Assay sensitivity and the interpretation of culture-based disc microbiology studies: a scoping review

Chronic lower back pain is a universal human experience, a heavy, persistent ache that can steal years of life and movement. For a specific group of patients, doctors can see a clue on an MRI scan: small patches of inflammation or fat in the bones where the spine meets the discs, known as Modic changes. For years, researchers have wondered if these changes are caused by a silent, low-level bacterial infection hiding deep inside the spine, rather than just wear and tear. The most common suspect is a bacterium called Cutibacterium acnes, a germ that usually lives harmlessly on human skin but can sometimes cause trouble when it gets deep inside the body. If this theory is true, the treatment could be as simple as antibiotics. But if the bacteria are not there, or if the tests used to find them are flawed, patients might be treated unnecessarily or left without relief. The central question has been difficult to answer because the bacteria, if present, are likely very few in number, scattered sparsely within the tough tissue of the spine.

A new review of scientific literature, led by researchers from several institutions including Persica Pharmaceuticals and Rockefeller University, suggests that the confusion surrounding this infection stems not from a lack of evidence, but from a lack of consistency in how scientists look for it. The team examined dozens of studies that tried to grow bacteria from human disc tissue. They found that the methods used to test these samples varied wildly, and many studies failed to report the specific details needed to know how sensitive their tests were. In simple terms, some researchers used tiny pieces of tissue and processed them in ways that would likely miss a small number of bacteria, while others used larger samples and more careful techniques. Because the bacterial load is expected to be so low, these differences in method are not minor details; they are the deciding factor between finding an infection and declaring a sample sterile.

The researchers mapped out the methods used in these studies and discovered a chaotic landscape of techniques. Some studies used tissue samples weighing only a few milligrams, while others used samples nearly a gram in weight. Some researchers chopped the tissue into tiny pieces and mixed them into a large volume of liquid, which dilutes any bacteria present, while others kept the volume small. Some studies only looked for bacteria for two or three days, whereas slow-growing bacteria might need a week or more to show up. The review highlighted that only a tiny fraction of the studies provided enough information to calculate the lowest number of bacteria their test could possibly detect. Without this number, a negative result—a test that says "no bacteria found"—cannot be trusted as proof that the tissue is clean. It might simply mean the test was not sensitive enough to see the few bacteria that were there.

One of the most striking findings was how these methodological differences could change the outcome of a study. The authors compared a high-sensitivity protocol, which used a large tissue sample, thorough mixing, and a long incubation period, against a low-sensitivity protocol that used a tiny sample and a large dilution. The math showed that the low-sensitivity method was roughly eighty times less likely to detect a low-level infection than the high-sensitivity one. If a study used the less sensitive method, it would likely miss the vast majority of infections that the better method would catch. This helps explain why some studies report finding bacteria in most patients, while others report finding none. The bacteria may have been present in both cases, but the tools used to find them were too blunt for the job in the less sensitive studies, making it impossible to distinguish true infection from a lack of detection.

The review also pointed out that many studies did not clearly report whether patients had received antibiotics before their surgery. These drugs are often given to prevent infection during the operation, but they can also kill the very bacteria researchers are trying to find. If a patient received a specific antibiotic that is effective against the suspected germ, and the tissue was taken shortly after, the bacteria might be dead or too weak to grow in a culture, leading researchers to falsely conclude the tissue was sterile. The authors noted that one common antibiotic, cefazolin, can actually penetrate deep into the disc tissue, meaning it could suppress bacterial growth even if the bacteria were present. Because many studies failed to track this variable, it remains impossible to know how much the use of these drugs influenced the results.

Perhaps the most significant issue identified was the way researchers decided what counted as an infection versus a contaminant. In a low-burden setting, where bacteria are few and far between, finding a small number of colonies on a petri dish is ambiguous. Some studies automatically dismissed low numbers as contamination from the skin or the environment, while others treated them as a sign of infection. The review found that these decisions were often based on arbitrary cutoffs rather than on a validated understanding of how many bacteria are needed to cause disease. Without a clear, standardized rule, the same result could be labeled as a disease in one study and a mistake in another. The authors argue that the field needs to move away from these guesswork thresholds and toward quantitative methods that can accurately count the bacteria and report the limits of their detection.

The authors conclude that the current body of research is too inconsistent to draw firm conclusions about the role of bacteria in chronic back pain. The variation in how samples are collected, processed, and analyzed makes it impossible to compare studies directly or to trust negative results as definitive proof of sterility. They suggest that future research must be more rigorous, requiring scientists to report exactly how much tissue they used, how they processed it, how long they waited for results, and what the lowest detectable limit of their test was. Until these standards are adopted, the question of whether a silent bacterial infection drives chronic back pain will remain clouded by methodological uncertainty. The path forward lies not in more studies with different methods, but in fewer, better-designed studies that can finally see what is hiding in the dark.