Robin Sealander was relieved with the results of her Pap smear.
For several months in late 1985, the 26-year-old Baldwin Park woman had suffered from heavy vaginal bleeding and irregular menstrual periods. She went to her gynecologist fearing that she had cancer.
The doctor took a Pap smear, a test used to detect cervical cancer. It came back negative, as had two previous tests.
Nevertheless, Sealander's symptoms didn't go away. Four months after the last lab test, Sealander consulted another physician. He found that she had cervical cancer. The effect of the disease was so severe that it was visible to the naked eye.
"He just looked at her and said, 'You have cancer,' " recalled William Sealander, Robin's father. "He knew right away."
Fifteen months later, Robin died.
"She was my only girl," Sealander said. "It's something you never get over."
What's shocking about Sealander's story--and thousands like it--is that Pap smears often fail to detect cervical cancer, which can be cured if caught in time. The industry's average error rate is an alarming 5%, and recent federal and state inspections revealed that some labs misdiagnose nearly 20% of the Pap smears they examine.
Nearly 70 years after the Pap smear was invented by Greek immigrant George Papanicolaou, the computer age has finally caught up with the test. Scientists and entrepreneurs, prompted by growing concerns among women and publicity about problems plaguing the Pap smear industry, are joining forces in an effort to improve the Pap smear's accuracy.
Biomedical companies around the United States have spotted opportunities. By year's end, companies will be selling computers that they claim can immediately detect cancer cells, kits that can spot a new virus that may cause cervical cancer and improved swabs designed to collect hard-to-reach cervical cells.
"All of this is way overdue," said Tenny Tse, co-owner of Medscand, a Florida medical supply company.
Pap smear diagnosis still requires a technician to stare through a microscope at hundreds of thousands of cells per slide to find a handful that might indicate the presence of cancer. Consequently, mistakes are frequent, and only recently has the government begun to impose quality standards.
Earlier this month, National Cancer Screening Laboratories was dropped from the Medicare and Medi-Cal programs after federal inspectors determined that the Fullerton lab had misread 19% of the slides in a sample survey. The company has shut down.
And a year ago this week, the state Department of Health Services shut down the nation's largest Pap smear lab, Central Pathology Services Medical Group in Tarzana, after a survey team reported that its error rate was 21%.
Nearly 60,000 women in the United States develop cervical cancer every year, and 7,000 die from it, according to the American Cancer Society.
"I have a cupboard full of cases of women who were diligent and went back for Pap smear after Pap smear and still died," said Dr. Dorothy Rosenthal, chief of cytology at UCLA Medical Center.
The cost of a Pap smear typically ranges from $20 to $50, and the market worldwide is tremendous. Fifty million women in the United States and 75 million in Europe spend billions of dollars a year getting Pap smears. About 125 million more smears are processed in Japan, South Africa and parts of South America.
Until recently, laboratories used the Pap smear as an inexpensive loss-leader to attract new business from doctors. Some labs charged as little as $1 per Pap smear--even though the test is done manually and requires heavy staffing--in return for orders for much more expensive tests, such as urinalyses and histologies.
Although some states require laboratories to bill patients directly, others, including California, still allow doctors to mark up laboratory tests. Physicians can boost their profits by seeking out a low-cost provider and then charging a higher price to the patient.
"Gynecologists have an incentive to shop around for the cheapest market," said Mark Rutenberg, co-inventor of a computerized Pap smear system named Papnet. "They buy laboratory services like they buy their carpeting or lights: It has to be inexpensive but look good."
Fewer and fewer labs, though, are using the Pap smear as a carrot. Federal regulations now place an 80-slide limit per day on a lab technician's workload. This has increased the demand for cytotechnologists--individuals trained to read Pap smears--at the same time that there is a growing shortage of such personnel.
The California Assn. of Cytotechnologists reports that the number of vacant positions has ballooned from 19 in 1988 to 150 last year. There are only about 1,000 cytotechnologists in the state, mostly because few people are interested in a job that can be tedious.
Salaries for qualified people have soared from about $20,000 two years ago to as much as $50,000 in the Los Angeles area. The cost of a Pap smear has risen accordingly.
Despite a crackdown on medical costs, few people think that patients or insurance companies will object to a slight rise in price for new products if those products yield better Pap smear results.
"For the first time (doctors) are willing to pay more to get an accurate Pap smear. In the past, the cheapest was the best," said Rosenthal. "Now there is a realization that Pap smears are important. They are not dipstick tests."
To many in the industry, automation appears to be the only solution to personnel shortages and the public clamor for accuracy.
Two tiny technology companies have emerged as the leading candidates in the drive to automate Pap smear diagnosis, and each is racing to finish first.
Papnet, manufactured by Neuromedical Systems Inc. in Suffern, N.Y., is expected to be the first computerized system to be installed in labs, starting this summer. Using image processing technology pioneered by Jet Propulsion Lab in Pasadena, Papnet picks the 64 worst-looking cells out of the hundreds of thousands that appear on a typical Pap smear slide.
Those suspicious cells are shown on a TV screen, which the cytotechnologist examines to make a diagnosis. Papnet eliminates eye strain and fatigue that result from trying to find the two or three abnormal cases out of every 100.
"It's a needle in a haystack problem," said Mark Rutenberg, co-inventor of Papnet. "Right now, the cytotechnologist has to find as few as 10 bad cells out of 300,000. The job with our machine is to find those 10 cells out of 64."
Neuromedical has notified the U.S. Food and Drug Administration of its intention to sell Papnet for research purposes. This is a step towards being able to introduce the machine commercially.
Pharmaceutical giant SmithKline Beecham has announced that it will install Papnet in two of its labs this summer. And UCLA Medical Center has said it intends to get one within the next eight weeks.
Neuromedical claims that during in-house tests, Papnet has never failed to identify a Pap smear containing cancerous cells. However, academic studies to confirm the machine's effectiveness are years away.
Dr. Leopold Koss, one of the world's foremost experts on Pap smears and chairman of the pathology department at Albert Einstein College of Medicine in the Bronx, has just begun reviewing Papnet's performance. "I wouldn't be wasting my time if the instrument were not promising," he said. I and my colleagues consider the machine very interesting."
Neuromedical, which plans to make a $20-million to $30-million public stock offering in a few months, expects to rent the machines commercially for $5 per Pap smear.
"On a slow day, we get calls from 30 labs wanting the machine," said Rutenberg.
Rosenthal at UCLA confirms that laboratories are anxious to find an automated system that works. "People who have seen this and other machines have practically signed contracts," she said. "They say, even before (the machines) are ready, 'We'll take five.' "
Neuromedical's chief competi tor is NeoPath Inc. The privately-held Bellevue, Wash., company is developing an image-processing system--the AutoPap 300--that will automatically sort slides into two piles: normal and suspicious. The cost is projected to be under $1 per slide. The company expects to begin FDA trials by year-end.
Financial analysts say whichever company first succeeds in automating Pap smear diagnoses stands to strike it rich.
"There are no (automated) systems out there today," said Wayne Wager, a general partner with the venture capital firm Cable & Howse. "The market is somewhere in the neighborhood of half a billion dollars in equipment for processing Pap smears."
Until now, the Pap smear has successfully resisted automation because of the complexity of the cells on each slide. They come in various shapes and sizes and often overlap like pancakes. Humans can distinguish boundaries, but computers often can't.
"Our eyes are more clever than the machine--so far," said Karen Andersen, president-elect of the California Assn. of Cytotechnologists.
Besides automation, biomedical companies are working on tests to detect a virus--human papilloma virus (HPV)--that some scientists think may be a precursor to cervical cancer. The tests are being used by researchers in conjunction with the Pap smear.
Life Technologies in Gaithersburg, Md., a publicly held company traded over the counter, is the only company to date that has received approval from the FDA to sell such a test for diagnostic purposes.
The company has sold about $4 million worth of its ViraPap product since the test was introduced last year.
Enzo Biochem in New York and Digene Diagnostics in Silver Spring, Md., are developing tests that they say are faster and easier to use than the ViraPap.
Even if a lab forgoes a new computer or the HPV test, it can greatly improve its diagnoses by using advanced swabs that cost less than 50 cents each. Studies show that many cancers and abnormalities are missed altogether on a Pap smear because the doctor has failed to collect an adequate sample of cells while inserting the swab into the cervix.
Two companies have introduced brushes that greatly increase the number of cells obtained during a smear, including hard-to-reach endocervical cells that are often the first to harbor cancer.
Medscand's Cytobrush was developed by a Swedish cytopathologist and is considered a vast improvement over the cotton applicators and wood spatulas still used by many physicians. The Cytobrush received FDA approval six years ago, and some 20 million have been sold so far at 20 cents apiece, compared to about 2 cents for a spatula or cotton swab.
Cytobrushes cost about 20 cents each, compared to about 2 cents for a cotton applicator or spatula.
The April issue of the journal Obstetrics & Gynecology contains an academic study led by Dr. Paul J. Murata at UCLA that says the use of the Cytobrush increased the number of cancers accurately diagnosed.
Connecticut-based Unimar has sold nearly a million of its Cervex-Brushes since it began marketing the product last year. The brush costs from 29 cents to 44 cents.
Poor sampling may have contributed to Robin Sealander's death from cervical cancer, according to a malpractice suit her father filed two years ago in U.S. Superior Court in Pomona against a medical lab and her former gynecologist.
Dr. Irene Davos, a pathologist at Cedars Sinai Hospital, examined Sealander's last Pap smear and testified that it "may have been an unsatisfactory specimen and (the lab) should have requested another test."
The case was settled out of court last year.
AIMING FOR BETTER PAP SMEAR ACCURACY
PAPNET: an image processing computer that selects the 64 worst-looking cells on a Pap smear and then projects them on a high-resolution television screen for diagnosis.
AutoPap 300: an image processing computer that sorts Pap smears into two categories, normal and suspicious. Only abnormal slides are reviewed by a cytotechnologist.
ViraPap: a DNA probe test used to detect the presence of human papilloma virus or HPV, which some scientists think may be a precursor to cervical cancer.
Cervex-Brush: a special swab designed to collect hard-to-reach endocervical cells, which are then put on a Pap smear slide.
Source: Product manufacturers IDENTIFYING ABNORMAL AND MALIGNANT CELLS
A technician uses a microscope to examine thousands of cells on a Pap smear, searching for the relative handful that may indicate the presence of cancer. Below are cell samples, as they appear through the microscope, showing normal, abnormal but not precancerous, precancerous and malignant cancer cells.
PROBLEM CYTOLOGY LABS
Federal health officials have terminated the participation of several cytology labs in reimbursement programs because of a high level of errors, mostly on Pap smear diagnoses.
Termination Facility Location date Uni Lab Inc. Westmont, Ill. 2/1/88 Central Medical Laboratory Inc. Southfield, Mi. 11/1/88 National Cyto Lab Inc. Harvey, Ill. 12/23/88 Pines Peak Diagnostic Services Inc. Colorado Springs, Co. 2/7/89 Central Pathology Services Med Group Tarzana 2/21/89 Plaza Health Laboratories Inc. Brooklyn, N.Y. 6/1/89 Medicon Diagnostic Lab New York 7/1/89 Cancer Cytology Medical Lab Rockville, Md. 3/1/90 National Cancer Screening Laboratories Fullerton 4/20/90
Source: U.S. Department of Health and Human Services.