HOUSTON, July 7 — Treatment-induced prostate shrinkage likely unmasked high-grade cancers, resulting in a detection bias in the finasteride (Proscar/Propecia) arm of the Prostate Cancer Prevention Trial (PCPT), data from a large patient series suggest.

• Explain to patients that a drug that decreases the size of the prostate might make cancer easier to find.
• The findings were based on a retrospective review of medical records, not a controlled clinical trial.

The value of prostate specific antigen (PSA) as a marker for prostate cancer declined steadily as prostate volume increased. A similar inverse relationship existed for high-grade cancer, Christopher S. Elliott, MD, of Stanford University, and colleagues reported in the July 15 issue of Clinical Cancer Research.

“Decreases in prostate volume over time and the resultant change in prostate-specific antigen performance characteristics may have contributed a bias toward the detection of high-grade disease in the finasteride arm of the Prostate Cancer Prevention Trial,” the authors concluded.

The observations could explain why the increased risk of high-grade cancer in the finasteride arm was limited to the subgroup of men who had symptom-driven biopsies during the trial, they added. The findings also are consistent with those from analyses that PCPT investigators have performed in the six years since the trial ended.

The PCPT involved 19,000 healthy men who were randomized to finasteride or placebo for seven years. The principal finding was a 25% reduction in prostate cancer incidence in the finasteride arm.

However, the beneficial effect has been overshadowed by the finding that finasteride-treated men had a small but statistically significant increase in the rate of high-grade cancer compared with the placebo group.

Closer examination of the PCPT data revealed inconsistencies in the occurrence of high-grade cancer. Specifically, the increase was limited to men who had “for-cause” biopsies, triggered by an abnormal digital rectal exam (DRE) or a rise in PSA. End-of-study biopsies showed an almost-identical incidence of high-grade cancer in the two treatment arms.

Prostate volume at the time of biopsy was 25% lower in the finasteride arm (25.5 cm³ versus 33.6 cm³), the authors noted.

Dr. Elliott and colleagues hypothesized that the increased rate of high-grade cancer in the finasteride arm resulted from the drug’s volume-reducing effect on the prostate. The shrinkage could have improved PSA’s performance characteristics for detecting prostate cancer.

To test their hypothesis, investigators retrospectively reviewed records on 1,304 men referred for an initial prostate biopsy. The referrals were prompted by a PSA value of 4 to 10 ng/mL or an abnormal digital rectal exam.

The study group had a median age of 66, median prostate volume of 42.9 cm³, and median PSA value of 5.5 ng/mL. Digital rectal exam was abnormal in 507 (38.9%).

The investigators calculated receiver-operator curves and positive predictive values for PSA, stratified by diagnosis and prostate volume.

For detection of any cancer, the area under the curve (AUC) decreased from 0.758 to 0.520 as prostate volume increased from <30 cm³ to >50 cm³. For detection of high-grade cancer, AUC decreased from 0.712 to 0.497 as organ volume increased.

A similar pattern emerged from calculations of positive predictive values.

For a prostate volume <30 cm³, the positive predictive value of a PSA of ?4 ng/mL was 25%, declining to 17.3% for a prostate volume >50 cm³. The differences increased for detection of high-grade cancer: 39% for a prostate volume <30 cm³ versus 10.7% for a volume >50 cm³.

Continued analysis of PCPT data has led to similar conclusions regarding the relationships among finasteride, prostate volume, and PSA performance characteristics for detection of prostate cancer, PCPT investigator Catherine Tangen, DrPH, of the Fred Hutchinson Cancer Research Center in Seattle, said in a statement.

Collectively, the data suggest that men should be offered finasteride, if they and their physicians agree that chemoprevention might be beneficial, she added.

Earlier this year, the American Society of Clinical Oncology and the American Urological Association recommended that healthy men ages 55 and older and with no signs of prostate cancer talk to their physicians about taking a 5-alpha reductase inhibitor to prevent prostate cancer. (See Men Encouraged to Consider Medication to Prevent Prostate Cancer)

Neither the authors nor Dr. Tangen reported any relevant disclosures.

Primary source: Clinical Cancer Research

Source reference:

Elliott CS, et al “The influence of prostate volume on prostate-specific antigen performance: implications for the Prostate Cancer Prevention Trial outcomes” Clin Cancer Res 2009; 15(14): 4694-99.

Related Article(s):

• Men Encouraged to Consider Medication to Prevent Prostate Cancer

TORONTO, July 7 — A vaccine against two oncogenic strains of human papillomavirus (HPV) is highly effective at preventing the lesions that lead to cervical cancer, researchers said.

• Explain to interested patients that cervical cancer is caused by the human papillomavirus, which has led to efforts to create a vaccine against the cancer-causing strains to the virus.
• Note that this study reports the final analysis of a major trial of a bivalent vaccine against the two main cancer-causing strains and explain that the vaccine was highly effective.

The final results of the so-called PATRICIA study — short for PApilloma TRIal against Cancer In young Adults — show more than 90% efficacy against HPV 16 and 18, the two main cancer-causing strains of the virus, according to Jorma Paavonen, MD, of the University of Helsinki, and colleagues.

The results of the randomized, placebo-controlled study extend an interim analysis indicating the vaccine was effective at preventing grade two cervical intraepithelial neoplasia (CIN2), Dr. Paavonen and colleagues reported online in The Lancet. (See New HPV Vaccine Found 90% Effective)

In the final analysis, based on three years of follow-up, the researchers also showed the vaccine prevented grade three cervical intraepithelial neoplasia (CIN3).

The vaccine — licensed in 90 countries as Cervarix — is not yet approved by the FDA. The drug’s manufacturer, GlaxoSmithKline , submitted data in March to support an approval application, which was rejected in 2007 pending more information.

If it’s eventually approved in the U.S., Cervarix would compete with Merck’s Gardasil vaccine — which targets four strains of HPV, including 16 and 18. The FDA approved Gardasil in 2006.

The PATRICIA trial enrolled more than 18,000 women in 14 countries. The primary endpoint was vaccine efficacy against CIN2 (or higher grade) associated with HPV-16 or -18 in women who were not infected with either at baseline, and who got all three injections of the vaccine.

At the final analysis, 60 cases of CIN2 or CIN3 were confirmed, of which 33 contained DNA from nonvaccine oncogenic HPV types as well as HPV-16 or -18.

Of those, four were in the vaccine group and 56 in the control group, the researchers found, for a vaccine efficacy rate of 92.9%.

When the researchers broke out results for each strain, they found two HPV-16 and two HPV-18 cases of CIN2 in the vaccine group, compared with 46 and 15, respectively, in the control group. (Some lesions in the control group contained both strains.)

For CIN3, there were two cases of HPV-16-associated lesions in the vaccine arm and 10 in the control group, the researchers found. There were no cases of CIN3 associated with HPV-18 in the vaccine arm and five in the control group.

Overall, for CIN3, the vaccine efficacy rate was 80% — 67.2% for HPV-16 and 100% for HPV-18.

Many women in the study were infected with nonvaccine oncogenic strains of HPV, and in a post hoc analysis, the researchers estimated a cross-protective efficacy that could represent between 11% and 16% additional protection against cervical cancer.

Dr. Paavonen and colleagues said the study’s strengths included its size and duration, as well as the diversity of the study population.

One major “unavoidable” limitation, they said, was that the high efficacy of the vaccine meant that more women in the control group than in the vaccine group were referred for colposcopy, which might have resulted in case ascertainment bias for detection of lesions associated with nonvaccine types.

The authors concluded that vaccination against HPV “has the potential to substantially reduce the incidence of cervical cancer and precancer and the numbers of colposcopy referrals and cervical excision procedures.”

However, in an accompanying comment article, two experts argued that current approaches to using HPV vaccines are too limited.

“Currently, the targets for HPV vaccination are girls and young women ages 11 to 26 years prior to sexual debut,” said Karin Michels, ScD, of Harvard Medical School, and Harald zur Hausen, DSc, MD, of the German Cancer Research Centre in Heidelberg, Germany.

“While good utilization of the program will reduce cervical cancer incidence in a couple of decades,” they argued, “this subgroup of the population at risk is too small to limit the spread of the virus.”

They argued that “the only efficient way to stop the virus” is to vaccinate boys and men as well.

“The goal to eradicate sexually transmitted carcinogenic viruses can be jointly carried by women and men and could be accomplished within a few decades,” they concluded.

The study was supported by GlaxoSmithKline Biologicals. Dr. Paavonen and several other authors reported honoraria, paid expert testimony, or travel grants from GlaxoSmithKline . Several other authors are employees of the company.

Primary source: The Lancet

Source reference:

Paavonen J, et al “Efficacy of human papillomavirus (HPV)-16/18 AS04-adjuvanted vaccine against cervical infection and precancer caused by oncogenic HPV types (PATRICIA): final analysis of a double-blind, randomised study in young women” Lancet 2009; DOI: 10.1016/S0140-6736(09)61248-4.

Additional source: The Lancet

Source reference:

Michels KB, zur Hausen H “HPV vaccine for all” Lancet 2009; DOI: 10.1016/S0140-6736(09)61247-2.

Related Article(s):

• New HPV Vaccine Found 90% Effective

HOUSTON, July 13 — Symptom patterns linked to ovarian cancer added minimal diagnostic information to transvaginal ultrasound but improved identification of benign tumors, helping some women avoid surgery, investigators reported.

• Explain to patients that a symptom index was not as accurate as ultrasound for diagnosing ovarian cancer.
• Note, though, that the symptom index improved the ability to distinguish benign tumors.

A six-symptom index correctly predicted malignancy in only 20% of patients compared with 73% for transvaginal ultrasound, Edward J. Pavlik, PhD, of the University of Kentucky in Lexington, and colleagues reported online in Cancer.

However, the symptom index had a higher specificity than did transvaginal ultrasound. Sequential use of the two diagnostic strategies resulted in worse sensitivity but improved specificity.

Clinical decisions based on use of either symptoms or ultrasound combined in parallel led to a small improvement in sensitivity but at the expense of a reduction in specificity.

“The current findings indicated that tumors that are negative by both ultrasound and a symptoms index are likely to be benign . . . and adding symptoms information that has weight equal to the weight of ultrasound only slightly improves the discrimination of malignancy,” the authors said.

“Thus, a major benefit in discriminating malignancy was achieved through ultrasound, whereas the absence of symptoms in conjunction with an abnormal ultrasound . . . indicated that the mass was benign and that surgery may not be required.”

Ovarian cancer has long had the reputation as a “silent killer” that has few specific symptoms that could aid in early diagnosis. However, recent studies have suggested that certain symptoms occur more often in women with ovarian cancer compared with women in the general population, the authors said.

The findings regarding ovarian cancer symptoms led to the development of a symptom index that has demonstrated some value for identifying early-stage disease, which often can be cured by conventional therapy, they continued.

The index comprises six symptoms:

• Pelvic pain
• Abdominal pain
• Increased abdominal size
• Bloating
• Feeling full
• Difficulty eating

The authors retrospectively evaluated the diagnostic utility of the symptom index in 272 women who participated in an annual transvaginal ultrasound screening project. All of the women had undergone surgery following an abnormal ultrasound.

Symptom results were compared with ultrasound and surgical pathology findings. The authors defined a positive symptom screen as the occurrence of one or more of the six symptoms for more than 12 days during the past year.

Criteria for an abnormal ultrasound screen consisted of ovarian volume more than two standard deviations above the mean for premenopausal and postmenopausal women and any ultrasound-detected abnormality. Abnormalities included women with malignant and benign tumors.

Using ultrasound findings, the investigators calculated each patient’s morphology index, which has demonstrated value in predicting the risk of malignancy. Ovarian size and structure were rated on a scale of 1 to 5, and the combined score (1 to 10) represented the index.

The results showed that transvaginal ultrasound had more than a threefold greater sensitivity for ovarian cancer compared with the symptom index (73.3% versus 20%). The symptom index had greater accuracy for identifying benign tumors (91.3% versus 74.4%).

Use of ultrasound and the symptom index in series resulted in a worse sensitivity (16.7%) but improved the ability to distinguish benign tumors (97.9% specificity).

Using receiver operating characteristic (ROC) curve analysis, investigators calculated a line of no discrimination (random guess or outcome) and found that a morphology index >4 to 5 resulted in the greatest distance from the line of no discrimination.

The authors examined clinical decision making on the basis of morphology index scores ?4 and for index scores ?5. They found that adding the symptom index to the ultrasound-derived morphology index increased sensitivity by 3.3% and reduced specificity by 5.8%.

Despite the modest performance of the symptom index, the authors said, “symptoms information cannot be ignored and is an important element in communication between patient and physician.”

“It is important to educate patients that informative symptoms should not be ignored and that the degree to which symptoms are a resultant indicator of early-stage ovarian malignancy has yet to be determined,” they added.

The study was supported by the Telford Foundation and by the Commonwealth of Kentucky Department of Health and Human Services. The authors reported no conflicts of interest.

Primary source: Cancer

Source reference:
Pavlik EJ, et al “The search for meaning-symptoms and transvaginal sonography screening for ovarian cancer” Cancer 2009; DOI: 10.1002/cncr.24407.