What Doctors Don't Tell YouI have lumpy breasts, but I am loathe to have a mammogram because of doubts about its safety and accuracy. My doctor has recommended that I go for a breast ultrasound scan. What’s your view? I’m 47. M D, Boston, Massachusetts.
These days, any gynecologist who feels a bump is likely to recommend a scan of some sort. According to our Alternatives columnist, Dr Harald Gaier, in the early stages, it’s highly difficult to tell with absolute certainty whether a lump is cancerous or benign, unless the lump is biopsied. (In this procedure, a needle is inserted into the lump and some cellular tissue extracted, in order to be analyzed in a lab.) Nevertheless, he says, it’s possible to get some idea of the sort of lump it is by feel. Pain, changes in size during your menstrual cycle, easy mobility, absence of hardness and the presence of multiple nodules probably means there is no cancer, he says, while malignant lumps are usually hard, irregular, non-tender and fixed. A lump that remains the same throughout your cycle or increases dimpling of the overlying skin or tethering to the skin above or muscle below the lump is slightly more likely to indicate cancer.
Discharges, says Dr Gaier, may indicate a number of things. A blood stained discharge from the nipple may not definitely indicate cancer but also be present in benign cystic mastitis. A greenish or yellowish discharge is invariably caused by mastitis; a watery one, early pregnancy; and a milky discharge (that is, if you’re not breastfeeding), an adverse drug reaction. Pain in the breast per se isn’t necessarily cause for alarm, he says (although it may sometimes prefigure the future development of breast cancer). Pain is often one of the the collection of symptoms of PMS, is present with breast abscesses or a candida albicans yeast overgrowth.
As we’ve written about in many issues of WDDTY, mammograms are increasingly falling into disrepute. Not only are they considered inaccurate and difficult to interpret, but downright harmful for certain women whose breasts may be sensitive to the carcinogenic potential of the strong x-rays (See WDDTY vol 3 no 10).
With doubts growing about mammograms and other forms of x-rays, researchers are turning to mammary ultrasound as well as ultrasound for many other diagnoses. As you know, ultrasound employs soundwaves to produce an image on what appears to be a television set.
As the instruments, including the transducer (the gadget producing the sound and “listening” for returning echoes) have grown far more sophisticated, ultrasound use has dramatically increased. These days, it is used to diagnose heart problems, a variety of tumours, circulatory problems, and to examine organs and body parts, including liver, spleen, uterus, placenta, brain and now breasts.
However, the success of ultrasound largely depends on the skill of the operator, as images can be hard to read and are open to misinterpretation. In particular, operators worry about visualizing “artifacts” that is, a ghosted image of something that isn’t there or mistaking something quite normal for something sinister. For instance, fetal hair has been mistaken for serious neural tube defects; bladders have been mixed up with pelvic tumours. This often happens when operators make errors in setting up scanning technique instruments or positioning the transducer (JAMA, March 6, 1991).
There’s also the problem of false echoes creating images on the screen that suggest things that aren’t there at all. This is a particular problem with curved, highly reflective surfaces such as the diaphragm or near large masses, such as the gallbladder or bladder. And problems in the accurate reflection of the sound beam can distort the size, shape, position and brightness of structures, miss real echoes and so miss important pathologic features.
With breast examinations, the most commonly used equipment is “real time” high resolution ultrasonography which means you are seeing on the screen exactly what the transducer is picking up at that moment.Typically, a test requires about 10 minutes of exposure. According to one study of 100 women with at least one breast nodule, the overall rate of accuracy of ultrasound was 74.8 per cent. This, of course means, that in one in four cases, the diagnosis was wrong. In 10 cases, the ultrasound diagnosed as cancerous benign breast cysts, and also missed altogether one breast cyst and one abscess (Radiation Med, 1994; 12 (5): 201-8).
The other type of ultrasound used is colour Doppler ultrasound, which measures the flow of blood, which in malignant tumours tends to be abnormal. The overall view of this technology is conflicting. In one study, overall accuracy for detecting breast tumours was 82 per cent (Anticancer Research, 1994; 14 (5B): 2249-51), but in another large scale study, the ability of Doppler colour ultrasound to specify which type of tumour was only 46.9 per cent (Radiologia Medica, 1994; 87 (1-2): 28-35) and in a third, 83 to 100 per cent of malignant tumours were correctly identified, but only 51-61 per cent of benign lesions correctly identified (Zentralblatt fur Gynakologie, 1993; 115 (11): 483-7).
The authors of one of the studies suggested that the colour Doppler ultrasound be used only to add further information to those obtained with conventional ultrasound (Radiologia Medica, 1994; 87 (1-2): 28-35. Other researchers have concluded that Doppler colour can reassure patients that lesions are benign (anything without any colour signals is deemed safe; those with colour signals merit a biopsy), although in another study cancerous lumps were only distinguished from benign in less than half of the cancer cases (Rofo, 1992; 156 (2): 142-5).
Nevertheless, the technology appears to be improving; presently the colour method is use by comparing a colour spectrum analysis with surrounding tissue; in cancerous tumours, the colour is typically more intense with sharp margins. In one study among 70 patients, this method only missed a single tumour (Geburt und Frau, 1994; 54 (8): 432-6).
To date, the most accurate diagnostic method is combining ultrasound with “high speed” punch biopsies, once lesions have been identified by ultrasound. In one facility in Germany, this technique reached an accuracy rate of close to 100 per cent (Geburt und Frau, 1994; 54 (10) 539-44).
According to Professor William Lees, director of radiology at UCL Hospitals Trust in London, the best ultrasound should have Doppler as part of the system and use the two types in tandem, which will boost an operator’s confidence about the accuracy of his diagnosis. Prof Lees also believes that a skilled operator will have a much higher accuracy rate than the studies demonstrate closer to 85 per cent.
It may be that ultrasound has a similar overall batting average with mammograms. In one review of 80 patients with both benign and malignant lesions, mammograms picked up five cancers missed by ultrasound, but ultrasound discovered nine cancers missed by mammograms. In yet another study, ultrasound picked up four cancers that weren’t yet palpable (Ultraschall in der Medizin, 1994; 15 (1): 20-3).
At the moment, ultrasound isn’t considered appropriate for screening not only because of the error rate but because accuracy depends so heavily on skilled operators, and there aren’t enough of them about.
The bottom line appears to be that ultrasound may be a better option than mammography to diagnose women with symptoms because it is safer and possibly more accurate than mammography for the under 50s. Although the technology is vastly improving and will probably eventually develop into a good tool, there are still some problems with accuracy.
Prof Lees agrees that it should not be used as a general population screening tool, but may be a better first line investigation of lumps felt during breast examination, because when combined with a needle biopsy done on the spot, it can be highly accurate.
A needle biopsy mostly causes bruising, although there have been very rare instances of lung puncture.
Mr Alan McKinna, a consultant breast cancer specialist, says that most doctors don’t like to rely on either ultrasound or mammograms alone, since ultrasound will pick up lumps you can feel but miss those you can’t; mammograms will pick up the invisible lumps but may miss the big ones you can feel. Many doctors are using both. It makes no sense to use ultrasound for screening since it won’t pick up lumps you can’t feel yourself.
This all adds up to the fact that you should only have a test performed with a highly trained operator highly skilled in all the latest equipment and equally well trained in artifacts and all the ghost images that ultrasound can produce.
The most important questions you should ask concern his/her expertise. Always opt for someone highly experienced, particularly in breast scans. Don’t be shy about asking his accuracy rate or if there have been any serious cases he’s missed. Also ask about the state of the equipment how new it is, how accurate, and when last serviced. Always seek out the best available equipment, specifically run by a radiologist specializing in ultrasound.
Finally, you may wish to avoid a scan without a good reason ie, an abnormal breast exam. Although ultrasound appears to be very safe for anyone other than a human fetus, Prof Lees confirms that no long term studies have been done on anyone but obstetric cases. Although we know that sound waves cause bubbling in cells, we don’t know if that means anything significant. It may be another generation before we know for sure.
Maryon Stewart
Tom Ferguson MD
Daniel Redwood DC