Many people wonder how likely it is that biopsies or surgical removal of cancerous tissue will cause the cancer to spread, either locally or by metastasis to lymph glands or through the blood stream. It can occur, but with proper care, the odds of it happening appear to be small.
We now have tools that can detect shed cancer cells in the blood stream. The technology has not been perfected, leaving uncertainty as to whether some of the tests are detecting cells capable of forming metastases, or even live, whole cells. Nevertheless, preliminary studies suggest that, in many cases, cancerous cells are present in the blood stream before, during, and even after surgery and can increase during some operations. It has not been determined whether that any increase due to the surgery is correlated with the development of metastases or poorer survival.
Pending direct evidence of this type, some observations can be made.
The prognosis with some types and stages of cancer is very good despite routine use of both biopsy and surgery, indicating that any contribution to cancer spread must be small. For example, untreated breast cancer is very prone to spread via lymph vessels to the axillary (armpit) lymph nodes and also via the blood stream to lungs, bones and elsewhere.
Yet following routine biopsy and surgical treatment only 6.7% of patients with breast cancers up to 3 cm (1.2 inches) in diameter develop axillary lymph node metastases within five years [1]. If bad prognosis cancers i.e. those with a high score on four parameters combined (larger tumor size, high tumour grade, laminin receptor positive, and HER2 over-expression) are excluded, the rate of development of lymph node metastases is only 2%. We know from other studies that many, possibly all, of these metastases would be present before any medical attention, but undetectable clinically.
The same applied to blood stream spread. Only 11% per cent of such patients went on to develop distant metastases within five years, and only 5% when those with a bad prognostic score were excluded. It is known that more aggressive breast cancers can give rise to occult (undetectable) blood-borne metastases at a very early stage, so that many or most of these metastases will be pre-existing. The biology of the cancer again appears to be the main determinant of outcomes. (These are medium-term results and will overestimate the final cure rate of breast cancer in these patients).
The same considerations apply to other cancers with high cure rates.
Surgical Risk
Well-conducted surgical operations need not spread cancer any more than would many other common activities. There will be little more squeezing or jostling of the cancer than has occurred over the 2-3 years that most cancers are present before diagnosis. Operations are designed to cut through tissues well away from the cancer, so that the cancer itself is not penetrated. A breast cancer is removed with a buffer of 1 cm or more of surrounding normal breast tissue whenever possible. With colon or stomach cancer, at least 5 cm of normal colon or stomach would be removed on either side of the cancer whenever anatomically possible. With a suspected melanoma of the skin, 1 cm of normal skin around the tumor, and the subcutaneous tissues down to the muscle sheath may be removed so as to be well clear of the cancer and any “roots”.
Where good clearance is not possible and the cancer is broken into, local seeding of the cancer is likely, with cells possibly also getting into the blood stream, local lymphatics, or drifting widely along tissue planes and spaces. The consequences of this would depend on the anatomical location, so it is difficult to generalise. In cases where cancer is being left behind, the spilled cells are unlikely to alter the outcome. In other cases, it may spread cancer through body cavities such as the abdomen and create additional problems for the patient. In some cases, radiation therapy is used to attack any cells that may have been spread during the surgery.
Biopsy Risk
There should be little risk of spread by a biopsy in which the cancerous lump is easily and completely removed. This is possible with superficial skin cancers including melanoma, lymphatic tumors, and small breast cancers. It is also unlikely that taking a 2-3 mm bite out of a fungating or ulcerating cancer anywhere in the gastrointestinal tract, respiratory tract, or bladder during endoscopic examinations, the usual mode of biopsy, is going to increase spread. These cancers are already subjected to significant trauma on a daily basis. Bits probably break off all the time and cause the bleeding that is their most common symptom.
Core and needle biopsies can occasionally implant cancer in the needle tract, but cases where this is a significant problem are rare. Wherever possible, for example after breast lump biopsy, care is taken to cut out the needle track when the patient has further surgery. I know of no evidence that such biopsies can induce blood stream spread, but it is a theoretical possibility. Fine needle aspiration cytology (FNA) is less traumatic than the larger needles used for core biopsies and should be safer, although the latter will provide more information about the cancer in some situations.
The main risks of serious spread will apply with incision biopsies, where a chunk is cut out of a cancer that is too large to completely remove in a simple diagnostic procedure and when a detailed diagnosis is required before definitive treatment can be planned. This would apply mainly with sarcomas—cancers of soft tissue and bone. However, the risk of implantation is well known, so biopsies of this type are carefully planned so that that whole area would be removed in any subsequent operation. The potential for blood stream spread is not yet known, but is theoretically possible or even likely. However, the risks of going without a diagnosis when effective treatment is available are much greater.
Reference
- Greco, M and others. Breast cancer patients treated without axillary surgery: Clinical implications and biologic analysis. Annals of Surgery 232:1-7, 2000.
This article was revised on April 16, 2007.
