As a basis for tumor diagnosis, the diagnostic error rate of surgical pathological diagnosis with obvious clinical significance is only 0.26% to 1.2% [1], while the reliability of physicochemical clinical, clinical, and postmortem inferences decreases in order. Therefore, the importance of pathological diagnosis cannot be overemphasized, but it seems to be biased to deduce that cancer treatment cannot be performed without pathological diagnosis for the following reasons. 1. There is no law or regulation in China which stipulates that cancer treatment cannot be carried out without pathological diagnosis. 2. Clinical practice and clinical scientific research should not be confused. In clinical practice, it is not uncommon that the diagnostic criteria are not fully met, otherwise there would be no distinction between excellent and mediocre doctors. The diagnosis of a disease depends to a large extent on the experience of the doctor and on the comprehensive judgment of the condition in a specific context. In contrast, clinical scientific research requires all patients to meet the established criteria (in clinical research of oncology, pathological diagnosis is required) and to be treated according to the established protocol, which is important to ensure the seriousness and comparability of scientific research, but it obviously cannot adapt to the intricate clinical practice. 3, compared with clinical diagnosis, pathological diagnosis is actually more of an empirical medicine. Putting aside the quality of pathological sections and equipment conditions, different pathologists often give completely different conclusions on the same pathological section, and even the same physician may reach contradictory conclusions when observing the same section several days or months apart. Clinically, there are often cases of conflicting diagnoses in different hospitals, and people are used to negate the opinions of “lower hospitals” by the so-called “higher hospitals”. In fact, there is no difference between the upper and lower levels of hospitals, except that hospitals with high academic standards are less likely to be wrong and more likely to be correct, but it is impossible for anyone or any unit to be correct all the time. The results of reading 460 difficult cases in Guangzhou between 1985 and 1995: only 65 cases (14.1%) were diagnosed by many experts, 110 cases with 2 diagnoses, 140 cases with 3 diagnoses, and 100 cases with 4 diagnoses [4]. This shows that hospitals with a large academic level, both clinical and pathologists, should avoid the style of easily dismissing or even blinking others from above (although this is not much), and it helps to avoid unnecessary medical disputes. 4, the possibility of pathological misdiagnosis also exists. Troxel reported that 22% of medical disputes due to melanoma were due to misdiagnosis of melanoma as Spitz nevus. Lymphomas from extra-nodal sites (skin, nasal cavity, pancreas, mediastinum and stomach) are particularly prone to false negatives, resulting in 43% of all medical disputes [1]. In many difficult cases, it is not uncommon for the clinician to make a correct diagnosis while the pathology is incorrect. Ignoring the role of clinicians and clinical data, and believing in “microscopic determination” will make the limitations of pathological diagnosis more obvious, and may even lead to misdiagnosis [5]. 5, new technologies and advances in pathology also have limitations. In recent years, immunohistochemistry and gene amplification technologies have been widely used in tumor pathology diagnosis, and many difficult cases have been clearly diagnosed and guided by clinical treatment. However, it must be recognized that they are only a supplement to conventional pathological histology and should be obeyed in case of contradiction between them, because the former is more likely to be false positive or false negative, and most of the tumor markers are only relatively specific, for example, T-cell lymphoma can express CD20, metastatic undifferentiated carcinoma and neuroendocrine carcinoma can also express LCA. Clinicians must be cautious when interpreting these pathology reports. The diagnosis of benign and malignant of many tumors cannot be made by pathology alone, but must be combined with clinical findings. For example, in the case of gastrointestinal mesenchymal tumors, the size of the tumor, the integrity of the envelope, and the presence of metastases from other sites must be combined to determine the benignity and malignancy of the tumor, and pathological morphology itself is often unable to help [6]. This is also true for thymoma and adrenal tumors. 7. Some tumors can be diagnosed with physical and chemical clinical basis, the most representative of which is primary liver cancer. The clinical diagnosis of primary hepatocellular carcinoma can be made with the presence of hepatic occupancy on imaging, together with significantly elevated AFP and exclusion of benign disease with elevated AFP [7]. It is also not necessary to have pathological basis for multiple myeloma. 8. Pathological specimens are not always available. The prerequisite for pathologic diagnosis is to obtain sufficient focal tissue or cellular specimens, and for superficial lesions, removal of a small piece of tissue is easily accepted by patients. For thoracoabdominal and intracranial lesions, ultrasound, X-ray, CT, and other imaging techniques guided by puncture largely improve the availability of specimens, but if the lesion is small or treacherously located, puncture is difficult and risky, and repeated punctures may incur patient resentment or refusal. When a patient already has a typical pulmonary occupancy with an occupying lesion in the skull or bone, most are able to make a clinical diagnosis of lung cancer; exceptions must be rare, and insisting on surgical biopsy for such patients may expose the patient to unnecessary treatment and unnecessary risk. Endless examinations and observations have the risk of delaying the disease and increasing the cost of examination and treatment, which is no less likely to lead to medical disputes than clinical diagnosis. Doing nothing in front of the disease will likewise lead to the prosecution of patients. 9. The stage, diversity and complexity of tumors often make pathological diagnosis powerless. As we all know, lesions are dynamic and developing, different stages of lesion development can affect the morphological structure of lesion cells, and a biopsy diagnosis can only reflect the pathological changes at a certain stage of the disease process. In this stage, the tumor does not necessarily show obvious malignancy, but the patient’s condition cannot wait, otherwise it may lead to serious consequences. We once had a patient with enlarged cervical lymph nodes 20 years ago. 7 domestic experts repeatedly consulted his tissue sections, and only 2 of them considered malignant lymphoma. After corresponding chemotherapy and radiotherapy, the patient suspected the diagnosis of tumor because he had been surviving without disease, but tumor and even systemic metastasis reappeared in the same area 20 years later. 10. Different lesions have similar or even the same morphological features hindering pathological diagnosis. There is often no absolute demarcation between benign and malignant of the same type of tumor. Granulomatous lesions may be tuberculosis, nodular disease, leprosy or even Hodgkin’s disease; focal nodular hyperplasia of the liver, hepatocellular adenoma and highly differentiated hepatocellular carcinoma are often morphologically similar to each other, and it is sometimes difficult to clearly distinguish between reactive hyperplasia of lymph nodes and early non-Hodgkin’s lymphoma. In these cases, pathology can only make an ambiguous diagnosis, but the clinician must make a decision on the diagnosis and treatment of the patient, and the only thing that can work at this point may be the experience and wisdom of the physician. 11. Changes in pathological diagnostic criteria and the correctness of diagnosis. Each diagnostic standard only reflects the medical understanding of a disease at that time, with obvious imprints of the times. With the development of medical science, people’s understanding of the nature of the disease will deepen, change or even deny the original opinion, and the corresponding diagnostic criteria will also change, which makes pathological diagnosis and the diagnostic criteria it relies on “time-sensitive”. Therefore, neither the past nor the present should be used to judge the present. In summary, the authors believe that pathologic diagnosis is not always necessary, that pathologic diagnosis is also subject to error, and that pathologic diagnosis does not necessarily negate clinical diagnosis. It is neither realistic nor possible for clinicians to put the responsibility of diagnosis entirely on pathologists. The reference to “no cancer treatment without a pathological diagnosis” is biased and is actually an unnecessary overreaction by physicians in the current situation of increasing medical disputes. Of course, in the absence of a pathological diagnosis, it is important to keep in mind and explain to the patient the possibility of an incorrect diagnosis, to proceed with caution and to closely observe changes in the course of treatment, and to revise or confirm the diagnosis at any time, with the patient’s full knowledge and approval.