Compare that to human health where a 1% disease rate is considered high risk. Inherited canine disease affects almost every body system: blood, cardiovascular, endocrine, eye, gastrointestinal, immune, musculoskeletal, nervous system, respiratory, skin, reproductive, and CANCER.

When the AKC CHF asks breed clubs to compile a list of the top ten diseases, cancer is always at the top of the list. It is the leading cause of death in pet dogs; about 25% will develop cancer, and about 50% of dogs over the age of 10 will die of cancer. Of the 42,000,000 vet visits per year in the US, about 10% of them result in a diagnosis of cancer. The presentation, histology and biology of a number of spontaneous canine cancers closely parallel those of human malignancies. Dogs and humans share similar genomes, and are exposed to the same environment. A number of canine cancers have been hypothesized to be appropriate models for human disease: osteosarcoma, non-Hodgkin's lymphoma, glial tumors, oral melanoma, mammary carcinoma, and prostate cancer.

There are breed predispositions for cancer and some breeds develop numerous cancers. The important associations in Dr. Breen's work are the Golden Retriever (lymphoma, osteosarcoma, soft tissue tumors, hemangiosarcoma, and histiocytic malignancies), and the Berner (osteosarcoma, soft tissue tumors, hemangiosarcoma, and histiocytic malignancies).

The Bernese Mountain Dog was created in the early 1900's from a small number of "pure" individuals. In the USA, the breed has increased 108 fold in the past 46 years (31 new AKC registrations in 1966 compared to 3,338 in 2008). The breed has a small breeding population, and thus a relatively small gene pool. Most breeding is line breeding with very little out crossing. Most BMDs are related to each other through a small number of common ancestors. While this gives greater consistency in the breed and more competitiveness at the show level, it means that disease genes may be highly prevalent in the population.

The top cancers in the breed in order based on the two most recent health surveys are:

2000 2005
histiocytic malignancies 54% 47%
lymphoma 24% 29%
hemangiosarcoma 9% 8%
osteosarcoma 7% 10%
mast cell 6% 6%

About 75% of all Berner cancer deaths are a result of histiocytic malignancies or lymphomas.

In Bernese Mountain Dogs recruited into Dr. Breen's study the common presentation of patients with HM (histiocytic malignancies) are:

Age of diagnosis
Youngest = 18 months
Oldest = 13 years
Average is about 6 years
Clinical signs
Lethargy, anorexia, weight loss
Enlargement of the spleen and/or liver and/or lymph nodes
Common organ involvement
Spleen, liver, lung
Kidney, lymph node, bone marrow, brain
Less than one hour up to a couple of months
Usually with a primary tumor in the spleen and removed early

[Pat Long note: I've heard of it as young as 7 months. We have a number of cases that have been detected early with a tumor in the lung, found on x-ray in preparation for a surgery in most cases. Lomustine (CCNU) has been the treatment of choice, and anecdotal evidence has seen survival of up to 18 months. Survival without any treatment has been as long as 18 months. I do know, however, that dogs that I would have only given days or weeks based on my experience have lasted several months with chemotherapy.]

Several of the dogs that were recruited for Dr. Breen''s study as unaffected siblings of an affected Berner have subsequently been diagnosed with HM within a year. The most common misdiagnosis of HM in Berners is either lymphoma or hemangiosarcoma.

HM is the 'Honey Badger' of tumors; the Honey Badger is a weasel like animal known for its aggressive fighting spirit - once it latches on, it doesn't let go. HM is aggressive and not typically responsive to treatment. The vocabulary of HM has varied over the years, and there is still a wide variety of terms used by pathologists. Dr. Breen uses the following terminology:
- Histiocytic sarcoma consists of an initial single tumor, perhaps present on a limb. It is diagnosed more in Flat Coated Retrievers (FCR) than in Berners. It presents as either a mass on a limb (which has a better prognosis) or as a tumor on a single internal organ (which has a poor prognosis).
- Malignant histiocytosis consists of tumors that arise in multiple tissues/organs simultaneously, not metastasis. It usually involves internal organs such as the lung, spleen, liver. It is diagnosed more in Berners than in FCR. Diagnosis is more likely if an internal mass is observed.
Note: it can only be diagnosed if a pathologist is given tumor tissue from multiple organs. This is the only way that a pathologist can determine if the tumors arose from the organ or were metastasized from a single organ.
- NOTE: Even if a dog has multiple histiocytic malignancies throughout its body, if only one mass is evaluated the pathology diagnosis will be a HS. Disseminated disease can be confirmed only if multiple tissues are evaluated.

Lymphoma is the most widely studied cancer across all dog breeds. Consequently researchers have made more headway with gene discovery. It can also be treatable and is often responsive to chemotherapy.

Hemangiosarcoma is a silent killer. There are often no signs of problems until the highly vascularized tumor ruptures and the dog bleeds out.

Cancer is a genetic disease, it is caused by mutations in the DNA at the cell level. The DNA is made up of the four component parts termed A, C, G, and T. One A looks just like all the other A's, so finding the specific genes responsible for disease is very difficult. Having a detailed road map of those genes is the first step.

The canine genome was sequenced and assembled at the Broad Institute at MIT. It was a collaborative effort, and Dr. Breen's lab contributed significantly with their development of cytogenetic tools to help anchor the genome. The sequencing took about 35 million reads of genetic material over a period of 7 months. The genome is approximately 2,400,000,000 base pairs of DNA, and it is estimated that there are about 19,000 genes.

The dog has 78 chromosomes, and during the cell division process each one is replicated, not always exactly. With cancer, those replication errors are of key importance. If we think of a chromosome as a filing cabinet, and the genes in that chromosome as the folders, sometimes those folders get duplicated, or lost (each of these are examples of unbalanced changes), or just rearranged during the replication process (a balanced change). These aberrations are hallmarks of gene deregulation and genome instability. Some aberrations have been shown to be present only in specific types of cancer, making them valuable diagnostic tools. In addition, the presence of some aberrations are tightly associated with response to certain therapies and so offer prognostic value and guide the type of treatment most likely to combat the cancer.

In looking at a grouping of the chromosomes in a canine cell, it is difficult for even a trained eye to group all of the chromosomes into their pairs (see Determining whether material on those chromosomes has been lost, duplicated, or rearranged is quite impossible without some sort of assistance. Reagents have been developed in human gene studies to assist with this research. The reagents available for canine genomic study are much more sparse. Fluorescence in situ hybridization, or FISH, was used at North Carolina State University in Dr. Breen's lab to assist with canine genomic research. FISH uses fluorescently labeled/detected DNA fragments to mark pieces of the genome ranging from a few thousand base pairs to an entire chromosome. Think of a yellow colored tab on one file folder in that filing cabinet, a quick glance will show if the new cabinet has that same yellow tab, or more than one, or none at all. With the use of the chromosome identification tools developed by Dr. Breen's lab, detecting aberrations becomes much more simple (see

Another technique used is called array comparative genomic hybridization (aCGH). This uses whole genome fluorescent labeling of tumor and non-tumor samples to compare directly the relative amount of DNA in each sample at intervals throughout the genome – the smaller the intervals the higher the resolution of the array. If there are not differences in the amount of DNA in a tumor and a non-tumor sample, the amount of fluorescence from each is the same. Where the tumor sample has missing, or extra DNA, the arrays reveal the genome location of this aberrations. (see Technology continues to advance; in 2005 there would have been one observation every 10,000,000 base pairs, but now the lab is evaluating genome staus with an observation every 2,500 base pairs. It becomes much easier to see the specific part of the chromosomes that have the aberrations. We see the same increase in clarity with a digital image when we increase the number of pixels.

To look for new structural chromosome changes in canine tumors requires viable cells. The submissions have to be fresh tumor biopsies, collected using sterile techniques. They need to be shipped and processed quickly. This is not always possible, the dog must come first. But if participation is possible, having a vet willing to follow the simple procedures carefully is key. Once the samples are received, the sterile biopsy cells are put into a culture and live on for a short while. Dr. Breen's lab is well aware of the fact that our dogs live on in their study, as evidenced by his request for photos that are displayed on the lab wall.

Research of cancer in humans is marked by the wide variation of genetic material among humans. Canine cancer research, however, is marked by very little variation, making it much easier to detect the specific genetic aberrations associated with cancers.

As a good example, meningiomas are marked by either a partial or total deletion of chromosome 22 in humans, a piece of the genome that contains over 550 genes. This human chromosome shares DNA sequence with regions of three dog chromosomes, numbers 10, 26 and 27. In the dog we don't see deletion of chromosomes 10 and 26, but we see that most cases are missing a copy of chromosome 27. The actual size of the genome sequence shared between human 22 and dog 27 is only very small and contains fewer than a dozen genes. This tells us that if human and dog meningioma are caused by the same genes (which we believe they are) the hunt for the key genes in both species should concentrate on just those 12 shared genes, not the other 500+ genes that are not involved with dog meningioma. The findings in the canine research enabled human researchers to focus on the equivalent areas on the human genome. Areas that they had not planned to study for a number of years and a great many research dollars later.

By studying cancers in dogs and people, we are able to focus on what they share and this accelerates the process of gene discovery. Cancer has subtypes, caused by aberrations of different genes.

These subtypes are often associated with vastly different responses to therapies, and with different prognoses.

The current NCSU BMD samples, as of March 31st, 2012

Total submissions 231 % of
191 diagnosed cancers

Confirmed MH/HS 117 61%
Hemangiosarcoma 13 7%
Lymphoma 13 7%
Genuine STS* 40 21%
Osteosarcoma 8 4%
Pending diagnosis (submitted as HS and HSA) 19
Unsuitable** 21

* includes fibrosarcoma, hemangiopericytoma, carcinoma
** necrotic tissue, non-tumor tissue, bad packaging

And blood from 110 healthy relatives

The age of diagnosis of HM is different between Berners and FCRs: it peaks at 7-8 years for Berners, and at 9-10 years for FCRs. The anatomical location varies by breed. Out of 146 HM samples, 68 US Berners, 33 French Berners, and 45 US FCRs:

One internal organ 23% 38%
Disseminated 26% 51%
Skin 8% 3%
Limb 38% 3%
Lymph Node 5% 5%

Genetically, the aberrations have some similarities and some differences between the two breeds as well. The findings summarized:

- chaotic genome
- highest recurrence of loss located on chromosomes 2, 11, 16, 22, and 31
  • Chr16 highest overall, with two regions at ~86% recurrence
    - 47-53Mb (telomeric end)
    - 41.8-44.2Mb
- Several genes of interest now being investigated:
  • PTEN
  • CDKN2A/B (p15, p16, ARF)
  • RB1
  • FAT1
  • MTAP
- Some noted differences between BMD/FCR

Looking at chromosome changes, hemangiosarcomas are much less complex than histiocytic malignancies; it is marked by far fewer DNA copy number aberrations. However, the genome-wide differences between hemangiosarcoma in BMDs and Golden retrievers show more complex changes in the Berners than in the Goldens.

What's next? Dr. Breen's lab has developed an assay to look at the function of genes using archival tissue samples. They have identified a list of genes that their data suggest are drivers of MH, but it's still quite a long list. That list needs to be narrowed, and then they need to assess the role of those genes in patients.

Dr. Breen has a unique opportunity to partner with a human pediatric oncologist. Dr. Joshua Shiffman owned a Berner that died of histio, and he works at the Huntsman Cancer Institute at the University of Utah. Comparing genetic changes in Berners and humans will assist in identifying what is shared. It will allow the list of candidate genes to be narrowed.

Tests on the horizon

Dogs were recruited for a study of lymphoma treatments. The genetics of the lymphoma subtypes were studied and compared to the treatment outcomes. A genetic test has been developed that identifies duration of the first remission of lymphoma patients treated with chemotherapy. That test is anticipated to be commercially available within a year.

Genetic comparison of HM, lymphoma and hemangiosarcoma is also possible, and a commercial test is feasible within two years. The real importance of this test is evidenced by the number of presumed HM samples Dr. Breen's lab has received that were determined to be lymphoma instead. Always ask your veterinarian to send in tumor biopsies as coming from a Yorkshire Terrier rather than from a Berner. Too many pathologists see Berners and assume histiocytic disease. Yorkies almost never get cancer, so it would enable a pathologist to work without having any preconceived notions.


Dr Breen and his lab expressed their sincere thanks to all the BMD owners, breeders and vets who have had the courage to submit samples from their beloved companions to help their research projects move forward. He also paid tribute to Joye Neff's fundraising efforts, particularly the mountain of tickets she got for the Visit to Dr. Breen's lab, won by Wendy Djang.

Wendy Djang's prize from the 2011 Cancer Fundraiser
"Spend a day working with the Breen Lab at NCSU"

He will be donating that offer again, and Wendy has offered her hospitality to future winners.

From Pat Long: I would like to give a very special thank you to Tessa. She works with each person who submits a sample to the research lab. She is dealing with the most emotionally trying of circumstances, and she does it with grace and care. Dr. Breen met with the people who had come for the Berner-L get together Monday night in Gettysburg. He claimed that Tessa knew the name of each person and each dog that has participated in the research. Several of us raised eyebrows - we know there have been hundreds of participants. He asked Lori Simidian the name of her girl who had died in 2008 as an example. He called Tessa and asked if she remembered Lori Simidian's girl. on speaker phone, no less. Yes, her name was Mallomar, and she participated in the study four years ago. To say that we were all blown away is an understatement.

Lori Simidian;s
Takari’s Cold Weather Cookie
11/24/01 - 9/22/08
B-G Dog ID = 1888

Dr. Breen may provide the brains of the lab, but Tessa provides the heart. While HM is one of the worst aspects of our breed, we are truly lucky to have the participation of researchers like those at NCSU.

Dr. Breen's Biography

See also this recent article