How cancer is tested

In brief

What a pathology report is

Almost every solid-tumor diagnosis is made on tissue that has traveled the same road. It is removed from the patient and fixed, which means soaked in a chemical that stops it decaying and locks its structure in place. It is then dehydrated, cleared (its alcohol replaced by a solvent that mixes with wax), soaked in molten paraffin wax, set in a wax block, sliced on a precision blade called a microtome, stained, and read under a light microscope ¹. The two workhorse chemicals of that road, formaldehyde and xylene, are hazardous enough that engineers have built processors designed to do without them ¹. Around the slide sits a growing ring of further tests: antibodies that stain one chosen molecule, chromosome studies, and DNA sequencing ^3,15,16.

A pathology report is best understood as a set of answers to questions somebody chose in advance. Is this cancer, and what kind? How large is it, how far has it spread, are the cut edges of the removed tissue clear, and which drug targets does it carry? The College of American Pathologists (CAP), the specialty's professional body in the United States, defined the content of the checklists that now structure most American cancer resection reports ^4,23, and its guidelines require every antibody test to be validated so that it measures 'the analyte of interest', the one molecule it was chosen to find ³. For the questions they were built to answer, the answers are good: 115 American pathologists reading single breast-biopsy slides agreed with an expert reference on 96% of invasive cancers ¹⁰.

This page follows a specimen through that chain and asks, at each step, how a parasite could be in the tissue and never appear in the report. The answers come in two strengths, and the page keeps them apart. Some failures are documented: a study measured them, or a published case shows one happening. Others are plausible: they follow from how a step works, but nobody has counted them. The counterweight belongs at the start as well. When parasites are present and someone looks, pathologists name them, from pinworms in removed appendices to a tapeworm cyst inside what imaging had shown only as a pancreatic mass ^24,25.

The full history

The method's oldest parts predate the microscope itself. Hematoxylin, the blue dye that colors cell nuclei on nearly every diagnostic slide, comes from the logwood tree, which Spanish explorers found in the Yucatán in 1502; amateur microscopists first used it to stain cells in the mid-1800s, and it remains the most popular nuclear stain in histology, the microscopic study of tissue ³². Formalin entered anatomy and histology after Ferdinand Blum discovered its fixing effect by accident in 1893 ³³; he, Ferdinand Julius Cohn and others put formaldehyde solutions to work preserving soft tissue ³⁴. Special stains that pick out microorganisms, scarring and abnormal deposits had been part of tissue diagnosis for nearly a hundred years by 2000 ⁹.

Pathology also found the first parasite-cancer links. In 1911 A. R. Ferguson, professor of pathology in Cairo, reported from 40 autopsies a likely association between bladder cancer and the granulomas, inflammatory nodules, that schistosome eggs provoke ³⁵. It also produced the field's most famous error. In 1926 Johannes Fibiger received the Nobel Prize for a 'Spiroptera carcinoma' said to be caused by a worm in rats ³⁶; a 1952 re-examination set the worm against vitamin A deficiency in the rat forestomach ³⁷, and a 2004 study of the Nobel archives calls it 'a wrong Nobel Prize' ³⁸.

The twentieth century added tests that each ask about one named thing. In 1956 Tjio and Levan established that a human cell carries 46 chromosomes, and the published recollections of how it happened differ strikingly from one another ³⁹. In 1974 Taylor and Mason showed that an enzyme-labeled antibody could find one specific protein, immunoglobulin, inside cells in formalin-fixed paraffin sections ⁴⁰. In 1991 a team whose first author worked at a reagent company showed that heating slides in a microwave oven, to as much as 100 °C in metal salt solutions, improved staining for 39 of 52 antibodies, including in long-fixed tissue that had failed to stain by older methods ⁴¹. DNA sequencing came last. By 2017 one cancer center had sequenced the tumors of more than 10,000 patients with a clinical gene panel ⁴², and the same year brought standards for the software that interprets the sequence ⁴³.

In the 1990s the profession measured itself in audits like these. A 1994 audit across 233 small hospitals found that 1.8% of the frozen-section diagnoses not deferred, the rapid answers given while the patient is still in surgery, disagreed with the final slides, most often because a tumor was underdiagnosed through block or tissue sampling error ⁴⁴. At one referral hospital, second review of outside cases changed 1.4% of diagnoses enough to alter treatment or prognosis ⁴⁵. Guidelines followed, admitting that up to 20% of estrogen- and progesterone-receptor tests worldwide might be wrong ⁴⁶, and then setting out how every antibody test must be validated ^3,47.

Eight reversals run through this history, and each came from someone measuring what routine practice had assumed. The human chromosome count proved to be 46, not the 48 accepted for more than 30 years ⁴⁸. Fibiger's Nobel-winning worm cancer did not survive re-examination ^37,38. Underfixation, not the long-feared overfixation, proved the bigger threat to antibody staining ⁴⁹. Hidden metastases found by extra cuts and an extra stain proved to matter little to survival ¹⁸. A single biopsy proved not to represent a tumor's genetics ⁵⁰. Molds named by shape proved, on DNA testing, sometimes to be other molds ^7,51. A tapeworm proved able to grow like a cancer inside a person while looking nothing like a tapeworm on the slide ²⁹. And the most prominent claim that tumors carry readable microbial signatures was withdrawn, with later reanalysis widening the correction ^21,22,52.

The journey of a specimen, step by step

The first choice is how much of the patient to take. A resection removes an organ or part of one. A biopsy takes a sample: by forceps through an endoscope, by a hollow core needle, or by freezing a piece onto a probe, which is called cryobiopsy. Cytology collects loose cells by fine-needle suction, brushing or drained fluid, and gives up the tissue's architecture. Quantity changes answers. In 300 early breast cancers, the grade on core biopsy, a score of how abnormal the cells look, matched the removed tumor in 73% overall, and in 83% when the cores totaled at least 50 mm against 68% when they did not, an association across cases rather than a trial ⁵⁷. Across 132,352 American biopsies of the duodenum, the first part of the small intestine, the recommended four or more pieces arrived in only 35%, and celiac disease, an immune reaction to gluten that damages the intestinal lining, was newly diagnosed in 1.8% of those against 0.7% of the rest, an association in retrospective data ⁵⁸.

Once tissue loses its blood supply, its cells keep working and begin digesting themselves until fixative reaches them. Warm ischemia time is the stretch while the tissue is still in the body with its blood supply clamped; cold ischemia time runs from removal to fixative. The breast-cancer guidelines of the American Society of Clinical Oncology (ASCO) and CAP ask for fixative within one hour and fixation for 6 to 72 hours ⁶. In 277 Japanese breast operations in 2018 and 2019, the median warm ischemia was 23 minutes, cold ischemia 37 minutes and fixation 43 hours, and 91.7% of cases met the guideline for cold ischemia and 94.9% for fixation time ⁵⁹. Practice elsewhere is rougher: in an earlier retrospective audit of 64 breast specimens at one Cameroonian laboratory, restated in a 2026 study there, 85.9% underwent what the authors call prolonged fixation, with mastectomies left in formalin for up to 147 hours, largely because specimens accumulated over weekends and holidays ⁶. Some molecules do not wait at all: in 93 breast cancers, several signaling proteins carrying phosphate tags, the switches cells use to pass on messages, lost staining within one to two hours, and such proteins have been reported to degrade within 30 minutes ⁶⁰.

The chemistry of fixation has its own page on formalin; the short version is this. Routine '10% formalin' is a one-in-ten dilution of a stock solution that is itself about 37% formaldehyde, so it holds 3.7% formaldehyde in water with about 1% methanol, and almost all of that formaldehyde exists as methylene glycol, a hydrated and less reactive form, so formalin soaks in quickly but fixes slowly: the first cross-links form over 24 to 48 hours, and fully stable ones may take about 30 days ⁵. Cross-linking means welding neighboring molecules together with small carbon bridges, which preserves shape and damages molecules. Prostate tissue fixed for a day or more stained more reliably than tissue rushed through the same day ⁴⁹. Bone needs decalcification, the removal of calcium so a blade can cut it: hydrochloric acid and long formic-acid soaks caused false-negative antibody and DNA results in 35 samples, while EDTA (ethylenediaminetetraacetic acid), a chemical that gently pulls calcium out, did not ³¹.

Grossing is the bench step in which a pathologist or a pathologists' assistant describes the specimen by eye, slices it, and chooses which pieces go into cassettes, the small plastic cages that carry tissue through processing. One laboratory chills mastectomies to −80 °C for 20 minutes so that they can be sliced evenly at 4-mm intervals ². Only the chosen pieces are ever seen under a microscope. How much that choice matters can be measured wherever someone has embedded everything: a benign gallbladder lesion called adenomyoma appeared in 9.3% of 203 gallbladders submitted whole, but in only 3.3% of a separate series of 2,347 routinely sampled ones, a comparison between two groups of patients rather than a test on the same organs ²⁶. How much is submitted is set by rules, and the rules are under pressure. When a hysterectomy follows a cone biopsy of the cervix, protocol calls for the entire cervix to be submitted, an average of 16 blocks per case, with a range of 4 to 41; a 2025 study proposed cutting this to one representative section per quadrant, saving about two to six hours of laboratory work per case ⁶¹. Less tissue submitted means fewer chances to come across anything nobody was looking for.

Processing replaces the water in tissue with wax so it can be sliced thin. The cassettes pass through graded alcohols, a series of increasingly pure ethanol baths that draw water out, then xylene, a solvent that displaces the alcohol and mixes with wax, then molten paraffin, and the tissue is finally embedded in a wax block ^1,6. Embedding sets the tissue's orientation in the block, and so which face the blade will cut; thin, narrow pieces curl during processing and are hard to orient, so they can end up cut at a slant ⁶². The chemicals are more forgiving of time than their reputation suggests: in tissue fixed for three to five weeks, an extra 24 hours in absolute ethanol, xylene, molten paraffin or on a heated embedding station caused no visible change in structure or routine staining, and most antibody tests were preserved ⁶³. The tissue shrinks on the way: 34 kidney tumors measured 11.4% smaller on the slide than when fresh, enough to move two of them into a lower size stage ⁸. The DNA is damaged as well: DNA extracted from formalin-fixed tissue is fragmented and carries chemical lesions that can masquerade as mutations ³⁰. It remains usable for targeted tests: one laboratory amplified 300- to 400-base-pair stretches from formalin-fixed, paraffin-embedded tissue, the standard archival state known as FFPE, and judged the DNA suitable for molecular testing ², a base pair being one rung of the DNA ladder.

The block is sliced on the microtome into ribbons 4 to 5 micrometers thick, a micrometer being a thousandth of a millimeter, and each ribbon is floated onto a glass slide ^6,7. The slide is stained with hematoxylin and eosin, H&E, the blue-and-pink pair: eosin is the pink dye that colors proteins in the cell body and the surrounding tissue. The hematoxylin acts as hemalum, a red complex of aluminum and hematein, the oxidized form of hematoxylin, that is drawn electrostatically to the phosphate backbone of DNA and turns blue and insoluble when the slide is rinsed at a pH of 5.5 to 8.5, near neutral on the 0-to-14 scale of acidity ⁶⁴. Everything else, from special stains to antibodies to sequencing, is called ancillary, meaning supplementary to that first look ²⁷.

How much of a tumor reaches a slide

The arithmetic is simple, and it is the heart of the matter. A single section 5 micrometers thick, cut from a piece of tissue a centimeter thick, contains one two-thousandth of that thickness, or 0.05%. From a 3-mm piece in a cassette, one section is about one six-hundredth. A core from a 14-gauge needle is about 1 mm across and 5 to 15 mm long ⁶; a 15-mm core from a tumor 2 cm across holds about 0.28% of the tumor's volume, and one section along that core holds about one part in 56,000 of the tumor. These figures are this page's arithmetic from published dimensions, not measurements, and real practice cuts several sections from several blocks. They make one point: a slide is a sample of a sample of a sample.

Geometry decides whether something small is caught. An object 100 micrometers across, sitting at a random depth in a 3-mm piece, is crossed by any one section plane with a probability of about 100 in 3,000, roughly 3%, again by arithmetic, for a single object; many eggs scattered through a piece raise the odds, which is why sections succeeded more often when eggs were dense ¹². That is why pathologists cut deeper 'levels' when the first slide is unrevealing, and the yield is real. In a randomized breast-cancer trial, sentinel lymph nodes, the first nodes draining a tumor, that had been reported negative were re-cut at two widely spaced deeper levels and stained both routinely and with an antibody for cytokeratin. Hidden metastases appeared in 15.9% of 3,887 patients, but the first look had been designed only to catch deposits larger than 2 mm, and five-year survival differed by just 1.2 percentage points ¹⁸. In the audit of 18,532 frozen sections, the commonest disagreement was an underdiagnosed tumor, usually through block or tissue sampling, and the authors advised deeper cuts or more tissue when a first section is negative or unrevealing ⁴⁴.

The best measurement of this loss for a parasite is old, small and unambiguous. In 228 women living where Schistosoma haematobium is common in Tanzania, eggs in cervical tissue were found in 112 (49%) by the quantitative compressed biopsy technique, which squashes the fresh piece flat between glass and looks through all of it; in 40 (18%) by routine histological sections; and in 6 (3%) by cervical smears ¹². Relative to the squash, sections found about 36% of the positive women and smears about 5%, by this page's arithmetic from the published counts. Sections succeeded more often when eggs were dense, which is what the geometry predicts ¹².

Tumors are not uniform, which compounds the problem. Sequencing several regions of the same kidney cancers found that 63% to 69% of all mutations could not be detected in every region ⁵⁰. A benign gallbladder lesion was recorded in 9.3% of gallbladders embedded whole against 3.3% of a separate, routinely sampled series, a comparison between two groups of patients rather than a test on the same organs ²⁶. The pattern across these studies is consistent: for small lesions and whole organisms the documented losses come from sampling, and they fall hardest on whatever is small, sparse or unevenly spread ^12,18,26; no study found has measured a comparable loss from chemistry.

Tests that answer only the question asked

Everything after H&E is a question about a named thing. An antibody stain asks whether one chosen protein is present; a gene panel asks whether chosen stretches of human DNA are altered; a chromosome study asks whether the human chromosomes are the right number and shape. The rules that govern these tests make the point themselves. CAP requires each antibody test to be validated with cases chosen so that it 'accurately measures the analyte of interest' in that laboratory's own specimens ³, and the 2024 update extends those rules to newer scoring systems and to cytology ⁴⁷. A negative result is therefore an answer to its own question and to nothing else.

That design is a strength as well as a limit. The workup for a metastatic cancer with no obvious origin runs as an algorithm that begins with two keratins, structural proteins of lining cells, called CK7 and CK20, and branches to organ-specific antibodies ²⁷. Even so, immunohistochemistry, the antibody method, pins a single tissue of origin in only 10.8% to 51% of these cancers of unknown primary ⁶⁷. Gene panels almost always find at least one change in the tumor's own DNA: a commercial panel run on 200 such tumors reported at least one alteration in 192 (96%), in a study whose 17 authors all listed Foundation Medicine, the company that sells it ¹⁶. Every one of these questions is about a human molecule ^16,27.

• H&E histologyHematoxylin and eosin on a 4- to 5-micrometer section: the standard first slide ^6,7,64.On essentially every specimen.96% agreement with experts for invasive breast cancer, 48% for atypia ¹⁰.Anything outside the plane cut: eggs were found in 18% of women by sections against 49% by squash ¹².
• Special stains for organismsChemical stains such as the acid-fast stain, which colors the waxy-walled bacteria of tuberculosis, and silver stains, used among other things for fungi ^9,68; a laboratory keeps 20 to 25 special stains ⁹.When someone suspects an infection.Poor for small numbers of organisms: in 100 lung washes from a study of colonization, three classic stains found Pneumocystis in none, an antibody test in 13 and a DNA test in 16 ⁶⁸.Organisms nobody asked about, since each stain must be requested.
• ImmunohistochemistryAn antibody against one chosen protein, validated for that target ^3,47.Tumor type, origin and drug targets.Names a single tissue of origin in 10.8% to 51% of cancers of unknown primary ⁶⁷; up to 20% of receptor tests may be inaccurate ⁴⁶.Every protein not on the panel. Antibodies against parasites, such as the laboratory-made Em2G11 for Echinococcus multilocularis, exist in specialist laboratories and must be specifically requested ⁶⁹.
• Tumor gene panel (next-generation sequencing)Reads DNA fragments and matches them to the human reference genome ^43,70.Choosing targeted therapy.At least one human alteration in 192 of 200 cancers of unknown primary, in a manufacturer-authored study ¹⁶.One clinical laboratory discarded the reads that did not match the human genome until researchers mined them, finding 1.9% non-human ⁷¹; no survey across laboratories was found, and formalin fragments and damages DNA ³⁰.
• Metagenomic sequencing of the block (mNGS)Sequences all DNA in an archived sample and looks for any DNA pathogen ¹³.Infectious pathology cases, not routine tumors.A plausible pathogen in 36.8% of 623 samples, 9 of them parasites; 9.6% uninterpretable ¹³.Older blocks: a parasite-specific DNA-copying test (PCR) worked in all 6 blocks under six years old but in only 3 of the 9 older ones ⁶⁹; no comparable figure was found for mNGS.

What the report is built to say, and how it is checked

Cancer reports have moved from free narrative to synoptic format, a fixed checklist of required items filled in for every case. CAP defined the scientifically validated content of those checklists, and when one Canadian province put them into use over three years, both the use of synoptic reports and their completeness rose ⁴. A systematic review of 33 studies found that all but one recorded more complete reports after the change, with gains in items such as resection margins, the cut edges of the removed tissue, and invasion of blood vessels, lymphatic channels and nerves ⁶⁶. Adoption is incomplete. Across the US Department of Veterans Affairs, the largest integrated health system in the country, 48% of 1,618 cancer pathology reports from 2019 to 2021 were synoptic: 77% of resections but only 19% of biopsies ²³.

A checklist records what it asks for. Its items, in the studies reviewed, describe the human tumor: its type, size, location, depth of invasion, margins and lymph nodes ⁶⁶. Each CAP cancer protocol lists required, or core, data elements, which the protocol authors describe as 'the minimum set of evidence-based, clinically actionable parameters' for diagnosis, prognosis and treatment ⁷⁶; the protocols have guided American cancer reporting for more than 35 years ⁷⁷. Required items can reach beyond the tumor when a finding changes care: since 2010 the kidney-cancer protocol has required evaluation of the non-cancerous kidney ⁷⁸. What a report asks also follows treatment: once a drug became available for breast cancers with low HER2 protein, the 2023 HER2 update made the line between 'IHC 0' and '1+' newly important to report ⁷⁹. Cancer registries inherit the tumor's shape. IARC's 2024 estimates take national counts by anatomical site and, for subtypes such as squamous cell carcinoma of the bladder, a cancer of flat lining cells, by the tumor-type codes of the international oncology classification, ICD-O-3; the share caused by an infection is then estimated mainly from the prevalence of that infection measured in people with the cancer ¹⁷. Registries record the tumor, not the organism; the infection share is supplied by studies that tested people with the cancer, so it can reflect only the organisms somebody tested for ¹⁷. No required item this page found asks for a coexisting parasite. Whether the optional findings lists in the current bladder and intrahepatic bile duct protocols, in their June 2025 versions, name schistosomiasis or liver flukes still has to be checked in the protocol documents, which could not be opened for this review.

Quality assurance in pathology is serious, and it is measured. Agreement depends sharply on the question: on single breast-biopsy slides, 115 pathologists matched an expert reference on 96% of invasive cancers but only 48% of atypia, the gray zone before cancer, and the three experts agreed unanimously among themselves only 75% of the time ¹⁰. Mandatory review of outside cases at one referral hospital changed 1.4% of diagnoses in a way that mattered for treatment or prognosis ⁴⁵. At one large academic center, 0.33% of final reports were amended between 2021 and 2025, within the benchmarks of CAP's quality programs ¹¹. And for molecules every laboratory hunts deliberately, the 2010 guideline panel accepted that up to 20% of receptor tests worldwide might be wrong ⁴⁶; the 2020 update added an 'ER Low Positive' reporting category for tumors with 1% to 10% of cells staining ⁸⁰.

Every one of these checks compares pathologists with each other, or tests with each other, on questions already being asked ^10,45. None of them can measure how often something is missed that no one is looking for, except by going back and asking: an unasked question has no second reader. The few measurements that exist were made that way, by re-testing archived tissue for a parasite ^81,82 or by mining sequence that had been discarded ⁷¹. That limit is a matter of design rather than of evidence, and it is why this page could find no measurement of how often parasite material in tumor specimens goes unreported.

Every way a parasite can be missed

Put the steps together and there are six distinct ways a parasite can be in a tumor specimen and absent from the report. They differ in evidence. Two are documented directly for parasites: the organism was not sampled, or it was not recognized. One is documented for molecules but not for organisms: damage in processing. Two are documented outside the tumor question: the specific test was never ordered, documented in infectious pathology, and sequence reads were set aside, documented in one clinical sequencing laboratory ⁷¹. One is plausible and unmeasured: the organism was seen but not recorded, because the report had no place for it. The table keeps them apart; the paragraphs give the evidence.

Failure to sample is the best documented. Routine sections found schistosome eggs in 40 women where the squash found them in 112, about a third as many ¹². A lung fluke's eggs were found on cryobiopsy in a patient who had both cryobiopsy and the smaller forceps biopsy, and the authors argue the larger specimen gives a better chance of containing eggs ⁶⁵; surgeons describing a lung fluke in the abdomen noted in general that eggs in biopsy specimens may be difficult to detect 'due to an insufficient amount' ⁸³. Failure to recognize is documented case by case. Tapeworm-derived tumor cells were 'unrecognizable as tapeworm tissue' until a DNA test named them ²⁹; a brain cysticercus, the larval cyst of the pork tapeworm, was diagnosed as malignant glioma, a brain cancer, on needle biopsy ¹⁹; and an urgent frozen section, read during the operation, that took lung fluke disease for glandular cancer led to a resection, after which the patient died of a pulmonary embolism, a blood clot in the lungs, on the second day after surgery ²⁰. The pattern continues. In 2025, a frozen section of a bladder mass in a 10-year-old Sudanese immigrant in the United States was read as a rhabdoid malignancy, an aggressive childhood cancer; the permanent sections showed numerous schistosome eggs and no cancer ⁸⁴. In northeastern Italy, five tonsil swellings that mimicked cancer between 2014 and 2024 proved to be leishmaniasis, an infection with a single-celled parasite, which was seen on histology in four and confirmed by PCR in all five ⁸⁵.

Recognition fails for an understandable reason: pathologists rarely see these organisms. Organisms can sit on routine slides for decades before anyone names them. Whipple's disease had been associated for 85 years with a bacillus that no one could grow or identify, until DNA copied from five patients' tissue with broad-range primers named it Tropheryma whippelii in 1992 ⁵⁴. In 1983 Warren and Marshall reported 'unidentified curved bacilli' on the stomach lining in gastritis ⁵³; that organism, Helicobacter pylori, is now credited with 760,000 cancers in 2024, more than any other infection ¹⁷. A 2025 review states that parasitic cysts and pseudocysts, cyst-like spaces without a true lining, are rarely encountered by pathologists, carry many look-alike pitfalls, and, when the diagnosis is difficult, should be reviewed by an infectious-disease pathologist or a parasitologist used to tissue sections ²⁸. Even organisms pathologists see often can mislead by shape: DNA testing found a different mold, Fusarium, in five archived specimens that histology had called aspergillosis, infection with the mold Aspergillus ⁷. On 111 archived fungal specimens, metagenomic sequencing named the species in 81.1%, against 50.5% to 64.9% for two targeted DNA tests, and caught two rare fungi misdiagnosed as Aspergillus ⁵¹. A tapeworm larva in a man's hip was treated as cysticercosis, without effect, until sequencing the archived block named Spirometra mansoni, in a report coauthored by employees of a commercial laboratory ⁸⁶.

Software now reads digitized slides, but it looks for what it was trained to find. A 2025 deep-learning tool trained to find Helicobacter pylori on scanned gastric biopsies scored 0.62 for the organisms themselves on the F1 scale, a 0-to-1 measure combining hits and false alarms ⁸⁷, and another group noted that some scanned images lack the resolution to show the bacteria ⁸⁸. This review found no published tool that screens tumor slides for parasites.

Failure to order the test is documented wherever someone went back and ordered it. A parasite-specific PCR on archived bowel specimens that showed only the tissue reaction to Angiostrongylus costaricensis, a roundworm that causes abdominal disease, found the worm's DNA in 4 of 20 ⁸¹. Outside pathology, seropositive migrant women in the Barcelona area had attended a sexual and reproductive health unit a median of 41 times before schistosomiasis was diagnosed ⁸⁹. Setting reads aside is documented in at least one clinical laboratory, where reads from a cancer gene panel that did not match the human genome had been 'previously discarded' ⁷¹. The published validation standards are written for detecting genomic alterations ⁴³, research pipelines that do look for non-human reads must first strip out human sequence against several human references ⁷⁰, and no survey across laboratories was found. In research data the opposite error is documented, in which microbial signals were reported that were not there ²¹.

The counter-evidence is just as real. When organisms are present and the task includes looking for them, they are found: in cervical cytology, Trichomonas vaginalis, a single-celled parasite, was identified in 10 of 11 positive samples on both preparations compared ⁹⁰. Routine histology of 846 children's appendices identified pinworms in 12 and Taenia tapeworm in 2, although the paper's printed rate of 0.39% does not match its own counts ²⁴. A pancreatic mass, removed along with the tail of the pancreas and the spleen, proved on histology to be a hydatid cyst, the larval stage of a tapeworm ²⁵. The errors also run the other way. A brown pigment that unbuffered formalin forms with blood 'may simulate microorganisms', and fungi, fibers or fragments of another patient's tissue can contaminate a slide during processing ⁶². Formalin and wax do not routinely erase a worm, since worms and cysts are named in fixed tissue ^24,25; the documented losses are a thin plane and an unfamiliar shape ^12,28, although no study found has measured how recognizable fixed helminth material remains.

What is done when someone suspects an organism

When a pathologist or clinician does suspect a parasite, the tools are good and getting better. The first is expertise: the 2025 review recommends slide review by an infectious-disease pathologist or a parasitologist accustomed to tissue sections, completed where possible by a combined tissue-and-molecular diagnosis ²⁸. The second is a targeted test. A PCR for Echinococcus based on its 12S ribosomal DNA, a gene for part of the parasite's protein-making machinery whose sequence differs between species, detected as few as 50 parasite cells per specimen and worked on archived blocks; of 15 blocks with living lesions of alveolar echinococcosis, the liver disease caused by that tapeworm's larva, 9 were positive, including all 6 stored for less than six years, and the authors recommend blocks under five years old ⁶⁹. For the smallest particles of parasite material the antibody stain did better: 11 of 15 such samples were negative by PCR ⁶⁹. A PCR for abdominal angiostrongyliasis had 55% sensitivity, meaning it caught 55% of true cases, and 100% specificity, meaning it raised no false alarms in the comparison specimens ⁸¹.

The broadest tool is unbiased sequencing of all the DNA in a block, called metagenomic next-generation sequencing, or mNGS. At University Hospital Basel, between November 2021 and April 2025, 623 archived tissue samples were tested this way as part of routine infectious-disease pathology: 229 (36.8%) yielded at least one plausible pathogen, 334 (53.6%) were negative, and 60 (9.6%) could not be interpreted because of quality failures or suspected contamination; 9 of the positive results were parasites ¹³. That is a working clinical service, and it was applied to infectious-pathology cases rather than to unselected tumor resections ¹³. On archived surgical specimens showing granulomas, the same approach detected all 31 confirmed mycobacterial infections, with 88% specificity, and found all 19 tuberculosis cases against 47% for the standard molecular test on the same blocks ⁹¹.

Some of the most effective measures are not laboratory tests at all. For eggs in tissue, looking through a whole squashed fragment found nearly three times as many infections as sections did ¹², and a larger biopsy found eggs, its authors argue because it was larger ⁶⁵. History changes the question too. The child whose cysticercosis was read as glioma had been born in Korea, which the team did not know at the time ¹⁹, and the surgeons who found a lung fluke in the abdomen urged clinicians to take a thorough history and to remember that an abdominal mass may be a parasite ⁸³. A test that is never ordered finds nothing.

What was believed and is not

The first eight cards below are reversals inside pathology and its laboratories, each found by someone who measured what routine practice had assumed. The last two concern ideas that come up whenever parasites and cancer are discussed together. One of them, the show's own working idea that every test in a tumor workup was built for a different question than a parasite raises, holds up. The other, that formalin and paraffin make parasites invisible, does not, at least as it is usually stated.

The reversals share a shape. In each, routine practice had stopped at the first plausible answer: one fixation time, one biopsy, one plane through a lymph node, one reading of a shape, one pass of software over sequence data. Each correction came from doing more than the routine did: a count redone with better preparations ⁴⁸, a worm set against a vitamin deficiency in the same organ ³⁷, more days in formalin ⁴⁹, more regions sequenced ⁵⁰, more levels cut and an extra stain ¹⁸, a DNA test where the slide was ambiguous ^7,29,51, a fresh analysis of someone else's reads ²¹. That is the same move this page argues for with parasites, and it carries the same warning, because looking harder finds more, and only some of what it finds turns out to matter ¹⁸.

What is still unknown

The central number does not exist. No study that this page could find has taken a consecutive series of tumor resections and re-examined them, by expert review and by broad DNA testing, to count how often parasite material was present and unreported. The nearest attempts are a pan-pathogen microarray, PathoChip, built with probes for viruses, bacteria, fungi, parasites and helminths and shown to work on formalin-fixed tumor tissue ⁹³. Its developers then reported 'parasitic signatures' in oral and oropharyngeal, ovarian and breast cancers ^94,95,96. Those reports come from one group, have not been independently replicated, and came before the reanalyses that showed how contamination and database errors create false microbial signals ^21,52. They are leads, not counts. In Thailand, where the liver fluke is common, a targeted PCR found Opisthorchis viverrini DNA in 5 of 6 resected cholangiocarcinomas, but also in 7 of 12 liver-cell cancers, so finding the DNA did not by itself mark the cancer the fluke causes ⁸². The nearest figures for what routine work misses are the 18%-against-49% gap for schistosome eggs in cervical tissue ¹² and the 20% of presumed angiostrongyliasis specimens in which only PCR found the worm ⁸¹, and neither can be carried over to tumors. A sequencing service like the one in Basel ¹³, applied to tumor blocks under five years old ⁶⁹, with rigorous removal of human reads ⁷⁰ and independent reanalysis of the kind that overturned the retracted microbiome work ²¹, would answer it.

Several smaller facts are also missing from the published record, as far as the searches behind this page could find. No published study has run a clinical cancer gene panel against purified parasite DNA. The claim that a panel built for human genes would miss it is an inference from how such panels work, not a measurement: amplicon panels copy only the stretches of DNA flanked by primers, short DNA starters matched to human sequence, while capture panels fish out human target regions with matching probes. One clinical laboratory has described discarding the reads that match nothing human ⁷¹, but no survey across accredited tumor-sequencing laboratories was found. No study has fixed known helminth (parasitic worm) material, processed it to paraffin, and scored how recognizable it remains. No study has counted how often a uniformly negative antibody panel is followed by an organism stain. This page could not source the temperature of the paraffin bath. And this page could not open the current CAP protocol documents to check whether any carries a place to record a coexisting infection.

Two practical limits constrain any search of old tissue. Archived blocks age: parasite DNA was recovered from all 6 alveolar echinococcosis blocks stored for less than six years, but, by subtraction from the published counts, from only 3 of the 9 older ones ⁶⁹. And one widely quoted timing rule could not be checked at its source: the 6-to-72-hour fixation window is cited here from a 2026 study restating the ASCO/CAP guidelines ⁶, because the abstract of the receptor guideline itself does not state it ⁴⁶.

How this connects to parasites and cancer

The parasite-cancer links that are established are counted through infection measured in people who have the cancer. IARC's 2024 accounting attributes 5,900 squamous cell carcinomas of the bladder to Schistosoma haematobium and 3,800 intrahepatic cholangiocarcinomas, cancers of the bile ducts inside the liver, to two liver flukes, Opisthorchis viverrini and Clonorchis sinensis, out of 2.3 million cancers attributed to infection worldwide ¹⁷. Those fractions were estimated mainly from how often the infection is found in people who have the cancer, and the parasite fractions were carried over unchanged from IARC's 2018 estimates ¹⁷. Whatever method measured infection in those patients, whether eggs, antibodies or tissue, therefore sets a ceiling on what can be counted: an infection nobody tested for never enters the count.

What this page supports is narrow and solid. A negative pathology report is evidence about the questions that were asked of the tissue that reached the glass ^3,12. Routine sections demonstrably miss focal parasite material that a fuller look finds ^12,65, failures of recognition are documented ^28,29, and the tools to look properly exist and work on archived tissue ^13,69. That is a measurement gap, and it is real.

What this page does not support is any estimate of how often parasites are actually present in tumors. The published cases found for this page mostly run the other way: parasitic lesions suspected or diagnosed as cancer and then corrected by excision, a larger biopsy or autopsy ^19,20,65. Where denominators exist, they depend on place. In a Tanzanian referral pathology department, routine reports recorded Schistosoma haematobium eggs in 25.2% of 481 bladder cancers over ten years, most often with squamous cell carcinoma ¹⁴. By contrast, lung fluke disease accounted for 0.1% of 1,699 cases of pleurisy, inflammation of the lining around the lungs, diagnosed by surgical biopsy in Japan ⁹⁷. A systematic review found parasites beyond the established three inside cancers or tumors, among them the tapeworm Echinococcus, the threadworm Strongyloides and the liver fluke Fasciola, and concluded that cause and effect had not been shown ⁹⁸. The tapeworm cancer remains a single case ²⁹.

So the honest position is the season's question, stated precisely. The established numbers are real, measured by people who looked, and they are also, by construction, the numbers of what anyone has looked for ¹⁷. Whether looking harder would raise them a little, a lot or not at all is unknown. The experiment that would tell us has become technically feasible ^13,70: broad sequencing of a consecutive series of recent tumor blocks, with careful removal of human reads, contamination controls and independent reanalysis. Small, uncontrolled versions exist ^82,93, but the decisive version has not been done.

Where it connects

Sources

98 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Haghbin N, Oveisi B, Banitaba AP. Automated variable power cold microwave tissue processing: A novel universal tissue processing protocol without using formaldehyde and xylene.Acta Histochem · 2022 · 124(4):151880doi:10.1016/j.acthis.2022.151880 · PMID 35344896

   Lays out the routine chain from fixation to slide and names formaldehyde and xylene as hazards; the 97-minute processor it describes is a prototype, not routine practice.

2. 2
   Kumarapeli AR, Bellamy W, Olgaard E, et al. Short-Duration Rapid Chilling of Mastectomy Specimens Does Not Interfere With Breast Cancer Biomarker and Molecular Testing.Arch Pathol Lab Med · 2019 · 143(1):92-98doi:10.5858/arpa.2017-0377-OA · PMID 29932859

   Gives two working numbers: mastectomies sliced at 4-mm intervals, and FFPE DNA that amplified successfully in 300- to 400-base-pair stretches, offered as evidence the DNA was well preserved rather than as a measure of fragmentation; it is one laboratory's experience.

3. 3
   Fitzgibbons PL, Bradley LA, Fatheree LA, et al. Principles of analytic validation of immunohistochemical assays: Guideline from the College of American Pathologists Pathology and Laboratory Quality Center.Arch Pathol Lab Med · 2014 · 138(11):1432-43doi:10.5858/arpa.2013-0610-CP · PMID 24646069

   Requires every antibody test to be validated for its analyte before clinical use; it says nothing about organisms, and one panelist lists a commercial laboratory affiliation.

4. 4
   Srigley JR, McGowan T, Maclean A, et al. Standardized synoptic cancer pathology reporting: a population-based approach.J Surg Oncol · 2009 · 99(8):517-24doi:10.1002/jso.21282 · PMID 19466743

   Describes CAP-defined checklist content and its province-wide rollout in Canada with rising completeness; it does not list the checklist items.

5. 5
   Thavarajah R, Mudimbaimannar VK, Elizabeth J, et al. Chemical and physical basics of routine formaldehyde fixation.J Oral Maxillofac Pathol · 2012 · 16(3):400-5doi:10.4103/0973-029X.102496 · PMID 23248474

   A narrative review compiling the chemistry of formalin (3.7% formaldehyde, the methylene glycol equilibrium, 24 to 48 hours for initial cross-links); its figures are gathered from earlier work rather than newly measured.

6. 6
   Ndengue CP, Atangana PJA, Ateba GR, et al. Pre-analytical variables affecting breast cancer biomarker expression: A controlled single-specimen study of fixation duration, cold ischemia time, and fixative preparation in a low-resource setting.PLoS One · 2026 · 21(6):e0343185doi:10.1371/journal.pone.0343185 · PMID 42247401

   Restates the ASCO/CAP timing rules (fixative within 1 hour, 6 to 72 hours of fixation) and measures marker loss after 2 hours of delay; it is one tumor cut into 50 cores, so it is a technical benchmark, not population evidence.

7. 7
   Salehi E, Hedayati MT, Zoll J, et al. Discrimination of Aspergillosis, Mucormycosis, Fusariosis, and Scedosporiosis in Formalin-Fixed Paraffin-Embedded Tissue Specimens by Use of Multiple Real-Time Quantitative PCR Assays.J Clin Microbiol · 2016 · 54(11):2798-2803doi:10.1128/JCM.01185-16 · PMID 27605714

   DNA testing of 102 archived specimens reassigned several fungal diagnoses made by shape, and used two 4- to 5-micrometer sections each; molds, not parasites.

8. 8
   Tran T, Sundaram CP, Bahler CD, et al. Correcting the Shrinkage Effects of Formalin Fixation and Tissue Processing for Renal Tumors: toward Standardization of Pathological Reporting of Tumor Size.J Cancer · 2015 · 6(8):759-66doi:10.7150/jca.12094 · PMID 26185538

   Measures 11.4% shrinkage from fresh tissue to slide in 34 kidney tumors; it does not measure parasites, whose identification can depend on size.

9. 9
   Grogan T, Reinhardt K, Jaramillo M, et al. An update on "special stain" histochemistry with emphasis on automation.Adv Anat Pathol · 2000 · 7(2):110-22doi:10.1097/00125480-200007020-00006 · PMID 10721418

   Describes the 20 to 25 special stains a surgical pathology laboratory keeps, including stains for microorganisms; it does not say how often they are ordered on tumors.

10. 10
    Elmore JG, Longton GM, Carney PA, et al. Diagnostic concordance among pathologists interpreting breast biopsy specimens.JAMA · 2015 · 313(11):1122-32doi:10.1001/jama.2015.1405 · PMID 25781441

    115 pathologists agreed with an expert reference on 96% of invasive cancers but 48% of atypia, reading one slide per case; it measures interpretation, not sampling.

11. 11
    Fridland S, Mehta V, Jager L, et al. Not all errors are created equal: assessment of amended diagnoses at a major academic center.Am J Clin Pathol · 2026 · 165(3)doi:10.1093/ajcp/aqag005 · PMID 41807328

    0.33% of final surgical pathology reports were amended over five years at one center, within CAP benchmarks; amendments catch errors someone noticed, not unasked questions.

12. 12
    Poggensee G, Sahebali S, Van Marck E, et al. Diagnosis of genital cervical schistosomiasis: comparison of cytological, histopathological and parasitological examination.Am J Trop Med Hyg · 2001 · 65(3):233-6doi:10.4269/ajtmh.2001.65.233 · PMID 11561710

    The clearest measurement of sampling loss for a parasite: in the same 228 women, eggs were found in 49% by squash, 18% by sections and 3% by smears; it concerns cervical tissue, not tumors.

13. 13
    Hamelin B, Hosch S, Neidhöfer C, et al. Unbiased DNA Pathogen Detection in Tissues: Real-World Experience With Metagenomic Sequencing in Pathology.Lab Invest · 2026 · 106(1):104254doi:10.1016/j.labinv.2025.104254 · PMID 41167475

    A routine service sequencing 623 archived infectious-pathology samples in Basel (36.8% positive, 9 parasites, 9.6% uninterpretable); it was not applied to unselected tumors.

14. 14
    Yohana C, Bakuza JS, Kinung'hi SM, Nyundo BA, Rambau PF. The trend of schistosomiasis related bladder cancer in the lake zone, Tanzania: a retrospective review over 10 years period.Infect Agent Cancer · 2023 · 18(1):10doi:10.1186/s13027-023-00491-1 · PMID 36800971

    Schistosome eggs recorded in 25.2% of 481 histologically confirmed bladder cancers at one referral center over ten years, most often with squamous cell carcinoma; the reports were read retrospectively, and how thoroughly eggs were sought is not stated.

15. 15
    Haley L, Parimi V, Jiang L, et al. Diagnostic Utility of Gene Fusion Panel to Detect Gene Fusions in Fresh and Formalin-Fixed, Paraffin-Embedded Cancer Specimens.J Mol Diagn · 2021 · 23(10):1343-1358doi:10.1016/j.jmoldx.2021.07.015 · PMID 34358677

    A 329-specimen validation in which 141 solid tumors were tested as fixed tissue and karyotyping served as one comparator; it is about fusion genes, not organisms.

16. 16
    Ross JS, Wang K, Gay L, et al. Comprehensive Genomic Profiling of Carcinoma of Unknown Primary Site: New Routes to Targeted Therapies.JAMA Oncol · 2015 · 1(1):40-49doi:10.1001/jamaoncol.2014.216 · PMID 26182302

    A commercial panel found at least one human alteration in 192 of 200 tumors; all 17 authors listed Foundation Medicine, which sells the assay.

17. 17
    Rumgay H, Georges D, Huang Y, et al. Global burden of cancer attributable to infections in 2024: a worldwide incidence analysis.Lancet Oncol · 2026 · 27(10):1237-1248doi:10.1016/S1470-2045(26)00307-4 · PMID 42805198

    IARC's 2024 count: 5,900 bladder and 3,800 bile-duct cancers attributed to parasites, with attributable fractions estimated mainly from infection prevalence in people with cancer; the full text available here strips organism names, so the parasite assigned to each figure is read from the paper's agent list.

18. 18
    Weaver DL, Ashikaga T, Krag DN, et al. Effect of occult metastases on survival in node-negative breast cancer.N Engl J Med · 2011 · 364(5):412-21doi:10.1056/NEJMoa1008108 · PMID 21247310

    Two deeper levels through 'negative' sentinel nodes, examined with routine and cytokeratin antibody stains after a first look designed to catch deposits larger than 2 mm, found hidden cancer in 15.9% of 3,887 patients, with a 1.2-point survival difference; it measures what one plane and one stain miss, not parasites.

19. 19
    Silver SA, Erozan YS, Hruban RH. Cerebral cysticercosis mimicking malignant glioma: a case report.Acta Cytol · 1996 · 40(2):351-7doi:10.1159/000333767 · PMID 8629426

    A needle biopsy read as malignant glioma that proved to be cysticercosis on excision; a single case.

20. 20
    Ermilov VV, Smirnov AV, Snigur GL, et al. [Pulmonary larval paragonimiasis mimicking lung cancer].Arkh Patol · 2018 · 80(2):60-63doi:10.17116/patol201880260-63 · PMID 29697674

    A frozen section read as glandular cancer led to resection, and lung fluke disease was diagnosed after the patient's death; a single case, in Russian.

21. 21
    Gihawi A, Ge Y, Lu J, et al. Major data analysis errors invalidate cancer microbiome findings.mBio · 2023 · 14(5):e0160723doi:10.1128/mbio.01607-23 · PMID 37811944

    The reanalysis showing most reported cancer microbes were absent from the samples and more than a dozen follow-up studies likely invalid; it addresses bacteria, not parasites.

22. 22
    Poore GD, Kopylova E, Zhu Q, et al. Retraction Note: Microbiome analyses of blood and tissues suggest cancer diagnostic approach.Nature · 2024 · 631(8021):694doi:10.1038/s41586-024-07656-x · PMID 38926587

    The formal retraction notice for the 2020 paper; PubMed holds no abstract for it.

23. 23
    Ould Ismail AA, Kale S, McGonagle K, et al. Adherence to Synoptic Cancer Pathology Reporting Among Pathologists in the National Department of Veterans Affairs Health Care System.Arch Pathol Lab Med · 2025 · 149(8):727-733doi:10.5858/arpa.2024-0229-OA · PMID 39514644

    Across the US Department of Veterans Affairs, 48% of 1,618 cancer reports from 2019 to 2021 were synoptic: 77% of resections but 19% of biopsies; four cancer types in one health system.

24. 24
    Yıldız T, İlçe Z, Turan G, et al. [Parasites in the Etiology of Pediatric Appendicitis].Turkiye Parazitol Derg · 2015 · 39(3):190-3doi:10.5152/tpd.2015.3737 · PMID 26470923

    Routine histology found pinworms or Taenia in 14 of 846 children's appendices; the abstract's printed 0.39% does not match its own counts.

25. 25
    Al Laham O, Abdul Khalek G, Alboushi H, et al. An incidentally diagnosed primary pancreatic body hydatid cyst: A case report and literature review.Int J Surg Case Rep · 2024 · 116:109392doi:10.1016/j.ijscr.2024.109392 · PMID 38367420

    A pancreatic body mass seen on imaging and resected with the tail and spleen, which histology showed to be a hydatid cyst; a single case, and evidence that the pipeline does name parasites it sees.

26. 26
    Dursun N, Memis B, Pehlivanoglu B, et al. Adenomyomas of the Gallbladder: An Analysis of Frequency, Clinicopathologic Associations, and Relationship to Carcinoma of a Malformative Lesion.Arch Pathol Lab Med · 2024 · 148(2):206-214doi:10.5858/arpa.2022-0379-OA · PMID 37134225

    A lesion appeared in 9.3% of 203 gallbladders embedded whole but 3.3% of a separate series of 2,347 routinely sampled ones, a comparison across cohorts rather than within the same specimens; a benign lesion, not a parasite.

27. 27
    Selves J, Long-Mira E, Mathieu MC, et al. Immunohistochemistry for Diagnosis of Metastatic Carcinomas of Unknown Primary Site.Cancers (Basel) · 2018 · 10(4):108doi:10.3390/cancers10040108 · PMID 29621151

    Describes the CK7/CK20-led antibody algorithm for cancers of unknown origin; a review of human markers, with no accuracy figures of its own.

28. 28
    Trecourt A, Radobonirina M. [Infectious cysts and pseudocysts: When parasites want to mimic tumors!].Ann Pathol · 2025 · 45(6):513-522doi:10.1016/j.annpat.2025.04.003 · PMID 40328550

    A 2025 French review stating that parasitic cysts are rarely encountered by pathologists and full of pitfalls, and recommending expert and molecular review; it gives no rates.

29. 29
    Muehlenbachs A, Bhatnagar J, Agudelo CA, et al. Malignant Transformation of Hymenolepis nana in a Human Host.N Engl J Med · 2015 · 373(19):1845-52doi:10.1056/NEJMoa1505892 · PMID 26535513

    Tumor-like cells in a man with HIV, unrecognizable as tapeworm on the slide, identified by a PCR targeting all eukaryotes; a single case, without published replication.

30. 30
    Do H, Dobrovic A. Sequence artifacts in DNA from formalin-fixed tissues: causes and strategies for minimization.Clin Chem · 2015 · 61(1):64-71doi:10.1373/clinchem.2014.223040 · PMID 25421801

    A mini-review stating that DNA extracted from formalin-fixed tissue is fragmented and carries lesions, such as deaminated cytosines and abasic sites, that are sources of sequence artifacts; written for human mutation calling, not organisms.

31. 31
    Miquelestorena-Standley E, Jourdan ML, Collin C, et al. Effect of decalcification protocols on immunohistochemistry and molecular analyses of bone samples.Mod Pathol · 2020 · 33(8):1505-1517doi:10.1038/s41379-020-0503-6 · PMID 32094425

    Hydrochloric acid and long formic-acid decalcification caused false-negative antibody and DNA results in 35 bone samples, while EDTA did not; it did not study organisms.

32. 32
    Titford M. The long history of hematoxylin.Biotech Histochem · 2005 · 80(2):73-8doi:10.1080/10520290500138372 · PMID 16195172

    A historical review: logwood found in 1502, used by microscopists from the mid-1800s; it is a history, not a study.

33. 33
    Puchtler H, Meloan SN. On the chemistry of formaldehyde fixation and its effects on immunohistochemical reactions.Histochemistry · 1985 · 82(3):201-4doi:10.1007/BF00501395 · PMID 3997553

    States that formalin fixation was discovered accidentally by F. Blum in 1893, and that formaldehyde's methylene cross-links resist washing; chemistry, and it predates modern retrieval methods.

34. 34
    Musiał A, Gryglewski RW, Kielczewski S, et al. Formalin use in anatomical and histological science in the 19th and 20th centuries.Folia Med Cracov · 2016 · 56(3):31-40PMID 28275269

    A history of formalin naming Ferdinand Blum among those who brought formaldehyde into tissue fixation; the abstract gives no exact year for his work.

35. 35
    Berry A, Iriart X, Fillaux J, Magnaval JF. [Urinary schistosomiasis and cancer].Bull Soc Pathol Exot · 2017 · 110(1):68-75doi:10.1007/s13149-017-0547-4 · PMID 28185084

    A review, in French with an English abstract, recording that A. R. Ferguson, professor of pathology in Cairo, reported from 40 autopsies in 1911 a likely association between bladder cancer and schistosome granulomas; history, not a new measurement.

36. 36
    Tsoucalas G, Laios K, Karamanou M, Gennimata V, Androutsos G. The fascinating germ theories on cancer pathogenesis.J BUON · 2014 · 19(1):319-23PMID 24659685

    A history of the germ theories of cancer, recording the 1926 Nobel Prize to Johannes Fibiger for the nematode Spiroptera as a cause of cancer; a historical review.

37. 37
    Hitchcock CR, Bell ET. Studies on the nematode parasite, Gongylonema neoplasticum (spiroptera neoplasticum), and avitaminosis A in the forestomach of rats: comparison with Fibiger's results.J Natl Cancer Inst · 1952 · 12(6):1345-87PMID 14939031

    The re-examination that set Fibiger's worm against vitamin A deficiency in the rat forestomach; PubMed holds no abstract, so this page reports only what the title states.

38. 38
    Stolt CM, Klein G, Jansson ATR. An analysis of a wrong Nobel Prize-Johannes Fibiger, 1926: a study in the Nobel archives.Adv Cancer Res · 2004 · 92:1-12doi:10.1016/S0065-230X(04)92001-5 · PMID 15530554

    A study of the Nobel archives whose title calls Fibiger's 1926 prize a wrong one; PubMed holds no abstract for it.

39. 39
    Arnason U. 50 years after--examination of some circumstances around the establishment of the correct chromosome number of man.Hereditas · 2006 · 143(2006):202-11doi:10.1111/j.2006.0018-0661.01963.x · PMID 17362356

    Checks the 1956 discovery of 46 human chromosomes against the laboratory logbook and finds the participants' accounts inconsistent; it is history, not cytogenetic method.

40. 40
    Taylor CR, Mason DY. The immunohistological detection of intracellular immunoglobulin in formalin-paraffin sections from multiple myeloma and related conditions using the immunoperoxidase technique.Clin Exp Immunol · 1974 · 18(3):417-29PMID 4219910

    An early demonstration that an enzyme-labeled antibody can find a chosen protein in formalin-fixed paraffin sections, in 33 plasma cell tumors.

41. 41
    Shi SR, Key ME, Kalra KL. Antigen retrieval in formalin-fixed, paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections.J Histochem Cytochem · 1991 · 39(6):741-8doi:10.1177/39.6.1709656 · PMID 1709656

    Introduced heat-based antigen retrieval (39 of 52 antibodies improved); the first author's listed affiliation is BioGenex Laboratories, a reagent company, a commercial interest the paper's readers should weigh.

42. 42
    Zehir A, Benayed R, Shah RH, et al. Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients.Nat Med · 2017 · 23(6):703-713doi:10.1038/nm.4333 · PMID 28481359

    More than 10,000 patients' tumors sequenced prospectively with one center's clinical gene panel; human alterations only, and it says nothing about non-human reads.

43. 43
    Roy S, Coldren C, Karunamurthy A, et al. Standards and Guidelines for Validating Next-Generation Sequencing Bioinformatics Pipelines: A Joint Recommendation of the Association for Molecular Pathology and the College of American Pathologists.J Mol Diagn · 2018 · 20(1):4-27doi:10.1016/j.jmoldx.2017.11.003 · PMID 29154853

    Seventeen consensus recommendations for validating clinical sequencing software, written because laboratories varied widely in how they did it; several authors list commercial testing companies (Invitae, Color Genomics, Sema4, PathGroup) or ARUP Laboratories, a University of Utah reference laboratory, and its abstract does not address non-human reads.

44. 44
    Novis DA, Gephardt GN, Zarbo RJ. Interinstitutional comparison of frozen section consultation in small hospitals: a College of American Pathologists Q-Probes study of 18,532 frozen section consultation diagnoses in 233 small hospitals.Arch Pathol Lab Med · 1996 · 120(12):1087-93PMID 15456172

    A 1994 audit in which 1.8% of non-deferred frozen-section diagnoses (316 of 17,673) disagreed with final slides, mostly underdiagnosed tumors through sampling error; it concerns frozen sections in small hospitals only.

45. 45
    Kronz JD, Westra WH, Epstein JI. Mandatory second opinion surgical pathology at a large referral hospital.Cancer · 1999 · 86(11):2426-35doi:10.1002/(SICI)1097-0142(19991201)86:11<2426::AID-CNCR34>3.0.CO;2-3 · PMID 10590387

    Second review of 6,171 outside cases changed 86 diagnoses (1.4%) in ways that altered treatment or prognosis; it measures disagreement about questions already asked.

46. 46
    Hammond ME, Hayes DF, Dowsett M, et al. American Society of Clinical Oncology/College Of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer.J Clin Oncol · 2010 · 28(16):2784-95doi:10.1200/JCO.2009.25.6529 · PMID 20404251

    The guideline panel's statement that up to 20% of receptor tests worldwide may be inaccurate, mostly from handling, thresholds and interpretation; panel disclosures were not read for this page, and the abstract does not state the fixation window.

47. 47
    Goldsmith JD, Troxell ML, Roy-Chowdhuri S, et al. Principles of Analytic Validation of Immunohistochemical Assays: Guideline Update.Arch Pathol Lab Med · 2024 · 148(6):e111-e153doi:10.5858/arpa.2023-0483-CP · PMID 38391878

    The 2024 update to CAP's antibody-validation rules, extended to new scoring systems and cytology; some panelists list consulting firms, and organisms are not its subject.

48. 48
    Harper PS. The discovery of the human chromosome number in Lund, 1955-1956.Hum Genet · 2006 · 119(1-2):226-32doi:10.1007/s00439-005-0121-x · PMID 16463025

    A history recording that the correct number, 46, ended more than 30 years in which the human number was thought to be 48, and that previously accepted conclusions shaped how earlier results were read.

49. 49
    De Marzo AM, Fedor HH, Gage WR, et al. Inadequate formalin fixation decreases reliability of p27 immunohistochemical staining: probing optimal fixation time using high-density tissue microarrays.Hum Pathol · 2002 · 33(7):756-60doi:10.1053/hupa.2002.126187 · PMID 12196928

    The reversal on fixation time: underfixed prostate tissue stained worse than tissue fixed for up to 8 days; it tested one protein, p27, in one organ.

50. 50
    Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.N Engl J Med · 2012 · 366(10):883-892doi:10.1056/NEJMoa1113205 · PMID 22397650

    63% to 69% of mutations were not detectable in every region of the same kidney cancers; a small number of tumors, and human mutations only.

51. 51
    Che J, Du J, Piao Y, et al. Fungal species identification in FFPE tissues: A comparative evaluation of droplet digital PCR, ITS sequencing, and metagenomic next-generation sequencing.Med Mycol · 2026 · 64(6)doi:10.1093/mmy/myag049 · PMID 42133463

    On 111 archived fungal specimens, metagenomic sequencing named the species in 81.1% against 64.9% and 50.5% for two targeted DNA tests, and caught two rare fungi misdiagnosed as Aspergillus; fungi, not parasites.

52. 52
    Ge Y, Lu J, Puiu D, Revsine M, Salzberg SL. Comprehensive analysis of microbial content in whole-genome sequencing samples from The Cancer Genome Atlas project.Sci Transl Med · 2025 · 17(814):eads6335doi:10.1126/scitranslmed.ads6335 · PMID 40901923

    Reanalysis of all 5,734 whole-genome samples in TCGA, across 25 cancer types, finding microbial presence far smaller than reported and many named species possibly absent; bacteria, viruses, archaea and fungi, not parasites.

53. 53
    Warren JR, Marshall B. Unidentified curved bacilli on gastric epithelium in active chronic gastritis.Lancet · 1983 · 1(8336):1273-5PMID 6134060

    The letter reporting unidentified curved bacilli on the stomach lining; PubMed holds no abstract, and the organism was not named in it.

54. 54
    Relman DA, Schmidt TM, MacDermott RP, Falkow S. Identification of the uncultured bacillus of Whipple's disease.N Engl J Med · 1992 · 327(5):293-301doi:10.1056/NEJM199207303270501 · PMID 1377787

    A bacillus associated with Whipple's disease for 85 years, never cultured, was named from DNA copied out of five patients' tissue with broad-range primers; bacteria, by the method this page argues for with parasites.

55. 55
    Feng H, Shuda M, Chang Y, et al. Clonal integration of a polyomavirus in human Merkel cell carcinoma.Science · 2008 · 319(5866):1096-100doi:10.1126/science.1152586 · PMID 18202256

    Found a new virus in 8 of 10 Merkel cell carcinomas by subtracting human sequence from tumor transcripts; it shows what a research pipeline can see, not what clinical pipelines do.

56. 56
    Poore GD, Kopylova E, Zhu Q, et al. Microbiome analyses of blood and tissues suggest cancer diagnostic approach.Nature · 2020 · 579(7800):567-574doi:10.1038/s41586-020-2095-1 · PMID 32214244

    RETRACTED in 2024. Claimed cancer-type microbial signatures in 18,116 TCGA tissue and blood samples from patients with 33 cancer types; one coauthor listed a commercial genomics affiliation (Clarity Genomics). Cited only as the claim that was withdrawn.

57. 57
    Focke CM, Decker T, van Diest PJ. The reliability of histological grade in breast cancer core needle biopsies depends on biopsy size: a comparative study with subsequent surgical excisions.Histopathology · 2016 · 69(6):1047-1054doi:10.1111/his.13036 · PMID 27417415

    More millimeters of core gave grades closer to the excised tumor (83% against 68%) in 300 breast cancers; observational, and about grade, not organisms.

58. 58
    Lebwohl B, Kapel RC, Neugut AI, et al. Adherence to biopsy guidelines increases celiac disease diagnosis.Gastrointest Endosc · 2011 · 74(1):103-9doi:10.1016/j.gie.2011.03.1236 · PMID 21601201

    Submitting four or more duodenal pieces was linked to more than double the diagnosis rate (1.8% against 0.7%) in 132,352 patients; an association in retrospective data, for celiac disease.

59. 59
    Suganuma N, Matsubara Y, Takahashi A, et al. Impact of Warm Ischemia Time on HER2 Expression in Breast Cancer Surgical Specimens.Anticancer Res · 2024 · 44(12):5225-5230doi:10.21873/anticanres.17350 · PMID 39626910

    Real handling times in 277 breast operations at one Japanese center (median cold ischemia 37 minutes, fixation 43 hours); it does not measure anything beyond receptor staining.

60. 60
    Vassilakopoulou M, Parisi F, Siddiqui S, et al. Preanalytical variables and phosphoepitope expression in FFPE tissue: quantitative epitope assessment after variable cold ischemic time.Lab Invest · 2015 · 95(3):334-41doi:10.1038/labinvest.2014.139 · PMID 25418580

    Shows loss of antigenicity for several phosphoproteins within 1 to 2 hours of delay in 93 breast cancers, and cites earlier reports of degradation within 30 minutes; it concerns labile molecules, not whole organisms.

61. 61
    Al-Nattah S, Martin A, Normington L, et al. Reduced cervical sampling of hysterectomy specimens with negative margins on conization: An opportunity to improve resource utilization.Am J Clin Pathol · 2025 · 163(3):435-438doi:10.1093/ajcp/aqae139 · PMID 39441180

    Submitting the whole cervix by protocol averaged 16 blocks per case (range 4 to 41); the authors propose one representative section per quadrant, saving about two to six hours of laboratory work; a small study arguing for less tissue, not more.

62. 62
    Bindhu P, Krishnapillai R, Thomas P, Jayanthi P. Facts in artifacts.J Oral Maxillofac Pathol · 2013 · 17(3):397-401doi:10.4103/0973-029X.125206 · PMID 24574659

    A review of processing artifacts: acid formalin hematin, a brown pigment formed from unbuffered formalin and blood, 'may simulate microorganisms'; fungi, fibers and fragments of another patient's tissue can contaminate a slide; and thin, narrow specimens curl during processing, so they are hard to orient when embedded and end up cut at a slant.

63. 63
    Honeywell R, Donner M, Criswell S. Thorough formalin fixation confers a protective effect for histochemical and immunohistochemical testing on tissues exposed to prolonged time in ethanol, xylene, heat, or paraffin.Histochem Cell Biol · 2026 · 164(1)doi:10.1007/s00418-026-02545-y · PMID 42825935

    In tissue fixed for three to five weeks, an extra 24 hours in absolute ethanol, xylene, molten paraffin or on a heated embedding station caused no visible change in structure or routine staining, and most antibody tests were preserved; it does not give the temperature of the paraffin bath.

64. 64
    Kiernan JA. Does progressive nuclear staining with hemalum (alum hematoxylin) involve DNA, and what is the nature of the dye-chromatin complex?Biotech Histochem · 2018 · 93(2):133-148doi:10.1080/10520295.2017.1399466 · PMID 29320873

    Experiments showing that the hematoxylin dye complex binds the phosphate of DNA electrostatically; it explains the blue of H&E, not what a reader will notice.

65. 65
    Kim KE, Jung SS, Park HS, et al. The first case report of Paragonimus westermani infection diagnosed by transbronchial lung cryobiopsy.Int J Infect Dis · 2023 · 128:184-186doi:10.1016/j.ijid.2022.12.041 · PMID 36608785

    Fluke eggs found by cryobiopsy in a patient who also had a forceps biopsy; the authors credit the larger specimen. A single case.

66. 66
    Sluijter CE, van Lonkhuijzen LR, van Slooten HJ, et al. The effects of implementing synoptic pathology reporting in cancer diagnosis: a systematic review.Virchows Arch · 2016 · 468(6):639-49doi:10.1007/s00428-016-1935-8 · PMID 27097810

    Thirty-two of 33 studies found checklist reports more complete, especially for margins and invasion; the items reviewed are human tumor features, and clinical benefit was not shown.

67. 67
    Rassy E, Pavlidis N. The diagnostic challenges of patients with carcinoma of unknown primary.Expert Rev Anticancer Ther · 2020 · 20(9):775-783doi:10.1080/14737140.2020.1807948 · PMID 32779501

    Antibody panels name a single tissue of origin in 10.8% to 51% of cancers of unknown primary; a review, so its ranges come from other studies.

68. 68
    Özmen A, Mıstık R, Alver O, et al. [The Pneumocystis jirovecii colonization in bronchoalveolar lavage (BAL) and bronchial washing and the comparison of methods which are used in diagnosis].Tuberk Toraks · 2013 · 61(4):303-11doi:10.5578/tt.2954 · PMID 24506746

    Three classic stains found the organism in none of 100 lung-wash samples while antibody and DNA tests found it in 13 and 16; a fungus in fluid, not a parasite in tissue.

69. 69
    Grimm J, Krickl J, Beck A, et al. Establishing and evaluation of a polymerase chain reaction for the detection of Echinococcus multilocularis in human tissue.PLoS Negl Trop Dis · 2021 · 15(2):e0009155doi:10.1371/journal.pntd.0009155 · PMID 33630840

    A validated parasite PCR and antibody that work on archived blocks: all 6 viable-lesion blocks stored under six years were positive, against 3 of the 9 older ones by subtraction from the counts; one parasite, one disease.

70. 70
    Frolova M, Maguire B, Duessmann H, et al. HumanFilt: a multi-reference host depletion pipeline improves Fusobacterium detection accuracy in tumor WGS data sets.mSystems · 2026 · 11(8):e0055026doi:10.1128/msystems.00550-26 · PMID 42446235

    Shows that clean recovery of non-human reads from tumor genomes needed three human references and removed over 99.9% of human reads; ten tumors, bacteria only.

71. 71
    Cimino PJ, Zhao G, Wang D, et al. Detection of viral pathogens in high grade gliomas from unmapped next-generation sequencing data.Exp Mol Pathol · 2014 · 96(3):310-5doi:10.1016/j.yexmp.2014.03.010 · PMID 24704430

    One academic clinical laboratory mined the previously discarded non-human reads of its 151-gene FFPE cancer panel, averaging about 38,000 reads per case (1.9%); the Epstein-Barr virus hits were not confirmed by in situ hybridization in any of the four cases with material for it; viruses only, 21 tumors.

72. 72
    Zapatka M, Borozan I, Brewer DS, et al. The landscape of viral associations in human cancers.Nat Genet · 2020 · 52(3):320-330doi:10.1038/s41588-019-0558-9 · PMID 32025001

    A research consortium found viral sequence in 382 of 2,658 cancer genomes using three pipelines; viruses, in research data, not clinical reports.

73. 73
    Nejman D, Livyatan I, Fuks G, et al. The human tumor microbiome is composed of tumor type-specific intracellular bacteria.Science · 2020 · 368(6494):973-980doi:10.1126/science.aay9189 · PMID 32467386

    Bacteria found inside both cancer and immune cells in 1,526 tumors across seven cancer types; bacteria only, and the authors note that the low biomass makes this work hard.

74. 74
    Battaglia TW, Mimpen IL, Traets JJH, et al. A pan-cancer analysis of the microbiome in metastatic cancer.Cell · 2024 · 187(9):2324-2335.e19doi:10.1016/j.cell.2024.03.021 · PMID 38599211

    Microbial communities in 4,160 metastatic tumor biopsies, with Fusobacterium tied to resistance to immunotherapy in lung cancer; one author lists the Hartwig Medical Foundation, which holds the sequencing data, and the abstract does not report parasites.

75. 75
    Neveling K, Mantere T, Vermeulen S, et al. Next-generation cytogenetics: Comprehensive assessment of 52 hematological malignancy genomes by optical genome mapping.Am J Hum Genet · 2021 · 108(8):1423-1435doi:10.1016/j.ajhg.2021.06.001 · PMID 34237281

    States that full structural analysis of blood cancers still combines karyotype, FISH and arrays, and tests a newer method in 52 patients; it does not concern solid tumors.

76. 76
    Campbell WS, Rous BA, Dubois S, et al. Advancements in Interoperability: Achieving Anatomic Pathology Reports That Adhere to International Standards and Are Both Human-Readable and Readily Computable.JCO Clin Cancer Inform · 2025 · 9:e2400180doi:10.1200/CCI-24-00180 · PMID 39908464

    Describes the required data elements of cancer protocols as 'the minimum set of evidence-based, clinically actionable parameters'; it is about coding and interoperability, and does not list the optional fields.

77. 77
    Torous VF, Simpson RW, Balani JP, et al. College of American Pathologists Cancer Protocols: From Optimizing Cancer Patient Care to Facilitating Interoperable Reporting and Downstream Data Use.JCO Clin Cancer Inform · 2021 · 5:47-55doi:10.1200/CCI.20.00104 · PMID 33439728

    A review stating that the CAP cancer protocols have guided American cancer reporting for more than 35 years; it describes the system rather than auditing its use.

78. 78
    Paik J, Ellis CL, Henriksen KJ, Chang A. An International Survey of Genitourinary and Renal Pathologists Regarding Evaluation of the Non-Neoplastic Parenchyma in Kidney Cancer Specimens.Int J Surg Pathol · 2023 · 32(2):273-278doi:10.1177/10668969231177408 · PMID 37226477

    Records that CAP's 2010 update required evaluation of the non-cancerous kidney in tumor nephrectomies, evidence that a required item can reach beyond the tumor; the rest is a survey of practice.

79. 79
    Wolff AC, Somerfield MR, Dowsett M, et al. Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: ASCO-College of American Pathologists Guideline Update.J Clin Oncol · 2023 · 41(22):3867-3872doi:10.1200/JCO.22.02864 · PMID 37284804

    Affirms the 2018 HER2 recommendations but makes the line between 'IHC 0' and '1+' newly relevant, because a drug was approved for tumors in that range; an example of a report question changing with treatment.

80. 80
    Allison KH, Hammond MEH, Dowsett M, et al. Estrogen and Progesterone Receptor Testing in Breast Cancer: ASCO/CAP Guideline Update.J Clin Oncol · 2020 · 38(12):1346-1366doi:10.1200/JCO.19.02309 · PMID 31928404

    The 2020 update that added an 'ER Low Positive' reporting category for tumors with 1% to 10% of nuclei staining; a guideline about one receptor pair, not about organisms.

81. 81
    Rodriguez R, da Silva AC, Müller CA, et al. PCR for the diagnosis of abdominal angiostrongyliasis in formalin-fixed paraffin-embedded human tissue.PLoS One · 2014 · 9(4):e93658doi:10.1371/journal.pone.0093658 · PMID 24705328

    PCR found worm DNA in 4 of 20 bowel specimens where sections showed only the tissue reaction, with 55% sensitivity and 100% specificity; small numbers, one worm.

82. 82
    Suksumek N, Leelawat K, Leelawat S, Russell B, Lek-Uthai U. TaqMan real-time PCR assay for specific detection of Opisthorchis viverrini DNA in Thai patients with hepatocellular carcinoma and cholangiocarcinoma.Exp Parasitol · 2008 · 119(2):217-24doi:10.1016/j.exppara.2008.01.018 · PMID 18329641

    Fluke DNA in 5 of 6 cholangiocarcinomas and 7 of 12 liver-cell cancers from resections, with no association found between cancer type and the parasite; 18 tumors in all.

83. 83
    Roh CK, Jung MJ. Laparoscopic excision for ectopic peritoneal paragonimiasis mimicking a gastric duplication cyst: A case report.Ann Med Surg (Lond) · 2021 · 69:102754doi:10.1016/j.amsu.2021.102754 · PMID 34484726

    A lung fluke in the abdomen, taken for a cyst on imaging and diagnosed on excision; the authors note in general that eggs in sputum and biopsy specimens may be hard to detect 'due to an insufficient amount'. A single case.

84. 84
    Mainland N, Madiraju S, Nwogu S, et al. Phimosis of foreskin uncovering schistosomiasis of the urinary bladder: a case report.Transl Androl Urol · 2025 · 14(6):1827-1831doi:10.21037/tau-2025-182 · PMID 40687665

    A frozen section of a bladder mass in a 10-year-old Sudanese immigrant was read as rhabdoid malignancy; the permanent sections showed numerous schistosome eggs and no cancer. A single case.

85. 85
    Querzoli G, Ortalli M, Varani S, et al. Tonsillar Leishmaniasis: A Rare Clinical Entity Mimicking Malignancy in the Oropharynx - A Case Series from Northeastern Italy.Head Neck Pathol · 2025 · 19(1):39doi:10.1007/s12105-025-01773-3 · PMID 40138028

    Five tonsil swellings that mimicked cancer between 2014 and 2024 proved to be leishmaniasis, seen on histology in four and confirmed by PCR in all five; a case series from one region.

86. 86
    Hu D, Jin W, Ding H, et al. Spirometra mansoni sparganosis identified by metagenomic next-generation sequencing: a case report.Int J Infect Dis · 2023 · 128:128-131doi:10.1016/j.ijid.2022.12.038 · PMID 36592686

    Sequencing of an archived block corrected a cysticercosis diagnosis to sparganosis; a single case, and two authors work for KingMed Diagnostics, a commercial laboratory.

87. 87
    Khor LY, Neo CC, Prathaban K, et al. Deep learning model for automated detection of Helicobacter pylori and intestinal metaplasia on gastric biopsy digital whole slide images.Am J Clin Pathol · 2025 · 164(5):830-839doi:10.1093/ajcp/aqaf110 · PMID 40996022

    A tool trained to find Helicobacter pylori on scanned gastric biopsies scored 0.617 on the F1 scale for the organisms themselves, against 0.861 for intestinal metaplasia; one organism, the one it was trained for, on 180 slides.

88. 88
    Parra-Medina R, Zambrano-Betancourt C, Peña-Rojas S, et al. Detection of Helicobacter pylori Infection in Histopathological Gastric Biopsies Using Deep Learning Models.J Imaging · 2025 · 11(7):226doi:10.3390/jimaging11070226 · PMID 40710613

    Deep-learning detection of Helicobacter pylori on 100 whole-slide images, noting that some scanned images lack the resolution to show the bacteria; the PubMed title drops the italicized organism name.

89. 89
    Roure S, Pérez-Quílez O, Vallès X, et al. Schistosomiasis Among Female Migrants in Non-endemic Countries: Neglected Among the Neglected? A Pilot Study.Front Public Health · 2022 · 10:778110doi:10.3389/fpubh.2022.778110 · PMID 35372213

    Seropositive migrant women in Barcelona had a median of 41 clinic visits before schistosomiasis was diagnosed; a pilot study of 51 women, about clinical, not pathological, delay.

90. 90
    Tawfik O, Davis M, Dillon S, et al. Whole Slide Imaging of Pap Cell Block Preparations versus Liquid-Based Thin-Layer Cervical Cytology: A Comparative Study Evaluating the Detection of Organisms and Nonneoplastic Findings.Acta Cytol · 2014 · 58(4):388-97doi:10.1159/000365046 · PMID 25033897

    Trichomonas identified in 10 of 11 positive cervical samples when organisms were part of what was assessed; a small set, and cytology rather than tumor tissue.

91. 91
    Sun WW, Dong ZW, Zhou YM, et al. Improving the identification and diagnostic efficiency of Metagenomic Next-Generation Sequencing for mycobacterial granuloma on postoperative formalin-fixed paraffin-embedded specimens.Microbes Infect · 2023 · 25(8):105185doi:10.1016/j.micinf.2023.105185 · PMID 37453490

    Broad sequencing of archived surgical granuloma blocks detected all 31 confirmed mycobacterial infections, with 88% specificity, and all 19 tuberculosis cases against 47% for the standard molecular test on the same blocks; mycobacteria, 65 cases.

92. 92
    Sitthirak S, Wangwiwatsin A, Jusakul A, et al. Whole exome sequencing of multi-regions reveals tumor heterogeneity in Opisthorchis viverrini-associated cholangiocarcinoma.Sci Rep · 2025 · 15(1):10886doi:10.1038/s41598-025-95142-3 · PMID 40157958

    Multiregion sequencing of 52 samples from 13 liver-fluke-associated bile-duct cancers found marked genetic differences between regions of the same tumor; 13 patients, human mutations only.

93. 93
    Baldwin DA, Feldman M, Alwine JC, Robertson ES. Metagenomic assay for identification of microbial pathogens in tumor tissues.mBio · 2014 · 5(5):e01714-14doi:10.1128/mBio.01714-14 · PMID 25227467

    Describes PathoChip, a microarray carrying probes for viruses, bacteria, fungi, parasites and helminths, and shows it working on formalin-fixed tumor tissue; a platform paper, with no count of parasites in tumors.

94. 94
    Banerjee S, Tian T, Wei Z, et al. Microbial Signatures Associated with Oropharyngeal and Oral Squamous Cell Carcinomas.Sci Rep · 2017 · 7(1):4036doi:10.1038/s41598-017-03466-6 · PMID 28642609

    PathoChip signatures, parasite signatures among them, reported in oral and oropharyngeal cancers; one group, no independent replication, and a signature is not an organism seen in tissue.

95. 95
    Banerjee S, Tian T, Wei Z, et al. The ovarian cancer oncobiome.Oncotarget · 2017 · 8(22):36225-36245doi:10.18632/oncotarget.16717 · PMID 28410234

    PathoChip signatures of viruses, bacteria, fungi and parasites reported in ovarian cancers; the same group as the oral and breast reports, with small comparison sets.

96. 96
    Banerjee S, Tian T, Wei Z, et al. Distinct Microbial Signatures Associated With Different Breast Cancer Types.Front Microbiol · 2018 · 9:951doi:10.3389/fmicb.2018.00951 · PMID 29867857

    PathoChip signatures, parasites among them, reported across four breast cancer types; the same group again, and published before the reanalyses that showed how contamination creates false microbial signals.

97. 97
    Hara K, Yamasaki K, Tahara M, et al. Epidemiologic evaluation of pleurisy diagnosed by surgical pleural biopsy using data from a nationwide administrative database.Thorac Cancer · 2022 · 13(8):1136-1142doi:10.1111/1759-7714.14368 · PMID 35243795

    In 1,699 Japanese patients with pleurisy diagnosed by surgical biopsy, lung fluke disease was 0.1%; an administrative database, so the diagnoses are coded, not reviewed.

98. 98
    Machicado C, Marcos LA. Carcinogenesis associated with parasites other than Schistosoma, Opisthorchis and Clonorchis: A systematic review.Int J Cancer · 2016 · 138(12):2915-21doi:10.1002/ijc.30028 · PMID 26840624

    Found parasites such as Echinococcus, Strongyloides and Fasciola inside cancers or tumors, but concluded cause and effect was not established; 19 studies from 1,266 screened.