On March 23, Nature Genetics published 3 related papers reporting the finding that SMARCA4 is frequently mutated in a rare ovarian cancer type, small cell carcinoma of the ovary, hypercalcemic type (SCCOHT) [Jelinic et al 2014, Ramos et al 2014, Witkowski et al 2014]
The fact that 3 independent research efforts made virtually the same discovery is, in a sense, remarkable, but it is also a reflection of just how critical this mutation is to the development of SCCOHT.
SCCOHT is a bit of a misnomer. Despite the fact that it is called a “carcinoma,” the World Health Organization classifies it as a “miscellaneous tumor.” As Dr. William Foulkes, senior author of one of the papers [Witkowski et al 2014], says:
“Like the Holy Roman Empire (not Holy, not Roman and not an Empire) not all cases of SCCOHT contained small cells, the term carcinoma was somewhat arbitrary and a third of patients never developed hypercalcaemia.”
SCCOHT is an extremely rare, very aggressive ovarian tumor that is most common in young women and girls. Regardless of the fact that most patients present at an early disease stage, the majority die within 2 years of diagnosis. You can find more about this tumor type at the Small Cell Ovarian Cancer Foundation’s website. Because SCCOHT is so rare, each of the three studies relied on archival samples (formalin-fixed paraffin-embedded tissue) and used different strategies to identify the genetic landscape of the tumor.
Douglas Levine’s group at Memorial Sloan-Kettering Cancer Center in New York used a candidate gene approach, sequencing the coding regions of 279 cancer-related genes in 12 pairs of tumor/normal tissue. They found biallelic inactivating mutations in SMARCA4 in each of the 12 tumor samples. Even more interesting, there were virtually no other recurrent mutations in any cancer-related gene in the tumors. Where suitable tissue was available, they were also able to show that the protein product of SMARCA4 was absent in the tumors. To test the functional consequences of SMARCA4 loss, the group introduced the protein into a lung cancer cell line that lacks it. By reintroducing SMARCA4, they were able to suppress cell growth. Depleting SMARCA4 transcript from another cell line had the opposite effect. (Read more about this study here and here)
William Foulkes’ group at McGill University started with familial cases of SCCOHT and used whole-exome sequencing to identify mutations. Although they suspected SMARCA4 already from previous work, they used the whole-exome approach to, paradoxically, obtain higher quality tumor sequence data for SMARCA4 than they were able to by traditional Sanger sequencing. This had the effect of not only identifying mutations in SMARCA4 in all 4 affected families, but also allowed the researchers to conclude, as with Dr. Levine’s group, that there were no other strong candidate genes that could be drivers of SCCOHT. Luckily, whole-exome sequencing turned out to be possible with these archival samples. They extended their sequencing efforts to additional non-familial samples and looked for protein loss as well. In all, 38/40 tumors showed loss of SMARCA4 protein. Similar to the other studies, they found mutations throughout the entire length of the gene and found that nearly all samples carried SMARCA4 mutations.
An interesting aspect of the paper by Witkowski et al. is that the authors propose a re-classification of SCCOHT to extra-cranial rhabdoid tumors instead of carcinomas based on both histology and the finding that SMARCA4 is the driving mutation. Dr. Foulkes characterized the findings as both a “game changer” for SCCOHT and a “name changer.” (read more about this study here and here)
Finally, a collaborative group led by David Huntsman and Jeffry Trent at TGen in Phoenix, AZ also identified germline and somatic mutations (as in Witkowski) in 75% (9/12) of their tumor samples and loss of SMARCA4 protein in 14/17 samples. The group used whole-genome sequencing of both tumor and blood to identify mutations causing SCCOHT. Even with extensive genome sequence data, they found that SMARCA4 is the only significantly mutated gene in this tumor type. They also demonstrate that SMARCA4 mutation is very specific to SCCOHT–only 0.4% of other tumor types carry this mutation (a similar discovery was also reported by Witkowski et al). (Read more about this study here)
The significance of these studies is two-fold. First, they pinpoint the driving mutation behind this extremely aggressive tumor type and give clinicians a new tool to diagnose them. Second, they open the door for development of specific therapies targeted to SCCOHT, of which there currently are none.
In the end, these types of studies have a single goal in mind: to give people with terminal illnesses a chance at survival. To close, I’d like to share this story from Dr. Foulkes related to the SCCOHT study:
“One of the most satisfying parts of the project was working directly with women and their families from all around the world. When we found the familial mutation in family 3, the aunt of the proband decided to have a preventive oophorectomy, as her identical twin sisters both died of SCCOHT. While this might not be sensible for all women at risk, you can appreciate her concern. Now we have to try to find better treatments. If you talk to Maren Petersen, from the small cell ovarian carcinoma foundation, she will tell you how far things have moved on since her daughter was diagnosed. Unfortunately, she did not survive. We have to hope this work will in the future help other women to escape her fate.”
Author: Brooke LaFlamme, assistant editor Nature Genetics
Follow me on Twitter: @Brooke_LaFlamme