Why Would a Biopsy Be Sent for a Second Opinion?

Biopsies get sent for second opinions because pathology diagnosis is, at its core, a human interpretation of tissue under a microscope, and that interpretation can vary from one pathologist to another in ways that change your treatment. Across thousands of reviewed cases, studies consistently find that somewhere between about 2% and 6% of second opinions reveal a major disagreement that alters what happens next for the patient. The rate of any discrepancy, including minor ones that don’t change treatment, is considerably higher. When the stakes involve surgery, chemotherapy, or watchful waiting, even a small chance of a wrong call is worth catching.

How Often Pathologists Disagree

When people hear that a biopsy needs a second look, they sometimes worry that something went badly wrong the first time. In reality, second-opinion review is a routine quality-control step in modern medicine, and the disagreement rates reflect the genuine difficulty of the work, not carelessness. A review of more than 4,200 second-opinion cases found that about 3.7% had a major discordance that didn’t change the treatment plan, and about 1% had a major discordance that did change treatment.¹ That 1% sounds small in the abstract, but when you’re the patient whose cancer would have been under-treated or over-treated, it’s enormous.

Other institutions report higher numbers. A study at a large Korean medical center found discrepancies in roughly a quarter of referred surgical cases, with about 6% classified as major, meaning the diagnosis itself changed or the type of malignancy was reclassified.² The biggest single driver of major disagreements was a straightforward difference in how two pathologists interpreted the same tissue, accounting for over 80% of major discrepancies. Additional stains or clinical context explained the rest.

The spread between these studies isn’t a contradiction. It reflects differences in which cases get referred, how aggressively the reviewing institution reclassifies, and what counts as “major.” Referral cancer centers tend to see trickier cases, so their discordance rates run higher than community hospitals that send everything for review. The consistent finding across studies is that a meaningful fraction of second opinions turn up something the first pathologist missed or interpreted differently, and in a subset of those, the patient’s treatment changes as a result.

What Counts as a Major Versus Minor Discrepancy

Pathology second-opinion studies draw a line between discrepancies that alter clinical management and those that don’t. A major discrepancy is one that changes what your doctor does next: a benign diagnosis gets reclassified as malignant, or a low-grade tumor turns out to be high-grade, triggering a shift from surveillance to surgery or chemotherapy.³ A minor discrepancy, by contrast, might refine the diagnosis in a way that adds useful prognostic information but doesn’t change the treatment plan itself.⁴

An example helps here. Suppose the first pathologist calls a breast lesion “atypical ductal hyperplasia” (a borderline finding that usually leads to close monitoring or minor surgery), and the second pathologist calls it “ductal carcinoma in situ” (a pre-cancer that might need wider excision or radiation). That’s a major discrepancy. If instead both call it DCIS but disagree on whether it’s low-grade or intermediate-grade, that’s typically a minor discrepancy since the treatment plan stays roughly the same. Minor discrepancies are far more common than major ones, but even they matter over time because grading and subtyping feed into long-term risk calculations.

Diagnoses That Trip Up Pathologists the Most

Not all tissue types are equally hard to read. Certain diagnoses sit in gray zones where even experienced pathologists routinely disagree, and these are the cases most likely to be sent for second opinions.

Breast Lesions

Breast pathology is the poster child for diagnostic variability. The spectrum from normal tissue through atypical hyperplasia to ductal carcinoma in situ is a continuum, and drawing a line between categories is genuinely subjective. Agreement between general pathologists and breast specialists on atypical ductal hyperplasia and low-grade DCIS is only moderate, with inter-observer agreement scores hovering around 0.44 to 0.47 on a scale where 1.0 means perfect agreement.⁵ Put plainly, if you showed the same breast biopsy slide to several pathologists, they’d agree on the diagnosis less often than you’d expect for something so consequential.

Atypical ductal hyperplasia is a particularly tricky call. It carries the highest risk of underestimation among borderline breast lesions found on needle biopsy, and its features differ from DCIS only by size.⁶ That’s a thin line to draw under a microscope. Many breast pathology labs have internal policies requiring a second review for these diagnoses precisely because the variability is so well documented. A survey of laboratory practices found that 65% of labs mandated a second opinion for invasive breast cancer and 56% for DCIS, but only 36% for atypical ductal hyperplasia, despite its being one of the most variable diagnoses.⁷

Skin Lesions and Melanoma

Melanocytic skin lesions are another well-known trouble spot. Distinguishing a moderately abnormal mole from early melanoma can be remarkably difficult. A large study modeling second-opinion strategies found that after a single pathologist’s interpretation, the misclassification rate for moderately dysplastic nevi was about 75%, and for severely dysplastic nevi or melanoma in situ it was around 60%.⁸ That number is startling, and it doesn’t mean three out of four pathologists are incompetent. It means that these middle-of-the-spectrum lesions genuinely look different to different trained eyes, and no single interpretation is reliably “correct.” Benign nevi, at the harmless end, and thicker invasive melanomas, at the clearly dangerous end, had much lower misclassification rates. The gray zone is in the middle, right where the treatment decisions are most consequential.

Prostate Biopsies

Prostate cancer grading has its own layer of subjectivity. A study comparing initial biopsy grades to second-opinion grades found full agreement in about 62% of cases.⁹ The grading of prostate cancer matters enormously because it determines whether a man is offered active surveillance, radiation, or surgery. Some institutions now require a mandatory second opinion on prostate needle biopsies before proceeding to radical prostatectomy, since even small grading shifts can redirect care.¹⁰

Why a Pathologist Might Request a Second Opinion Themselves

It’s easy to picture second opinions as something patients demand when they’re skeptical. In practice, pathologists themselves are the most common initiators. A pathologist looking at a slide might recognize that the tissue is unusual, poorly differentiated, or sits right on the boundary between two diagnoses. In those cases, the responsible move is to get another set of eyes on it, often a subspecialist who sees that tissue type hundreds of times a year.

Malpractice anxiety also plays a role, and pathologists are candid about this. A nationwide survey of breast pathologists found that 88% reported using at least one “assurance behavior” driven by malpractice concerns, including ordering extra stains, recommending additional tissue sampling, getting a second review, or choosing the more severe diagnosis when a case was borderline.¹¹ That last strategy, favoring the more severe interpretation, is a form of defensive medicine that can lead to overtreatment, but from the pathologist’s perspective, missing a cancer is a career-ending lawsuit while overcalling one is a less visible harm. Second-opinion review helps resolve borderline cases on their merits rather than through defensive overcalling.

The causes of diagnostic errors in pathology are split between genuine interpretive disagreement and problems with the specimen itself, like poor tissue quality or sampling that didn’t capture the most abnormal area. One analysis found that errors attributed to misinterpretation ranged from 5% to over 50% depending on the institution, with the remainder caused by clinical sampling issues.¹² That wide range across institutions underscores why referral centers with subspecialized pathologists tend to catch discrepancies that community labs miss: it’s partly about expertise and partly about having the infrastructure for additional testing.

When Patients Seek a Second Opinion

Patients don’t always wait for their pathologist or oncologist to suggest a second look. A systematic review of patient-initiated second consultations found that the most common motivations were wanting more information about the diagnosis, exploring treatment options, and seeking reassurance that the first diagnosis was correct.¹³ Dissatisfaction with how the first doctor communicated, including feeling rushed, was reported by roughly a fifth to half of patients who sought second opinions. Family and friends also frequently encouraged the decision, sometimes recommending a specific doctor they trusted.

The personality profile of people who seek second opinions skews toward people who actively want to be involved in their medical decisions and who are high information seekers.¹⁴ That’s worth knowing because it pushes back against the stereotype that second-opinion seekers are just anxious or difficult. They tend to be people who process a serious diagnosis by gathering as much data as possible, and in oncology, that instinct frequently turns out to be well-founded given the discrepancy rates described above.

If you’re considering requesting a second opinion on a biopsy, there are a few practical things to know. You typically don’t need a new biopsy. The original tissue sample is preserved in paraffin blocks and on glass slides, and your hospital’s pathology department can send those materials to the consulting institution. Your doctor can arrange the referral, or many academic cancer centers accept cases sent directly by patients. Insurance coverage varies, but most plans cover second-opinion pathology review when a cancer diagnosis or a borderline finding is involved. The turnaround time ranges from a few days to a few weeks depending on the complexity of the case and whether additional staining or molecular testing is needed.

How Institutions Build Second Opinions Into Their Workflow

At major cancer centers, second-opinion pathology isn’t an exception; it’s part of the system. Multidisciplinary team meetings, where surgeons, oncologists, radiologists, and pathologists sit together to discuss each new case, routinely include a review of pathology from outside hospitals. A five-year study of gynecological oncology cases reviewed at these team meetings found that the process reliably caught problematic diagnoses, particularly for lesions that were ambiguous or rare.¹⁵ The review process also identified cases where additional stains or molecular tests could resolve the ambiguity, something the original lab might not have had the resources to do.

This institutional layer of quality control is one reason referral cancer centers report better outcomes for complex cancers. It’s not that their surgeons’ hands are steadier; it’s that the diagnosis has been vetted by multiple experts before anyone picks up a scalpel. A combination of routine second opinions, structured peer review, and participation in external quality assurance programs has been shown to reduce diagnostic risk across the board.¹⁶

Digital Slides and Remote Consultations

One of the practical barriers to second opinions has always been logistics. Glass slides are fragile and irreplaceable, and mailing them across the country takes time and introduces the risk of loss or damage. Whole-slide imaging, which scans a glass slide into a high-resolution digital file, is changing that equation. Digital slides can be transmitted securely to a specialist anywhere in the world, giving patients in rural or underserved areas access to expert opinions that would previously have required traveling to an academic center or waiting weeks for a mailed consultation.¹⁷

The regulatory picture adds complexity. In the United States, pathologists are licensed by state, and the rules for practicing across state lines vary widely. A comprehensive survey of licensing laws found significant heterogeneity between states, meaning a pathologist in one state consulting on a case from another state might need additional licensure depending on the jurisdictions involved.¹⁸ This patchwork of regulations hasn’t stopped digital consultations from expanding, but it does mean the infrastructure is still catching up to the technology.

Artificial Intelligence as a Safety Net

AI is beginning to enter the second-opinion conversation, not as a replacement for human pathologists but as a parallel screening tool. In a trial of AI-assisted quality control for gastrointestinal biopsies, an AI system was able to double-check nearly 100% of slides, compared to the small fraction typically re-reviewed in conventional monthly quality-control rounds. The system detected roughly 17 to 30 times as many potential human errors as the standard process, and flagged them within about a day on average.¹⁹ When the AI classified slides as either suspicious for malignancy or not, its accuracy exceeded 97% for colorectal specimens.

These numbers are promising, but they come with caveats. The AI was tested in a controlled laboratory setting on a specific tissue type, and its performance on rare or ambiguous cases, the exact cases where second opinions matter most, is less well characterized. For now, AI tools function best as a first-pass filter that catches obvious discrepancies and flags borderline cases for human re-review. They don’t yet handle the subtlety of distinguishing, say, atypical ductal hyperplasia from low-grade DCIS, which is a judgment call that trips up experienced human pathologists. But as a way to catch the clearest misses, the ones where a malignant slide was initially read as benign, AI quality control looks like a meaningful addition to the safety net.

When Your Pet’s Biopsy Needs a Second Look

The same diagnostic uncertainties that affect human pathology show up in veterinary medicine, and the discrepancy rates are actually higher. A retrospective study of 430 veterinary oncology cases found full agreement between first and second opinions in 70% of cases, partial disagreement (usually a change in grade or margin status) in 20%, and complete diagnostic disagreement in 10%, including cases where a benign diagnosis was reclassified as malignant or vice versa.²⁰ A smaller prospective study found even lower initial agreement, with full concordance in only 52% of cases and complete disagreement in 19%.²¹

The most common reasons veterinary oncologists requested second opinions were that the tumor looked atypical or poorly differentiated under the microscope, or that the pathology report didn’t match what the clinician was seeing in the animal, like a biopsy report calling something benign in a patient whose tumor was behaving aggressively.²² For pet owners facing decisions about surgery or chemotherapy for a dog or cat, a second pathology opinion can be just as consequential as it is in human medicine. Many veterinary oncology referral centers now treat it as standard practice for complex tumors.