Dr. Teja M. Kapoor

As a clinician researcher in the Lupus Center in the Division of Rheumatology at NewYork-Presbyterian/Columbia University Medical Center, Teja M. Kapoor, MD, spends a good deal of her time on the trail of opportunistic infections that can develop in patients with systemic lupus erythematosus (SLE).

“Lupus is a very complex disease,” says Dr. Kapoor, who previously completed a fellowship in rheumatology at Columbia and joined the faculty last July. “It tends to affect young women who are in the prime of their lives. Along with coping with this lifelong disease, they are also prone to developing infections, sometimes from the medications used to treat lupus.”

Powerful immunosuppressants typically prescribed for SLE increase the risk for infections, a major cause of morbidity and mortality. One of the best-studied opportunistic infections is Pneumocystis carinii pneumonia (PCP); however, the prevalence of PCP in SLE is not clearly defined. “Rheumatologists have differing opinions about using PCP prophylaxis, and the medication that is most commonly used may cause trouble for our lupus patients,” says Dr. Kapoor. She and her Columbia colleagues set out to analyze the prevalence of the infection, with a focus on validating the PCP and SLE diagnoses using an electronic medical record database.

Working with a team at Columbia University’s Clinical Data Warehouse, the researchers analyzed electronic medical record data of patients treated at Columbia from 2000 to 2014 using ICD-9 diagnostic codes. “We then created a cohort of 2,013 patients identified as true SLE based on methods we validated,” says Dr. Kapoor.

“The strength of having an electronic medical record-based database, rather than an administrative database, is it gives us access to the actual medical records themselves,” says Dr. Kapoor. “We can ‘dive in’ and gather very specific information on laboratory data, tissue cultures, and biopsies and confirm whether a diagnosis was an actual diagnosis.”

With chart confirmation, overall the investigators determined that the prevalence for PCP in lupus in this cohort was only 0.45 percent – a very low number – while the prevalence of PCP in the patients with AIDS cohort was higher at 5.98 percent. “We were able to screen out patients who were billed as SLE or having PCP infection, but actually did not have the disease or the infection at all,” she says. “We also found out that only one patient had actual PCP on histology. The other patients were based on a clinical and radiographic diagnosis for PCP. Remarkably, the one patient with actual PCP had coexisting HIV/AIDS, which is more likely to have put the patient at risk for PCP, rather than the lupus itself.”

These results, says Dr. Kapoor, make a strong point against the routine use of PCP prophylaxis in the treatment of lupus patients. “While it is important to prevent infections in patients with lupus, weighing the risks of medications is also key. Their risk for PCP is low, while at the same time, there is a high risk to develop other side effects from the drug.”

According to Dr. Kapoor, patients at high risk for PCP are prescribed a PCP prophylaxis – generally trimethoprim/sulfamethoxazole (TMP- SMX), a sulfur-based antibiotic. “The reason why the study results are significant to our care of lupus patients is because a number of lupus patients can have an allergic reaction to medicines with sulfa. It can also lead to a lupus flare in 21 percent of these patients. There are alternatives to TMP-SMX, but the other options can cause gastrointestinal distress, are more expensive, or may not work as well as TMP- SMX,” she notes. “Considering the risks and benefits when using TMP-SMX, we find the benefit very low. From our data we’ve shown that even when they are immune-compromised, there is a low prevalence of PCP in lupus patients overall, and that they may not need PCP prophylaxis unless they have HIV/AIDS.”

Historically drug therapies for lupus have come from medications used in other specialties. “The most common are cyclophosphamide and rituximab, typically used in hematological malignancies,” says Dr. Kapoor. “We also borrow from transplant, using mycophenolate mofetil or tacrolimus and cyclosporine. There is a hypothesis that hydroxychloroquine, an anti-malarial agent used very commonly in lupus, could be a protective medication against PCP.”

In another study, Dr. Kapoor and the Columbia team examined progressive multifocal leukoencephalopathy (PML), a potentially fatal demyelinating disease in the brain caused by the John Cunningham virus, to evaluate the prevalence of PML in adult and pediatric SLE patients. The researchers focused on validating PML and SLE diagnoses with clinical information again obtained from corresponding medical records in order to better define the risk of PML in SLE. They concluded that the prevalence of PML in adult lupus patients is less than 3/10,000 patients and could not find any PML in children.

Dr. Kapoor undertook another retrospective cohort study of Columbia electronic medical records to establish the prevalence of herpes viruses, both varicella-zoster virus and herpes simplex virus, in SLE patients. “These viruses are very common in immune-compromised patients, and we are finding a high prevalence of these infections in lupus – especially with varicella zoster virus, which causes shingles. In our cohort, not only is there a high rate of shingles, but it is also occurring at a much younger age in lupus patients and causing organ-threatening and life-threatening complications. It highlights the need for having the shingles vaccine for this group, too. Even though they might not meet the age requirement that’s now set for the shingles vaccine, lupus patients are at higher risk of developing shingles. Looking into prophylaxis for this infection may be beneficial for our lupus patients.

“We are also studying patients who are in the beginning stages of developing lupus,” Dr. Kapoor says. “They are usually categorized as having undifferentiated connective tissue disease. Fewer than 20 percent progress to an autoimmune disease; about 30 percent go into clinical remission; and 50 percent usually continue with just undifferentiated connective tissue disease. We are looking to prevent the progression of lupus by analyzing the biomarkers in the very high-risk patients. If we’re able to identify who is in this high-risk group, we can try to treat them earlier and more aggressively to see if we can prevent them from ever developing lupus.”

Even with dramatically increased understanding, SLE seems to traffic in a kind of paradox that requires medical detective work to confront it. “As we learn more and more about this disease,” says Dr. Kapoor, “we will be better able to create a more targeted treatment approach. There are newer medications that are being tested. Newer research is helping us to clarify who lupus affects and how it affects certain patients, and also to identify the risks of some of these medications.”