Genetics in hospital medicine

Obtaining and applying genetic information can play a vital role in diagnosis and treatment.

It's harder than ever to ignore the molecular discoveries that are filling our journals.

A common pattern is that the discovery of a gene associated with a disorder also identifies a previously unrecognized pathophysiologic pathway. This may suggest additional genes involved in the condition, related to other components in the pathway, and offer potential targets to pharmaceutical developers for “magic bullet” therapies that act at the molecular level.

Courtesy of Dr Shah
Courtesy of Dr. Shah.

But serendipity has now repeatedly also given sufferers the added fortune of existing drugs, which are sometimes found to act on the pathway and may be immediately applied to interrupt the disease process. These are the fruits of basic science research that many have long awaited. Still, gene discovery and all that can derive from it must start with diagnosis. One cannot start a gene hunt without defining the phenotype.

As hospitalists, we have charged ourselves with looking after some of the sickest patients, with the rarest to the commonest presentations. But while one may argue that the likelihood is low that a genetic diagnosis will impact a stable outpatient returning for scheduled follow-up for her mild and manageable comorbidities, it's a different story when someone requires hospitalization. If I had to list the best hints that genetics may have a role in someone's presentation, severity is right up there with age of onset, frequency of recurrences and, of course, family history.

When was the last time you took a family history in order to construct the differential diagnosis? Maybe you have an assistant do it, or maybe there simply isn't enough time when the ED is paging with the next three admissions. Yet isn't diagnosis the very reason we were taught to ask for that information?

Take a patient with class IV heart failure with clean coronaries and stable chronic kidney disease and a documented decrease in ejection fraction from years ago. This is quite possibly familial amyloidosis—and the family history can be the quickest means to affirm it. The history provides the justification to finally obtain a tissue diagnosis (from neither kidney nor heart, but merely subcutaneous fat) for this cardiorenal syndrome that is usually simply managed symptomatically. If the true connection had been recognized early enough, liver transplant may have been the best, but otherwise totally unimaginable, treatment.

Or, if the history tells you your 51-year-old-with-chest-pain patient's dad and two uncles had their first myocardial infarctions at ages 47-49, your investigations and management ought to radically change. Or, take the 48-year-old Jewish woman with ovarian cancer who is under medical treatment of malignant ascites. If she had a BRCA1 mutation, she and her offspring might be at risk of breast, ovarian and other cancers. Her risk of having mutation could be estimated just by inputting simple family demographic data to a free website. Whatever her prognosis, she may choose to manage her breast cancer risk. In addition, she may also want to hear that while she may instinctively assume her daughters' risk of dying from cancer is 100%, the risk of them having the mutation is not greater than 50%, and knowing they are at risk will open the door to options that reduce the risk of cancer by more than 90%.

These are not just cardiology or women's health issues: Correct diagnosis and treatment are the foundations of internal medicine. Molecular medicine may be gaining an ability to assist, but to employ it we need to start with broad differentials in mind.

How many of us have offered the self-fulfilling prophecy that ‘‘genetic etiologies are rare’’? Most geneticists are wary of prevalence numbers because what is formally diagnosed is but the tip of the iceberg. On the other hand, reduced penetrance is an extremely confusing concept. A person may have the family history, and even the very genetic alteration that led to the disease in their relatives—for example hemochromatosis or Factor V Leiden—but her individual risk of experiencing clinical consequences is very low and it may be debatable whether to even pay it any attention at all. Many conditions are common in particular ethnicities yet uncommon generally—that does not excuse missing the diagnosis in those who have it. The more we hope to “individualize” medicine, the more we need to pay attention to the very things that make us all unique.

I hope I have illustrated that the details of obtaining and applying genetic information can be vital. No two chest pain episodes or pulmonary embolisms are exactly alike. Appreciating the different factors that brought them about is the first step to correctly treating them then and there, and allowing specific treatments to prevent progression or recurrence for the individual, as well as their relatives.

Whole-genome sequencing has begun to demonstrate the degree to which we are all predisposed to various diseases. This should be no surprise; there is no one alive who is immune to all illness. But we are now reaching a turning point in being able to intervene and this will inevitably be done by us, as hospitalists, more than anyone else.