Brian F. Mandell, MD, PhD, MACP, was very clear about which test to order when attempting to distinguish crystal disease from infection in acute monoarticular arthritis during his session at Hospital Medicine 2015 on “Effective Testing in Rheumatic Diseases.”
“The best, and perhaps only, test to order is a synovial fluid analysis,” said Dr. Mandell, professor and chairman of academic medicine in the department of rheumatic and immunologic disease at the Cleveland Clinic in Ohio. “Everything else is so indirect it is not going to answer the question that you need to answer about that patient: Crystal disease or infection?”
Although most patients who present in a hospital with acute monoarticular arthritis have crystal disease, Dr. Mandell said, a study found that 17% have documented bacterial septic arthritis. That possibility, with its 10% to 12% mortality rate, should dictate clinicians' evaluation. “Quite clearly, any delay in diagnosis leads to a worse outcome, in terms of not only morbidity but also in true mortality,” he said.
Radiographs may detect prior joint damage and osteomyelitis but “there's no way that radiographs will distinguish crystals from bugs,” Dr. Mandell said. MRIs won't do so either. Ultrasound, which is more sensitive than physical exam for effusion and tenosynovitis, can be used to direct aspiration and can reveal urate deposits but won't exclude co-infection, he noted.
“The real issue here is that imaging ... is not going to give you any diagnostic information, and essentially is a waste of time in acute monoarticular arthritis, other than if it's a deep joint in an unresponsive patient that may help you detect unrecognized trauma or fracture, such as the hip or the shoulder,” Dr. Mandell said.
The same goes for most laboratory tests, Dr. Mandell noted. Many laboratory findings, including elevated erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), leukocytosis, and Gram stain are also “distinctly insensitive” for septic arthritis, he said. Presence of a fever above 38 °C is not a good predictor, either.
“The fact is, when patients initially present with acute monoarticular arthritis, [these lab tests are] not sensitive enough to really let you know if they're sick or not sick,” he said. Patients with an elevated ESR and elevated CRP have a higher likelihood ratio for septic arthritis than those with normal levels, Dr. Mandell said, “but there's so much insensitivity and nonspecificity here that it's not going to help you make that distinction.”
Dr. Mandell also stressed that CRP and ESR are “useless, absolutely useless” in patients taking immunosuppressive drugs that blunt the acute-phase reaction, such as tocilizumab, and noted that clinicians should always be wary of a monoarticular flare in patients with rheumatoid arthritis. In addition, CRP, which changes more rapidly than ESR, is also not helpful immediately after surgery because the trauma to tissue will elevate it, he noted. “Again, sensitive, not very specific to distinguish the types of arthritis that might be going on,” Dr. Mandell said.
Clinicians may consider serologic tests for Lyme disease in patients with nonspecific musculoskeletal symptoms, but Dr. Mandell advised against that. Such tests are ordered appropriately less than 40% of the time in Lyme-endemic areas and less than 10% of the time in nonendemic areas, he noted. To be useful, Lyme tests should be performed only in patients who report the potential for exposure and an appropriate clinical picture; patients who present only with nonspecific symptoms, such as chronic fatigue, should not be tested for Lyme disease, he said.
A urate level might seem like a good tool to distinguish other forms of arthritis from gout, but Dr. Mandell pointed to 2 studies showing that 50% or more of patients with septic arthritis or calcium pyrophosphate deposition disease had hyperuricemia (urate >8 mg/dL). “In fact, it's a coin toss if you just look at urate,” he said. “I would go all out and say it is of zero value in the setting of acute monoarticular or even oligoarticular arthritis.”
Further, a low urate level doesn't necessarily mean gout is less likely, Dr. Mandell noted. “This is kind of one of those myths that exist, that if your urate is low you're not going to get an attack of gout,” he said. The literature shows that normal or even low serum urate levels are common at the time of an acute gout attack. “One of the reasons for that is when you first start people on urate-lowering therapy, you're likely to induce attacks of gout,” he said. “There's really not much reason to get a serum urate when evaluating a patient who has acute arthritis, thinking that that will help you say it's gout or not gout.”
He outlined what he called a “1-drop” approach to evaluating synovial fluid for acute monoarticular arthritis. He noted, “In real life most of us in the hospital do not have access to a microscope in order to do this, but conceptually, this is the way I think that it should be done, and it's useful to think about it in this way.”
Wet microscopic analysis of 1 drop of synovial fluid can yield as accurate a white blood cell count as a tube of fluid, Dr. Mandell said, because it will be fresh and the cells will not settle. Crystal analysis of this fluid is also possible, he noted. A cell count of greater than 7, 500 cells/mm3 or more than 85% or 90% polymorphonuclear leukocytes indicates a significant inflammatory reaction. “That's inflammatory fluid that needs an explanation. That's not osteoarthritis, that's inflammatory fluid,” he said.
Once disease is proven not to be inflammatory, radiographs could be useful because the cause of the patient's symptoms could be something more chronic, Dr. Mandell said. If no crystals are seen but inflammatory fluid is noted in an acutely inflamed joint, that indicates an infected joint until proven otherwise, he said. And “If you see crystals, treat for crystals. You have a diagnosis,” Dr. Mandell stressed. “Is it OK to take a culture of that fluid as well? Absolutely. But I think if you have a diagnosis, there's no reason not to treat.”
Rheumatoid arthritis and lupus rarely cause monoarticular arthritis, so testing for them is usually a waste of money in that setting, Dr. Mandell said. In addition, the results probably won't come back until the patient has been discharged. “Don't get a rheumatoid factor. Don't get a CCP [cyclic citrullinated peptide]. Don't get an ANA [antinuclear antibody]. There's no reason to do that” in the patient with acute monoarticular arthritis, he said.
In addition to evaluating acute monoarticular arthritis, Dr. Mandell also discussed how to distinguish liver from muscle injury.
Testing for aspartate aminotransferase and alanine aminotransferase is not helpful because both are present in liver and muscle, he said. Aldolase, meanwhile, has the same organ distribution as lactate dehydrogenase (LDH) and is not specific for muscle, he emphasized. Creatine kinase is the most specific test for muscle disease versus a liver source, Dr. Mandell said. But, he noted, clinicians should be aware that it loses specificity for cardiac muscle in the setting of regenerating skeletal muscle.
When using acute-phase reactants to evaluate systemic syndromes, Dr. Mandell reminded attendees that CRP changes more rapidly than ESR. In lupus patients with a fever, high CRP suggests infection or serositis, not active lupus. ESR, meanwhile, is elevated by paraproteins and hypergammaglobulinemia, decreases in disseminated intravascular coagulation, and increases with age, but changes minimally with anemia, except in severe cases.
High CRP or ESR rarely suggests a specific diagnosis, Dr. Mandell said. Either may be used to distinguish inflammatory from noninflammatory conditions, “but that's not uniformly true. I'm sure you've all encountered cases where there's a mismatch between [ESR] or CRP and the patient,” he said. “The classic, high [ESR], low CRP, that suggests a paraprotein,” he said.
In chronic infections like endocarditis, Dr. Mandell said, ANAs, anti-neutrophil cytoplasmic antibodies (ANCAs), and rheumatoid factor are often positive and therefore do not suggest a specific diagnosis. “You just cannot count on those tests alone, so you really should be comfortable with what your pretest likelihood is before ordering those specific serologies,” he said.
For example, a study found that almost 10% of 194 patients with PR3+ C-ANCA, which is thought to be the most specific test for granulomatosis with polyangiitis or Wegener's granulomatosis, did not in fact have primary vasculitis, Dr. Mandell stressed. He also noted that drug-induced ANCAs are fairly common and will yield the highest positive results for perinuclear ANCAs with antimyeloperoxidase antibodies.
Dr. Mandell said that his overall goal was not to make his audience experts in ANCAs but rather to help them realize that these tests are not totally specific and that in fact there is no specific blood test that can be used to diagnose vasculitis.
“I talk to the housestaff about this and they say, ‘Well, we'll get the vasculitis panel.’ I'm a vasculitis guy, and I don't know about a vasculitis panel. I'm glad they do, but I don't,” he said. “You really need to be very wary about the indiscriminate use of serologies to diagnose primary vasculitis, lupus, or other systemic autoimmune diseases.”